nld-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9049346879535558,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.294069576263428,
      "epoch": 0.00045355587808418,
      "grad_norm": 1.0,
      "learning_rate": 2e-06,
      "loss": 6.8082,
      "mean_token_accuracy": 0.10324694961309433,
      "num_tokens": 10827.0,
      "step": 5
    },
    {
      "entropy": 7.281773281097412,
      "epoch": 0.00090711175616836,
      "grad_norm": 1.2109375,
      "learning_rate": 4.5e-06,
      "loss": 6.82,
      "mean_token_accuracy": 0.09896889626979828,
      "num_tokens": 21025.0,
      "step": 10
    },
    {
      "entropy": 7.273117637634277,
      "epoch": 0.00136066763425254,
      "grad_norm": 1.1328125,
      "learning_rate": 7e-06,
      "loss": 6.7665,
      "mean_token_accuracy": 0.10093757957220077,
      "num_tokens": 31678.0,
      "step": 15
    },
    {
      "entropy": 7.247605180740356,
      "epoch": 0.00181422351233672,
      "grad_norm": 1.0390625,
      "learning_rate": 9.5e-06,
      "loss": 6.7109,
      "mean_token_accuracy": 0.11424258798360824,
      "num_tokens": 41025.0,
      "step": 20
    },
    {
      "entropy": 7.302770519256592,
      "epoch": 0.0022677793904208998,
      "grad_norm": 0.984375,
      "learning_rate": 1.2e-05,
      "loss": 6.8182,
      "mean_token_accuracy": 0.10542535930871963,
      "num_tokens": 51270.0,
      "step": 25
    },
    {
      "entropy": 7.143268346786499,
      "epoch": 0.00272133526850508,
      "grad_norm": 0.98828125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.7184,
      "mean_token_accuracy": 0.09942657127976418,
      "num_tokens": 62142.0,
      "step": 30
    },
    {
      "entropy": 7.138898038864136,
      "epoch": 0.00317489114658926,
      "grad_norm": 0.921875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.8522,
      "mean_token_accuracy": 0.1027273491024971,
      "num_tokens": 72709.0,
      "step": 35
    },
    {
      "entropy": 7.083880186080933,
      "epoch": 0.00362844702467344,
      "grad_norm": 0.9140625,
      "learning_rate": 1.95e-05,
      "loss": 6.691,
      "mean_token_accuracy": 0.1096431165933609,
      "num_tokens": 82313.0,
      "step": 40
    },
    {
      "entropy": 7.065870761871338,
      "epoch": 0.00408200290275762,
      "grad_norm": 0.9921875,
      "learning_rate": 2.2e-05,
      "loss": 6.709,
      "mean_token_accuracy": 0.11063554510474205,
      "num_tokens": 92450.0,
      "step": 45
    },
    {
      "entropy": 7.089556646347046,
      "epoch": 0.0045355587808417995,
      "grad_norm": 0.98828125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.7247,
      "mean_token_accuracy": 0.09662125557661057,
      "num_tokens": 103978.0,
      "step": 50
    },
    {
      "entropy": 7.111036014556885,
      "epoch": 0.00498911465892598,
      "grad_norm": 0.8046875,
      "learning_rate": 2.7e-05,
      "loss": 6.6995,
      "mean_token_accuracy": 0.11020737215876579,
      "num_tokens": 114134.0,
      "step": 55
    },
    {
      "entropy": 7.1653160572052,
      "epoch": 0.00544267053701016,
      "grad_norm": 0.9375,
      "learning_rate": 2.95e-05,
      "loss": 6.8947,
      "mean_token_accuracy": 0.09984557479619979,
      "num_tokens": 125037.0,
      "step": 60
    },
    {
      "entropy": 7.146482801437378,
      "epoch": 0.00589622641509434,
      "grad_norm": 1.0234375,
      "learning_rate": 3.2e-05,
      "loss": 6.7116,
      "mean_token_accuracy": 0.11133431643247604,
      "num_tokens": 134721.0,
      "step": 65
    },
    {
      "entropy": 7.114493370056152,
      "epoch": 0.00634978229317852,
      "grad_norm": 0.89453125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.6489,
      "mean_token_accuracy": 0.10958910807967186,
      "num_tokens": 144448.0,
      "step": 70
    },
    {
      "entropy": 7.126422929763794,
      "epoch": 0.006803338171262699,
      "grad_norm": 0.921875,
      "learning_rate": 3.7e-05,
      "loss": 6.7058,
      "mean_token_accuracy": 0.10741295143961907,
      "num_tokens": 154343.0,
      "step": 75
    },
    {
      "entropy": 7.110815334320068,
      "epoch": 0.00725689404934688,
      "grad_norm": 0.8671875,
      "learning_rate": 3.95e-05,
      "loss": 6.7713,
      "mean_token_accuracy": 0.10833674892783166,
      "num_tokens": 165650.0,
      "step": 80
    },
    {
      "entropy": 7.209178161621094,
      "epoch": 0.007710449927431059,
      "grad_norm": 1.0,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.9573,
      "mean_token_accuracy": 0.10000077486038209,
      "num_tokens": 177098.0,
      "step": 85
    },
    {
      "entropy": 7.122071313858032,
      "epoch": 0.00816400580551524,
      "grad_norm": 1.0078125,
      "learning_rate": 4.45e-05,
      "loss": 6.6834,
      "mean_token_accuracy": 0.11381161436438561,
      "num_tokens": 187903.0,
      "step": 90
    },
    {
      "entropy": 7.1197929859161375,
      "epoch": 0.00861756168359942,
      "grad_norm": 0.9453125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.7278,
      "mean_token_accuracy": 0.10795715600252151,
      "num_tokens": 198077.0,
      "step": 95
    },
    {
      "entropy": 7.142152261734009,
      "epoch": 0.009071117561683599,
      "grad_norm": 0.87890625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.8489,
      "mean_token_accuracy": 0.10571764856576919,
      "num_tokens": 208641.0,
      "step": 100
    },
    {
      "entropy": 7.104328584671021,
      "epoch": 0.009524673439767779,
      "grad_norm": 0.9609375,
      "learning_rate": 5.2e-05,
      "loss": 6.7348,
      "mean_token_accuracy": 0.114326973259449,
      "num_tokens": 218398.0,
      "step": 105
    },
    {
      "entropy": 7.09408917427063,
      "epoch": 0.00997822931785196,
      "grad_norm": 0.9296875,
      "learning_rate": 5.45e-05,
      "loss": 6.6445,
      "mean_token_accuracy": 0.10643699690699578,
      "num_tokens": 229059.0,
      "step": 110
    },
    {
      "entropy": 7.079065132141113,
      "epoch": 0.01043178519593614,
      "grad_norm": 0.85546875,
      "learning_rate": 5.7e-05,
      "loss": 6.7188,
      "mean_token_accuracy": 0.10808592811226844,
      "num_tokens": 239775.0,
      "step": 115
    },
    {
      "entropy": 7.048356246948242,
      "epoch": 0.01088534107402032,
      "grad_norm": 0.87109375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.604,
      "mean_token_accuracy": 0.10746061727404595,
      "num_tokens": 249954.0,
      "step": 120
    },
    {
      "entropy": 7.094142627716065,
      "epoch": 0.011338896952104499,
      "grad_norm": 0.88671875,
      "learning_rate": 6.2e-05,
      "loss": 6.7139,
      "mean_token_accuracy": 0.1050918959081173,
      "num_tokens": 260056.0,
      "step": 125
    },
    {
      "entropy": 7.079626655578613,
      "epoch": 0.01179245283018868,
      "grad_norm": 0.98828125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.6984,
      "mean_token_accuracy": 0.1138601079583168,
      "num_tokens": 269904.0,
      "step": 130
    },
    {
      "entropy": 7.136023092269897,
      "epoch": 0.01224600870827286,
      "grad_norm": 0.96484375,
      "learning_rate": 6.7e-05,
      "loss": 6.6975,
      "mean_token_accuracy": 0.10696214884519577,
      "num_tokens": 280900.0,
      "step": 135
    },
    {
      "entropy": 7.108730602264404,
      "epoch": 0.01269956458635704,
      "grad_norm": 0.984375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.7283,
      "mean_token_accuracy": 0.1071628138422966,
      "num_tokens": 290844.0,
      "step": 140
    },
    {
      "entropy": 7.061822891235352,
      "epoch": 0.013153120464441219,
      "grad_norm": 0.94140625,
      "learning_rate": 7.2e-05,
      "loss": 6.7121,
      "mean_token_accuracy": 0.10810653567314148,
      "num_tokens": 301345.0,
      "step": 145
    },
    {
      "entropy": 7.080287647247315,
      "epoch": 0.013606676342525399,
      "grad_norm": 0.84765625,
      "learning_rate": 7.45e-05,
      "loss": 6.7084,
      "mean_token_accuracy": 0.10700314119458199,
      "num_tokens": 312681.0,
      "step": 150
    },
    {
      "entropy": 7.125663185119629,
      "epoch": 0.01406023222060958,
      "grad_norm": 0.8984375,
      "learning_rate": 7.7e-05,
      "loss": 6.6748,
      "mean_token_accuracy": 0.10948033705353737,
      "num_tokens": 323492.0,
      "step": 155
    },
    {
      "entropy": 7.086132001876831,
      "epoch": 0.01451378809869376,
      "grad_norm": 0.98828125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.7303,
      "mean_token_accuracy": 0.10884013846516609,
      "num_tokens": 334521.0,
      "step": 160
    },
    {
      "entropy": 7.097598886489868,
      "epoch": 0.014967343976777939,
      "grad_norm": 0.97265625,
      "learning_rate": 8.2e-05,
      "loss": 6.7547,
      "mean_token_accuracy": 0.11234977021813393,
      "num_tokens": 345468.0,
      "step": 165
    },
    {
      "entropy": 7.158268976211548,
      "epoch": 0.015420899854862119,
      "grad_norm": 1.0546875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.6848,
      "mean_token_accuracy": 0.11324930489063263,
      "num_tokens": 355911.0,
      "step": 170
    },
    {
      "entropy": 7.074994707107544,
      "epoch": 0.0158744557329463,
      "grad_norm": 1.0078125,
      "learning_rate": 8.7e-05,
      "loss": 6.7033,
      "mean_token_accuracy": 0.10836530178785324,
      "num_tokens": 367537.0,
      "step": 175
    },
    {
      "entropy": 7.234744691848755,
      "epoch": 0.01632801161103048,
      "grad_norm": 1.046875,
      "learning_rate": 8.95e-05,
      "loss": 6.8874,
      "mean_token_accuracy": 0.10362846553325653,
      "num_tokens": 377918.0,
      "step": 180
    },
    {
      "entropy": 7.165385055541992,
      "epoch": 0.01678156748911466,
      "grad_norm": 0.91796875,
      "learning_rate": 9.2e-05,
      "loss": 6.7471,
      "mean_token_accuracy": 0.10905997231602668,
      "num_tokens": 388461.0,
      "step": 185
    },
    {
      "entropy": 7.085760498046875,
      "epoch": 0.01723512336719884,
      "grad_norm": 1.078125,
      "learning_rate": 9.45e-05,
      "loss": 6.7294,
      "mean_token_accuracy": 0.1090123176574707,
      "num_tokens": 399771.0,
      "step": 190
    },
    {
      "entropy": 7.09751558303833,
      "epoch": 0.01768867924528302,
      "grad_norm": 0.95703125,
      "learning_rate": 9.7e-05,
      "loss": 6.6457,
      "mean_token_accuracy": 0.11132576763629913,
      "num_tokens": 410254.0,
      "step": 195
    },
    {
      "entropy": 7.104769086837768,
      "epoch": 0.018142235123367198,
      "grad_norm": 1.0,
      "learning_rate": 9.95e-05,
      "loss": 6.6448,
      "mean_token_accuracy": 0.10658138245344162,
      "num_tokens": 419792.0,
      "step": 200
    },
    {
      "entropy": 7.064562034606934,
      "epoch": 0.018595791001451378,
      "grad_norm": 1.015625,
      "learning_rate": 0.000102,
      "loss": 6.6915,
      "mean_token_accuracy": 0.11256536692380906,
      "num_tokens": 429264.0,
      "step": 205
    },
    {
      "entropy": 7.033247804641723,
      "epoch": 0.019049346879535557,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.6445,
      "mean_token_accuracy": 0.11356118023395538,
      "num_tokens": 439624.0,
      "step": 210
    },
    {
      "entropy": 7.0533287525177,
      "epoch": 0.01950290275761974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000107,
      "loss": 6.6641,
      "mean_token_accuracy": 0.10742003992199897,
      "num_tokens": 449556.0,
      "step": 215
    },
    {
      "entropy": 7.127027368545532,
      "epoch": 0.01995645863570392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001095,
      "loss": 6.6976,
      "mean_token_accuracy": 0.1103837251663208,
      "num_tokens": 460243.0,
      "step": 220
    },
    {
      "entropy": 7.084058809280395,
      "epoch": 0.0204100145137881,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000112,
      "loss": 6.7234,
      "mean_token_accuracy": 0.10724443346261978,
      "num_tokens": 469767.0,
      "step": 225
    },
    {
      "entropy": 7.108950233459472,
      "epoch": 0.02086357039187228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001145,
      "loss": 6.6963,
      "mean_token_accuracy": 0.1088191345334053,
      "num_tokens": 480329.0,
      "step": 230
    },
    {
      "entropy": 7.037991762161255,
      "epoch": 0.02131712626995646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.6476,
      "mean_token_accuracy": 0.10804546251893044,
      "num_tokens": 490096.0,
      "step": 235
    },
    {
      "entropy": 7.215046215057373,
      "epoch": 0.02177068214804064,
      "grad_norm": 1.0625,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.6471,
      "mean_token_accuracy": 0.10496766492724419,
      "num_tokens": 501231.0,
      "step": 240
    },
    {
      "entropy": 6.999786281585694,
      "epoch": 0.022224238026124818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000122,
      "loss": 6.6219,
      "mean_token_accuracy": 0.11307241767644882,
      "num_tokens": 510722.0,
      "step": 245
    },
    {
      "entropy": 7.085280036926269,
      "epoch": 0.022677793904208998,
      "grad_norm": 1.0,
      "learning_rate": 0.0001245,
      "loss": 6.6711,
      "mean_token_accuracy": 0.10972743034362793,
      "num_tokens": 521452.0,
      "step": 250
    },
    {
      "entropy": 7.075028562545777,
      "epoch": 0.023131349782293177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000127,
      "loss": 6.6852,
      "mean_token_accuracy": 0.10879904478788376,
      "num_tokens": 532397.0,
      "step": 255
    },
    {
      "entropy": 7.080456399917603,
      "epoch": 0.02358490566037736,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001295,
      "loss": 6.6625,
      "mean_token_accuracy": 0.10898046344518661,
      "num_tokens": 542271.0,
      "step": 260
    },
    {
      "entropy": 7.04259991645813,
      "epoch": 0.02403846153846154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000132,
      "loss": 6.6673,
      "mean_token_accuracy": 0.11369868144392967,
      "num_tokens": 552411.0,
      "step": 265
    },
    {
      "entropy": 7.1117438793182375,
      "epoch": 0.02449201741654572,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.6571,
      "mean_token_accuracy": 0.11307868435978889,
      "num_tokens": 561518.0,
      "step": 270
    },
    {
      "entropy": 7.074210357666016,
      "epoch": 0.0249455732946299,
      "grad_norm": 1.171875,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.7211,
      "mean_token_accuracy": 0.11085665598511696,
      "num_tokens": 571361.0,
      "step": 275
    },
    {
      "entropy": 7.007254123687744,
      "epoch": 0.02539912917271408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001395,
      "loss": 6.615,
      "mean_token_accuracy": 0.11170280948281289,
      "num_tokens": 580926.0,
      "step": 280
    },
    {
      "entropy": 7.089535856246949,
      "epoch": 0.025852685050798258,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.7083,
      "mean_token_accuracy": 0.10579465702176094,
      "num_tokens": 592282.0,
      "step": 285
    },
    {
      "entropy": 7.124145221710205,
      "epoch": 0.026306240928882438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001445,
      "loss": 6.7235,
      "mean_token_accuracy": 0.10901360288262367,
      "num_tokens": 602767.0,
      "step": 290
    },
    {
      "entropy": 7.125961971282959,
      "epoch": 0.026759796806966617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000147,
      "loss": 6.6398,
      "mean_token_accuracy": 0.11005486473441124,
      "num_tokens": 612763.0,
      "step": 295
    },
    {
      "entropy": 6.99384126663208,
      "epoch": 0.027213352685050797,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0001495,
      "loss": 6.6203,
      "mean_token_accuracy": 0.11310217007994652,
      "num_tokens": 622796.0,
      "step": 300
    },
    {
      "entropy": 6.990754795074463,
      "epoch": 0.027666908563134977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000152,
      "loss": 6.5983,
      "mean_token_accuracy": 0.1153037615120411,
      "num_tokens": 633136.0,
      "step": 305
    },
    {
      "entropy": 7.018349838256836,
      "epoch": 0.02812046444121916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.6196,
      "mean_token_accuracy": 0.11451629996299743,
      "num_tokens": 643347.0,
      "step": 310
    },
    {
      "entropy": 7.088607692718506,
      "epoch": 0.02857402031930334,
      "grad_norm": 1.25,
      "learning_rate": 0.000157,
      "loss": 6.6067,
      "mean_token_accuracy": 0.1193146750330925,
      "num_tokens": 653096.0,
      "step": 315
    },
    {
      "entropy": 6.883705759048462,
      "epoch": 0.02902757619738752,
      "grad_norm": 1.203125,
      "learning_rate": 0.0001595,
      "loss": 6.6439,
      "mean_token_accuracy": 0.1120058111846447,
      "num_tokens": 663557.0,
      "step": 320
    },
    {
      "entropy": 7.102012681961059,
      "epoch": 0.0294811320754717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000162,
      "loss": 6.7247,
      "mean_token_accuracy": 0.1098653145134449,
      "num_tokens": 674395.0,
      "step": 325
    },
    {
      "entropy": 7.06192626953125,
      "epoch": 0.029934687953555878,
      "grad_norm": 1.109375,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.6559,
      "mean_token_accuracy": 0.1131248153746128,
      "num_tokens": 684776.0,
      "step": 330
    },
    {
      "entropy": 7.036684370040893,
      "epoch": 0.030388243831640058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.6097,
      "mean_token_accuracy": 0.11125744804739952,
      "num_tokens": 694886.0,
      "step": 335
    },
    {
      "entropy": 7.014414644241333,
      "epoch": 0.030841799709724237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.6881,
      "mean_token_accuracy": 0.11227559223771096,
      "num_tokens": 705426.0,
      "step": 340
    },
    {
      "entropy": 7.103907060623169,
      "epoch": 0.03129535558780842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.6959,
      "mean_token_accuracy": 0.10788564383983612,
      "num_tokens": 716271.0,
      "step": 345
    },
    {
      "entropy": 7.0990581035614015,
      "epoch": 0.0317489114658926,
      "grad_norm": 0.953125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.672,
      "mean_token_accuracy": 0.1117681197822094,
      "num_tokens": 726875.0,
      "step": 350
    },
    {
      "entropy": 6.9609973430633545,
      "epoch": 0.032202467343976776,
      "grad_norm": 1.015625,
      "learning_rate": 0.000177,
      "loss": 6.6149,
      "mean_token_accuracy": 0.11049453616142273,
      "num_tokens": 737825.0,
      "step": 355
    },
    {
      "entropy": 7.027775573730469,
      "epoch": 0.03265602322206096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001795,
      "loss": 6.5764,
      "mean_token_accuracy": 0.11501729264855384,
      "num_tokens": 747424.0,
      "step": 360
    },
    {
      "entropy": 7.032524490356446,
      "epoch": 0.033109579100145135,
      "grad_norm": 1.140625,
      "learning_rate": 0.000182,
      "loss": 6.6177,
      "mean_token_accuracy": 0.11091509833931923,
      "num_tokens": 757649.0,
      "step": 365
    },
    {
      "entropy": 7.0149918556213375,
      "epoch": 0.03356313497822932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001845,
      "loss": 6.6171,
      "mean_token_accuracy": 0.11376268118619919,
      "num_tokens": 768556.0,
      "step": 370
    },
    {
      "entropy": 7.010801887512207,
      "epoch": 0.034016690856313495,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000187,
      "loss": 6.607,
      "mean_token_accuracy": 0.1097978562116623,
      "num_tokens": 778331.0,
      "step": 375
    },
    {
      "entropy": 6.967943716049194,
      "epoch": 0.03447024673439768,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 6.6039,
      "mean_token_accuracy": 0.1124206118285656,
      "num_tokens": 788065.0,
      "step": 380
    },
    {
      "entropy": 7.05260066986084,
      "epoch": 0.03492380261248186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000192,
      "loss": 6.6727,
      "mean_token_accuracy": 0.11531179398298264,
      "num_tokens": 798912.0,
      "step": 385
    },
    {
      "entropy": 7.010505723953247,
      "epoch": 0.03537735849056604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001945,
      "loss": 6.5932,
      "mean_token_accuracy": 0.11802476570010186,
      "num_tokens": 809580.0,
      "step": 390
    },
    {
      "entropy": 7.041739130020142,
      "epoch": 0.03583091436865022,
      "grad_norm": 1.09375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.698,
      "mean_token_accuracy": 0.11100108623504638,
      "num_tokens": 820207.0,
      "step": 395
    },
    {
      "entropy": 6.895339822769165,
      "epoch": 0.036284470246734396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.5155,
      "mean_token_accuracy": 0.11325377970933914,
      "num_tokens": 831038.0,
      "step": 400
    },
    {
      "entropy": 7.01606822013855,
      "epoch": 0.03673802612481858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000202,
      "loss": 6.619,
      "mean_token_accuracy": 0.11496082544326783,
      "num_tokens": 842144.0,
      "step": 405
    },
    {
      "entropy": 7.006781196594238,
      "epoch": 0.037191582002902755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.6274,
      "mean_token_accuracy": 0.11621586382389068,
      "num_tokens": 851315.0,
      "step": 410
    },
    {
      "entropy": 6.954495906829834,
      "epoch": 0.03764513788098694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000207,
      "loss": 6.6407,
      "mean_token_accuracy": 0.11225903108716011,
      "num_tokens": 861233.0,
      "step": 415
    },
    {
      "entropy": 6.989374589920044,
      "epoch": 0.038098693759071114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002095,
      "loss": 6.6035,
      "mean_token_accuracy": 0.11319806352257729,
      "num_tokens": 871131.0,
      "step": 420
    },
    {
      "entropy": 6.961730718612671,
      "epoch": 0.0385522496371553,
      "grad_norm": 1.109375,
      "learning_rate": 0.000212,
      "loss": 6.6019,
      "mean_token_accuracy": 0.10756484866142273,
      "num_tokens": 882195.0,
      "step": 425
    },
    {
      "entropy": 7.033092260360718,
      "epoch": 0.03900580551523948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002145,
      "loss": 6.5797,
      "mean_token_accuracy": 0.11503332033753395,
      "num_tokens": 891806.0,
      "step": 430
    },
    {
      "entropy": 6.988789701461792,
      "epoch": 0.03945936139332366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6323,
      "mean_token_accuracy": 0.10891624465584755,
      "num_tokens": 901350.0,
      "step": 435
    },
    {
      "entropy": 7.0045853614807125,
      "epoch": 0.03991291727140784,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002195,
      "loss": 6.5539,
      "mean_token_accuracy": 0.11177553907036782,
      "num_tokens": 912061.0,
      "step": 440
    },
    {
      "entropy": 6.900162267684936,
      "epoch": 0.040366473149492016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000222,
      "loss": 6.5201,
      "mean_token_accuracy": 0.11248114556074143,
      "num_tokens": 922220.0,
      "step": 445
    },
    {
      "entropy": 6.956882047653198,
      "epoch": 0.0408200290275762,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002245,
      "loss": 6.5593,
      "mean_token_accuracy": 0.11303049996495247,
      "num_tokens": 932319.0,
      "step": 450
    },
    {
      "entropy": 6.911965465545654,
      "epoch": 0.041273584905660375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.58,
      "mean_token_accuracy": 0.11454110518097878,
      "num_tokens": 943254.0,
      "step": 455
    },
    {
      "entropy": 6.996125650405884,
      "epoch": 0.04172714078374456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6029,
      "mean_token_accuracy": 0.10846876502037048,
      "num_tokens": 953921.0,
      "step": 460
    },
    {
      "entropy": 6.936468839645386,
      "epoch": 0.042180696661828734,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.526,
      "mean_token_accuracy": 0.11692283749580383,
      "num_tokens": 963832.0,
      "step": 465
    },
    {
      "entropy": 6.945839929580688,
      "epoch": 0.04263425253991292,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6208,
      "mean_token_accuracy": 0.10841431692242623,
      "num_tokens": 974066.0,
      "step": 470
    },
    {
      "entropy": 6.90062198638916,
      "epoch": 0.0430878084179971,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000237,
      "loss": 6.616,
      "mean_token_accuracy": 0.11435005068778992,
      "num_tokens": 984402.0,
      "step": 475
    },
    {
      "entropy": 6.956665706634522,
      "epoch": 0.04354136429608128,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002395,
      "loss": 6.5122,
      "mean_token_accuracy": 0.11360828503966332,
      "num_tokens": 993555.0,
      "step": 480
    },
    {
      "entropy": 6.95721468925476,
      "epoch": 0.04399492017416546,
      "grad_norm": 1.140625,
      "learning_rate": 0.000242,
      "loss": 6.5271,
      "mean_token_accuracy": 0.12218032553791999,
      "num_tokens": 1002920.0,
      "step": 485
    },
    {
      "entropy": 6.885728740692139,
      "epoch": 0.044448476052249636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002445,
      "loss": 6.608,
      "mean_token_accuracy": 0.10632575079798698,
      "num_tokens": 1013611.0,
      "step": 490
    },
    {
      "entropy": 6.892173099517822,
      "epoch": 0.04490203193033382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000247,
      "loss": 6.4836,
      "mean_token_accuracy": 0.11247801780700684,
      "num_tokens": 1024121.0,
      "step": 495
    },
    {
      "entropy": 7.059810018539428,
      "epoch": 0.045355587808417995,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002495,
      "loss": 6.7215,
      "mean_token_accuracy": 0.1127111479640007,
      "num_tokens": 1033845.0,
      "step": 500
    },
    {
      "entropy": 6.917029905319214,
      "epoch": 0.04580914368650218,
      "grad_norm": 1.015625,
      "learning_rate": 0.000252,
      "loss": 6.5315,
      "mean_token_accuracy": 0.11300075054168701,
      "num_tokens": 1044642.0,
      "step": 505
    },
    {
      "entropy": 6.874821615219116,
      "epoch": 0.046262699564586354,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002545,
      "loss": 6.6032,
      "mean_token_accuracy": 0.10996716171503067,
      "num_tokens": 1054728.0,
      "step": 510
    },
    {
      "entropy": 6.894815063476562,
      "epoch": 0.04671625544267054,
      "grad_norm": 1.03125,
      "learning_rate": 0.000257,
      "loss": 6.558,
      "mean_token_accuracy": 0.10969225391745567,
      "num_tokens": 1065743.0,
      "step": 515
    },
    {
      "entropy": 6.828075170516968,
      "epoch": 0.04716981132075472,
      "grad_norm": 1.125,
      "learning_rate": 0.0002595,
      "loss": 6.4864,
      "mean_token_accuracy": 0.11496817469596862,
      "num_tokens": 1076798.0,
      "step": 520
    },
    {
      "entropy": 6.936053895950318,
      "epoch": 0.047623367198838897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000262,
      "loss": 6.6064,
      "mean_token_accuracy": 0.11198530569672585,
      "num_tokens": 1087445.0,
      "step": 525
    },
    {
      "entropy": 6.88051404953003,
      "epoch": 0.04807692307692308,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5258,
      "mean_token_accuracy": 0.11618097051978112,
      "num_tokens": 1098379.0,
      "step": 530
    },
    {
      "entropy": 6.968464517593384,
      "epoch": 0.048530478955007256,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6118,
      "mean_token_accuracy": 0.11241324171423912,
      "num_tokens": 1108016.0,
      "step": 535
    },
    {
      "entropy": 6.838916206359864,
      "epoch": 0.04898403483309144,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.4859,
      "mean_token_accuracy": 0.1130056694149971,
      "num_tokens": 1117887.0,
      "step": 540
    },
    {
      "entropy": 6.830152797698974,
      "epoch": 0.049437590711175615,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.4995,
      "mean_token_accuracy": 0.106851976364851,
      "num_tokens": 1128497.0,
      "step": 545
    },
    {
      "entropy": 6.847278594970703,
      "epoch": 0.0498911465892598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002745,
      "loss": 6.5224,
      "mean_token_accuracy": 0.11530316844582558,
      "num_tokens": 1138168.0,
      "step": 550
    },
    {
      "entropy": 6.883500242233277,
      "epoch": 0.050344702467343974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000277,
      "loss": 6.5128,
      "mean_token_accuracy": 0.11332335695624352,
      "num_tokens": 1148357.0,
      "step": 555
    },
    {
      "entropy": 6.851631021499633,
      "epoch": 0.05079825834542816,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002795,
      "loss": 6.5414,
      "mean_token_accuracy": 0.11176767796278,
      "num_tokens": 1159964.0,
      "step": 560
    },
    {
      "entropy": 6.893160057067871,
      "epoch": 0.05125181422351233,
      "grad_norm": 1.125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.5865,
      "mean_token_accuracy": 0.11133491396903991,
      "num_tokens": 1170683.0,
      "step": 565
    },
    {
      "entropy": 6.922141647338867,
      "epoch": 0.051705370101596516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002845,
      "loss": 6.552,
      "mean_token_accuracy": 0.11003944724798202,
      "num_tokens": 1181680.0,
      "step": 570
    },
    {
      "entropy": 6.8009681224823,
      "epoch": 0.0521589259796807,
      "grad_norm": 1.109375,
      "learning_rate": 0.000287,
      "loss": 6.4299,
      "mean_token_accuracy": 0.11521938815712929,
      "num_tokens": 1191619.0,
      "step": 575
    },
    {
      "entropy": 6.7351405143737795,
      "epoch": 0.052612481857764876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002895,
      "loss": 6.4436,
      "mean_token_accuracy": 0.1167814739048481,
      "num_tokens": 1202264.0,
      "step": 580
    },
    {
      "entropy": 6.800995492935181,
      "epoch": 0.05306603773584906,
      "grad_norm": 1.171875,
      "learning_rate": 0.000292,
      "loss": 6.4535,
      "mean_token_accuracy": 0.11420593708753586,
      "num_tokens": 1212119.0,
      "step": 585
    },
    {
      "entropy": 6.896333932876587,
      "epoch": 0.053519593613933235,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002945,
      "loss": 6.5814,
      "mean_token_accuracy": 0.11330544948577881,
      "num_tokens": 1222317.0,
      "step": 590
    },
    {
      "entropy": 6.882574033737183,
      "epoch": 0.05397314949201742,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000297,
      "loss": 6.5286,
      "mean_token_accuracy": 0.11739532127976418,
      "num_tokens": 1232914.0,
      "step": 595
    },
    {
      "entropy": 6.691092920303345,
      "epoch": 0.054426705370101594,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002995,
      "loss": 6.3838,
      "mean_token_accuracy": 0.11833471432328224,
      "num_tokens": 1244214.0,
      "step": 600
    },
    {
      "entropy": 6.853950929641724,
      "epoch": 0.05488026124818578,
      "grad_norm": 1.171875,
      "learning_rate": 0.000302,
      "loss": 6.4824,
      "mean_token_accuracy": 0.11780562847852707,
      "num_tokens": 1254626.0,
      "step": 605
    },
    {
      "entropy": 6.790469121932984,
      "epoch": 0.05533381712626995,
      "grad_norm": 1.25,
      "learning_rate": 0.0003045,
      "loss": 6.5719,
      "mean_token_accuracy": 0.11432634964585305,
      "num_tokens": 1265461.0,
      "step": 610
    },
    {
      "entropy": 6.8617181301116945,
      "epoch": 0.055787373004354136,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000307,
      "loss": 6.4572,
      "mean_token_accuracy": 0.1158202901482582,
      "num_tokens": 1275810.0,
      "step": 615
    },
    {
      "entropy": 6.72202844619751,
      "epoch": 0.05624092888243832,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003095,
      "loss": 6.4716,
      "mean_token_accuracy": 0.1191383220255375,
      "num_tokens": 1285170.0,
      "step": 620
    },
    {
      "entropy": 6.832190465927124,
      "epoch": 0.056694484760522496,
      "grad_norm": 1.015625,
      "learning_rate": 0.000312,
      "loss": 6.4065,
      "mean_token_accuracy": 0.11602196991443633,
      "num_tokens": 1295460.0,
      "step": 625
    },
    {
      "entropy": 6.798471832275391,
      "epoch": 0.05714804063860668,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003145,
      "loss": 6.5401,
      "mean_token_accuracy": 0.11370082646608352,
      "num_tokens": 1306656.0,
      "step": 630
    },
    {
      "entropy": 6.835934972763061,
      "epoch": 0.057601596516690855,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000317,
      "loss": 6.5059,
      "mean_token_accuracy": 0.11711395084857941,
      "num_tokens": 1318019.0,
      "step": 635
    },
    {
      "entropy": 6.653035640716553,
      "epoch": 0.05805515239477504,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003195,
      "loss": 6.4287,
      "mean_token_accuracy": 0.11944135650992393,
      "num_tokens": 1328120.0,
      "step": 640
    },
    {
      "entropy": 6.86923885345459,
      "epoch": 0.058508708272859214,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000322,
      "loss": 6.527,
      "mean_token_accuracy": 0.11324021890759468,
      "num_tokens": 1338354.0,
      "step": 645
    },
    {
      "entropy": 6.790145397186279,
      "epoch": 0.0589622641509434,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.4831,
      "mean_token_accuracy": 0.11808143481612206,
      "num_tokens": 1348666.0,
      "step": 650
    },
    {
      "entropy": 6.80904107093811,
      "epoch": 0.05941582002902757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5404,
      "mean_token_accuracy": 0.1175168700516224,
      "num_tokens": 1359713.0,
      "step": 655
    },
    {
      "entropy": 6.7852167129516605,
      "epoch": 0.059869375907111756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.4064,
      "mean_token_accuracy": 0.12213529795408248,
      "num_tokens": 1369507.0,
      "step": 660
    },
    {
      "entropy": 6.834307098388672,
      "epoch": 0.06032293178519594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5583,
      "mean_token_accuracy": 0.11135573908686638,
      "num_tokens": 1380726.0,
      "step": 665
    },
    {
      "entropy": 6.805551242828369,
      "epoch": 0.060776487663280115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4852,
      "mean_token_accuracy": 0.11822790279984474,
      "num_tokens": 1391423.0,
      "step": 670
    },
    {
      "entropy": 6.7903587341308596,
      "epoch": 0.0612300435413643,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000337,
      "loss": 6.4651,
      "mean_token_accuracy": 0.1208954781293869,
      "num_tokens": 1402492.0,
      "step": 675
    },
    {
      "entropy": 6.778665447235108,
      "epoch": 0.061683599419448475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003395,
      "loss": 6.5073,
      "mean_token_accuracy": 0.11749565601348877,
      "num_tokens": 1412497.0,
      "step": 680
    },
    {
      "entropy": 6.856817674636841,
      "epoch": 0.06213715529753266,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000342,
      "loss": 6.4803,
      "mean_token_accuracy": 0.1222834125161171,
      "num_tokens": 1423214.0,
      "step": 685
    },
    {
      "entropy": 6.691046953201294,
      "epoch": 0.06259071117561683,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4779,
      "mean_token_accuracy": 0.1166260376572609,
      "num_tokens": 1434044.0,
      "step": 690
    },
    {
      "entropy": 6.79459776878357,
      "epoch": 0.06304426705370102,
      "grad_norm": 1.109375,
      "learning_rate": 0.000347,
      "loss": 6.4429,
      "mean_token_accuracy": 0.11843409463763237,
      "num_tokens": 1444387.0,
      "step": 695
    },
    {
      "entropy": 6.704948234558105,
      "epoch": 0.0634978229317852,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003495,
      "loss": 6.3861,
      "mean_token_accuracy": 0.11568867340683937,
      "num_tokens": 1455157.0,
      "step": 700
    },
    {
      "entropy": 6.735180950164795,
      "epoch": 0.06395137880986937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000352,
      "loss": 6.477,
      "mean_token_accuracy": 0.12313677221536637,
      "num_tokens": 1466145.0,
      "step": 705
    },
    {
      "entropy": 6.7879307746887205,
      "epoch": 0.06440493468795355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003545,
      "loss": 6.4737,
      "mean_token_accuracy": 0.1173794686794281,
      "num_tokens": 1476713.0,
      "step": 710
    },
    {
      "entropy": 6.802341651916504,
      "epoch": 0.06485849056603774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000357,
      "loss": 6.482,
      "mean_token_accuracy": 0.12098045200109482,
      "num_tokens": 1486908.0,
      "step": 715
    },
    {
      "entropy": 6.710763883590698,
      "epoch": 0.06531204644412192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003595,
      "loss": 6.4084,
      "mean_token_accuracy": 0.1253804959356785,
      "num_tokens": 1498185.0,
      "step": 720
    },
    {
      "entropy": 6.715804719924927,
      "epoch": 0.0657656023222061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000362,
      "loss": 6.4454,
      "mean_token_accuracy": 0.11805666983127594,
      "num_tokens": 1508376.0,
      "step": 725
    },
    {
      "entropy": 6.765777683258056,
      "epoch": 0.06621915820029027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003645,
      "loss": 6.5331,
      "mean_token_accuracy": 0.11824854388833046,
      "num_tokens": 1519953.0,
      "step": 730
    },
    {
      "entropy": 6.677792072296143,
      "epoch": 0.06667271407837445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000367,
      "loss": 6.4204,
      "mean_token_accuracy": 0.11954097524285316,
      "num_tokens": 1529793.0,
      "step": 735
    },
    {
      "entropy": 6.716114711761475,
      "epoch": 0.06712626995645864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003695,
      "loss": 6.4162,
      "mean_token_accuracy": 0.12043775469064713,
      "num_tokens": 1540859.0,
      "step": 740
    },
    {
      "entropy": 6.67792387008667,
      "epoch": 0.06757982583454282,
      "grad_norm": 1.0625,
      "learning_rate": 0.000372,
      "loss": 6.3369,
      "mean_token_accuracy": 0.127581225335598,
      "num_tokens": 1549783.0,
      "step": 745
    },
    {
      "entropy": 6.635021018981933,
      "epoch": 0.06803338171262699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003745,
      "loss": 6.4073,
      "mean_token_accuracy": 0.12199747264385223,
      "num_tokens": 1560522.0,
      "step": 750
    },
    {
      "entropy": 6.677228021621704,
      "epoch": 0.06848693759071117,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000377,
      "loss": 6.3805,
      "mean_token_accuracy": 0.12109958082437515,
      "num_tokens": 1570668.0,
      "step": 755
    },
    {
      "entropy": 6.705593681335449,
      "epoch": 0.06894049346879536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003795,
      "loss": 6.4978,
      "mean_token_accuracy": 0.1242837056517601,
      "num_tokens": 1581523.0,
      "step": 760
    },
    {
      "entropy": 6.739868640899658,
      "epoch": 0.06939404934687954,
      "grad_norm": 1.140625,
      "learning_rate": 0.000382,
      "loss": 6.4013,
      "mean_token_accuracy": 0.1199745386838913,
      "num_tokens": 1591356.0,
      "step": 765
    },
    {
      "entropy": 6.6305756092071535,
      "epoch": 0.06984760522496372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003845,
      "loss": 6.3821,
      "mean_token_accuracy": 0.12265411987900735,
      "num_tokens": 1601779.0,
      "step": 770
    },
    {
      "entropy": 6.71905837059021,
      "epoch": 0.07030116110304789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3932,
      "mean_token_accuracy": 0.12244714498519897,
      "num_tokens": 1611995.0,
      "step": 775
    },
    {
      "entropy": 6.672496271133423,
      "epoch": 0.07075471698113207,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.4211,
      "mean_token_accuracy": 0.1178926706314087,
      "num_tokens": 1621834.0,
      "step": 780
    },
    {
      "entropy": 6.678148460388184,
      "epoch": 0.07120827285921626,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4579,
      "mean_token_accuracy": 0.11523651331663132,
      "num_tokens": 1632881.0,
      "step": 785
    },
    {
      "entropy": 6.615700531005859,
      "epoch": 0.07166182873730044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4437,
      "mean_token_accuracy": 0.11981857419013978,
      "num_tokens": 1642870.0,
      "step": 790
    },
    {
      "entropy": 6.719856405258179,
      "epoch": 0.07211538461538461,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3202,
      "mean_token_accuracy": 0.12486634775996208,
      "num_tokens": 1653390.0,
      "step": 795
    },
    {
      "entropy": 6.682466459274292,
      "epoch": 0.07256894049346879,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003995,
      "loss": 6.446,
      "mean_token_accuracy": 0.12393867447972298,
      "num_tokens": 1663382.0,
      "step": 800
    },
    {
      "entropy": 6.698669910430908,
      "epoch": 0.07302249637155298,
      "grad_norm": 1.046875,
      "learning_rate": 0.000402,
      "loss": 6.449,
      "mean_token_accuracy": 0.1184344507753849,
      "num_tokens": 1673601.0,
      "step": 805
    },
    {
      "entropy": 6.570560455322266,
      "epoch": 0.07347605224963716,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004045,
      "loss": 6.3681,
      "mean_token_accuracy": 0.126424128562212,
      "num_tokens": 1683565.0,
      "step": 810
    },
    {
      "entropy": 6.632832860946655,
      "epoch": 0.07392960812772134,
      "grad_norm": 1.828125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.3367,
      "mean_token_accuracy": 0.12308898121118546,
      "num_tokens": 1693636.0,
      "step": 815
    },
    {
      "entropy": 6.7626114845275875,
      "epoch": 0.07438316400580551,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004095,
      "loss": 6.4517,
      "mean_token_accuracy": 0.11969177201390266,
      "num_tokens": 1703415.0,
      "step": 820
    },
    {
      "entropy": 6.699670362472534,
      "epoch": 0.0748367198838897,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000412,
      "loss": 6.4305,
      "mean_token_accuracy": 0.12528733313083648,
      "num_tokens": 1713579.0,
      "step": 825
    },
    {
      "entropy": 6.586101007461548,
      "epoch": 0.07529027576197388,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004145,
      "loss": 6.4246,
      "mean_token_accuracy": 0.11557562425732612,
      "num_tokens": 1724583.0,
      "step": 830
    },
    {
      "entropy": 6.649460124969482,
      "epoch": 0.07574383164005806,
      "grad_norm": 1.125,
      "learning_rate": 0.000417,
      "loss": 6.349,
      "mean_token_accuracy": 0.12887958213686942,
      "num_tokens": 1735293.0,
      "step": 835
    },
    {
      "entropy": 6.688171482086181,
      "epoch": 0.07619738751814223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004195,
      "loss": 6.366,
      "mean_token_accuracy": 0.123409453779459,
      "num_tokens": 1745623.0,
      "step": 840
    },
    {
      "entropy": 6.583874368667603,
      "epoch": 0.07665094339622641,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000422,
      "loss": 6.4085,
      "mean_token_accuracy": 0.12038984745740891,
      "num_tokens": 1755713.0,
      "step": 845
    },
    {
      "entropy": 6.691501235961914,
      "epoch": 0.0771044992743106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004245,
      "loss": 6.3932,
      "mean_token_accuracy": 0.12171499505639076,
      "num_tokens": 1765264.0,
      "step": 850
    },
    {
      "entropy": 6.6751635551452635,
      "epoch": 0.07755805515239478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000427,
      "loss": 6.4043,
      "mean_token_accuracy": 0.12203977331519127,
      "num_tokens": 1775769.0,
      "step": 855
    },
    {
      "entropy": 6.57877140045166,
      "epoch": 0.07801161103047896,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004295,
      "loss": 6.3398,
      "mean_token_accuracy": 0.12160504087805749,
      "num_tokens": 1786245.0,
      "step": 860
    },
    {
      "entropy": 6.531408166885376,
      "epoch": 0.07846516690856313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000432,
      "loss": 6.3694,
      "mean_token_accuracy": 0.1285658784210682,
      "num_tokens": 1796040.0,
      "step": 865
    },
    {
      "entropy": 6.592545700073242,
      "epoch": 0.07891872278664731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004345,
      "loss": 6.3967,
      "mean_token_accuracy": 0.12193747386336326,
      "num_tokens": 1806007.0,
      "step": 870
    },
    {
      "entropy": 6.7914567470550535,
      "epoch": 0.0793722786647315,
      "grad_norm": 1.15625,
      "learning_rate": 0.000437,
      "loss": 6.4675,
      "mean_token_accuracy": 0.11332392543554307,
      "num_tokens": 1816935.0,
      "step": 875
    },
    {
      "entropy": 6.508628511428833,
      "epoch": 0.07982583454281568,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004395,
      "loss": 6.3239,
      "mean_token_accuracy": 0.1232911117374897,
      "num_tokens": 1827419.0,
      "step": 880
    },
    {
      "entropy": 6.673329305648804,
      "epoch": 0.08027939042089985,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000442,
      "loss": 6.4276,
      "mean_token_accuracy": 0.11844546273350716,
      "num_tokens": 1837933.0,
      "step": 885
    },
    {
      "entropy": 6.630283260345459,
      "epoch": 0.08073294629898403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004445,
      "loss": 6.4793,
      "mean_token_accuracy": 0.12463728785514831,
      "num_tokens": 1848551.0,
      "step": 890
    },
    {
      "entropy": 6.692006587982178,
      "epoch": 0.08118650217706821,
      "grad_norm": 1.125,
      "learning_rate": 0.000447,
      "loss": 6.3544,
      "mean_token_accuracy": 0.11869508177042007,
      "num_tokens": 1858380.0,
      "step": 895
    },
    {
      "entropy": 6.5042383670806885,
      "epoch": 0.0816400580551524,
      "grad_norm": 1.125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2822,
      "mean_token_accuracy": 0.12670984119176865,
      "num_tokens": 1868291.0,
      "step": 900
    },
    {
      "entropy": 6.616396141052246,
      "epoch": 0.08209361393323658,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.386,
      "mean_token_accuracy": 0.12300045192241668,
      "num_tokens": 1879219.0,
      "step": 905
    },
    {
      "entropy": 6.637605619430542,
      "epoch": 0.08254716981132075,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3695,
      "mean_token_accuracy": 0.12218734100461007,
      "num_tokens": 1889809.0,
      "step": 910
    },
    {
      "entropy": 6.52483057975769,
      "epoch": 0.08300072568940493,
      "grad_norm": 1.125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3383,
      "mean_token_accuracy": 0.12149963527917862,
      "num_tokens": 1899986.0,
      "step": 915
    },
    {
      "entropy": 6.640835094451904,
      "epoch": 0.08345428156748912,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3212,
      "mean_token_accuracy": 0.12286782488226891,
      "num_tokens": 1909314.0,
      "step": 920
    },
    {
      "entropy": 6.6047581195831295,
      "epoch": 0.0839078374455733,
      "grad_norm": 1.078125,
      "learning_rate": 0.000462,
      "loss": 6.3726,
      "mean_token_accuracy": 0.12332175076007842,
      "num_tokens": 1918443.0,
      "step": 925
    },
    {
      "entropy": 6.563626766204834,
      "epoch": 0.08436139332365747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004645,
      "loss": 6.316,
      "mean_token_accuracy": 0.12443615049123764,
      "num_tokens": 1928683.0,
      "step": 930
    },
    {
      "entropy": 6.458373975753784,
      "epoch": 0.08481494920174165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000467,
      "loss": 6.233,
      "mean_token_accuracy": 0.12297066450119018,
      "num_tokens": 1938293.0,
      "step": 935
    },
    {
      "entropy": 6.509970378875733,
      "epoch": 0.08526850507982583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004695,
      "loss": 6.3464,
      "mean_token_accuracy": 0.12207101732492447,
      "num_tokens": 1948233.0,
      "step": 940
    },
    {
      "entropy": 6.600475120544433,
      "epoch": 0.08572206095791002,
      "grad_norm": 1.125,
      "learning_rate": 0.000472,
      "loss": 6.2553,
      "mean_token_accuracy": 0.12516278997063637,
      "num_tokens": 1958611.0,
      "step": 945
    },
    {
      "entropy": 6.593886280059815,
      "epoch": 0.0861756168359942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745,
      "loss": 6.4088,
      "mean_token_accuracy": 0.1253935366868973,
      "num_tokens": 1968957.0,
      "step": 950
    },
    {
      "entropy": 6.540861320495606,
      "epoch": 0.08662917271407837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000477,
      "loss": 6.2584,
      "mean_token_accuracy": 0.12826553955674172,
      "num_tokens": 1978503.0,
      "step": 955
    },
    {
      "entropy": 6.460687065124512,
      "epoch": 0.08708272859216255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004795,
      "loss": 6.2557,
      "mean_token_accuracy": 0.12641479223966598,
      "num_tokens": 1988458.0,
      "step": 960
    },
    {
      "entropy": 6.554324579238892,
      "epoch": 0.08753628447024674,
      "grad_norm": 1.09375,
      "learning_rate": 0.000482,
      "loss": 6.3931,
      "mean_token_accuracy": 0.12135342583060264,
      "num_tokens": 1999984.0,
      "step": 965
    },
    {
      "entropy": 6.4496111392974855,
      "epoch": 0.08798984034833092,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845,
      "loss": 6.2431,
      "mean_token_accuracy": 0.13175339475274087,
      "num_tokens": 2010645.0,
      "step": 970
    },
    {
      "entropy": 6.600621032714844,
      "epoch": 0.08844339622641509,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000487,
      "loss": 6.36,
      "mean_token_accuracy": 0.12460424080491066,
      "num_tokens": 2020196.0,
      "step": 975
    },
    {
      "entropy": 6.53763222694397,
      "epoch": 0.08889695210449927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895,
      "loss": 6.2527,
      "mean_token_accuracy": 0.12676869034767152,
      "num_tokens": 2030101.0,
      "step": 980
    },
    {
      "entropy": 6.473430156707764,
      "epoch": 0.08935050798258345,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000492,
      "loss": 6.3426,
      "mean_token_accuracy": 0.1231530986726284,
      "num_tokens": 2040454.0,
      "step": 985
    },
    {
      "entropy": 6.448872709274292,
      "epoch": 0.08980406386066764,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945,
      "loss": 6.2774,
      "mean_token_accuracy": 0.12086713090538978,
      "num_tokens": 2050378.0,
      "step": 990
    },
    {
      "entropy": 6.557286500930786,
      "epoch": 0.09025761973875182,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497,
      "loss": 6.3341,
      "mean_token_accuracy": 0.12634406611323357,
      "num_tokens": 2060282.0,
      "step": 995
    },
    {
      "entropy": 6.511825323104858,
      "epoch": 0.09071117561683599,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995,
      "loss": 6.3212,
      "mean_token_accuracy": 0.12492988184094429,
      "num_tokens": 2070563.0,
      "step": 1000
    },
    {
      "entropy": 6.540437269210815,
      "epoch": 0.09116473149492017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999999985110217,
      "loss": 6.3931,
      "mean_token_accuracy": 0.12742957547307016,
      "num_tokens": 2080153.0,
      "step": 1005
    },
    {
      "entropy": 6.470778942108154,
      "epoch": 0.09161828737300436,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999999924620471,
      "loss": 6.236,
      "mean_token_accuracy": 0.12746582329273223,
      "num_tokens": 2090590.0,
      "step": 1010
    },
    {
      "entropy": 6.5432836532592775,
      "epoch": 0.09207184325108854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999999817600155,
      "loss": 6.318,
      "mean_token_accuracy": 0.11900549307465554,
      "num_tokens": 2101945.0,
      "step": 1015
    },
    {
      "entropy": 6.5527753829956055,
      "epoch": 0.09252539912917271,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999664049268,
      "loss": 6.3201,
      "mean_token_accuracy": 0.12768215388059617,
      "num_tokens": 2112514.0,
      "step": 1020
    },
    {
      "entropy": 6.417323160171509,
      "epoch": 0.09297895500725689,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999463967816,
      "loss": 6.2586,
      "mean_token_accuracy": 0.12710441127419472,
      "num_tokens": 2122296.0,
      "step": 1025
    },
    {
      "entropy": 6.639082574844361,
      "epoch": 0.09343251088534107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049999992173558,
      "loss": 6.4053,
      "mean_token_accuracy": 0.11305732280015945,
      "num_tokens": 2134512.0,
      "step": 1030
    },
    {
      "entropy": 6.5072770595550535,
      "epoch": 0.09388606676342526,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999998924213228,
      "loss": 6.3261,
      "mean_token_accuracy": 0.12832158356904982,
      "num_tokens": 2143934.0,
      "step": 1035
    },
    {
      "entropy": 6.44057936668396,
      "epoch": 0.09433962264150944,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999998584540105,
      "loss": 6.2249,
      "mean_token_accuracy": 0.12571750357747077,
      "num_tokens": 2155730.0,
      "step": 1040
    },
    {
      "entropy": 6.528116464614868,
      "epoch": 0.09479317851959361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998198336437,
      "loss": 6.2518,
      "mean_token_accuracy": 0.12236737087368965,
      "num_tokens": 2166422.0,
      "step": 1045
    },
    {
      "entropy": 6.461041498184204,
      "epoch": 0.09524673439767779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999997765602233,
      "loss": 6.3449,
      "mean_token_accuracy": 0.12111935913562774,
      "num_tokens": 2177842.0,
      "step": 1050
    },
    {
      "entropy": 6.55608811378479,
      "epoch": 0.09570029027576198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999997286337502,
      "loss": 6.425,
      "mean_token_accuracy": 0.12338349744677543,
      "num_tokens": 2188614.0,
      "step": 1055
    },
    {
      "entropy": 6.552207946777344,
      "epoch": 0.09615384615384616,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999996760542254,
      "loss": 6.2222,
      "mean_token_accuracy": 0.12883713990449905,
      "num_tokens": 2197886.0,
      "step": 1060
    },
    {
      "entropy": 6.3705360889434814,
      "epoch": 0.09660740203193033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049999961882165,
      "loss": 6.2774,
      "mean_token_accuracy": 0.12635389864444732,
      "num_tokens": 2209930.0,
      "step": 1065
    },
    {
      "entropy": 6.534476470947266,
      "epoch": 0.09706095791001451,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999995569360251,
      "loss": 6.2868,
      "mean_token_accuracy": 0.12291447669267655,
      "num_tokens": 2219767.0,
      "step": 1070
    },
    {
      "entropy": 6.428853178024292,
      "epoch": 0.0975145137880987,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499999490397352,
      "loss": 6.2117,
      "mean_token_accuracy": 0.1293780729174614,
      "num_tokens": 2230082.0,
      "step": 1075
    },
    {
      "entropy": 6.476440477371216,
      "epoch": 0.09796806966618288,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999994192056321,
      "loss": 6.2369,
      "mean_token_accuracy": 0.13364363461732864,
      "num_tokens": 2240066.0,
      "step": 1080
    },
    {
      "entropy": 6.405725526809692,
      "epoch": 0.09842162554426705,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999993433608669,
      "loss": 6.2701,
      "mean_token_accuracy": 0.12502679005265235,
      "num_tokens": 2250732.0,
      "step": 1085
    },
    {
      "entropy": 6.442107105255127,
      "epoch": 0.09887518142235123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999992628630579,
      "loss": 6.1593,
      "mean_token_accuracy": 0.12542710900306703,
      "num_tokens": 2261452.0,
      "step": 1090
    },
    {
      "entropy": 6.452964830398559,
      "epoch": 0.09932873730043541,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999991777122069,
      "loss": 6.2658,
      "mean_token_accuracy": 0.1284552037715912,
      "num_tokens": 2271929.0,
      "step": 1095
    },
    {
      "entropy": 6.414393424987793,
      "epoch": 0.0997822931785196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999990879083156,
      "loss": 6.2473,
      "mean_token_accuracy": 0.12654093131422997,
      "num_tokens": 2282987.0,
      "step": 1100
    },
    {
      "entropy": 6.367011642456054,
      "epoch": 0.10023584905660378,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999989934513857,
      "loss": 6.2291,
      "mean_token_accuracy": 0.13299606069922448,
      "num_tokens": 2292796.0,
      "step": 1105
    },
    {
      "entropy": 6.531451177597046,
      "epoch": 0.10068940493468795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999988943414193,
      "loss": 6.2776,
      "mean_token_accuracy": 0.127064548432827,
      "num_tokens": 2303057.0,
      "step": 1110
    },
    {
      "entropy": 6.33807897567749,
      "epoch": 0.10114296081277213,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999987905784184,
      "loss": 6.2576,
      "mean_token_accuracy": 0.1281444862484932,
      "num_tokens": 2314168.0,
      "step": 1115
    },
    {
      "entropy": 6.440323400497436,
      "epoch": 0.10159651669085631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999986821623853,
      "loss": 6.2274,
      "mean_token_accuracy": 0.12680768072605134,
      "num_tokens": 2325241.0,
      "step": 1120
    },
    {
      "entropy": 6.514708232879639,
      "epoch": 0.1020500725689405,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999985690933219,
      "loss": 6.2553,
      "mean_token_accuracy": 0.12574926167726516,
      "num_tokens": 2334866.0,
      "step": 1125
    },
    {
      "entropy": 6.322805023193359,
      "epoch": 0.10250362844702467,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999984513712311,
      "loss": 6.2558,
      "mean_token_accuracy": 0.12497551888227462,
      "num_tokens": 2345422.0,
      "step": 1130
    },
    {
      "entropy": 6.5371935844421385,
      "epoch": 0.10295718432510885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999983289961146,
      "loss": 6.2772,
      "mean_token_accuracy": 0.12496376484632492,
      "num_tokens": 2356285.0,
      "step": 1135
    },
    {
      "entropy": 6.343741798400879,
      "epoch": 0.10341074020319303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999982019679755,
      "loss": 6.174,
      "mean_token_accuracy": 0.13649375587701798,
      "num_tokens": 2366861.0,
      "step": 1140
    },
    {
      "entropy": 6.431583738327026,
      "epoch": 0.10386429608127722,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999980702868164,
      "loss": 6.2109,
      "mean_token_accuracy": 0.13263476118445397,
      "num_tokens": 2376851.0,
      "step": 1145
    },
    {
      "entropy": 6.428322172164917,
      "epoch": 0.1043178519593614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999979339526396,
      "loss": 6.2835,
      "mean_token_accuracy": 0.13185692876577376,
      "num_tokens": 2386857.0,
      "step": 1150
    },
    {
      "entropy": 6.469877624511719,
      "epoch": 0.10477140783744557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999977929654484,
      "loss": 6.2088,
      "mean_token_accuracy": 0.13108245208859443,
      "num_tokens": 2397035.0,
      "step": 1155
    },
    {
      "entropy": 6.404506397247315,
      "epoch": 0.10522496371552975,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999976473252453,
      "loss": 6.2786,
      "mean_token_accuracy": 0.1281719334423542,
      "num_tokens": 2408177.0,
      "step": 1160
    },
    {
      "entropy": 6.543879461288452,
      "epoch": 0.10567851959361393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999974970320338,
      "loss": 6.2436,
      "mean_token_accuracy": 0.12375756427645683,
      "num_tokens": 2418445.0,
      "step": 1165
    },
    {
      "entropy": 6.355627489089966,
      "epoch": 0.10613207547169812,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999973420858165,
      "loss": 6.2404,
      "mean_token_accuracy": 0.1271333783864975,
      "num_tokens": 2428990.0,
      "step": 1170
    },
    {
      "entropy": 6.4137852668762205,
      "epoch": 0.10658563134978229,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999971824865968,
      "loss": 6.268,
      "mean_token_accuracy": 0.12645680457353592,
      "num_tokens": 2439643.0,
      "step": 1175
    },
    {
      "entropy": 6.5276916980743405,
      "epoch": 0.10703918722786647,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999970182343782,
      "loss": 6.311,
      "mean_token_accuracy": 0.12705472484230995,
      "num_tokens": 2450691.0,
      "step": 1180
    },
    {
      "entropy": 6.504130935668945,
      "epoch": 0.10749274310595065,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999968493291638,
      "loss": 6.2423,
      "mean_token_accuracy": 0.12769789844751359,
      "num_tokens": 2460706.0,
      "step": 1185
    },
    {
      "entropy": 6.226739454269409,
      "epoch": 0.10794629898403484,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999966757709573,
      "loss": 6.1367,
      "mean_token_accuracy": 0.13683899343013764,
      "num_tokens": 2470947.0,
      "step": 1190
    },
    {
      "entropy": 6.458954811096191,
      "epoch": 0.10839985486211902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999964975597622,
      "loss": 6.1051,
      "mean_token_accuracy": 0.13537428453564643,
      "num_tokens": 2481132.0,
      "step": 1195
    },
    {
      "entropy": 6.338390970230103,
      "epoch": 0.10885341074020319,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999963146955821,
      "loss": 6.138,
      "mean_token_accuracy": 0.1305425487458706,
      "num_tokens": 2491701.0,
      "step": 1200
    },
    {
      "entropy": 6.297045183181763,
      "epoch": 0.10930696661828737,
      "grad_norm": 1.0,
      "learning_rate": 0.000499996127178421,
      "loss": 6.2305,
      "mean_token_accuracy": 0.12452783957123756,
      "num_tokens": 2502360.0,
      "step": 1205
    },
    {
      "entropy": 6.401448202133179,
      "epoch": 0.10976052249637155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999959350082825,
      "loss": 6.1111,
      "mean_token_accuracy": 0.1346064828336239,
      "num_tokens": 2512140.0,
      "step": 1210
    },
    {
      "entropy": 6.388574886322021,
      "epoch": 0.11021407837445574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999957381851709,
      "loss": 6.2384,
      "mean_token_accuracy": 0.1266896404325962,
      "num_tokens": 2522721.0,
      "step": 1215
    },
    {
      "entropy": 6.324961376190186,
      "epoch": 0.1106676342525399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999955367090901,
      "loss": 6.1038,
      "mean_token_accuracy": 0.13158604130148888,
      "num_tokens": 2531786.0,
      "step": 1220
    },
    {
      "entropy": 6.4173516750335695,
      "epoch": 0.11112119013062409,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999953305800441,
      "loss": 6.2082,
      "mean_token_accuracy": 0.12343020662665367,
      "num_tokens": 2542895.0,
      "step": 1225
    },
    {
      "entropy": 6.308479452133179,
      "epoch": 0.11157474600870827,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999951197980374,
      "loss": 6.2113,
      "mean_token_accuracy": 0.12858483716845512,
      "num_tokens": 2554120.0,
      "step": 1230
    },
    {
      "entropy": 6.494872283935547,
      "epoch": 0.11202830188679246,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999949043630744,
      "loss": 6.2613,
      "mean_token_accuracy": 0.126002337038517,
      "num_tokens": 2563935.0,
      "step": 1235
    },
    {
      "entropy": 6.2588221549987795,
      "epoch": 0.11248185776487664,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999946842751593,
      "loss": 6.1536,
      "mean_token_accuracy": 0.13200209736824037,
      "num_tokens": 2574465.0,
      "step": 1240
    },
    {
      "entropy": 6.344051551818848,
      "epoch": 0.11293541364296081,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999944595342968,
      "loss": 6.1119,
      "mean_token_accuracy": 0.13319186195731164,
      "num_tokens": 2584694.0,
      "step": 1245
    },
    {
      "entropy": 6.293050384521484,
      "epoch": 0.11338896952104499,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999942301404916,
      "loss": 6.0892,
      "mean_token_accuracy": 0.13468018174171448,
      "num_tokens": 2594540.0,
      "step": 1250
    },
    {
      "entropy": 6.27103443145752,
      "epoch": 0.11384252539912917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999939960937484,
      "loss": 6.0265,
      "mean_token_accuracy": 0.13721646815538407,
      "num_tokens": 2604896.0,
      "step": 1255
    },
    {
      "entropy": 6.413160896301269,
      "epoch": 0.11429608127721336,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999937573940721,
      "loss": 6.1974,
      "mean_token_accuracy": 0.14260305911302568,
      "num_tokens": 2614143.0,
      "step": 1260
    },
    {
      "entropy": 6.228365230560303,
      "epoch": 0.11474963715529753,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999935140414674,
      "loss": 6.1018,
      "mean_token_accuracy": 0.13245780095458032,
      "num_tokens": 2623999.0,
      "step": 1265
    },
    {
      "entropy": 6.427315664291382,
      "epoch": 0.11520319303338171,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999932660359395,
      "loss": 6.3045,
      "mean_token_accuracy": 0.1259209007024765,
      "num_tokens": 2636208.0,
      "step": 1270
    },
    {
      "entropy": 6.353607034683227,
      "epoch": 0.11565674891146589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999930133774936,
      "loss": 6.1813,
      "mean_token_accuracy": 0.13700075224041938,
      "num_tokens": 2645854.0,
      "step": 1275
    },
    {
      "entropy": 6.392856884002685,
      "epoch": 0.11611030478955008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999927560661348,
      "loss": 6.1641,
      "mean_token_accuracy": 0.13880657628178597,
      "num_tokens": 2655877.0,
      "step": 1280
    },
    {
      "entropy": 6.396235799789428,
      "epoch": 0.11656386066763426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999924941018685,
      "loss": 6.1871,
      "mean_token_accuracy": 0.1347483940422535,
      "num_tokens": 2666157.0,
      "step": 1285
    },
    {
      "entropy": 6.304231357574463,
      "epoch": 0.11701741654571843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999922274847,
      "loss": 6.1112,
      "mean_token_accuracy": 0.13436754122376443,
      "num_tokens": 2676647.0,
      "step": 1290
    },
    {
      "entropy": 6.259562253952026,
      "epoch": 0.11747097242380261,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499991956214635,
      "loss": 6.1475,
      "mean_token_accuracy": 0.14410557746887206,
      "num_tokens": 2686417.0,
      "step": 1295
    },
    {
      "entropy": 6.3692951679229735,
      "epoch": 0.1179245283018868,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499991680291679,
      "loss": 6.1398,
      "mean_token_accuracy": 0.12797631248831748,
      "num_tokens": 2697540.0,
      "step": 1300
    },
    {
      "entropy": 6.3081450939178465,
      "epoch": 0.11837808417997098,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999913997158378,
      "loss": 6.1925,
      "mean_token_accuracy": 0.12314079701900482,
      "num_tokens": 2709397.0,
      "step": 1305
    },
    {
      "entropy": 6.354034662246704,
      "epoch": 0.11883164005805515,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999911144871169,
      "loss": 6.1452,
      "mean_token_accuracy": 0.12890623807907103,
      "num_tokens": 2720430.0,
      "step": 1310
    },
    {
      "entropy": 6.274476385116577,
      "epoch": 0.11928519593613933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999908246055226,
      "loss": 6.1047,
      "mean_token_accuracy": 0.128923050314188,
      "num_tokens": 2730648.0,
      "step": 1315
    },
    {
      "entropy": 6.303538846969604,
      "epoch": 0.11973875181422351,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999905300710607,
      "loss": 6.0876,
      "mean_token_accuracy": 0.13420135006308556,
      "num_tokens": 2739979.0,
      "step": 1320
    },
    {
      "entropy": 6.31509165763855,
      "epoch": 0.1201923076923077,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999902308837373,
      "loss": 6.2133,
      "mean_token_accuracy": 0.12678968757390977,
      "num_tokens": 2750329.0,
      "step": 1325
    },
    {
      "entropy": 6.295187377929688,
      "epoch": 0.12064586357039188,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999899270435584,
      "loss": 6.0895,
      "mean_token_accuracy": 0.14072596430778503,
      "num_tokens": 2760535.0,
      "step": 1330
    },
    {
      "entropy": 6.30014271736145,
      "epoch": 0.12109941944847605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999896185505307,
      "loss": 6.1235,
      "mean_token_accuracy": 0.133646260201931,
      "num_tokens": 2770470.0,
      "step": 1335
    },
    {
      "entropy": 6.295889663696289,
      "epoch": 0.12155297532656023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999893054046603,
      "loss": 6.1483,
      "mean_token_accuracy": 0.13114681392908095,
      "num_tokens": 2781506.0,
      "step": 1340
    },
    {
      "entropy": 6.219323301315308,
      "epoch": 0.12200653120464441,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999889876059537,
      "loss": 6.0122,
      "mean_token_accuracy": 0.14254361912608146,
      "num_tokens": 2791245.0,
      "step": 1345
    },
    {
      "entropy": 6.264402437210083,
      "epoch": 0.1224600870827286,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999886651544176,
      "loss": 6.0352,
      "mean_token_accuracy": 0.13747782483696938,
      "num_tokens": 2801680.0,
      "step": 1350
    },
    {
      "entropy": 6.328179216384887,
      "epoch": 0.12291364296081277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999883380500584,
      "loss": 6.1313,
      "mean_token_accuracy": 0.13885134533047677,
      "num_tokens": 2812805.0,
      "step": 1355
    },
    {
      "entropy": 6.274612092971802,
      "epoch": 0.12336719883889695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999880062928831,
      "loss": 6.1525,
      "mean_token_accuracy": 0.13258814290165902,
      "num_tokens": 2822834.0,
      "step": 1360
    },
    {
      "entropy": 6.32262659072876,
      "epoch": 0.12382075471698113,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999876698828985,
      "loss": 6.093,
      "mean_token_accuracy": 0.1388528026640415,
      "num_tokens": 2832312.0,
      "step": 1365
    },
    {
      "entropy": 6.264431667327881,
      "epoch": 0.12427431059506532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999873288201116,
      "loss": 6.1074,
      "mean_token_accuracy": 0.14017656743526458,
      "num_tokens": 2842998.0,
      "step": 1370
    },
    {
      "entropy": 6.334399604797364,
      "epoch": 0.1247278664731495,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004999869831045294,
      "loss": 6.1358,
      "mean_token_accuracy": 0.13416288197040557,
      "num_tokens": 2854218.0,
      "step": 1375
    },
    {
      "entropy": 6.257809448242187,
      "epoch": 0.12518142235123367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999866327361589,
      "loss": 6.0218,
      "mean_token_accuracy": 0.14412764012813567,
      "num_tokens": 2864484.0,
      "step": 1380
    },
    {
      "entropy": 6.278401279449463,
      "epoch": 0.12563497822931785,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999862777150077,
      "loss": 6.1335,
      "mean_token_accuracy": 0.13539522439241408,
      "num_tokens": 2873800.0,
      "step": 1385
    },
    {
      "entropy": 6.327728033065796,
      "epoch": 0.12608853410740203,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999859180410829,
      "loss": 6.1425,
      "mean_token_accuracy": 0.13279536589980126,
      "num_tokens": 2884247.0,
      "step": 1390
    },
    {
      "entropy": 6.269402170181275,
      "epoch": 0.12654208998548622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999855537143919,
      "loss": 6.0613,
      "mean_token_accuracy": 0.13359882831573486,
      "num_tokens": 2894739.0,
      "step": 1395
    },
    {
      "entropy": 6.197747993469238,
      "epoch": 0.1269956458635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999851847349425,
      "loss": 6.0982,
      "mean_token_accuracy": 0.13437446355819702,
      "num_tokens": 2904376.0,
      "step": 1400
    },
    {
      "entropy": 6.377307605743408,
      "epoch": 0.12744920174165458,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999848111027419,
      "loss": 6.1661,
      "mean_token_accuracy": 0.12998344823718072,
      "num_tokens": 2914304.0,
      "step": 1405
    },
    {
      "entropy": 6.292350578308105,
      "epoch": 0.12790275761973874,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999844328177983,
      "loss": 6.0278,
      "mean_token_accuracy": 0.14133364632725715,
      "num_tokens": 2924400.0,
      "step": 1410
    },
    {
      "entropy": 6.371039295196534,
      "epoch": 0.12835631349782292,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999840498801191,
      "loss": 6.274,
      "mean_token_accuracy": 0.13513783514499664,
      "num_tokens": 2935040.0,
      "step": 1415
    },
    {
      "entropy": 6.21938738822937,
      "epoch": 0.1288098693759071,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999836622897126,
      "loss": 6.0552,
      "mean_token_accuracy": 0.13850467056035995,
      "num_tokens": 2945536.0,
      "step": 1420
    },
    {
      "entropy": 6.202126550674438,
      "epoch": 0.1292634252539913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999832700465865,
      "loss": 5.9793,
      "mean_token_accuracy": 0.13881825804710388,
      "num_tokens": 2954969.0,
      "step": 1425
    },
    {
      "entropy": 6.261548900604248,
      "epoch": 0.12971698113207547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999828731507491,
      "loss": 6.1059,
      "mean_token_accuracy": 0.13205914497375487,
      "num_tokens": 2966218.0,
      "step": 1430
    },
    {
      "entropy": 6.315013360977173,
      "epoch": 0.13017053701015965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999824716022085,
      "loss": 6.1289,
      "mean_token_accuracy": 0.12892259508371354,
      "num_tokens": 2976640.0,
      "step": 1435
    },
    {
      "entropy": 6.221342897415161,
      "epoch": 0.13062409288824384,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999820654009731,
      "loss": 6.0923,
      "mean_token_accuracy": 0.13740533366799354,
      "num_tokens": 2985883.0,
      "step": 1440
    },
    {
      "entropy": 6.278382968902588,
      "epoch": 0.13107764876632802,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999816545470513,
      "loss": 6.0839,
      "mean_token_accuracy": 0.12974392548203467,
      "num_tokens": 2996695.0,
      "step": 1445
    },
    {
      "entropy": 6.26499662399292,
      "epoch": 0.1315312046444122,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999812390404514,
      "loss": 6.1001,
      "mean_token_accuracy": 0.13423352390527726,
      "num_tokens": 3007500.0,
      "step": 1450
    },
    {
      "entropy": 6.136911249160766,
      "epoch": 0.13198476052249636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999808188811823,
      "loss": 5.9549,
      "mean_token_accuracy": 0.13566315323114395,
      "num_tokens": 3017299.0,
      "step": 1455
    },
    {
      "entropy": 6.368114423751831,
      "epoch": 0.13243831640058054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999803940692524,
      "loss": 6.239,
      "mean_token_accuracy": 0.12817587703466415,
      "num_tokens": 3028389.0,
      "step": 1460
    },
    {
      "entropy": 6.223598575592041,
      "epoch": 0.13289187227866472,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999799646046707,
      "loss": 6.0568,
      "mean_token_accuracy": 0.13951559141278266,
      "num_tokens": 3038883.0,
      "step": 1465
    },
    {
      "entropy": 6.345055007934571,
      "epoch": 0.1333454281567489,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999795304874461,
      "loss": 6.1808,
      "mean_token_accuracy": 0.13515384122729301,
      "num_tokens": 3049904.0,
      "step": 1470
    },
    {
      "entropy": 6.184125995635986,
      "epoch": 0.1337989840348331,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999790917175873,
      "loss": 6.0163,
      "mean_token_accuracy": 0.13716932758688927,
      "num_tokens": 3059788.0,
      "step": 1475
    },
    {
      "entropy": 6.27112979888916,
      "epoch": 0.13425253991291727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999786482951036,
      "loss": 6.0095,
      "mean_token_accuracy": 0.1407728001475334,
      "num_tokens": 3070413.0,
      "step": 1480
    },
    {
      "entropy": 6.198488378524781,
      "epoch": 0.13470609579100146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999782002200041,
      "loss": 6.1247,
      "mean_token_accuracy": 0.13304170593619347,
      "num_tokens": 3081093.0,
      "step": 1485
    },
    {
      "entropy": 6.299404525756836,
      "epoch": 0.13515965166908564,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999777474922982,
      "loss": 6.0758,
      "mean_token_accuracy": 0.13359509408473969,
      "num_tokens": 3090594.0,
      "step": 1490
    },
    {
      "entropy": 6.272291612625122,
      "epoch": 0.13561320754716982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999772901119951,
      "loss": 6.1602,
      "mean_token_accuracy": 0.1321714386343956,
      "num_tokens": 3101061.0,
      "step": 1495
    },
    {
      "entropy": 6.250188732147217,
      "epoch": 0.13606676342525398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999768280791043,
      "loss": 6.0805,
      "mean_token_accuracy": 0.13871695548295976,
      "num_tokens": 3111715.0,
      "step": 1500
    },
    {
      "entropy": 6.187767601013183,
      "epoch": 0.13652031930333816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999763613936352,
      "loss": 6.0033,
      "mean_token_accuracy": 0.14132409542798996,
      "num_tokens": 3121378.0,
      "step": 1505
    },
    {
      "entropy": 6.173904228210449,
      "epoch": 0.13697387518142234,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999758900555979,
      "loss": 5.9421,
      "mean_token_accuracy": 0.142991504073143,
      "num_tokens": 3131229.0,
      "step": 1510
    },
    {
      "entropy": 6.184138822555542,
      "epoch": 0.13742743105950653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999754140650017,
      "loss": 6.033,
      "mean_token_accuracy": 0.1338609889149666,
      "num_tokens": 3140859.0,
      "step": 1515
    },
    {
      "entropy": 6.279017591476441,
      "epoch": 0.1378809869375907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999749334218567,
      "loss": 6.0272,
      "mean_token_accuracy": 0.13679298311471938,
      "num_tokens": 3151692.0,
      "step": 1520
    },
    {
      "entropy": 6.150258111953735,
      "epoch": 0.1383345428156749,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999744481261726,
      "loss": 6.1035,
      "mean_token_accuracy": 0.13335202634334564,
      "num_tokens": 3161879.0,
      "step": 1525
    },
    {
      "entropy": 6.205261278152466,
      "epoch": 0.13878809869375908,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999739581779597,
      "loss": 5.9614,
      "mean_token_accuracy": 0.1442497506737709,
      "num_tokens": 3171282.0,
      "step": 1530
    },
    {
      "entropy": 6.204788684844971,
      "epoch": 0.13924165457184326,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499973463577228,
      "loss": 6.0154,
      "mean_token_accuracy": 0.14128000661730766,
      "num_tokens": 3181398.0,
      "step": 1535
    },
    {
      "entropy": 6.210041856765747,
      "epoch": 0.13969521044992744,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999729643239877,
      "loss": 6.1509,
      "mean_token_accuracy": 0.1372975580394268,
      "num_tokens": 3191431.0,
      "step": 1540
    },
    {
      "entropy": 6.246035623550415,
      "epoch": 0.1401487663280116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999724604182492,
      "loss": 6.1423,
      "mean_token_accuracy": 0.1305283486843109,
      "num_tokens": 3202147.0,
      "step": 1545
    },
    {
      "entropy": 6.317409181594849,
      "epoch": 0.14060232220609578,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999719518600229,
      "loss": 6.0187,
      "mean_token_accuracy": 0.14262037947773934,
      "num_tokens": 3212428.0,
      "step": 1550
    },
    {
      "entropy": 6.198874998092651,
      "epoch": 0.14105587808417996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999714386493192,
      "loss": 6.0638,
      "mean_token_accuracy": 0.1392004132270813,
      "num_tokens": 3222099.0,
      "step": 1555
    },
    {
      "entropy": 6.23433141708374,
      "epoch": 0.14150943396226415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999709207861489,
      "loss": 6.0113,
      "mean_token_accuracy": 0.1394033081829548,
      "num_tokens": 3232079.0,
      "step": 1560
    },
    {
      "entropy": 6.08824315071106,
      "epoch": 0.14196298984034833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999703982705227,
      "loss": 5.928,
      "mean_token_accuracy": 0.14834348112344742,
      "num_tokens": 3241762.0,
      "step": 1565
    },
    {
      "entropy": 6.164975643157959,
      "epoch": 0.1424165457184325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999698711024513,
      "loss": 6.0104,
      "mean_token_accuracy": 0.1406618483364582,
      "num_tokens": 3252571.0,
      "step": 1570
    },
    {
      "entropy": 6.212284135818481,
      "epoch": 0.1428701015965167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999693392819456,
      "loss": 6.0603,
      "mean_token_accuracy": 0.13540706485509874,
      "num_tokens": 3263542.0,
      "step": 1575
    },
    {
      "entropy": 6.246869134902954,
      "epoch": 0.14332365747460088,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999688028090166,
      "loss": 6.1386,
      "mean_token_accuracy": 0.13631665483117103,
      "num_tokens": 3273295.0,
      "step": 1580
    },
    {
      "entropy": 6.328453207015992,
      "epoch": 0.14377721335268506,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999682616836755,
      "loss": 6.1427,
      "mean_token_accuracy": 0.13787241727113725,
      "num_tokens": 3283605.0,
      "step": 1585
    },
    {
      "entropy": 6.191803979873657,
      "epoch": 0.14423076923076922,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999677159059334,
      "loss": 5.9952,
      "mean_token_accuracy": 0.14498581141233444,
      "num_tokens": 3293837.0,
      "step": 1590
    },
    {
      "entropy": 6.072175979614258,
      "epoch": 0.1446843251088534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999671654758016,
      "loss": 5.9983,
      "mean_token_accuracy": 0.14269243702292442,
      "num_tokens": 3304131.0,
      "step": 1595
    },
    {
      "entropy": 6.235024261474609,
      "epoch": 0.14513788098693758,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999666103932915,
      "loss": 6.0234,
      "mean_token_accuracy": 0.13487293496727942,
      "num_tokens": 3313543.0,
      "step": 1600
    },
    {
      "entropy": 6.185160875320435,
      "epoch": 0.14559143686502177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999660506584145,
      "loss": 6.0201,
      "mean_token_accuracy": 0.13741163462400435,
      "num_tokens": 3324487.0,
      "step": 1605
    },
    {
      "entropy": 6.239445686340332,
      "epoch": 0.14604499274310595,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999654862711823,
      "loss": 6.0262,
      "mean_token_accuracy": 0.137547168135643,
      "num_tokens": 3335094.0,
      "step": 1610
    },
    {
      "entropy": 6.134017467498779,
      "epoch": 0.14649854862119013,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999649172316065,
      "loss": 5.9642,
      "mean_token_accuracy": 0.13992214426398278,
      "num_tokens": 3344252.0,
      "step": 1615
    },
    {
      "entropy": 6.197755432128906,
      "epoch": 0.14695210449927432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499964343539699,
      "loss": 6.0429,
      "mean_token_accuracy": 0.13646679371595383,
      "num_tokens": 3354265.0,
      "step": 1620
    },
    {
      "entropy": 6.2219359397888185,
      "epoch": 0.1474056603773585,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999637651954714,
      "loss": 6.125,
      "mean_token_accuracy": 0.1394253857433796,
      "num_tokens": 3364616.0,
      "step": 1625
    },
    {
      "entropy": 6.27594027519226,
      "epoch": 0.14785921625544268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999631821989358,
      "loss": 6.0927,
      "mean_token_accuracy": 0.14138601943850518,
      "num_tokens": 3373650.0,
      "step": 1630
    },
    {
      "entropy": 6.101440382003784,
      "epoch": 0.14831277213352684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999625945501043,
      "loss": 5.9886,
      "mean_token_accuracy": 0.1461173690855503,
      "num_tokens": 3384590.0,
      "step": 1635
    },
    {
      "entropy": 6.26271767616272,
      "epoch": 0.14876632801161102,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499962002248989,
      "loss": 6.0205,
      "mean_token_accuracy": 0.13982762396335602,
      "num_tokens": 3395616.0,
      "step": 1640
    },
    {
      "entropy": 6.153381443023681,
      "epoch": 0.1492198838896952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999614052956024,
      "loss": 6.0239,
      "mean_token_accuracy": 0.13940254971385002,
      "num_tokens": 3405810.0,
      "step": 1645
    },
    {
      "entropy": 6.254888677597046,
      "epoch": 0.1496734397677794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999608036899563,
      "loss": 6.0591,
      "mean_token_accuracy": 0.1363350972533226,
      "num_tokens": 3415932.0,
      "step": 1650
    },
    {
      "entropy": 6.151538419723511,
      "epoch": 0.15012699564586357,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999601974320636,
      "loss": 6.0116,
      "mean_token_accuracy": 0.14114982038736343,
      "num_tokens": 3425626.0,
      "step": 1655
    },
    {
      "entropy": 6.121203470230102,
      "epoch": 0.15058055152394775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999595865219367,
      "loss": 6.0208,
      "mean_token_accuracy": 0.13342469930648804,
      "num_tokens": 3434837.0,
      "step": 1660
    },
    {
      "entropy": 6.156308937072754,
      "epoch": 0.15103410740203194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999589709595881,
      "loss": 6.0332,
      "mean_token_accuracy": 0.14258277863264085,
      "num_tokens": 3445238.0,
      "step": 1665
    },
    {
      "entropy": 6.166830587387085,
      "epoch": 0.15148766328011612,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999583507450308,
      "loss": 6.0043,
      "mean_token_accuracy": 0.1305796168744564,
      "num_tokens": 3456738.0,
      "step": 1670
    },
    {
      "entropy": 6.070332002639771,
      "epoch": 0.1519412191582003,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999577258782774,
      "loss": 5.8838,
      "mean_token_accuracy": 0.13995402976870536,
      "num_tokens": 3466296.0,
      "step": 1675
    },
    {
      "entropy": 6.266418838500977,
      "epoch": 0.15239477503628446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499957096359341,
      "loss": 6.087,
      "mean_token_accuracy": 0.13597871735692024,
      "num_tokens": 3476989.0,
      "step": 1680
    },
    {
      "entropy": 6.111058521270752,
      "epoch": 0.15284833091436864,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999564621882343,
      "loss": 5.9321,
      "mean_token_accuracy": 0.141887167096138,
      "num_tokens": 3487050.0,
      "step": 1685
    },
    {
      "entropy": 6.225695085525513,
      "epoch": 0.15330188679245282,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999558233649708,
      "loss": 6.139,
      "mean_token_accuracy": 0.1312915213406086,
      "num_tokens": 3497777.0,
      "step": 1690
    },
    {
      "entropy": 6.17779574394226,
      "epoch": 0.153755442670537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999551798895636,
      "loss": 6.0475,
      "mean_token_accuracy": 0.13619913384318352,
      "num_tokens": 3509066.0,
      "step": 1695
    },
    {
      "entropy": 6.141854095458984,
      "epoch": 0.1542089985486212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999545317620258,
      "loss": 5.9776,
      "mean_token_accuracy": 0.1392234146595001,
      "num_tokens": 3519567.0,
      "step": 1700
    },
    {
      "entropy": 6.124539947509765,
      "epoch": 0.15466255442670537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999538789823711,
      "loss": 6.0076,
      "mean_token_accuracy": 0.1419394336640835,
      "num_tokens": 3529402.0,
      "step": 1705
    },
    {
      "entropy": 6.27455096244812,
      "epoch": 0.15511611030478956,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999532215506127,
      "loss": 6.0228,
      "mean_token_accuracy": 0.14713949486613273,
      "num_tokens": 3539805.0,
      "step": 1710
    },
    {
      "entropy": 6.072864151000976,
      "epoch": 0.15556966618287374,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999525594667645,
      "loss": 5.881,
      "mean_token_accuracy": 0.14898346215486527,
      "num_tokens": 3550477.0,
      "step": 1715
    },
    {
      "entropy": 6.0882893085479735,
      "epoch": 0.15602322206095792,
      "grad_norm": 1.0,
      "learning_rate": 0.00049995189273084,
      "loss": 5.9843,
      "mean_token_accuracy": 0.1400413043797016,
      "num_tokens": 3560729.0,
      "step": 1720
    },
    {
      "entropy": 6.1748028755187985,
      "epoch": 0.15647677793904208,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499951221342853,
      "loss": 6.0394,
      "mean_token_accuracy": 0.1325826108455658,
      "num_tokens": 3571538.0,
      "step": 1725
    },
    {
      "entropy": 6.174065685272216,
      "epoch": 0.15693033381712626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999505453028174,
      "loss": 6.0644,
      "mean_token_accuracy": 0.13689751103520392,
      "num_tokens": 3583128.0,
      "step": 1730
    },
    {
      "entropy": 6.2016395092010494,
      "epoch": 0.15738388969521044,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999498646107472,
      "loss": 5.9871,
      "mean_token_accuracy": 0.1429124280810356,
      "num_tokens": 3594358.0,
      "step": 1735
    },
    {
      "entropy": 6.129647302627563,
      "epoch": 0.15783744557329463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999491792666566,
      "loss": 5.9292,
      "mean_token_accuracy": 0.14557284563779832,
      "num_tokens": 3604345.0,
      "step": 1740
    },
    {
      "entropy": 6.089707612991333,
      "epoch": 0.1582910014513788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999484892705597,
      "loss": 5.9717,
      "mean_token_accuracy": 0.1409473091363907,
      "num_tokens": 3616129.0,
      "step": 1745
    },
    {
      "entropy": 6.162917709350586,
      "epoch": 0.158744557329463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999477946224705,
      "loss": 6.0177,
      "mean_token_accuracy": 0.14049824178218842,
      "num_tokens": 3627458.0,
      "step": 1750
    },
    {
      "entropy": 6.149751949310303,
      "epoch": 0.15919811320754718,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999470953224037,
      "loss": 5.9031,
      "mean_token_accuracy": 0.1436824843287468,
      "num_tokens": 3636047.0,
      "step": 1755
    },
    {
      "entropy": 6.0508606910705565,
      "epoch": 0.15965166908563136,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999463913703735,
      "loss": 5.8869,
      "mean_token_accuracy": 0.1433118313550949,
      "num_tokens": 3646930.0,
      "step": 1760
    },
    {
      "entropy": 6.124081325531006,
      "epoch": 0.16010522496371554,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999456827663948,
      "loss": 5.9919,
      "mean_token_accuracy": 0.1459735155105591,
      "num_tokens": 3657279.0,
      "step": 1765
    },
    {
      "entropy": 6.052764081954956,
      "epoch": 0.1605587808417997,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499944969510482,
      "loss": 5.9001,
      "mean_token_accuracy": 0.13991431295871734,
      "num_tokens": 3667070.0,
      "step": 1770
    },
    {
      "entropy": 6.181622266769409,
      "epoch": 0.16101233671988388,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999442516026498,
      "loss": 5.9141,
      "mean_token_accuracy": 0.14285601824522018,
      "num_tokens": 3677560.0,
      "step": 1775
    },
    {
      "entropy": 6.027710103988648,
      "epoch": 0.16146589259796806,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999435290429133,
      "loss": 5.8862,
      "mean_token_accuracy": 0.14947912245988845,
      "num_tokens": 3686701.0,
      "step": 1780
    },
    {
      "entropy": 6.117189884185791,
      "epoch": 0.16191944847605225,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999428018312872,
      "loss": 5.9719,
      "mean_token_accuracy": 0.14543041959404945,
      "num_tokens": 3696338.0,
      "step": 1785
    },
    {
      "entropy": 6.091988801956177,
      "epoch": 0.16237300435413643,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999420699677867,
      "loss": 5.9492,
      "mean_token_accuracy": 0.1432087704539299,
      "num_tokens": 3705752.0,
      "step": 1790
    },
    {
      "entropy": 6.203689908981323,
      "epoch": 0.1628265602322206,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999413334524269,
      "loss": 6.0384,
      "mean_token_accuracy": 0.13643145337700843,
      "num_tokens": 3716225.0,
      "step": 1795
    },
    {
      "entropy": 6.111277675628662,
      "epoch": 0.1632801161103048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999405922852229,
      "loss": 5.9098,
      "mean_token_accuracy": 0.14482017382979392,
      "num_tokens": 3726593.0,
      "step": 1800
    },
    {
      "entropy": 6.028970766067505,
      "epoch": 0.16373367198838898,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999398464661901,
      "loss": 5.9708,
      "mean_token_accuracy": 0.13964075967669487,
      "num_tokens": 3737563.0,
      "step": 1805
    },
    {
      "entropy": 6.246074914932251,
      "epoch": 0.16418722786647316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999390959953441,
      "loss": 6.0176,
      "mean_token_accuracy": 0.14332932382822036,
      "num_tokens": 3747752.0,
      "step": 1810
    },
    {
      "entropy": 6.053060626983642,
      "epoch": 0.16464078374455732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999383408727002,
      "loss": 5.9539,
      "mean_token_accuracy": 0.14173785969614983,
      "num_tokens": 3758795.0,
      "step": 1815
    },
    {
      "entropy": 6.167564058303833,
      "epoch": 0.1650943396226415,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999375810982741,
      "loss": 5.9315,
      "mean_token_accuracy": 0.14446532130241393,
      "num_tokens": 3768690.0,
      "step": 1820
    },
    {
      "entropy": 6.0776159286499025,
      "epoch": 0.16554789550072568,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999368166720815,
      "loss": 6.014,
      "mean_token_accuracy": 0.14304881766438485,
      "num_tokens": 3779018.0,
      "step": 1825
    },
    {
      "entropy": 6.147589683532715,
      "epoch": 0.16600145137880987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999360475941382,
      "loss": 5.9118,
      "mean_token_accuracy": 0.14721195623278618,
      "num_tokens": 3788985.0,
      "step": 1830
    },
    {
      "entropy": 6.131836366653443,
      "epoch": 0.16645500725689405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049993527386446,
      "loss": 6.0394,
      "mean_token_accuracy": 0.1322140447795391,
      "num_tokens": 3800787.0,
      "step": 1835
    },
    {
      "entropy": 6.089927768707275,
      "epoch": 0.16690856313497823,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999344954830633,
      "loss": 5.9208,
      "mean_token_accuracy": 0.1421339437365532,
      "num_tokens": 3811381.0,
      "step": 1840
    },
    {
      "entropy": 6.0987998962402346,
      "epoch": 0.16736211901306242,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999337124499638,
      "loss": 6.0339,
      "mean_token_accuracy": 0.14644376039505005,
      "num_tokens": 3822055.0,
      "step": 1845
    },
    {
      "entropy": 6.147271633148193,
      "epoch": 0.1678156748911466,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999329247651777,
      "loss": 5.9932,
      "mean_token_accuracy": 0.1378726117312908,
      "num_tokens": 3831900.0,
      "step": 1850
    },
    {
      "entropy": 6.1165355205535885,
      "epoch": 0.16826923076923078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999321324287213,
      "loss": 5.9383,
      "mean_token_accuracy": 0.14189887940883636,
      "num_tokens": 3842702.0,
      "step": 1855
    },
    {
      "entropy": 6.1751471042633055,
      "epoch": 0.16872278664731494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999313354406114,
      "loss": 5.9187,
      "mean_token_accuracy": 0.1416791319847107,
      "num_tokens": 3852690.0,
      "step": 1860
    },
    {
      "entropy": 5.9985252857208256,
      "epoch": 0.16917634252539912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499930533800864,
      "loss": 5.9101,
      "mean_token_accuracy": 0.1507345899939537,
      "num_tokens": 3863040.0,
      "step": 1865
    },
    {
      "entropy": 6.10200867652893,
      "epoch": 0.1696298984034833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999297275094958,
      "loss": 6.0044,
      "mean_token_accuracy": 0.14036339968442918,
      "num_tokens": 3873613.0,
      "step": 1870
    },
    {
      "entropy": 5.999226474761963,
      "epoch": 0.1700834542815675,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999289165665235,
      "loss": 5.8075,
      "mean_token_accuracy": 0.14994008243083953,
      "num_tokens": 3883936.0,
      "step": 1875
    },
    {
      "entropy": 5.995340251922608,
      "epoch": 0.17053701015965167,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999281009719639,
      "loss": 5.7634,
      "mean_token_accuracy": 0.15587486177682877,
      "num_tokens": 3893788.0,
      "step": 1880
    },
    {
      "entropy": 6.001563358306885,
      "epoch": 0.17099056603773585,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999272807258339,
      "loss": 5.8376,
      "mean_token_accuracy": 0.15261975079774856,
      "num_tokens": 3903434.0,
      "step": 1885
    },
    {
      "entropy": 6.05434021949768,
      "epoch": 0.17144412191582004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999264558281504,
      "loss": 5.8409,
      "mean_token_accuracy": 0.1521665245294571,
      "num_tokens": 3913476.0,
      "step": 1890
    },
    {
      "entropy": 5.971798706054687,
      "epoch": 0.17189767779390422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999256262789305,
      "loss": 5.8007,
      "mean_token_accuracy": 0.151979461312294,
      "num_tokens": 3923301.0,
      "step": 1895
    },
    {
      "entropy": 6.028662157058716,
      "epoch": 0.1723512336719884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999247920781913,
      "loss": 5.9603,
      "mean_token_accuracy": 0.14160882011055947,
      "num_tokens": 3933335.0,
      "step": 1900
    },
    {
      "entropy": 6.160859155654907,
      "epoch": 0.17280478955007256,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999239532259502,
      "loss": 5.978,
      "mean_token_accuracy": 0.1399615041911602,
      "num_tokens": 3944171.0,
      "step": 1905
    },
    {
      "entropy": 6.11788387298584,
      "epoch": 0.17325834542815674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999231097222244,
      "loss": 6.056,
      "mean_token_accuracy": 0.14640697762370108,
      "num_tokens": 3954160.0,
      "step": 1910
    },
    {
      "entropy": 6.125342941284179,
      "epoch": 0.17371190130624092,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999222615670312,
      "loss": 5.8442,
      "mean_token_accuracy": 0.15440410375595093,
      "num_tokens": 3964110.0,
      "step": 1915
    },
    {
      "entropy": 6.047020721435547,
      "epoch": 0.1741654571843251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999214087603885,
      "loss": 5.9347,
      "mean_token_accuracy": 0.14503437876701356,
      "num_tokens": 3974293.0,
      "step": 1920
    },
    {
      "entropy": 6.03211555480957,
      "epoch": 0.1746190130624093,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999205513023135,
      "loss": 5.8896,
      "mean_token_accuracy": 0.14775249734520912,
      "num_tokens": 3984346.0,
      "step": 1925
    },
    {
      "entropy": 5.992477941513061,
      "epoch": 0.17507256894049347,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999196891928245,
      "loss": 5.8312,
      "mean_token_accuracy": 0.15544059574604036,
      "num_tokens": 3992897.0,
      "step": 1930
    },
    {
      "entropy": 6.042530679702759,
      "epoch": 0.17552612481857766,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999188224319388,
      "loss": 5.8655,
      "mean_token_accuracy": 0.15182426571846008,
      "num_tokens": 4002575.0,
      "step": 1935
    },
    {
      "entropy": 6.01702356338501,
      "epoch": 0.17597968069666184,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999179510196746,
      "loss": 5.9547,
      "mean_token_accuracy": 0.1407930850982666,
      "num_tokens": 4013752.0,
      "step": 1940
    },
    {
      "entropy": 6.041807699203491,
      "epoch": 0.17643323657474602,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999170749560498,
      "loss": 5.833,
      "mean_token_accuracy": 0.1515166938304901,
      "num_tokens": 4023718.0,
      "step": 1945
    },
    {
      "entropy": 6.038792991638184,
      "epoch": 0.17688679245283018,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999161942410827,
      "loss": 5.9323,
      "mean_token_accuracy": 0.13799159675836564,
      "num_tokens": 4034210.0,
      "step": 1950
    },
    {
      "entropy": 6.134712362289429,
      "epoch": 0.17734034833091436,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999153088747913,
      "loss": 5.9265,
      "mean_token_accuracy": 0.14558120667934418,
      "num_tokens": 4044868.0,
      "step": 1955
    },
    {
      "entropy": 5.951734399795532,
      "epoch": 0.17779390420899854,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499914418857194,
      "loss": 5.9128,
      "mean_token_accuracy": 0.14710975885391236,
      "num_tokens": 4055049.0,
      "step": 1960
    },
    {
      "entropy": 6.109003973007202,
      "epoch": 0.17824746008708273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999135241883092,
      "loss": 5.9693,
      "mean_token_accuracy": 0.14337438270449637,
      "num_tokens": 4065889.0,
      "step": 1965
    },
    {
      "entropy": 6.076221990585327,
      "epoch": 0.1787010159651669,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999126248681555,
      "loss": 5.9986,
      "mean_token_accuracy": 0.14153247475624084,
      "num_tokens": 4077288.0,
      "step": 1970
    },
    {
      "entropy": 6.059194612503052,
      "epoch": 0.1791545718432511,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999117208967513,
      "loss": 5.9359,
      "mean_token_accuracy": 0.14241613298654557,
      "num_tokens": 4087402.0,
      "step": 1975
    },
    {
      "entropy": 6.074989175796508,
      "epoch": 0.17960812772133528,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999108122741154,
      "loss": 5.9911,
      "mean_token_accuracy": 0.1394512176513672,
      "num_tokens": 4098150.0,
      "step": 1980
    },
    {
      "entropy": 5.997276258468628,
      "epoch": 0.18006168359941946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999098990002666,
      "loss": 5.8489,
      "mean_token_accuracy": 0.14526972770690919,
      "num_tokens": 4108134.0,
      "step": 1985
    },
    {
      "entropy": 6.0917863845825195,
      "epoch": 0.18051523947750364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999089810752237,
      "loss": 5.988,
      "mean_token_accuracy": 0.1419826477766037,
      "num_tokens": 4118251.0,
      "step": 1990
    },
    {
      "entropy": 6.037380886077881,
      "epoch": 0.1809687953555878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499908058499006,
      "loss": 5.8256,
      "mean_token_accuracy": 0.1488537833094597,
      "num_tokens": 4129228.0,
      "step": 1995
    },
    {
      "entropy": 5.989283561706543,
      "epoch": 0.18142235123367198,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999071312716321,
      "loss": 5.8905,
      "mean_token_accuracy": 0.15052457451820372,
      "num_tokens": 4139391.0,
      "step": 2000
    },
    {
      "entropy": 6.038852834701538,
      "epoch": 0.18187590711175616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999061993931215,
      "loss": 5.8198,
      "mean_token_accuracy": 0.15358539968729018,
      "num_tokens": 4148891.0,
      "step": 2005
    },
    {
      "entropy": 6.235394716262817,
      "epoch": 0.18232946298984035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999052628634934,
      "loss": 6.1728,
      "mean_token_accuracy": 0.14022321105003357,
      "num_tokens": 4159551.0,
      "step": 2010
    },
    {
      "entropy": 6.079950952529908,
      "epoch": 0.18278301886792453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999043216827671,
      "loss": 5.9696,
      "mean_token_accuracy": 0.14237845912575722,
      "num_tokens": 4170055.0,
      "step": 2015
    },
    {
      "entropy": 5.993285369873047,
      "epoch": 0.1832365747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999033758509621,
      "loss": 5.8201,
      "mean_token_accuracy": 0.14752258732914925,
      "num_tokens": 4180767.0,
      "step": 2020
    },
    {
      "entropy": 6.0369898796081545,
      "epoch": 0.1836901306240929,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499902425368098,
      "loss": 5.9485,
      "mean_token_accuracy": 0.14603547155857086,
      "num_tokens": 4190654.0,
      "step": 2025
    },
    {
      "entropy": 6.0134642124176025,
      "epoch": 0.18414368650217708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999014702341944,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1434512048959732,
      "num_tokens": 4200753.0,
      "step": 2030
    },
    {
      "entropy": 6.070598554611206,
      "epoch": 0.18459724238026126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999005104492712,
      "loss": 5.9491,
      "mean_token_accuracy": 0.14099996089935302,
      "num_tokens": 4210994.0,
      "step": 2035
    },
    {
      "entropy": 6.0680242538452145,
      "epoch": 0.18505079825834542,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998995460133481,
      "loss": 5.8913,
      "mean_token_accuracy": 0.14143144562840462,
      "num_tokens": 4220883.0,
      "step": 2040
    },
    {
      "entropy": 6.077210140228272,
      "epoch": 0.1855043541364296,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998985769264451,
      "loss": 5.9245,
      "mean_token_accuracy": 0.13767858743667602,
      "num_tokens": 4231269.0,
      "step": 2045
    },
    {
      "entropy": 5.974451160430908,
      "epoch": 0.18595791001451378,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998976031885821,
      "loss": 5.93,
      "mean_token_accuracy": 0.1419460207223892,
      "num_tokens": 4241559.0,
      "step": 2050
    },
    {
      "entropy": 6.0086987018585205,
      "epoch": 0.18641146589259797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998966247997794,
      "loss": 5.8188,
      "mean_token_accuracy": 0.14645475745201111,
      "num_tokens": 4250870.0,
      "step": 2055
    },
    {
      "entropy": 5.967829656600952,
      "epoch": 0.18686502177068215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998956417600572,
      "loss": 5.7746,
      "mean_token_accuracy": 0.15297217816114425,
      "num_tokens": 4260848.0,
      "step": 2060
    },
    {
      "entropy": 5.986316537857055,
      "epoch": 0.18731857764876633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998946540694358,
      "loss": 5.9036,
      "mean_token_accuracy": 0.15208268463611602,
      "num_tokens": 4270973.0,
      "step": 2065
    },
    {
      "entropy": 6.132760667800904,
      "epoch": 0.18777213352685052,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998936617279357,
      "loss": 5.9802,
      "mean_token_accuracy": 0.14554873257875442,
      "num_tokens": 4281211.0,
      "step": 2070
    },
    {
      "entropy": 5.941362285614014,
      "epoch": 0.1882256894049347,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998926647355773,
      "loss": 5.8306,
      "mean_token_accuracy": 0.150546196103096,
      "num_tokens": 4291425.0,
      "step": 2075
    },
    {
      "entropy": 6.098439311981201,
      "epoch": 0.18867924528301888,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998916630923813,
      "loss": 5.9406,
      "mean_token_accuracy": 0.14438463374972343,
      "num_tokens": 4301079.0,
      "step": 2080
    },
    {
      "entropy": 6.045752716064453,
      "epoch": 0.18913280116110304,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998906567983685,
      "loss": 5.9215,
      "mean_token_accuracy": 0.1422693431377411,
      "num_tokens": 4312374.0,
      "step": 2085
    },
    {
      "entropy": 6.030663347244262,
      "epoch": 0.18958635703918722,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998896458535595,
      "loss": 5.8334,
      "mean_token_accuracy": 0.15475437194108962,
      "num_tokens": 4322810.0,
      "step": 2090
    },
    {
      "entropy": 5.984178876876831,
      "epoch": 0.1900399129172714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998886302579752,
      "loss": 5.8383,
      "mean_token_accuracy": 0.14914876371622085,
      "num_tokens": 4332082.0,
      "step": 2095
    },
    {
      "entropy": 6.012246084213257,
      "epoch": 0.19049346879535559,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499887610011637,
      "loss": 5.9249,
      "mean_token_accuracy": 0.1479165866971016,
      "num_tokens": 4342632.0,
      "step": 2100
    },
    {
      "entropy": 5.986341905593872,
      "epoch": 0.19094702467343977,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998865851145656,
      "loss": 5.8567,
      "mean_token_accuracy": 0.14595670849084855,
      "num_tokens": 4353544.0,
      "step": 2105
    },
    {
      "entropy": 5.9498518943786625,
      "epoch": 0.19140058055152395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998855555667822,
      "loss": 5.981,
      "mean_token_accuracy": 0.1430116519331932,
      "num_tokens": 4363636.0,
      "step": 2110
    },
    {
      "entropy": 6.076320600509644,
      "epoch": 0.19185413642960814,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998845213683084,
      "loss": 5.831,
      "mean_token_accuracy": 0.1475878320634365,
      "num_tokens": 4373672.0,
      "step": 2115
    },
    {
      "entropy": 5.910988473892212,
      "epoch": 0.19230769230769232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998834825191652,
      "loss": 5.8096,
      "mean_token_accuracy": 0.14788725823163987,
      "num_tokens": 4383199.0,
      "step": 2120
    },
    {
      "entropy": 6.013854932785034,
      "epoch": 0.1927612481857765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998824390193743,
      "loss": 5.886,
      "mean_token_accuracy": 0.1449464038014412,
      "num_tokens": 4394446.0,
      "step": 2125
    },
    {
      "entropy": 6.0993109226226805,
      "epoch": 0.19321480406386066,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998813908689573,
      "loss": 5.9237,
      "mean_token_accuracy": 0.14512224718928338,
      "num_tokens": 4405882.0,
      "step": 2130
    },
    {
      "entropy": 5.93023910522461,
      "epoch": 0.19366835994194484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998803380679359,
      "loss": 5.7867,
      "mean_token_accuracy": 0.1542275607585907,
      "num_tokens": 4415280.0,
      "step": 2135
    },
    {
      "entropy": 5.99576849937439,
      "epoch": 0.19412191582002902,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998792806163317,
      "loss": 5.9946,
      "mean_token_accuracy": 0.1401658058166504,
      "num_tokens": 4425755.0,
      "step": 2140
    },
    {
      "entropy": 5.925059509277344,
      "epoch": 0.1945754716981132,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998782185141667,
      "loss": 5.8523,
      "mean_token_accuracy": 0.14792224615812302,
      "num_tokens": 4436259.0,
      "step": 2145
    },
    {
      "entropy": 6.045332574844361,
      "epoch": 0.1950290275761974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998771517614629,
      "loss": 5.881,
      "mean_token_accuracy": 0.14263954162597656,
      "num_tokens": 4446423.0,
      "step": 2150
    },
    {
      "entropy": 6.00939302444458,
      "epoch": 0.19548258345428157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998760803582422,
      "loss": 5.8288,
      "mean_token_accuracy": 0.15420894026756288,
      "num_tokens": 4456678.0,
      "step": 2155
    },
    {
      "entropy": 5.945328330993652,
      "epoch": 0.19593613933236576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998750043045269,
      "loss": 5.8269,
      "mean_token_accuracy": 0.15192546397447587,
      "num_tokens": 4467031.0,
      "step": 2160
    },
    {
      "entropy": 6.024298143386841,
      "epoch": 0.19638969521044994,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998739236003392,
      "loss": 5.9066,
      "mean_token_accuracy": 0.14100065529346467,
      "num_tokens": 4477771.0,
      "step": 2165
    },
    {
      "entropy": 6.0035483837127686,
      "epoch": 0.1968432510885341,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998728382457014,
      "loss": 5.9222,
      "mean_token_accuracy": 0.1434047520160675,
      "num_tokens": 4488426.0,
      "step": 2170
    },
    {
      "entropy": 6.103923177719116,
      "epoch": 0.19729680696661828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998717482406362,
      "loss": 5.9056,
      "mean_token_accuracy": 0.14213445484638215,
      "num_tokens": 4499735.0,
      "step": 2175
    },
    {
      "entropy": 6.009974479675293,
      "epoch": 0.19775036284470246,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998706535851657,
      "loss": 5.8373,
      "mean_token_accuracy": 0.14943975061178208,
      "num_tokens": 4509500.0,
      "step": 2180
    },
    {
      "entropy": 5.964470767974854,
      "epoch": 0.19820391872278664,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998695542793129,
      "loss": 5.8475,
      "mean_token_accuracy": 0.14806249961256981,
      "num_tokens": 4520431.0,
      "step": 2185
    },
    {
      "entropy": 6.013790941238403,
      "epoch": 0.19865747460087083,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998684503231003,
      "loss": 5.9367,
      "mean_token_accuracy": 0.1468561165034771,
      "num_tokens": 4532294.0,
      "step": 2190
    },
    {
      "entropy": 5.979345083236694,
      "epoch": 0.199111030478955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499867341716551,
      "loss": 5.8657,
      "mean_token_accuracy": 0.1456448383629322,
      "num_tokens": 4542488.0,
      "step": 2195
    },
    {
      "entropy": 5.9495869159698485,
      "epoch": 0.1995645863570392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998662284596878,
      "loss": 5.8013,
      "mean_token_accuracy": 0.14156620278954507,
      "num_tokens": 4552933.0,
      "step": 2200
    },
    {
      "entropy": 6.036973190307617,
      "epoch": 0.20001814223512338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998651105525335,
      "loss": 5.7937,
      "mean_token_accuracy": 0.15192690938711167,
      "num_tokens": 4562373.0,
      "step": 2205
    },
    {
      "entropy": 5.927747917175293,
      "epoch": 0.20047169811320756,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998639879951115,
      "loss": 5.9131,
      "mean_token_accuracy": 0.14802779853343964,
      "num_tokens": 4572715.0,
      "step": 2210
    },
    {
      "entropy": 6.064771509170532,
      "epoch": 0.2009252539912917,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499862860787445,
      "loss": 5.7918,
      "mean_token_accuracy": 0.14533086642622947,
      "num_tokens": 4582782.0,
      "step": 2215
    },
    {
      "entropy": 5.980919456481933,
      "epoch": 0.2013788098693759,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998617289295571,
      "loss": 5.8965,
      "mean_token_accuracy": 0.13982278481125832,
      "num_tokens": 4594108.0,
      "step": 2220
    },
    {
      "entropy": 6.0538468837738035,
      "epoch": 0.20183236574746008,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998605924214714,
      "loss": 5.8628,
      "mean_token_accuracy": 0.1460525505244732,
      "num_tokens": 4604732.0,
      "step": 2225
    },
    {
      "entropy": 6.000491809844971,
      "epoch": 0.20228592162554426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998594512632114,
      "loss": 5.9955,
      "mean_token_accuracy": 0.13708511143922805,
      "num_tokens": 4614998.0,
      "step": 2230
    },
    {
      "entropy": 5.970520877838135,
      "epoch": 0.20273947750362845,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998583054548006,
      "loss": 5.8136,
      "mean_token_accuracy": 0.1484377056360245,
      "num_tokens": 4624665.0,
      "step": 2235
    },
    {
      "entropy": 6.00954236984253,
      "epoch": 0.20319303338171263,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998571549962628,
      "loss": 5.8101,
      "mean_token_accuracy": 0.15187345445156097,
      "num_tokens": 4633731.0,
      "step": 2240
    },
    {
      "entropy": 5.991653680801392,
      "epoch": 0.2036465892597968,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998559998876216,
      "loss": 5.8384,
      "mean_token_accuracy": 0.14615286365151406,
      "num_tokens": 4643421.0,
      "step": 2245
    },
    {
      "entropy": 5.987816905975341,
      "epoch": 0.204100145137881,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998548401289012,
      "loss": 5.9151,
      "mean_token_accuracy": 0.14586994647979737,
      "num_tokens": 4653795.0,
      "step": 2250
    },
    {
      "entropy": 6.008340549468994,
      "epoch": 0.20455370101596518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998536757201253,
      "loss": 5.8858,
      "mean_token_accuracy": 0.1444897711277008,
      "num_tokens": 4664840.0,
      "step": 2255
    },
    {
      "entropy": 6.039184617996216,
      "epoch": 0.20500725689404933,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998525066613181,
      "loss": 5.8979,
      "mean_token_accuracy": 0.15061988234519957,
      "num_tokens": 4675508.0,
      "step": 2260
    },
    {
      "entropy": 5.932389974594116,
      "epoch": 0.20546081277213352,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998513329525038,
      "loss": 5.8199,
      "mean_token_accuracy": 0.15319341719150542,
      "num_tokens": 4685995.0,
      "step": 2265
    },
    {
      "entropy": 6.023037481307983,
      "epoch": 0.2059143686502177,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998501545937066,
      "loss": 5.8348,
      "mean_token_accuracy": 0.14633184522390366,
      "num_tokens": 4696659.0,
      "step": 2270
    },
    {
      "entropy": 5.960624837875367,
      "epoch": 0.20636792452830188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499848971584951,
      "loss": 5.8072,
      "mean_token_accuracy": 0.15713879466056824,
      "num_tokens": 4706438.0,
      "step": 2275
    },
    {
      "entropy": 5.881855773925781,
      "epoch": 0.20682148040638607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998477839262611,
      "loss": 5.8119,
      "mean_token_accuracy": 0.15000793784856797,
      "num_tokens": 4716402.0,
      "step": 2280
    },
    {
      "entropy": 5.990855503082275,
      "epoch": 0.20727503628447025,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499846591617662,
      "loss": 5.8976,
      "mean_token_accuracy": 0.14913058578968047,
      "num_tokens": 4728488.0,
      "step": 2285
    },
    {
      "entropy": 5.952394151687622,
      "epoch": 0.20772859216255443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998453946591779,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1516925185918808,
      "num_tokens": 4738693.0,
      "step": 2290
    },
    {
      "entropy": 5.991802835464478,
      "epoch": 0.20818214804063861,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998441930508339,
      "loss": 5.8145,
      "mean_token_accuracy": 0.15066843181848527,
      "num_tokens": 4748895.0,
      "step": 2295
    },
    {
      "entropy": 6.047665977478028,
      "epoch": 0.2086357039187228,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998429867926547,
      "loss": 5.925,
      "mean_token_accuracy": 0.14194059297442435,
      "num_tokens": 4760135.0,
      "step": 2300
    },
    {
      "entropy": 6.035406303405762,
      "epoch": 0.20908925979680695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998417758846651,
      "loss": 5.9523,
      "mean_token_accuracy": 0.15069413632154466,
      "num_tokens": 4769935.0,
      "step": 2305
    },
    {
      "entropy": 6.0572274684906,
      "epoch": 0.20954281567489114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998405603268903,
      "loss": 5.8976,
      "mean_token_accuracy": 0.14732977300882338,
      "num_tokens": 4779520.0,
      "step": 2310
    },
    {
      "entropy": 5.93422679901123,
      "epoch": 0.20999637155297532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998393401193554,
      "loss": 5.8067,
      "mean_token_accuracy": 0.1515841841697693,
      "num_tokens": 4791256.0,
      "step": 2315
    },
    {
      "entropy": 5.881869649887085,
      "epoch": 0.2104499274310595,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998381152620857,
      "loss": 5.7539,
      "mean_token_accuracy": 0.15506275743246078,
      "num_tokens": 4801334.0,
      "step": 2320
    },
    {
      "entropy": 5.885318040847778,
      "epoch": 0.21090348330914369,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998368857551067,
      "loss": 5.7321,
      "mean_token_accuracy": 0.1555730938911438,
      "num_tokens": 4811361.0,
      "step": 2325
    },
    {
      "entropy": 5.969886207580567,
      "epoch": 0.21135703918722787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998356515984433,
      "loss": 5.807,
      "mean_token_accuracy": 0.15263233184814454,
      "num_tokens": 4821190.0,
      "step": 2330
    },
    {
      "entropy": 5.9508056640625,
      "epoch": 0.21181059506531205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998344127921213,
      "loss": 5.8241,
      "mean_token_accuracy": 0.1541541337966919,
      "num_tokens": 4830694.0,
      "step": 2335
    },
    {
      "entropy": 5.928486108779907,
      "epoch": 0.21226415094339623,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998331693361666,
      "loss": 5.8082,
      "mean_token_accuracy": 0.1508532479405403,
      "num_tokens": 4841344.0,
      "step": 2340
    },
    {
      "entropy": 5.951602220535278,
      "epoch": 0.21271770682148042,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998319212306046,
      "loss": 5.8021,
      "mean_token_accuracy": 0.14895136952400206,
      "num_tokens": 4851161.0,
      "step": 2345
    },
    {
      "entropy": 5.9326536655426025,
      "epoch": 0.21317126269956457,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499830668475461,
      "loss": 5.7841,
      "mean_token_accuracy": 0.14950372502207757,
      "num_tokens": 4860740.0,
      "step": 2350
    },
    {
      "entropy": 5.928380012512207,
      "epoch": 0.21362481857764876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499829411070762,
      "loss": 5.8168,
      "mean_token_accuracy": 0.15025952607393264,
      "num_tokens": 4871733.0,
      "step": 2355
    },
    {
      "entropy": 5.940114498138428,
      "epoch": 0.21407837445573294,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998281490165335,
      "loss": 5.8779,
      "mean_token_accuracy": 0.14290931671857834,
      "num_tokens": 4883617.0,
      "step": 2360
    },
    {
      "entropy": 6.0039795398712155,
      "epoch": 0.21453193033381712,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998268823128016,
      "loss": 5.8133,
      "mean_token_accuracy": 0.14657811671495438,
      "num_tokens": 4894519.0,
      "step": 2365
    },
    {
      "entropy": 5.913796472549438,
      "epoch": 0.2149854862119013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998256109595923,
      "loss": 5.7663,
      "mean_token_accuracy": 0.15133122503757476,
      "num_tokens": 4904965.0,
      "step": 2370
    },
    {
      "entropy": 5.905171871185303,
      "epoch": 0.2154390420899855,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998243349569323,
      "loss": 5.6806,
      "mean_token_accuracy": 0.16264960318803787,
      "num_tokens": 4914618.0,
      "step": 2375
    },
    {
      "entropy": 5.793045377731323,
      "epoch": 0.21589259796806967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998230543048477,
      "loss": 5.7376,
      "mean_token_accuracy": 0.15822856724262238,
      "num_tokens": 4924341.0,
      "step": 2380
    },
    {
      "entropy": 5.860754489898682,
      "epoch": 0.21634615384615385,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499821769003365,
      "loss": 5.7202,
      "mean_token_accuracy": 0.15276679992675782,
      "num_tokens": 4934231.0,
      "step": 2385
    },
    {
      "entropy": 5.90138463973999,
      "epoch": 0.21679970972423804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998204790525109,
      "loss": 5.7442,
      "mean_token_accuracy": 0.15233908742666244,
      "num_tokens": 4944303.0,
      "step": 2390
    },
    {
      "entropy": 5.892811250686646,
      "epoch": 0.2172532656023222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998191844523119,
      "loss": 5.8289,
      "mean_token_accuracy": 0.15439640134572982,
      "num_tokens": 4954571.0,
      "step": 2395
    },
    {
      "entropy": 5.912092924118042,
      "epoch": 0.21770682148040638,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499817885202795,
      "loss": 5.7943,
      "mean_token_accuracy": 0.15238530337810516,
      "num_tokens": 4964418.0,
      "step": 2400
    },
    {
      "entropy": 5.960055160522461,
      "epoch": 0.21816037735849056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998165813039869,
      "loss": 5.7833,
      "mean_token_accuracy": 0.1479422241449356,
      "num_tokens": 4975036.0,
      "step": 2405
    },
    {
      "entropy": 5.970535087585449,
      "epoch": 0.21861393323657474,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998152727559145,
      "loss": 5.8868,
      "mean_token_accuracy": 0.14599555134773254,
      "num_tokens": 4985272.0,
      "step": 2410
    },
    {
      "entropy": 6.004713773727417,
      "epoch": 0.21906748911465893,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499813959558605,
      "loss": 5.812,
      "mean_token_accuracy": 0.15417784452438354,
      "num_tokens": 4995187.0,
      "step": 2415
    },
    {
      "entropy": 5.99905571937561,
      "epoch": 0.2195210449927431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998126417120856,
      "loss": 5.8535,
      "mean_token_accuracy": 0.1530926562845707,
      "num_tokens": 5005013.0,
      "step": 2420
    },
    {
      "entropy": 5.994871044158936,
      "epoch": 0.2199746008708273,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998113192163835,
      "loss": 5.881,
      "mean_token_accuracy": 0.14475401639938354,
      "num_tokens": 5016789.0,
      "step": 2425
    },
    {
      "entropy": 5.921261167526245,
      "epoch": 0.22042815674891147,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499809992071526,
      "loss": 5.828,
      "mean_token_accuracy": 0.1518557384610176,
      "num_tokens": 5026913.0,
      "step": 2430
    },
    {
      "entropy": 5.9377799987792965,
      "epoch": 0.22088171262699566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998086602775406,
      "loss": 5.9104,
      "mean_token_accuracy": 0.14928861260414122,
      "num_tokens": 5037501.0,
      "step": 2435
    },
    {
      "entropy": 5.983516836166382,
      "epoch": 0.2213352685050798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998073238344547,
      "loss": 5.7602,
      "mean_token_accuracy": 0.15100034326314926,
      "num_tokens": 5048478.0,
      "step": 2440
    },
    {
      "entropy": 5.903777933120727,
      "epoch": 0.221788824383164,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998059827422962,
      "loss": 5.8255,
      "mean_token_accuracy": 0.14812486618757248,
      "num_tokens": 5059118.0,
      "step": 2445
    },
    {
      "entropy": 5.963100814819336,
      "epoch": 0.22224238026124818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998046370010926,
      "loss": 5.7848,
      "mean_token_accuracy": 0.15494603514671326,
      "num_tokens": 5070180.0,
      "step": 2450
    },
    {
      "entropy": 5.880324077606201,
      "epoch": 0.22269593613933236,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998032866108718,
      "loss": 5.7808,
      "mean_token_accuracy": 0.1510962054133415,
      "num_tokens": 5079912.0,
      "step": 2455
    },
    {
      "entropy": 5.994835710525512,
      "epoch": 0.22314949201741655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998019315716618,
      "loss": 5.8732,
      "mean_token_accuracy": 0.15168511420488356,
      "num_tokens": 5090549.0,
      "step": 2460
    },
    {
      "entropy": 5.909025812149048,
      "epoch": 0.22360304789550073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998005718834905,
      "loss": 5.7193,
      "mean_token_accuracy": 0.1588350161910057,
      "num_tokens": 5100880.0,
      "step": 2465
    },
    {
      "entropy": 5.910657835006714,
      "epoch": 0.2240566037735849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997992075463862,
      "loss": 5.8277,
      "mean_token_accuracy": 0.14884106069803238,
      "num_tokens": 5111796.0,
      "step": 2470
    },
    {
      "entropy": 5.975823974609375,
      "epoch": 0.2245101596516691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997978385603769,
      "loss": 5.8232,
      "mean_token_accuracy": 0.15248575657606125,
      "num_tokens": 5123468.0,
      "step": 2475
    },
    {
      "entropy": 5.874198150634766,
      "epoch": 0.22496371552975328,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997964649254911,
      "loss": 5.7465,
      "mean_token_accuracy": 0.15882757753133775,
      "num_tokens": 5132925.0,
      "step": 2480
    },
    {
      "entropy": 5.932957172393799,
      "epoch": 0.22541727140783743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997950866417572,
      "loss": 5.8278,
      "mean_token_accuracy": 0.14813223630189895,
      "num_tokens": 5143445.0,
      "step": 2485
    },
    {
      "entropy": 5.961171960830688,
      "epoch": 0.22587082728592162,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997937037092034,
      "loss": 5.745,
      "mean_token_accuracy": 0.14807531088590623,
      "num_tokens": 5153176.0,
      "step": 2490
    },
    {
      "entropy": 5.9957068920135494,
      "epoch": 0.2263243831640058,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997923161278587,
      "loss": 5.7973,
      "mean_token_accuracy": 0.15407470166683196,
      "num_tokens": 5162765.0,
      "step": 2495
    },
    {
      "entropy": 5.935280895233154,
      "epoch": 0.22677793904208998,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997909238977514,
      "loss": 5.8473,
      "mean_token_accuracy": 0.15373046398162843,
      "num_tokens": 5173390.0,
      "step": 2500
    },
    {
      "entropy": 5.851876974105835,
      "epoch": 0.22723149492017417,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997895270189108,
      "loss": 5.7455,
      "mean_token_accuracy": 0.15115862116217613,
      "num_tokens": 5184212.0,
      "step": 2505
    },
    {
      "entropy": 5.847887372970581,
      "epoch": 0.22768505079825835,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997881254913653,
      "loss": 5.7465,
      "mean_token_accuracy": 0.15100416243076326,
      "num_tokens": 5194311.0,
      "step": 2510
    },
    {
      "entropy": 5.867375993728638,
      "epoch": 0.22813860667634253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997867193151441,
      "loss": 5.7339,
      "mean_token_accuracy": 0.15452023446559907,
      "num_tokens": 5204708.0,
      "step": 2515
    },
    {
      "entropy": 5.844597864151001,
      "epoch": 0.22859216255442671,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997853084902763,
      "loss": 5.7152,
      "mean_token_accuracy": 0.15443815141916276,
      "num_tokens": 5214566.0,
      "step": 2520
    },
    {
      "entropy": 5.916435194015503,
      "epoch": 0.2290457184325109,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997838930167911,
      "loss": 5.7463,
      "mean_token_accuracy": 0.15494998544454575,
      "num_tokens": 5224762.0,
      "step": 2525
    },
    {
      "entropy": 5.9442017555236815,
      "epoch": 0.22949927431059505,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997824728947176,
      "loss": 5.8129,
      "mean_token_accuracy": 0.15142686367034913,
      "num_tokens": 5235146.0,
      "step": 2530
    },
    {
      "entropy": 5.965577602386475,
      "epoch": 0.22995283018867924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997810481240853,
      "loss": 5.786,
      "mean_token_accuracy": 0.1514734596014023,
      "num_tokens": 5243867.0,
      "step": 2535
    },
    {
      "entropy": 5.83701639175415,
      "epoch": 0.23040638606676342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997796187049238,
      "loss": 5.702,
      "mean_token_accuracy": 0.153061743080616,
      "num_tokens": 5254279.0,
      "step": 2540
    },
    {
      "entropy": 6.0492133617401125,
      "epoch": 0.2308599419448476,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997781846372623,
      "loss": 5.7978,
      "mean_token_accuracy": 0.1522805020213127,
      "num_tokens": 5264769.0,
      "step": 2545
    },
    {
      "entropy": 5.893232774734497,
      "epoch": 0.23131349782293179,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997767459211307,
      "loss": 5.7644,
      "mean_token_accuracy": 0.14629466086626053,
      "num_tokens": 5274291.0,
      "step": 2550
    },
    {
      "entropy": 5.868297386169433,
      "epoch": 0.23176705370101597,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997753025565587,
      "loss": 5.695,
      "mean_token_accuracy": 0.15659698024392127,
      "num_tokens": 5284366.0,
      "step": 2555
    },
    {
      "entropy": 5.889286661148072,
      "epoch": 0.23222060957910015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997738545435763,
      "loss": 5.8173,
      "mean_token_accuracy": 0.15293926745653152,
      "num_tokens": 5294494.0,
      "step": 2560
    },
    {
      "entropy": 5.899365377426148,
      "epoch": 0.23267416545718433,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997724018822133,
      "loss": 5.813,
      "mean_token_accuracy": 0.15529518127441405,
      "num_tokens": 5305606.0,
      "step": 2565
    },
    {
      "entropy": 5.90656361579895,
      "epoch": 0.23312772133526852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997709445724996,
      "loss": 5.8101,
      "mean_token_accuracy": 0.1502317950129509,
      "num_tokens": 5315677.0,
      "step": 2570
    },
    {
      "entropy": 5.9617547512054445,
      "epoch": 0.23358127721335267,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997694826144656,
      "loss": 5.866,
      "mean_token_accuracy": 0.14772987067699433,
      "num_tokens": 5326453.0,
      "step": 2575
    },
    {
      "entropy": 5.8899938583374025,
      "epoch": 0.23403483309143686,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997680160081414,
      "loss": 5.7592,
      "mean_token_accuracy": 0.1525429829955101,
      "num_tokens": 5337591.0,
      "step": 2580
    },
    {
      "entropy": 5.914120006561279,
      "epoch": 0.23448838896952104,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997665447535573,
      "loss": 5.7414,
      "mean_token_accuracy": 0.1570722669363022,
      "num_tokens": 5349400.0,
      "step": 2585
    },
    {
      "entropy": 5.9674485206604,
      "epoch": 0.23494194484760522,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997650688507438,
      "loss": 5.7325,
      "mean_token_accuracy": 0.15442833602428435,
      "num_tokens": 5359984.0,
      "step": 2590
    },
    {
      "entropy": 5.855543184280395,
      "epoch": 0.2353955007256894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997635882997314,
      "loss": 5.8149,
      "mean_token_accuracy": 0.16013004034757614,
      "num_tokens": 5369254.0,
      "step": 2595
    },
    {
      "entropy": 5.920786714553833,
      "epoch": 0.2358490566037736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997621031005507,
      "loss": 5.755,
      "mean_token_accuracy": 0.15076399594545364,
      "num_tokens": 5379648.0,
      "step": 2600
    },
    {
      "entropy": 5.818716669082642,
      "epoch": 0.23630261248185777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997606132532325,
      "loss": 5.7028,
      "mean_token_accuracy": 0.16014982685446738,
      "num_tokens": 5390321.0,
      "step": 2605
    },
    {
      "entropy": 5.850492286682129,
      "epoch": 0.23675616835994195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997591187578075,
      "loss": 5.7541,
      "mean_token_accuracy": 0.1600041940808296,
      "num_tokens": 5401078.0,
      "step": 2610
    },
    {
      "entropy": 5.8976398468017575,
      "epoch": 0.23720972423802614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997576196143067,
      "loss": 5.71,
      "mean_token_accuracy": 0.16146167367696762,
      "num_tokens": 5411977.0,
      "step": 2615
    },
    {
      "entropy": 5.871705007553101,
      "epoch": 0.2376632801161103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499756115822761,
      "loss": 5.7102,
      "mean_token_accuracy": 0.15028589516878127,
      "num_tokens": 5422967.0,
      "step": 2620
    },
    {
      "entropy": 5.950361585617065,
      "epoch": 0.23811683599419448,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997546073832015,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16250978112220765,
      "num_tokens": 5432260.0,
      "step": 2625
    },
    {
      "entropy": 5.842865753173828,
      "epoch": 0.23857039187227866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997530942956595,
      "loss": 5.7182,
      "mean_token_accuracy": 0.15647428780794143,
      "num_tokens": 5441706.0,
      "step": 2630
    },
    {
      "entropy": 5.885302925109864,
      "epoch": 0.23902394775036284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997515765601662,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1620200365781784,
      "num_tokens": 5452397.0,
      "step": 2635
    },
    {
      "entropy": 5.802298927307129,
      "epoch": 0.23947750362844702,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997500541767531,
      "loss": 5.6999,
      "mean_token_accuracy": 0.1544307217001915,
      "num_tokens": 5463762.0,
      "step": 2640
    },
    {
      "entropy": 5.895985698699951,
      "epoch": 0.2399310595065312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997485271454516,
      "loss": 5.7718,
      "mean_token_accuracy": 0.14597637057304383,
      "num_tokens": 5474622.0,
      "step": 2645
    },
    {
      "entropy": 5.872747230529785,
      "epoch": 0.2403846153846154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997469954662933,
      "loss": 5.7431,
      "mean_token_accuracy": 0.15389883369207383,
      "num_tokens": 5484697.0,
      "step": 2650
    },
    {
      "entropy": 5.795325708389282,
      "epoch": 0.24083817126269957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997454591393097,
      "loss": 5.744,
      "mean_token_accuracy": 0.15505960136651992,
      "num_tokens": 5496433.0,
      "step": 2655
    },
    {
      "entropy": 5.929238319396973,
      "epoch": 0.24129172714078376,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997439181645329,
      "loss": 5.8642,
      "mean_token_accuracy": 0.14926385432481765,
      "num_tokens": 5507917.0,
      "step": 2660
    },
    {
      "entropy": 6.066269302368164,
      "epoch": 0.2417452830188679,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997423725419945,
      "loss": 5.8473,
      "mean_token_accuracy": 0.15168920159339905,
      "num_tokens": 5519159.0,
      "step": 2665
    },
    {
      "entropy": 5.896298265457153,
      "epoch": 0.2421988388969521,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997408222717267,
      "loss": 5.7941,
      "mean_token_accuracy": 0.15869408696889878,
      "num_tokens": 5529180.0,
      "step": 2670
    },
    {
      "entropy": 5.829392433166504,
      "epoch": 0.24265239477503628,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997392673537612,
      "loss": 5.7138,
      "mean_token_accuracy": 0.1638789251446724,
      "num_tokens": 5539363.0,
      "step": 2675
    },
    {
      "entropy": 5.891768503189087,
      "epoch": 0.24310595065312046,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997377077881306,
      "loss": 5.8155,
      "mean_token_accuracy": 0.15023489147424698,
      "num_tokens": 5549841.0,
      "step": 2680
    },
    {
      "entropy": 5.889240789413452,
      "epoch": 0.24355950653120464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997361435748668,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16268197000026702,
      "num_tokens": 5560299.0,
      "step": 2685
    },
    {
      "entropy": 5.9125152111053465,
      "epoch": 0.24401306240928883,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997345747140024,
      "loss": 5.7689,
      "mean_token_accuracy": 0.15359243452548982,
      "num_tokens": 5571081.0,
      "step": 2690
    },
    {
      "entropy": 5.8776861190795895,
      "epoch": 0.244466618287373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997330012055695,
      "loss": 5.7108,
      "mean_token_accuracy": 0.15765073597431184,
      "num_tokens": 5581389.0,
      "step": 2695
    },
    {
      "entropy": 5.861866426467896,
      "epoch": 0.2449201741654572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997314230496011,
      "loss": 5.754,
      "mean_token_accuracy": 0.1500275418162346,
      "num_tokens": 5592347.0,
      "step": 2700
    },
    {
      "entropy": 5.839778137207031,
      "epoch": 0.24537373004354138,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997298402461296,
      "loss": 5.7228,
      "mean_token_accuracy": 0.15842826217412947,
      "num_tokens": 5603411.0,
      "step": 2705
    },
    {
      "entropy": 5.9260224342346195,
      "epoch": 0.24582728592162553,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997282527951876,
      "loss": 5.7162,
      "mean_token_accuracy": 0.15986051708459853,
      "num_tokens": 5614615.0,
      "step": 2710
    },
    {
      "entropy": 5.790437936782837,
      "epoch": 0.24628084179970972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997266606968081,
      "loss": 5.6868,
      "mean_token_accuracy": 0.1561066210269928,
      "num_tokens": 5624582.0,
      "step": 2715
    },
    {
      "entropy": 5.811624383926391,
      "epoch": 0.2467343976777939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997250639510241,
      "loss": 5.7,
      "mean_token_accuracy": 0.15437885671854018,
      "num_tokens": 5634310.0,
      "step": 2720
    },
    {
      "entropy": 5.889466571807861,
      "epoch": 0.24718795355587808,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997234625578684,
      "loss": 5.7947,
      "mean_token_accuracy": 0.14918309599161148,
      "num_tokens": 5644018.0,
      "step": 2725
    },
    {
      "entropy": 5.858435821533203,
      "epoch": 0.24764150943396226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997218565173742,
      "loss": 5.7233,
      "mean_token_accuracy": 0.15628414005041122,
      "num_tokens": 5653592.0,
      "step": 2730
    },
    {
      "entropy": 5.904895401000976,
      "epoch": 0.24809506531204645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997202458295748,
      "loss": 5.7632,
      "mean_token_accuracy": 0.15781814455986024,
      "num_tokens": 5662812.0,
      "step": 2735
    },
    {
      "entropy": 5.902950763702393,
      "epoch": 0.24854862119013063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997186304945034,
      "loss": 5.7511,
      "mean_token_accuracy": 0.15532326996326445,
      "num_tokens": 5672939.0,
      "step": 2740
    },
    {
      "entropy": 5.827338409423828,
      "epoch": 0.24900217706821481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997170105121935,
      "loss": 5.7313,
      "mean_token_accuracy": 0.15588268786668777,
      "num_tokens": 5683350.0,
      "step": 2745
    },
    {
      "entropy": 5.919451379776001,
      "epoch": 0.249455732946299,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997153858826786,
      "loss": 5.816,
      "mean_token_accuracy": 0.14666671603918074,
      "num_tokens": 5694516.0,
      "step": 2750
    },
    {
      "entropy": 5.843132162094117,
      "epoch": 0.24990928882438315,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997137566059922,
      "loss": 5.6435,
      "mean_token_accuracy": 0.15954998284578323,
      "num_tokens": 5705443.0,
      "step": 2755
    },
    {
      "entropy": 5.838101053237915,
      "epoch": 0.25036284470246734,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997121226821681,
      "loss": 5.742,
      "mean_token_accuracy": 0.15265777483582496,
      "num_tokens": 5715933.0,
      "step": 2760
    },
    {
      "entropy": 5.910187196731568,
      "epoch": 0.25081640058055155,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049971048411124,
      "loss": 5.8084,
      "mean_token_accuracy": 0.1447537824511528,
      "num_tokens": 5726630.0,
      "step": 2765
    },
    {
      "entropy": 5.9618613719940186,
      "epoch": 0.2512699564586357,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997088408932418,
      "loss": 5.7349,
      "mean_token_accuracy": 0.15529899746179582,
      "num_tokens": 5736890.0,
      "step": 2770
    },
    {
      "entropy": 5.754767274856567,
      "epoch": 0.25172351233671986,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997071930282076,
      "loss": 5.6394,
      "mean_token_accuracy": 0.16765589267015457,
      "num_tokens": 5747042.0,
      "step": 2775
    },
    {
      "entropy": 5.836727952957153,
      "epoch": 0.25217706821480407,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997055405161712,
      "loss": 5.7737,
      "mean_token_accuracy": 0.15657578855752946,
      "num_tokens": 5757367.0,
      "step": 2780
    },
    {
      "entropy": 5.892507696151734,
      "epoch": 0.2526306240928882,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997038833571671,
      "loss": 5.8198,
      "mean_token_accuracy": 0.15040036141872407,
      "num_tokens": 5769557.0,
      "step": 2785
    },
    {
      "entropy": 5.882242345809937,
      "epoch": 0.25308417997097243,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997022215512295,
      "loss": 5.7227,
      "mean_token_accuracy": 0.162361441552639,
      "num_tokens": 5780615.0,
      "step": 2790
    },
    {
      "entropy": 5.901854085922241,
      "epoch": 0.2535377358490566,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997005550983925,
      "loss": 5.7026,
      "mean_token_accuracy": 0.1581372618675232,
      "num_tokens": 5791271.0,
      "step": 2795
    },
    {
      "entropy": 5.816129541397094,
      "epoch": 0.2539912917271408,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499698883998691,
      "loss": 5.6376,
      "mean_token_accuracy": 0.15995124131441116,
      "num_tokens": 5801117.0,
      "step": 2800
    },
    {
      "entropy": 5.778412342071533,
      "epoch": 0.25444484760522496,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996972082521591,
      "loss": 5.6657,
      "mean_token_accuracy": 0.16171006560325624,
      "num_tokens": 5811696.0,
      "step": 2805
    },
    {
      "entropy": 5.8469995021820065,
      "epoch": 0.25489840348330917,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996955278588319,
      "loss": 5.6983,
      "mean_token_accuracy": 0.16263554394245147,
      "num_tokens": 5822207.0,
      "step": 2810
    },
    {
      "entropy": 5.802956390380859,
      "epoch": 0.2553519593613933,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996938428187437,
      "loss": 5.6452,
      "mean_token_accuracy": 0.1667325183749199,
      "num_tokens": 5831927.0,
      "step": 2815
    },
    {
      "entropy": 5.801891422271728,
      "epoch": 0.2558055152394775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996921531319297,
      "loss": 5.7182,
      "mean_token_accuracy": 0.15510134547948837,
      "num_tokens": 5842093.0,
      "step": 2820
    },
    {
      "entropy": 5.818565034866333,
      "epoch": 0.2562590711175617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996904587984246,
      "loss": 5.6452,
      "mean_token_accuracy": 0.16608070433139802,
      "num_tokens": 5850551.0,
      "step": 2825
    },
    {
      "entropy": 5.797169351577759,
      "epoch": 0.25671262699564584,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996887598182636,
      "loss": 5.6412,
      "mean_token_accuracy": 0.1570335805416107,
      "num_tokens": 5860651.0,
      "step": 2830
    },
    {
      "entropy": 5.70640926361084,
      "epoch": 0.25716618287373005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996870561914819,
      "loss": 5.6313,
      "mean_token_accuracy": 0.1594139203429222,
      "num_tokens": 5871500.0,
      "step": 2835
    },
    {
      "entropy": 5.810226488113403,
      "epoch": 0.2576197387518142,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996853479181145,
      "loss": 5.6415,
      "mean_token_accuracy": 0.15245891511440277,
      "num_tokens": 5881523.0,
      "step": 2840
    },
    {
      "entropy": 5.9306385040283205,
      "epoch": 0.2580732946298984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996836349981969,
      "loss": 5.7752,
      "mean_token_accuracy": 0.1541695550084114,
      "num_tokens": 5891162.0,
      "step": 2845
    },
    {
      "entropy": 5.793090677261352,
      "epoch": 0.2585268505079826,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996819174317645,
      "loss": 5.6667,
      "mean_token_accuracy": 0.15518372505903244,
      "num_tokens": 5901818.0,
      "step": 2850
    },
    {
      "entropy": 5.8290753841400145,
      "epoch": 0.2589804063860668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996801952188528,
      "loss": 5.7439,
      "mean_token_accuracy": 0.15479833632707596,
      "num_tokens": 5913889.0,
      "step": 2855
    },
    {
      "entropy": 5.848030471801758,
      "epoch": 0.25943396226415094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996784683594973,
      "loss": 5.7061,
      "mean_token_accuracy": 0.16026207208633422,
      "num_tokens": 5923429.0,
      "step": 2860
    },
    {
      "entropy": 5.953624963760376,
      "epoch": 0.2598875181422351,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996767368537338,
      "loss": 5.7536,
      "mean_token_accuracy": 0.15464293360710143,
      "num_tokens": 5933847.0,
      "step": 2865
    },
    {
      "entropy": 5.797641706466675,
      "epoch": 0.2603410740203193,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996750007015983,
      "loss": 5.6199,
      "mean_token_accuracy": 0.15784441232681273,
      "num_tokens": 5943874.0,
      "step": 2870
    },
    {
      "entropy": 5.836393356323242,
      "epoch": 0.26079462989840346,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996732599031264,
      "loss": 5.6901,
      "mean_token_accuracy": 0.15797710716724395,
      "num_tokens": 5953653.0,
      "step": 2875
    },
    {
      "entropy": 5.831515836715698,
      "epoch": 0.2612481857764877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996715144583543,
      "loss": 5.6401,
      "mean_token_accuracy": 0.16178645938634872,
      "num_tokens": 5964039.0,
      "step": 2880
    },
    {
      "entropy": 5.8883569717407225,
      "epoch": 0.26170174165457183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499669764367318,
      "loss": 5.7538,
      "mean_token_accuracy": 0.15600325912237167,
      "num_tokens": 5974083.0,
      "step": 2885
    },
    {
      "entropy": 5.775406360626221,
      "epoch": 0.26215529753265604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996680096300537,
      "loss": 5.711,
      "mean_token_accuracy": 0.15503483414649963,
      "num_tokens": 5984629.0,
      "step": 2890
    },
    {
      "entropy": 5.7855353355407715,
      "epoch": 0.2626088534107402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996662502465978,
      "loss": 5.6067,
      "mean_token_accuracy": 0.1710730493068695,
      "num_tokens": 5994295.0,
      "step": 2895
    },
    {
      "entropy": 5.8054187297821045,
      "epoch": 0.2630624092888244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996644862169864,
      "loss": 5.665,
      "mean_token_accuracy": 0.15672381818294526,
      "num_tokens": 6004156.0,
      "step": 2900
    },
    {
      "entropy": 5.815729665756225,
      "epoch": 0.26351596516690856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996627175412563,
      "loss": 5.7391,
      "mean_token_accuracy": 0.16197800040245056,
      "num_tokens": 6014188.0,
      "step": 2905
    },
    {
      "entropy": 5.904583549499511,
      "epoch": 0.2639695210449927,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499660944219444,
      "loss": 5.8105,
      "mean_token_accuracy": 0.15212338119745256,
      "num_tokens": 6024879.0,
      "step": 2910
    },
    {
      "entropy": 5.926459264755249,
      "epoch": 0.2644230769230769,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996591662515861,
      "loss": 5.8701,
      "mean_token_accuracy": 0.15266856625676156,
      "num_tokens": 6036374.0,
      "step": 2915
    },
    {
      "entropy": 5.832221269607544,
      "epoch": 0.2648766328011611,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996573836377193,
      "loss": 5.7056,
      "mean_token_accuracy": 0.15986677557229995,
      "num_tokens": 6046432.0,
      "step": 2920
    },
    {
      "entropy": 5.8731166362762455,
      "epoch": 0.2653301886792453,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996555963778806,
      "loss": 5.7218,
      "mean_token_accuracy": 0.15033943355083465,
      "num_tokens": 6056864.0,
      "step": 2925
    },
    {
      "entropy": 5.9467894554138185,
      "epoch": 0.26578374455732945,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499653804472107,
      "loss": 5.7396,
      "mean_token_accuracy": 0.1595652773976326,
      "num_tokens": 6067294.0,
      "step": 2930
    },
    {
      "entropy": 5.863904571533203,
      "epoch": 0.26623730043541366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996520079204353,
      "loss": 5.741,
      "mean_token_accuracy": 0.15215880274772645,
      "num_tokens": 6077587.0,
      "step": 2935
    },
    {
      "entropy": 5.880077457427978,
      "epoch": 0.2666908563134978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499650206722903,
      "loss": 5.6892,
      "mean_token_accuracy": 0.1558769851922989,
      "num_tokens": 6089306.0,
      "step": 2940
    },
    {
      "entropy": 5.7567362785339355,
      "epoch": 0.267144412191582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996484008795471,
      "loss": 5.6433,
      "mean_token_accuracy": 0.15309089198708534,
      "num_tokens": 6099651.0,
      "step": 2945
    },
    {
      "entropy": 5.811124467849732,
      "epoch": 0.2675979680696662,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499646590390405,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1572161540389061,
      "num_tokens": 6110063.0,
      "step": 2950
    },
    {
      "entropy": 5.807859134674072,
      "epoch": 0.26805152394775034,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996447752555142,
      "loss": 5.6777,
      "mean_token_accuracy": 0.15904018133878708,
      "num_tokens": 6120031.0,
      "step": 2955
    },
    {
      "entropy": 5.8083930015563965,
      "epoch": 0.26850507982583455,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996429554749122,
      "loss": 5.7183,
      "mean_token_accuracy": 0.15861095041036605,
      "num_tokens": 6131274.0,
      "step": 2960
    },
    {
      "entropy": 5.888201808929443,
      "epoch": 0.2689586357039187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996411310486367,
      "loss": 5.7658,
      "mean_token_accuracy": 0.15850751847028732,
      "num_tokens": 6140878.0,
      "step": 2965
    },
    {
      "entropy": 5.859344720840454,
      "epoch": 0.2694121915820029,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996393019767253,
      "loss": 5.7646,
      "mean_token_accuracy": 0.1520871713757515,
      "num_tokens": 6152229.0,
      "step": 2970
    },
    {
      "entropy": 5.737780809402466,
      "epoch": 0.26986574746008707,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996374682592158,
      "loss": 5.6579,
      "mean_token_accuracy": 0.15743956118822097,
      "num_tokens": 6162311.0,
      "step": 2975
    },
    {
      "entropy": 5.8878035068511965,
      "epoch": 0.2703193033381713,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996356298961463,
      "loss": 5.7654,
      "mean_token_accuracy": 0.15596626251935958,
      "num_tokens": 6172806.0,
      "step": 2980
    },
    {
      "entropy": 5.860615348815918,
      "epoch": 0.27077285921625543,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996337868875549,
      "loss": 5.7448,
      "mean_token_accuracy": 0.1540456607937813,
      "num_tokens": 6183479.0,
      "step": 2985
    },
    {
      "entropy": 5.865048408508301,
      "epoch": 0.27122641509433965,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996319392334792,
      "loss": 5.7043,
      "mean_token_accuracy": 0.15600668787956237,
      "num_tokens": 6194201.0,
      "step": 2990
    },
    {
      "entropy": 5.800745820999145,
      "epoch": 0.2716799709724238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499630086933958,
      "loss": 5.6423,
      "mean_token_accuracy": 0.1544359117746353,
      "num_tokens": 6204774.0,
      "step": 2995
    },
    {
      "entropy": 5.796499633789063,
      "epoch": 0.27213352685050796,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996282299890292,
      "loss": 5.6475,
      "mean_token_accuracy": 0.15237775593996047,
      "num_tokens": 6215056.0,
      "step": 3000
    },
    {
      "epoch": 0.27213352685050796,
      "eval_entropy": 5.64969411068912,
      "eval_loss": 5.721183776855469,
      "eval_mean_token_accuracy": 0.16399123540223992,
      "eval_num_tokens": 6215056.0,
      "eval_runtime": 20.9259,
      "eval_samples_per_second": 1689.772,
      "eval_steps_per_second": 211.221,
      "step": 3000
    },
    {
      "entropy": 5.928546047210693,
      "epoch": 0.27258708272859217,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996263683987316,
      "loss": 5.754,
      "mean_token_accuracy": 0.15922772735357285,
      "num_tokens": 6226264.0,
      "step": 3005
    },
    {
      "entropy": 5.842231750488281,
      "epoch": 0.2730406386066763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996245021631033,
      "loss": 5.6621,
      "mean_token_accuracy": 0.1633378177881241,
      "num_tokens": 6235137.0,
      "step": 3010
    },
    {
      "entropy": 5.790783786773682,
      "epoch": 0.27349419448476053,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499622631282183,
      "loss": 5.6825,
      "mean_token_accuracy": 0.16040071249008178,
      "num_tokens": 6245411.0,
      "step": 3015
    },
    {
      "entropy": 5.880749797821045,
      "epoch": 0.2739477503628447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996207557560095,
      "loss": 5.7469,
      "mean_token_accuracy": 0.15628764629364014,
      "num_tokens": 6255633.0,
      "step": 3020
    },
    {
      "entropy": 5.8452211856842045,
      "epoch": 0.2744013062409289,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996188755846216,
      "loss": 5.7219,
      "mean_token_accuracy": 0.15557182282209397,
      "num_tokens": 6265109.0,
      "step": 3025
    },
    {
      "entropy": 5.8197873592376705,
      "epoch": 0.27485486211901305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499616990768058,
      "loss": 5.7484,
      "mean_token_accuracy": 0.15053788274526597,
      "num_tokens": 6275241.0,
      "step": 3030
    },
    {
      "entropy": 5.869422435760498,
      "epoch": 0.27530841799709727,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996151013063579,
      "loss": 5.7928,
      "mean_token_accuracy": 0.1479799047112465,
      "num_tokens": 6286603.0,
      "step": 3035
    },
    {
      "entropy": 5.895780897140503,
      "epoch": 0.2757619738751814,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996132071995601,
      "loss": 5.6345,
      "mean_token_accuracy": 0.16945138126611708,
      "num_tokens": 6297219.0,
      "step": 3040
    },
    {
      "entropy": 5.763274097442627,
      "epoch": 0.2762155297532656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996113084477042,
      "loss": 5.6749,
      "mean_token_accuracy": 0.16423184126615525,
      "num_tokens": 6307694.0,
      "step": 3045
    },
    {
      "entropy": 5.769646883010864,
      "epoch": 0.2766690856313498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996094050508291,
      "loss": 5.604,
      "mean_token_accuracy": 0.16311236768960952,
      "num_tokens": 6317384.0,
      "step": 3050
    },
    {
      "entropy": 5.845623970031738,
      "epoch": 0.27712264150943394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996074970089741,
      "loss": 5.7395,
      "mean_token_accuracy": 0.15895291715860366,
      "num_tokens": 6327558.0,
      "step": 3055
    },
    {
      "entropy": 5.829644441604614,
      "epoch": 0.27757619738751815,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996055843221789,
      "loss": 5.6008,
      "mean_token_accuracy": 0.16788485497236252,
      "num_tokens": 6337828.0,
      "step": 3060
    },
    {
      "entropy": 5.797014570236206,
      "epoch": 0.2780297532656023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996036669904831,
      "loss": 5.6642,
      "mean_token_accuracy": 0.15934574156999587,
      "num_tokens": 6348678.0,
      "step": 3065
    },
    {
      "entropy": 5.835013437271118,
      "epoch": 0.2784833091436865,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996017450139261,
      "loss": 5.7507,
      "mean_token_accuracy": 0.1531164437532425,
      "num_tokens": 6359670.0,
      "step": 3070
    },
    {
      "entropy": 5.947513818740845,
      "epoch": 0.2789368650217707,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995998183925478,
      "loss": 5.7193,
      "mean_token_accuracy": 0.1559832513332367,
      "num_tokens": 6371034.0,
      "step": 3075
    },
    {
      "entropy": 5.747111892700195,
      "epoch": 0.2793904208998549,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995978871263881,
      "loss": 5.6698,
      "mean_token_accuracy": 0.1631323963403702,
      "num_tokens": 6380672.0,
      "step": 3080
    },
    {
      "entropy": 5.774804639816284,
      "epoch": 0.27984397677793904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995959512154868,
      "loss": 5.7757,
      "mean_token_accuracy": 0.1484546534717083,
      "num_tokens": 6390799.0,
      "step": 3085
    },
    {
      "entropy": 5.823790121078491,
      "epoch": 0.2802975326560232,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499594010659884,
      "loss": 5.6352,
      "mean_token_accuracy": 0.16301891803741456,
      "num_tokens": 6400965.0,
      "step": 3090
    },
    {
      "entropy": 5.842621088027954,
      "epoch": 0.2807510885341074,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995920654596199,
      "loss": 5.6585,
      "mean_token_accuracy": 0.15496362000703812,
      "num_tokens": 6411567.0,
      "step": 3095
    },
    {
      "entropy": 5.816281747817993,
      "epoch": 0.28120464441219156,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995901156147345,
      "loss": 5.7114,
      "mean_token_accuracy": 0.16414457112550734,
      "num_tokens": 6421193.0,
      "step": 3100
    },
    {
      "entropy": 5.827236461639404,
      "epoch": 0.2816582002902758,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995881611252684,
      "loss": 5.7181,
      "mean_token_accuracy": 0.16552949845790862,
      "num_tokens": 6430281.0,
      "step": 3105
    },
    {
      "entropy": 5.8368793487548825,
      "epoch": 0.28211175616835993,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995862019912619,
      "loss": 5.7229,
      "mean_token_accuracy": 0.15318924188613892,
      "num_tokens": 6441055.0,
      "step": 3110
    },
    {
      "entropy": 5.827468013763427,
      "epoch": 0.28256531204644414,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995842382127555,
      "loss": 5.638,
      "mean_token_accuracy": 0.1578447103500366,
      "num_tokens": 6451082.0,
      "step": 3115
    },
    {
      "entropy": 5.795475244522095,
      "epoch": 0.2830188679245283,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995822697897897,
      "loss": 5.7249,
      "mean_token_accuracy": 0.15724158138036728,
      "num_tokens": 6461617.0,
      "step": 3120
    },
    {
      "entropy": 5.884941148757934,
      "epoch": 0.2834724238026125,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995802967224056,
      "loss": 5.7196,
      "mean_token_accuracy": 0.15904805064201355,
      "num_tokens": 6472780.0,
      "step": 3125
    },
    {
      "entropy": 5.719329166412353,
      "epoch": 0.28392597968069666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995783190106434,
      "loss": 5.6335,
      "mean_token_accuracy": 0.15374166816473006,
      "num_tokens": 6484078.0,
      "step": 3130
    },
    {
      "entropy": 5.816254758834839,
      "epoch": 0.2843795355587808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995763366545445,
      "loss": 5.7225,
      "mean_token_accuracy": 0.15541643053293228,
      "num_tokens": 6494366.0,
      "step": 3135
    },
    {
      "entropy": 5.751023006439209,
      "epoch": 0.284833091436865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995743496541497,
      "loss": 5.6544,
      "mean_token_accuracy": 0.1647457867860794,
      "num_tokens": 6505461.0,
      "step": 3140
    },
    {
      "entropy": 5.815382051467895,
      "epoch": 0.2852866473149492,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995723580095001,
      "loss": 5.6527,
      "mean_token_accuracy": 0.16272574812173843,
      "num_tokens": 6516141.0,
      "step": 3145
    },
    {
      "entropy": 5.74614405632019,
      "epoch": 0.2857402031930334,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499570361720637,
      "loss": 5.6605,
      "mean_token_accuracy": 0.15768903791904448,
      "num_tokens": 6525536.0,
      "step": 3150
    },
    {
      "entropy": 5.821631193161011,
      "epoch": 0.28619375907111755,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995683607876015,
      "loss": 5.7061,
      "mean_token_accuracy": 0.15657218396663666,
      "num_tokens": 6535852.0,
      "step": 3155
    },
    {
      "entropy": 5.84665880203247,
      "epoch": 0.28664731494920176,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995663552104352,
      "loss": 5.6836,
      "mean_token_accuracy": 0.16722792834043504,
      "num_tokens": 6546353.0,
      "step": 3160
    },
    {
      "entropy": 5.704407453536987,
      "epoch": 0.2871008708272859,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995643449891793,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1609612762928009,
      "num_tokens": 6557782.0,
      "step": 3165
    },
    {
      "entropy": 5.824968910217285,
      "epoch": 0.2875544267053701,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995623301238757,
      "loss": 5.6597,
      "mean_token_accuracy": 0.16069735288619996,
      "num_tokens": 6568100.0,
      "step": 3170
    },
    {
      "entropy": 5.805774164199829,
      "epoch": 0.2880079825834543,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995603106145657,
      "loss": 5.7218,
      "mean_token_accuracy": 0.1530776634812355,
      "num_tokens": 6578698.0,
      "step": 3175
    },
    {
      "entropy": 5.830034828186035,
      "epoch": 0.28846153846153844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995582864612913,
      "loss": 5.7009,
      "mean_token_accuracy": 0.15772538930177687,
      "num_tokens": 6588908.0,
      "step": 3180
    },
    {
      "entropy": 5.755585670471191,
      "epoch": 0.28891509433962265,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995562576640943,
      "loss": 5.5465,
      "mean_token_accuracy": 0.16224001944065095,
      "num_tokens": 6599445.0,
      "step": 3185
    },
    {
      "entropy": 5.754920482635498,
      "epoch": 0.2893686502177068,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995542242230167,
      "loss": 5.6333,
      "mean_token_accuracy": 0.16428707391023636,
      "num_tokens": 6610246.0,
      "step": 3190
    },
    {
      "entropy": 5.780771970748901,
      "epoch": 0.289822206095791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995521861381006,
      "loss": 5.6527,
      "mean_token_accuracy": 0.15799290537834168,
      "num_tokens": 6620501.0,
      "step": 3195
    },
    {
      "entropy": 5.877050399780273,
      "epoch": 0.29027576197387517,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499550143409388,
      "loss": 5.7227,
      "mean_token_accuracy": 0.15273681730031968,
      "num_tokens": 6631206.0,
      "step": 3200
    },
    {
      "entropy": 5.770636701583863,
      "epoch": 0.2907293178519594,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995480960369211,
      "loss": 5.6805,
      "mean_token_accuracy": 0.15422717183828355,
      "num_tokens": 6641600.0,
      "step": 3205
    },
    {
      "entropy": 5.789561319351196,
      "epoch": 0.29118287373004353,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995460440207425,
      "loss": 5.7389,
      "mean_token_accuracy": 0.15444392561912537,
      "num_tokens": 6652319.0,
      "step": 3210
    },
    {
      "entropy": 5.84764404296875,
      "epoch": 0.29163642960812775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995439873608945,
      "loss": 5.6796,
      "mean_token_accuracy": 0.15282317996025085,
      "num_tokens": 6663702.0,
      "step": 3215
    },
    {
      "entropy": 5.774284696578979,
      "epoch": 0.2920899854862119,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995419260574196,
      "loss": 5.6669,
      "mean_token_accuracy": 0.1519273892045021,
      "num_tokens": 6673908.0,
      "step": 3220
    },
    {
      "entropy": 5.769118547439575,
      "epoch": 0.29254354136429606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995398601103603,
      "loss": 5.682,
      "mean_token_accuracy": 0.15996284335851668,
      "num_tokens": 6683237.0,
      "step": 3225
    },
    {
      "entropy": 5.814051771163941,
      "epoch": 0.29299709724238027,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995377895197597,
      "loss": 5.6143,
      "mean_token_accuracy": 0.16433483213186265,
      "num_tokens": 6693451.0,
      "step": 3230
    },
    {
      "entropy": 5.719491100311279,
      "epoch": 0.2934506531204644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995357142856602,
      "loss": 5.5676,
      "mean_token_accuracy": 0.16208960115909576,
      "num_tokens": 6703479.0,
      "step": 3235
    },
    {
      "entropy": 5.692001056671143,
      "epoch": 0.29390420899854863,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499533634408105,
      "loss": 5.5692,
      "mean_token_accuracy": 0.15959729105234147,
      "num_tokens": 6715321.0,
      "step": 3240
    },
    {
      "entropy": 5.770639801025391,
      "epoch": 0.2943577648766328,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499531549887137,
      "loss": 5.6309,
      "mean_token_accuracy": 0.15734803080558776,
      "num_tokens": 6725541.0,
      "step": 3245
    },
    {
      "entropy": 5.791307258605957,
      "epoch": 0.294811320754717,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995294607227993,
      "loss": 5.6795,
      "mean_token_accuracy": 0.16653743833303453,
      "num_tokens": 6735165.0,
      "step": 3250
    },
    {
      "entropy": 5.812675523757934,
      "epoch": 0.29526487663280115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995273669151351,
      "loss": 5.6685,
      "mean_token_accuracy": 0.16017006635665892,
      "num_tokens": 6744245.0,
      "step": 3255
    },
    {
      "entropy": 5.793766069412231,
      "epoch": 0.29571843251088537,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499525268464188,
      "loss": 5.6445,
      "mean_token_accuracy": 0.1621537297964096,
      "num_tokens": 6754894.0,
      "step": 3260
    },
    {
      "entropy": 5.82327790260315,
      "epoch": 0.2961719883889695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995231653700009,
      "loss": 5.7043,
      "mean_token_accuracy": 0.15255022346973418,
      "num_tokens": 6764442.0,
      "step": 3265
    },
    {
      "entropy": 5.78665943145752,
      "epoch": 0.2966255442670537,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995210576326176,
      "loss": 5.6535,
      "mean_token_accuracy": 0.156018228828907,
      "num_tokens": 6775054.0,
      "step": 3270
    },
    {
      "entropy": 5.80814037322998,
      "epoch": 0.2970791001451379,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995189452520816,
      "loss": 5.6367,
      "mean_token_accuracy": 0.17229300737380981,
      "num_tokens": 6784825.0,
      "step": 3275
    },
    {
      "entropy": 5.80222601890564,
      "epoch": 0.29753265602322204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995168282284366,
      "loss": 5.6766,
      "mean_token_accuracy": 0.16131959110498428,
      "num_tokens": 6795928.0,
      "step": 3280
    },
    {
      "entropy": 5.726570558547974,
      "epoch": 0.29798621190130625,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995147065617263,
      "loss": 5.6071,
      "mean_token_accuracy": 0.1610921859741211,
      "num_tokens": 6805456.0,
      "step": 3285
    },
    {
      "entropy": 5.799875020980835,
      "epoch": 0.2984397677793904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995125802519946,
      "loss": 5.6826,
      "mean_token_accuracy": 0.15558315366506575,
      "num_tokens": 6816215.0,
      "step": 3290
    },
    {
      "entropy": 5.817404937744141,
      "epoch": 0.2988933236574746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995104492992857,
      "loss": 5.7207,
      "mean_token_accuracy": 0.15770665407180787,
      "num_tokens": 6826545.0,
      "step": 3295
    },
    {
      "entropy": 5.704713821411133,
      "epoch": 0.2993468795355588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995083137036434,
      "loss": 5.6256,
      "mean_token_accuracy": 0.16132950484752656,
      "num_tokens": 6836184.0,
      "step": 3300
    },
    {
      "entropy": 5.845241355895996,
      "epoch": 0.299800435413643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995061734651121,
      "loss": 5.7128,
      "mean_token_accuracy": 0.15842559039592743,
      "num_tokens": 6846646.0,
      "step": 3305
    },
    {
      "entropy": 5.760690546035766,
      "epoch": 0.30025399129172714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995040285837358,
      "loss": 5.6189,
      "mean_token_accuracy": 0.16771125495433808,
      "num_tokens": 6857597.0,
      "step": 3310
    },
    {
      "entropy": 5.727470827102661,
      "epoch": 0.3007075471698113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995018790595589,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1587997391819954,
      "num_tokens": 6867523.0,
      "step": 3315
    },
    {
      "entropy": 5.822723150253296,
      "epoch": 0.3011611030478955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994997248926261,
      "loss": 5.7959,
      "mean_token_accuracy": 0.15733444988727568,
      "num_tokens": 6878190.0,
      "step": 3320
    },
    {
      "entropy": 5.769148635864258,
      "epoch": 0.30161465892597966,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994975660829817,
      "loss": 5.5958,
      "mean_token_accuracy": 0.17123553454875945,
      "num_tokens": 6888483.0,
      "step": 3325
    },
    {
      "entropy": 5.718564510345459,
      "epoch": 0.3020682148040639,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994954026306705,
      "loss": 5.5724,
      "mean_token_accuracy": 0.16064103171229363,
      "num_tokens": 6898337.0,
      "step": 3330
    },
    {
      "entropy": 5.827276992797851,
      "epoch": 0.30252177068214803,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994932345357371,
      "loss": 5.6561,
      "mean_token_accuracy": 0.15858807861804963,
      "num_tokens": 6909131.0,
      "step": 3335
    },
    {
      "entropy": 5.740424537658692,
      "epoch": 0.30297532656023224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994910617982264,
      "loss": 5.6382,
      "mean_token_accuracy": 0.15910502076148986,
      "num_tokens": 6919790.0,
      "step": 3340
    },
    {
      "entropy": 5.753969144821167,
      "epoch": 0.3034288824383164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994888844181833,
      "loss": 5.6574,
      "mean_token_accuracy": 0.1657829388976097,
      "num_tokens": 6930072.0,
      "step": 3345
    },
    {
      "entropy": 5.6898661136627195,
      "epoch": 0.3038824383164006,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994867023956529,
      "loss": 5.5388,
      "mean_token_accuracy": 0.16034141033887864,
      "num_tokens": 6940646.0,
      "step": 3350
    },
    {
      "entropy": 5.677322959899902,
      "epoch": 0.30433599419448476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994845157306803,
      "loss": 5.5421,
      "mean_token_accuracy": 0.16668861508369445,
      "num_tokens": 6951645.0,
      "step": 3355
    },
    {
      "entropy": 5.837560987472534,
      "epoch": 0.3047895500725689,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994823244233107,
      "loss": 5.703,
      "mean_token_accuracy": 0.15850953310728072,
      "num_tokens": 6963020.0,
      "step": 3360
    },
    {
      "entropy": 5.787491798400879,
      "epoch": 0.3052431059506531,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994801284735895,
      "loss": 5.6444,
      "mean_token_accuracy": 0.1636459544301033,
      "num_tokens": 6972895.0,
      "step": 3365
    },
    {
      "entropy": 5.849544525146484,
      "epoch": 0.3056966618287373,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994779278815619,
      "loss": 5.7795,
      "mean_token_accuracy": 0.14996607154607772,
      "num_tokens": 6985363.0,
      "step": 3370
    },
    {
      "entropy": 5.838080978393554,
      "epoch": 0.3061502177068215,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994757226472736,
      "loss": 5.7216,
      "mean_token_accuracy": 0.16052751541137694,
      "num_tokens": 6997040.0,
      "step": 3375
    },
    {
      "entropy": 5.793620634078979,
      "epoch": 0.30660377358490565,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994735127707702,
      "loss": 5.6421,
      "mean_token_accuracy": 0.16104835271835327,
      "num_tokens": 7007591.0,
      "step": 3380
    },
    {
      "entropy": 5.819001531600952,
      "epoch": 0.30705732946298986,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994712982520972,
      "loss": 5.6422,
      "mean_token_accuracy": 0.16898961067199708,
      "num_tokens": 7017858.0,
      "step": 3385
    },
    {
      "entropy": 5.713206720352173,
      "epoch": 0.307510885341074,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994690790913008,
      "loss": 5.6155,
      "mean_token_accuracy": 0.16000766903162003,
      "num_tokens": 7027365.0,
      "step": 3390
    },
    {
      "entropy": 5.717849445343018,
      "epoch": 0.3079644412191582,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994668552884265,
      "loss": 5.6084,
      "mean_token_accuracy": 0.16479909121990205,
      "num_tokens": 7038728.0,
      "step": 3395
    },
    {
      "entropy": 5.835485458374023,
      "epoch": 0.3084179970972424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004994646268435204,
      "loss": 5.5429,
      "mean_token_accuracy": 0.16616460382938386,
      "num_tokens": 7048736.0,
      "step": 3400
    },
    {
      "entropy": 5.722419404983521,
      "epoch": 0.30887155297532654,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994623937566286,
      "loss": 5.6685,
      "mean_token_accuracy": 0.16291871815919876,
      "num_tokens": 7059205.0,
      "step": 3405
    },
    {
      "entropy": 5.760222864151001,
      "epoch": 0.30932510885341075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994601560277974,
      "loss": 5.667,
      "mean_token_accuracy": 0.15780892074108124,
      "num_tokens": 7069513.0,
      "step": 3410
    },
    {
      "entropy": 5.734495067596436,
      "epoch": 0.3097786647314949,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499457913657073,
      "loss": 5.5936,
      "mean_token_accuracy": 0.1593753680586815,
      "num_tokens": 7079940.0,
      "step": 3415
    },
    {
      "entropy": 5.781790494918823,
      "epoch": 0.3102322206095791,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994556666445016,
      "loss": 5.5958,
      "mean_token_accuracy": 0.16559863835573196,
      "num_tokens": 7090792.0,
      "step": 3420
    },
    {
      "entropy": 5.704123544692993,
      "epoch": 0.31068577648766327,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994534149901299,
      "loss": 5.5471,
      "mean_token_accuracy": 0.16237700134515762,
      "num_tokens": 7101964.0,
      "step": 3425
    },
    {
      "entropy": 5.715693664550781,
      "epoch": 0.3111393323657475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994511586940044,
      "loss": 5.6225,
      "mean_token_accuracy": 0.15924402177333832,
      "num_tokens": 7113332.0,
      "step": 3430
    },
    {
      "entropy": 5.885712957382202,
      "epoch": 0.31159288824383163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994488977561717,
      "loss": 5.764,
      "mean_token_accuracy": 0.15799518972635268,
      "num_tokens": 7123481.0,
      "step": 3435
    },
    {
      "entropy": 5.681813049316406,
      "epoch": 0.31204644412191584,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994466321766786,
      "loss": 5.5308,
      "mean_token_accuracy": 0.16969044506549835,
      "num_tokens": 7133694.0,
      "step": 3440
    },
    {
      "entropy": 5.724149560928344,
      "epoch": 0.3125,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499444361955572,
      "loss": 5.6527,
      "mean_token_accuracy": 0.16183393001556395,
      "num_tokens": 7144491.0,
      "step": 3445
    },
    {
      "entropy": 5.882103443145752,
      "epoch": 0.31295355587808416,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994420870928988,
      "loss": 5.7266,
      "mean_token_accuracy": 0.15855416059494018,
      "num_tokens": 7154716.0,
      "step": 3450
    },
    {
      "entropy": 5.830023527145386,
      "epoch": 0.31340711175616837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994398075887059,
      "loss": 5.7259,
      "mean_token_accuracy": 0.15234277099370958,
      "num_tokens": 7164258.0,
      "step": 3455
    },
    {
      "entropy": 5.782030057907105,
      "epoch": 0.3138606676342525,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994375234430407,
      "loss": 5.589,
      "mean_token_accuracy": 0.1607864186167717,
      "num_tokens": 7175098.0,
      "step": 3460
    },
    {
      "entropy": 5.742347669601441,
      "epoch": 0.31431422351233673,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994352346559504,
      "loss": 5.6563,
      "mean_token_accuracy": 0.15974079370498656,
      "num_tokens": 7186480.0,
      "step": 3465
    },
    {
      "entropy": 5.664342737197876,
      "epoch": 0.3147677793904209,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994329412274821,
      "loss": 5.6363,
      "mean_token_accuracy": 0.1637357532978058,
      "num_tokens": 7197056.0,
      "step": 3470
    },
    {
      "entropy": 5.80676236152649,
      "epoch": 0.3152213352685051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994306431576835,
      "loss": 5.7374,
      "mean_token_accuracy": 0.15914391726255417,
      "num_tokens": 7206040.0,
      "step": 3475
    },
    {
      "entropy": 5.898715686798096,
      "epoch": 0.31567489114658925,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499428340446602,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1631159707903862,
      "num_tokens": 7216344.0,
      "step": 3480
    },
    {
      "entropy": 5.750396013259888,
      "epoch": 0.31612844702467346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994260330942851,
      "loss": 5.6801,
      "mean_token_accuracy": 0.15558458715677262,
      "num_tokens": 7226729.0,
      "step": 3485
    },
    {
      "entropy": 5.676659345626831,
      "epoch": 0.3165820029027576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994237211007808,
      "loss": 5.6213,
      "mean_token_accuracy": 0.16082685589790344,
      "num_tokens": 7237616.0,
      "step": 3490
    },
    {
      "entropy": 5.75124158859253,
      "epoch": 0.3170355587808418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994214044661367,
      "loss": 5.5436,
      "mean_token_accuracy": 0.16586074680089952,
      "num_tokens": 7247857.0,
      "step": 3495
    },
    {
      "entropy": 5.63331036567688,
      "epoch": 0.317489114658926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994190831904007,
      "loss": 5.5215,
      "mean_token_accuracy": 0.16701960563659668,
      "num_tokens": 7257858.0,
      "step": 3500
    },
    {
      "entropy": 5.611927175521851,
      "epoch": 0.31794267053701014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499416757273621,
      "loss": 5.539,
      "mean_token_accuracy": 0.16479599773883818,
      "num_tokens": 7268440.0,
      "step": 3505
    },
    {
      "entropy": 5.79567551612854,
      "epoch": 0.31839622641509435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994144267158454,
      "loss": 5.6125,
      "mean_token_accuracy": 0.16928928196430207,
      "num_tokens": 7278560.0,
      "step": 3510
    },
    {
      "entropy": 5.7546285629272464,
      "epoch": 0.3188497822931785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994120915171224,
      "loss": 5.6208,
      "mean_token_accuracy": 0.16289258599281312,
      "num_tokens": 7288422.0,
      "step": 3515
    },
    {
      "entropy": 5.8093061447143555,
      "epoch": 0.3193033381712627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994097516775002,
      "loss": 5.6798,
      "mean_token_accuracy": 0.16264722645282745,
      "num_tokens": 7298613.0,
      "step": 3520
    },
    {
      "entropy": 5.806758069992066,
      "epoch": 0.3197568940493469,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994074071970271,
      "loss": 5.7208,
      "mean_token_accuracy": 0.15486309304833412,
      "num_tokens": 7309509.0,
      "step": 3525
    },
    {
      "entropy": 5.782179450988769,
      "epoch": 0.3202104499274311,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994050580757515,
      "loss": 5.6159,
      "mean_token_accuracy": 0.16160835027694703,
      "num_tokens": 7319015.0,
      "step": 3530
    },
    {
      "entropy": 5.694505119323731,
      "epoch": 0.32066400580551524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994027043137223,
      "loss": 5.5828,
      "mean_token_accuracy": 0.16993721425533295,
      "num_tokens": 7329026.0,
      "step": 3535
    },
    {
      "entropy": 5.8108502388000485,
      "epoch": 0.3211175616835994,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994003459109879,
      "loss": 5.6928,
      "mean_token_accuracy": 0.1579549044370651,
      "num_tokens": 7339638.0,
      "step": 3540
    },
    {
      "entropy": 5.797452449798584,
      "epoch": 0.3215711175616836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993979828675972,
      "loss": 5.5935,
      "mean_token_accuracy": 0.16423679292201995,
      "num_tokens": 7349979.0,
      "step": 3545
    },
    {
      "entropy": 5.655317640304565,
      "epoch": 0.32202467343976776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499395615183599,
      "loss": 5.5311,
      "mean_token_accuracy": 0.1711718648672104,
      "num_tokens": 7359398.0,
      "step": 3550
    },
    {
      "entropy": 5.670080709457397,
      "epoch": 0.32247822931785197,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993932428590422,
      "loss": 5.5775,
      "mean_token_accuracy": 0.1649343878030777,
      "num_tokens": 7370418.0,
      "step": 3555
    },
    {
      "entropy": 5.7741282939910885,
      "epoch": 0.3229317851959361,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993908658939762,
      "loss": 5.6027,
      "mean_token_accuracy": 0.15476736575365066,
      "num_tokens": 7380461.0,
      "step": 3560
    },
    {
      "entropy": 5.758694982528686,
      "epoch": 0.32338534107402034,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993884842884496,
      "loss": 5.6153,
      "mean_token_accuracy": 0.1681668609380722,
      "num_tokens": 7390646.0,
      "step": 3565
    },
    {
      "entropy": 5.743930101394653,
      "epoch": 0.3238388969521045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993860980425121,
      "loss": 5.7011,
      "mean_token_accuracy": 0.16390874236822128,
      "num_tokens": 7402106.0,
      "step": 3570
    },
    {
      "entropy": 5.784041881561279,
      "epoch": 0.3242924528301887,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993837071562131,
      "loss": 5.6979,
      "mean_token_accuracy": 0.16273627281188965,
      "num_tokens": 7412936.0,
      "step": 3575
    },
    {
      "entropy": 5.752962303161621,
      "epoch": 0.32474600870827286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993813116296017,
      "loss": 5.5556,
      "mean_token_accuracy": 0.16882994323968886,
      "num_tokens": 7422303.0,
      "step": 3580
    },
    {
      "entropy": 5.709475040435791,
      "epoch": 0.325199564586357,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993789114627275,
      "loss": 5.6289,
      "mean_token_accuracy": 0.1605585351586342,
      "num_tokens": 7432829.0,
      "step": 3585
    },
    {
      "entropy": 5.85622272491455,
      "epoch": 0.3256531204644412,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993765066556403,
      "loss": 5.6422,
      "mean_token_accuracy": 0.16322125345468522,
      "num_tokens": 7443218.0,
      "step": 3590
    },
    {
      "entropy": 5.7688047885894775,
      "epoch": 0.3261066763425254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993740972083899,
      "loss": 5.6019,
      "mean_token_accuracy": 0.16230728030204772,
      "num_tokens": 7454531.0,
      "step": 3595
    },
    {
      "entropy": 5.730414962768554,
      "epoch": 0.3265602322206096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993716831210259,
      "loss": 5.6645,
      "mean_token_accuracy": 0.16234444826841354,
      "num_tokens": 7464993.0,
      "step": 3600
    },
    {
      "entropy": 5.69191312789917,
      "epoch": 0.32701378809869375,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993692643935983,
      "loss": 5.6913,
      "mean_token_accuracy": 0.1614596240222454,
      "num_tokens": 7475755.0,
      "step": 3605
    },
    {
      "entropy": 5.772009658813476,
      "epoch": 0.32746734397677796,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993668410261571,
      "loss": 5.6709,
      "mean_token_accuracy": 0.1652410715818405,
      "num_tokens": 7486778.0,
      "step": 3610
    },
    {
      "entropy": 5.873863935470581,
      "epoch": 0.3279208998548621,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993644130187525,
      "loss": 5.6487,
      "mean_token_accuracy": 0.16203359365463257,
      "num_tokens": 7497439.0,
      "step": 3615
    },
    {
      "entropy": 5.71992974281311,
      "epoch": 0.3283744557329463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993619803714347,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1641794115304947,
      "num_tokens": 7508274.0,
      "step": 3620
    },
    {
      "entropy": 5.724187231063842,
      "epoch": 0.3288280116110305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993595430842539,
      "loss": 5.6626,
      "mean_token_accuracy": 0.15977796465158461,
      "num_tokens": 7519315.0,
      "step": 3625
    },
    {
      "entropy": 5.81933822631836,
      "epoch": 0.32928156748911463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993571011572606,
      "loss": 5.6515,
      "mean_token_accuracy": 0.16541533768177033,
      "num_tokens": 7529910.0,
      "step": 3630
    },
    {
      "entropy": 5.755266094207764,
      "epoch": 0.32973512336719885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993546545905053,
      "loss": 5.5745,
      "mean_token_accuracy": 0.16873921304941178,
      "num_tokens": 7538966.0,
      "step": 3635
    },
    {
      "entropy": 5.755073690414429,
      "epoch": 0.330188679245283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993522033840386,
      "loss": 5.6146,
      "mean_token_accuracy": 0.158041512966156,
      "num_tokens": 7549683.0,
      "step": 3640
    },
    {
      "entropy": 5.700992918014526,
      "epoch": 0.3306422351233672,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499349747537911,
      "loss": 5.5953,
      "mean_token_accuracy": 0.1624542772769928,
      "num_tokens": 7560629.0,
      "step": 3645
    },
    {
      "entropy": 5.730870103836059,
      "epoch": 0.33109579100145137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993472870521736,
      "loss": 5.5385,
      "mean_token_accuracy": 0.17402430772781372,
      "num_tokens": 7570613.0,
      "step": 3650
    },
    {
      "entropy": 5.715100336074829,
      "epoch": 0.3315493468795356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993448219268771,
      "loss": 5.5211,
      "mean_token_accuracy": 0.16446274071931838,
      "num_tokens": 7580640.0,
      "step": 3655
    },
    {
      "entropy": 5.743779134750366,
      "epoch": 0.33200290275761973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993423521620726,
      "loss": 5.6224,
      "mean_token_accuracy": 0.16669037342071533,
      "num_tokens": 7590164.0,
      "step": 3660
    },
    {
      "entropy": 5.77204122543335,
      "epoch": 0.33245645863570394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993398777578109,
      "loss": 5.6652,
      "mean_token_accuracy": 0.16001658663153648,
      "num_tokens": 7600173.0,
      "step": 3665
    },
    {
      "entropy": 5.65158109664917,
      "epoch": 0.3329100145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993373987141435,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17389463186264037,
      "num_tokens": 7609958.0,
      "step": 3670
    },
    {
      "entropy": 5.784420680999756,
      "epoch": 0.33336357039187225,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993349150311214,
      "loss": 5.6059,
      "mean_token_accuracy": 0.16194465905427932,
      "num_tokens": 7620237.0,
      "step": 3675
    },
    {
      "entropy": 5.6676253318786625,
      "epoch": 0.33381712626995647,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993324267087962,
      "loss": 5.5129,
      "mean_token_accuracy": 0.1702829569578171,
      "num_tokens": 7629886.0,
      "step": 3680
    },
    {
      "entropy": 5.706976079940796,
      "epoch": 0.3342706821480406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993299337472192,
      "loss": 5.5823,
      "mean_token_accuracy": 0.16765174865722657,
      "num_tokens": 7640057.0,
      "step": 3685
    },
    {
      "entropy": 5.674705648422242,
      "epoch": 0.33472423802612483,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499327436146442,
      "loss": 5.5229,
      "mean_token_accuracy": 0.16337296962738038,
      "num_tokens": 7649766.0,
      "step": 3690
    },
    {
      "entropy": 5.715948963165284,
      "epoch": 0.335177793904209,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993249339065162,
      "loss": 5.607,
      "mean_token_accuracy": 0.16630489975214005,
      "num_tokens": 7660890.0,
      "step": 3695
    },
    {
      "entropy": 5.735551595687866,
      "epoch": 0.3356313497822932,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993224270274937,
      "loss": 5.63,
      "mean_token_accuracy": 0.16649477332830429,
      "num_tokens": 7670737.0,
      "step": 3700
    },
    {
      "entropy": 5.690478563308716,
      "epoch": 0.33608490566037735,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993199155094263,
      "loss": 5.5932,
      "mean_token_accuracy": 0.16307836025953293,
      "num_tokens": 7681407.0,
      "step": 3705
    },
    {
      "entropy": 5.706917667388916,
      "epoch": 0.33653846153846156,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993173993523658,
      "loss": 5.5156,
      "mean_token_accuracy": 0.17325208634138106,
      "num_tokens": 7691758.0,
      "step": 3710
    },
    {
      "entropy": 5.671240568161011,
      "epoch": 0.3369920174165457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993148785563643,
      "loss": 5.5515,
      "mean_token_accuracy": 0.16601790338754654,
      "num_tokens": 7702934.0,
      "step": 3715
    },
    {
      "entropy": 5.69557638168335,
      "epoch": 0.3374455732946299,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499312353121474,
      "loss": 5.5392,
      "mean_token_accuracy": 0.1636522337794304,
      "num_tokens": 7713001.0,
      "step": 3720
    },
    {
      "entropy": 5.736032581329345,
      "epoch": 0.3378991291727141,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993098230477471,
      "loss": 5.5958,
      "mean_token_accuracy": 0.1599409118294716,
      "num_tokens": 7724005.0,
      "step": 3725
    },
    {
      "entropy": 5.748565006256103,
      "epoch": 0.33835268505079824,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993072883352358,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1650003969669342,
      "num_tokens": 7733743.0,
      "step": 3730
    },
    {
      "entropy": 5.813916301727295,
      "epoch": 0.33880624092888245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993047489839927,
      "loss": 5.7438,
      "mean_token_accuracy": 0.1575533241033554,
      "num_tokens": 7743762.0,
      "step": 3735
    },
    {
      "entropy": 5.814930248260498,
      "epoch": 0.3392597968069666,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993022049940702,
      "loss": 5.6361,
      "mean_token_accuracy": 0.16483197808265687,
      "num_tokens": 7753758.0,
      "step": 3740
    },
    {
      "entropy": 5.688127851486206,
      "epoch": 0.3397133526850508,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992996563655209,
      "loss": 5.5699,
      "mean_token_accuracy": 0.165471251308918,
      "num_tokens": 7764692.0,
      "step": 3745
    },
    {
      "entropy": 5.683252620697021,
      "epoch": 0.340166908563135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992971030983975,
      "loss": 5.6076,
      "mean_token_accuracy": 0.15619706213474274,
      "num_tokens": 7776892.0,
      "step": 3750
    },
    {
      "entropy": 5.818699216842651,
      "epoch": 0.3406204644412192,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992945451927531,
      "loss": 5.6558,
      "mean_token_accuracy": 0.16157331615686416,
      "num_tokens": 7787143.0,
      "step": 3755
    },
    {
      "entropy": 5.6922383308410645,
      "epoch": 0.34107402031930334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992919826486402,
      "loss": 5.5713,
      "mean_token_accuracy": 0.16458574086427688,
      "num_tokens": 7798537.0,
      "step": 3760
    },
    {
      "entropy": 5.70722393989563,
      "epoch": 0.3415275761973875,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992894154661118,
      "loss": 5.5466,
      "mean_token_accuracy": 0.16931652128696442,
      "num_tokens": 7808434.0,
      "step": 3765
    },
    {
      "entropy": 5.727777719497681,
      "epoch": 0.3419811320754717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992868436452213,
      "loss": 5.6646,
      "mean_token_accuracy": 0.16223400980234146,
      "num_tokens": 7819592.0,
      "step": 3770
    },
    {
      "entropy": 5.751184320449829,
      "epoch": 0.34243468795355586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992842671860217,
      "loss": 5.5501,
      "mean_token_accuracy": 0.16743091493844986,
      "num_tokens": 7830602.0,
      "step": 3775
    },
    {
      "entropy": 5.776014566421509,
      "epoch": 0.34288824383164007,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992816860885662,
      "loss": 5.6971,
      "mean_token_accuracy": 0.16009064316749572,
      "num_tokens": 7840850.0,
      "step": 3780
    },
    {
      "entropy": 5.83824052810669,
      "epoch": 0.3433417997097242,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992791003529084,
      "loss": 5.6309,
      "mean_token_accuracy": 0.1558419495820999,
      "num_tokens": 7852371.0,
      "step": 3785
    },
    {
      "entropy": 5.6923058986663815,
      "epoch": 0.34379535558780844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992765099791017,
      "loss": 5.6581,
      "mean_token_accuracy": 0.15808210968971254,
      "num_tokens": 7863290.0,
      "step": 3790
    },
    {
      "entropy": 5.6762631893157955,
      "epoch": 0.3442489114658926,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992739149671995,
      "loss": 5.4947,
      "mean_token_accuracy": 0.168349090218544,
      "num_tokens": 7873927.0,
      "step": 3795
    },
    {
      "entropy": 5.698384523391724,
      "epoch": 0.3447024673439768,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992713153172556,
      "loss": 5.5329,
      "mean_token_accuracy": 0.16461328268051148,
      "num_tokens": 7884977.0,
      "step": 3800
    },
    {
      "entropy": 5.671432590484619,
      "epoch": 0.34515602322206096,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992687110293238,
      "loss": 5.6109,
      "mean_token_accuracy": 0.16780710071325303,
      "num_tokens": 7894620.0,
      "step": 3805
    },
    {
      "entropy": 5.681638097763061,
      "epoch": 0.3456095791001451,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992661021034579,
      "loss": 5.5377,
      "mean_token_accuracy": 0.17044852972030639,
      "num_tokens": 7904541.0,
      "step": 3810
    },
    {
      "entropy": 5.677857828140259,
      "epoch": 0.3460631349782293,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992634885397118,
      "loss": 5.5551,
      "mean_token_accuracy": 0.16660392731428147,
      "num_tokens": 7915435.0,
      "step": 3815
    },
    {
      "entropy": 5.675379896163941,
      "epoch": 0.3465166908563135,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992608703381396,
      "loss": 5.54,
      "mean_token_accuracy": 0.1684734731912613,
      "num_tokens": 7925870.0,
      "step": 3820
    },
    {
      "entropy": 5.698918628692627,
      "epoch": 0.3469702467343977,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992582474987955,
      "loss": 5.541,
      "mean_token_accuracy": 0.1659692794084549,
      "num_tokens": 7937046.0,
      "step": 3825
    },
    {
      "entropy": 5.677004718780518,
      "epoch": 0.34742380261248185,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992556200217337,
      "loss": 5.4992,
      "mean_token_accuracy": 0.1643098548054695,
      "num_tokens": 7947150.0,
      "step": 3830
    },
    {
      "entropy": 5.739584589004517,
      "epoch": 0.34787735849056606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992529879070085,
      "loss": 5.6016,
      "mean_token_accuracy": 0.1661156862974167,
      "num_tokens": 7956815.0,
      "step": 3835
    },
    {
      "entropy": 5.73980484008789,
      "epoch": 0.3483309143686502,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992503511546745,
      "loss": 5.6341,
      "mean_token_accuracy": 0.16282225549221038,
      "num_tokens": 7968114.0,
      "step": 3840
    },
    {
      "entropy": 5.691490125656128,
      "epoch": 0.3487844702467344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992477097647859,
      "loss": 5.5408,
      "mean_token_accuracy": 0.16325437575578688,
      "num_tokens": 7977840.0,
      "step": 3845
    },
    {
      "entropy": 5.76722149848938,
      "epoch": 0.3492380261248186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992450637373976,
      "loss": 5.6678,
      "mean_token_accuracy": 0.1663844257593155,
      "num_tokens": 7988663.0,
      "step": 3850
    },
    {
      "entropy": 5.655972480773926,
      "epoch": 0.34969158200290273,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992424130725642,
      "loss": 5.48,
      "mean_token_accuracy": 0.1633950486779213,
      "num_tokens": 7999693.0,
      "step": 3855
    },
    {
      "entropy": 5.750917959213257,
      "epoch": 0.35014513788098695,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992397577703406,
      "loss": 5.6568,
      "mean_token_accuracy": 0.15703431218862535,
      "num_tokens": 8010154.0,
      "step": 3860
    },
    {
      "entropy": 5.7925513744354244,
      "epoch": 0.3505986937590711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992370978307816,
      "loss": 5.6324,
      "mean_token_accuracy": 0.16514518857002258,
      "num_tokens": 8020921.0,
      "step": 3865
    },
    {
      "entropy": 5.6815894603729244,
      "epoch": 0.3510522496371553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992344332539424,
      "loss": 5.5597,
      "mean_token_accuracy": 0.1687205910682678,
      "num_tokens": 8031404.0,
      "step": 3870
    },
    {
      "entropy": 5.654696989059448,
      "epoch": 0.35150580551523947,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992317640398779,
      "loss": 5.57,
      "mean_token_accuracy": 0.16497284919023514,
      "num_tokens": 8041310.0,
      "step": 3875
    },
    {
      "entropy": 5.681327152252197,
      "epoch": 0.3519593613933237,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992290901886434,
      "loss": 5.559,
      "mean_token_accuracy": 0.16811044812202453,
      "num_tokens": 8051054.0,
      "step": 3880
    },
    {
      "entropy": 5.717649364471436,
      "epoch": 0.35241291727140783,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499226411700294,
      "loss": 5.5503,
      "mean_token_accuracy": 0.16839763820171355,
      "num_tokens": 8061063.0,
      "step": 3885
    },
    {
      "entropy": 5.685747528076172,
      "epoch": 0.35286647314949204,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992237285748855,
      "loss": 5.5605,
      "mean_token_accuracy": 0.1663712427020073,
      "num_tokens": 8071715.0,
      "step": 3890
    },
    {
      "entropy": 5.768758487701416,
      "epoch": 0.3533200290275762,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992210408124731,
      "loss": 5.6295,
      "mean_token_accuracy": 0.156692710518837,
      "num_tokens": 8082679.0,
      "step": 3895
    },
    {
      "entropy": 5.718517923355103,
      "epoch": 0.35377358490566035,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992183484131123,
      "loss": 5.6508,
      "mean_token_accuracy": 0.1636080786585808,
      "num_tokens": 8093785.0,
      "step": 3900
    },
    {
      "entropy": 5.654246377944946,
      "epoch": 0.35422714078374457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992156513768591,
      "loss": 5.5534,
      "mean_token_accuracy": 0.16666727364063263,
      "num_tokens": 8105822.0,
      "step": 3905
    },
    {
      "entropy": 5.649802350997925,
      "epoch": 0.3546806966618287,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499212949703769,
      "loss": 5.4627,
      "mean_token_accuracy": 0.17031579464673996,
      "num_tokens": 8115521.0,
      "step": 3910
    },
    {
      "entropy": 5.711194086074829,
      "epoch": 0.35513425253991293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992102433938981,
      "loss": 5.5528,
      "mean_token_accuracy": 0.1639755219221115,
      "num_tokens": 8125524.0,
      "step": 3915
    },
    {
      "entropy": 5.670215368270874,
      "epoch": 0.3555878084179971,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992075324473021,
      "loss": 5.4705,
      "mean_token_accuracy": 0.17687036842107773,
      "num_tokens": 8135552.0,
      "step": 3920
    },
    {
      "entropy": 5.7498579025268555,
      "epoch": 0.3560413642960813,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992048168640371,
      "loss": 5.5912,
      "mean_token_accuracy": 0.16968106478452682,
      "num_tokens": 8145617.0,
      "step": 3925
    },
    {
      "entropy": 5.6834397315979,
      "epoch": 0.35649492017416545,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992020966441595,
      "loss": 5.6008,
      "mean_token_accuracy": 0.1607098937034607,
      "num_tokens": 8156729.0,
      "step": 3930
    },
    {
      "entropy": 5.697573709487915,
      "epoch": 0.35694847605224966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991993717877254,
      "loss": 5.5865,
      "mean_token_accuracy": 0.16268351525068284,
      "num_tokens": 8167637.0,
      "step": 3935
    },
    {
      "entropy": 5.648408317565918,
      "epoch": 0.3574020319303338,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991966422947911,
      "loss": 5.5628,
      "mean_token_accuracy": 0.170132377743721,
      "num_tokens": 8177354.0,
      "step": 3940
    },
    {
      "entropy": 5.63679690361023,
      "epoch": 0.357855587808418,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991939081654132,
      "loss": 5.5381,
      "mean_token_accuracy": 0.1754870280623436,
      "num_tokens": 8187580.0,
      "step": 3945
    },
    {
      "entropy": 5.660647439956665,
      "epoch": 0.3583091436865022,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499191169399648,
      "loss": 5.6353,
      "mean_token_accuracy": 0.15820996463298798,
      "num_tokens": 8199382.0,
      "step": 3950
    },
    {
      "entropy": 5.687941408157348,
      "epoch": 0.35876269956458634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991884259975525,
      "loss": 5.5178,
      "mean_token_accuracy": 0.16903046071529387,
      "num_tokens": 8208484.0,
      "step": 3955
    },
    {
      "entropy": 5.708724451065064,
      "epoch": 0.35921625544267055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499185677959183,
      "loss": 5.5645,
      "mean_token_accuracy": 0.16452753990888597,
      "num_tokens": 8219056.0,
      "step": 3960
    },
    {
      "entropy": 5.7164100170135494,
      "epoch": 0.3596698113207547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991829252845966,
      "loss": 5.5797,
      "mean_token_accuracy": 0.1641779735684395,
      "num_tokens": 8230149.0,
      "step": 3965
    },
    {
      "entropy": 5.706824064254761,
      "epoch": 0.3601233671988389,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991801679738502,
      "loss": 5.6529,
      "mean_token_accuracy": 0.15674541145563126,
      "num_tokens": 8240803.0,
      "step": 3970
    },
    {
      "entropy": 5.706471729278564,
      "epoch": 0.3605769230769231,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991774060270008,
      "loss": 5.5946,
      "mean_token_accuracy": 0.16189683228731155,
      "num_tokens": 8251563.0,
      "step": 3975
    },
    {
      "entropy": 5.671058511734008,
      "epoch": 0.3610304789550073,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991746394441055,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1629596546292305,
      "num_tokens": 8261342.0,
      "step": 3980
    },
    {
      "entropy": 5.712836742401123,
      "epoch": 0.36148403483309144,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991718682252216,
      "loss": 5.6412,
      "mean_token_accuracy": 0.1584403395652771,
      "num_tokens": 8271540.0,
      "step": 3985
    },
    {
      "entropy": 5.757377099990845,
      "epoch": 0.3619375907111756,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991690923704063,
      "loss": 5.5587,
      "mean_token_accuracy": 0.16804730147123337,
      "num_tokens": 8280593.0,
      "step": 3990
    },
    {
      "entropy": 5.687913990020752,
      "epoch": 0.3623911465892598,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991663118797171,
      "loss": 5.5785,
      "mean_token_accuracy": 0.16617974787950515,
      "num_tokens": 8290126.0,
      "step": 3995
    },
    {
      "entropy": 5.699881505966187,
      "epoch": 0.36284470246734396,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991635267532113,
      "loss": 5.5149,
      "mean_token_accuracy": 0.1671944186091423,
      "num_tokens": 8301416.0,
      "step": 4000
    },
    {
      "entropy": 5.742449140548706,
      "epoch": 0.36329825834542817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991607369909468,
      "loss": 5.59,
      "mean_token_accuracy": 0.16540921181440355,
      "num_tokens": 8311034.0,
      "step": 4005
    },
    {
      "entropy": 5.688007831573486,
      "epoch": 0.3637518142235123,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991579425929811,
      "loss": 5.6005,
      "mean_token_accuracy": 0.16405065953731537,
      "num_tokens": 8322203.0,
      "step": 4010
    },
    {
      "entropy": 5.6758204936981205,
      "epoch": 0.36420537010159654,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499155143559372,
      "loss": 5.4789,
      "mean_token_accuracy": 0.17067255973815917,
      "num_tokens": 8330981.0,
      "step": 4015
    },
    {
      "entropy": 5.636501121520996,
      "epoch": 0.3646589259796807,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991523398901773,
      "loss": 5.5188,
      "mean_token_accuracy": 0.16740261018276215,
      "num_tokens": 8341294.0,
      "step": 4020
    },
    {
      "entropy": 5.790857219696045,
      "epoch": 0.3651124818577649,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991495315854553,
      "loss": 5.6633,
      "mean_token_accuracy": 0.1621541276574135,
      "num_tokens": 8351824.0,
      "step": 4025
    },
    {
      "entropy": 5.688350629806519,
      "epoch": 0.36556603773584906,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991467186452637,
      "loss": 5.5408,
      "mean_token_accuracy": 0.167211352288723,
      "num_tokens": 8361975.0,
      "step": 4030
    },
    {
      "entropy": 5.64109468460083,
      "epoch": 0.3660195936139332,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991439010696609,
      "loss": 5.482,
      "mean_token_accuracy": 0.16923303604125978,
      "num_tokens": 8372145.0,
      "step": 4035
    },
    {
      "entropy": 5.672000598907471,
      "epoch": 0.3664731494920174,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991410788587051,
      "loss": 5.5725,
      "mean_token_accuracy": 0.16295102685689927,
      "num_tokens": 8382542.0,
      "step": 4040
    },
    {
      "entropy": 5.721254110336304,
      "epoch": 0.3669267053701016,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991382520124547,
      "loss": 5.572,
      "mean_token_accuracy": 0.16192779392004014,
      "num_tokens": 8393450.0,
      "step": 4045
    },
    {
      "entropy": 5.723563814163208,
      "epoch": 0.3673802612481858,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991354205309681,
      "loss": 5.5075,
      "mean_token_accuracy": 0.16757548451423646,
      "num_tokens": 8404929.0,
      "step": 4050
    },
    {
      "entropy": 5.734829044342041,
      "epoch": 0.36783381712626995,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499132584414304,
      "loss": 5.5422,
      "mean_token_accuracy": 0.16685086190700532,
      "num_tokens": 8415703.0,
      "step": 4055
    },
    {
      "entropy": 5.728091859817505,
      "epoch": 0.36828737300435416,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991297436625207,
      "loss": 5.6072,
      "mean_token_accuracy": 0.16412289291620255,
      "num_tokens": 8426220.0,
      "step": 4060
    },
    {
      "entropy": 5.648223352432251,
      "epoch": 0.3687409288824383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991268982756774,
      "loss": 5.4911,
      "mean_token_accuracy": 0.16989713162183762,
      "num_tokens": 8436220.0,
      "step": 4065
    },
    {
      "entropy": 5.782150840759277,
      "epoch": 0.3691944847605225,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991240482538327,
      "loss": 5.6445,
      "mean_token_accuracy": 0.15544479340314865,
      "num_tokens": 8446724.0,
      "step": 4070
    },
    {
      "entropy": 5.689100742340088,
      "epoch": 0.3696480406386067,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991211935970455,
      "loss": 5.5028,
      "mean_token_accuracy": 0.16541081219911574,
      "num_tokens": 8457270.0,
      "step": 4075
    },
    {
      "entropy": 5.654791688919067,
      "epoch": 0.37010159651669083,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991183343053748,
      "loss": 5.5721,
      "mean_token_accuracy": 0.15914703905582428,
      "num_tokens": 8468137.0,
      "step": 4080
    },
    {
      "entropy": 5.730112028121948,
      "epoch": 0.37055515239477504,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049911547037888,
      "loss": 5.5037,
      "mean_token_accuracy": 0.17106323689222336,
      "num_tokens": 8477574.0,
      "step": 4085
    },
    {
      "entropy": 5.67306604385376,
      "epoch": 0.3710087082728592,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991126018176202,
      "loss": 5.5959,
      "mean_token_accuracy": 0.16015791594982148,
      "num_tokens": 8487268.0,
      "step": 4090
    },
    {
      "entropy": 5.792642593383789,
      "epoch": 0.3714622641509434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991097286216545,
      "loss": 5.6546,
      "mean_token_accuracy": 0.16796909868717194,
      "num_tokens": 8497785.0,
      "step": 4095
    },
    {
      "entropy": 5.751044225692749,
      "epoch": 0.37191582002902757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991068507910427,
      "loss": 5.5268,
      "mean_token_accuracy": 0.1661210760474205,
      "num_tokens": 8508169.0,
      "step": 4100
    },
    {
      "entropy": 5.6846167087554935,
      "epoch": 0.3723693759071118,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499103968325844,
      "loss": 5.4959,
      "mean_token_accuracy": 0.16699425280094146,
      "num_tokens": 8519201.0,
      "step": 4105
    },
    {
      "entropy": 5.7432060718536375,
      "epoch": 0.37282293178519593,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991010812261181,
      "loss": 5.6047,
      "mean_token_accuracy": 0.1590347096323967,
      "num_tokens": 8531093.0,
      "step": 4110
    },
    {
      "entropy": 5.601396274566651,
      "epoch": 0.37327648766328014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990981894919248,
      "loss": 5.474,
      "mean_token_accuracy": 0.17184109836816788,
      "num_tokens": 8541730.0,
      "step": 4115
    },
    {
      "entropy": 5.6851277351379395,
      "epoch": 0.3737300435413643,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990952931233238,
      "loss": 5.5322,
      "mean_token_accuracy": 0.16669057607650756,
      "num_tokens": 8552570.0,
      "step": 4120
    },
    {
      "entropy": 5.711943817138672,
      "epoch": 0.37418359941944845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499092392120375,
      "loss": 5.4496,
      "mean_token_accuracy": 0.16830459237098694,
      "num_tokens": 8562340.0,
      "step": 4125
    },
    {
      "entropy": 5.690545225143433,
      "epoch": 0.37463715529753266,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990894864831385,
      "loss": 5.6457,
      "mean_token_accuracy": 0.16141000092029573,
      "num_tokens": 8573240.0,
      "step": 4130
    },
    {
      "entropy": 5.74811692237854,
      "epoch": 0.3750907111756168,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990865762116743,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17035448253154756,
      "num_tokens": 8583194.0,
      "step": 4135
    },
    {
      "entropy": 5.673021078109741,
      "epoch": 0.37554426705370103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990836613060426,
      "loss": 5.5799,
      "mean_token_accuracy": 0.1630268007516861,
      "num_tokens": 8593603.0,
      "step": 4140
    },
    {
      "entropy": 5.630602216720581,
      "epoch": 0.3759978229317852,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990807417663036,
      "loss": 5.5842,
      "mean_token_accuracy": 0.1646374523639679,
      "num_tokens": 8603176.0,
      "step": 4145
    },
    {
      "entropy": 5.777580118179321,
      "epoch": 0.3764513788098694,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990778175925179,
      "loss": 5.596,
      "mean_token_accuracy": 0.1675265073776245,
      "num_tokens": 8614801.0,
      "step": 4150
    },
    {
      "entropy": 5.7292474746704105,
      "epoch": 0.37690493468795355,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990748887847457,
      "loss": 5.6252,
      "mean_token_accuracy": 0.16409169584512712,
      "num_tokens": 8624999.0,
      "step": 4155
    },
    {
      "entropy": 5.659530401229858,
      "epoch": 0.37735849056603776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990719553430477,
      "loss": 5.4784,
      "mean_token_accuracy": 0.17621796131134032,
      "num_tokens": 8635358.0,
      "step": 4160
    },
    {
      "entropy": 5.682961511611938,
      "epoch": 0.3778120464441219,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990690172674847,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1646416664123535,
      "num_tokens": 8645430.0,
      "step": 4165
    },
    {
      "entropy": 5.717040538787842,
      "epoch": 0.3782656023222061,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990660745581172,
      "loss": 5.4953,
      "mean_token_accuracy": 0.1789042666554451,
      "num_tokens": 8655115.0,
      "step": 4170
    },
    {
      "entropy": 5.636549854278565,
      "epoch": 0.3787191582002903,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990631272150062,
      "loss": 5.5041,
      "mean_token_accuracy": 0.16583775132894515,
      "num_tokens": 8665391.0,
      "step": 4175
    },
    {
      "entropy": 5.702080345153808,
      "epoch": 0.37917271407837444,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990601752382127,
      "loss": 5.5887,
      "mean_token_accuracy": 0.16532650887966155,
      "num_tokens": 8676819.0,
      "step": 4180
    },
    {
      "entropy": 5.778425836563111,
      "epoch": 0.37962626995645865,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990572186277975,
      "loss": 5.5227,
      "mean_token_accuracy": 0.171260467171669,
      "num_tokens": 8687052.0,
      "step": 4185
    },
    {
      "entropy": 5.689051246643066,
      "epoch": 0.3800798258345428,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990542573838221,
      "loss": 5.5942,
      "mean_token_accuracy": 0.1582357779145241,
      "num_tokens": 8697756.0,
      "step": 4190
    },
    {
      "entropy": 5.596316719055176,
      "epoch": 0.380533381712627,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990512915063475,
      "loss": 5.5045,
      "mean_token_accuracy": 0.1715959295630455,
      "num_tokens": 8708388.0,
      "step": 4195
    },
    {
      "entropy": 5.747104120254517,
      "epoch": 0.38098693759071117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990483209954352,
      "loss": 5.5915,
      "mean_token_accuracy": 0.1606937661767006,
      "num_tokens": 8719475.0,
      "step": 4200
    },
    {
      "entropy": 5.6669543266296385,
      "epoch": 0.3814404934687954,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990453458511464,
      "loss": 5.454,
      "mean_token_accuracy": 0.17628490179777145,
      "num_tokens": 8729395.0,
      "step": 4205
    },
    {
      "entropy": 5.690093278884888,
      "epoch": 0.38189404934687954,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990423660735428,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17369182407855988,
      "num_tokens": 8740351.0,
      "step": 4210
    },
    {
      "entropy": 5.721720170974732,
      "epoch": 0.3823476052249637,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499039381662686,
      "loss": 5.5275,
      "mean_token_accuracy": 0.1712048977613449,
      "num_tokens": 8749260.0,
      "step": 4215
    },
    {
      "entropy": 5.719943189620972,
      "epoch": 0.3828011611030479,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990363926186377,
      "loss": 5.6179,
      "mean_token_accuracy": 0.15975076854228973,
      "num_tokens": 8759604.0,
      "step": 4220
    },
    {
      "entropy": 5.750780868530273,
      "epoch": 0.38325471698113206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990333989414597,
      "loss": 5.6349,
      "mean_token_accuracy": 0.15848522633314133,
      "num_tokens": 8769815.0,
      "step": 4225
    },
    {
      "entropy": 5.646167278289795,
      "epoch": 0.38370827285921627,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990304006312138,
      "loss": 5.5614,
      "mean_token_accuracy": 0.16491910070180893,
      "num_tokens": 8780557.0,
      "step": 4230
    },
    {
      "entropy": 5.693453121185303,
      "epoch": 0.3841618287373004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990273976879622,
      "loss": 5.5145,
      "mean_token_accuracy": 0.17479519844055175,
      "num_tokens": 8790993.0,
      "step": 4235
    },
    {
      "entropy": 5.70437707901001,
      "epoch": 0.38461538461538464,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990243901117669,
      "loss": 5.5828,
      "mean_token_accuracy": 0.15901805460453033,
      "num_tokens": 8800657.0,
      "step": 4240
    },
    {
      "entropy": 5.6222734451293945,
      "epoch": 0.3850689404934688,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990213779026903,
      "loss": 5.5851,
      "mean_token_accuracy": 0.1631401777267456,
      "num_tokens": 8812983.0,
      "step": 4245
    },
    {
      "entropy": 5.734277868270874,
      "epoch": 0.385522496371553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990183610607944,
      "loss": 5.543,
      "mean_token_accuracy": 0.16615747958421706,
      "num_tokens": 8822980.0,
      "step": 4250
    },
    {
      "entropy": 5.662422704696655,
      "epoch": 0.38597605224963716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990153395861415,
      "loss": 5.5319,
      "mean_token_accuracy": 0.1691455915570259,
      "num_tokens": 8833821.0,
      "step": 4255
    },
    {
      "entropy": 5.62187876701355,
      "epoch": 0.3864296081277213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990123134787946,
      "loss": 5.4695,
      "mean_token_accuracy": 0.170258067548275,
      "num_tokens": 8843929.0,
      "step": 4260
    },
    {
      "entropy": 5.683696556091308,
      "epoch": 0.3868831640058055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990092827388159,
      "loss": 5.4591,
      "mean_token_accuracy": 0.17503607273101807,
      "num_tokens": 8854180.0,
      "step": 4265
    },
    {
      "entropy": 5.589134645462036,
      "epoch": 0.3873367198838897,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990062473662681,
      "loss": 5.463,
      "mean_token_accuracy": 0.1750032678246498,
      "num_tokens": 8865018.0,
      "step": 4270
    },
    {
      "entropy": 5.705612516403198,
      "epoch": 0.3877902757619739,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499003207361214,
      "loss": 5.6611,
      "mean_token_accuracy": 0.16093496978282928,
      "num_tokens": 8875423.0,
      "step": 4275
    },
    {
      "entropy": 5.781583690643311,
      "epoch": 0.38824383164005805,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990001627237165,
      "loss": 5.5854,
      "mean_token_accuracy": 0.1674569606781006,
      "num_tokens": 8887850.0,
      "step": 4280
    },
    {
      "entropy": 5.6495767593383786,
      "epoch": 0.38869738751814226,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989971134538386,
      "loss": 5.4845,
      "mean_token_accuracy": 0.17036610543727876,
      "num_tokens": 8897429.0,
      "step": 4285
    },
    {
      "entropy": 5.62213659286499,
      "epoch": 0.3891509433962264,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989940595516432,
      "loss": 5.5079,
      "mean_token_accuracy": 0.16780147105455398,
      "num_tokens": 8907048.0,
      "step": 4290
    },
    {
      "entropy": 5.6579169750213625,
      "epoch": 0.3896044992743106,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989910010171938,
      "loss": 5.5632,
      "mean_token_accuracy": 0.16277251839637757,
      "num_tokens": 8917987.0,
      "step": 4295
    },
    {
      "entropy": 5.714656162261963,
      "epoch": 0.3900580551523948,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989879378505532,
      "loss": 5.4548,
      "mean_token_accuracy": 0.16808317601680756,
      "num_tokens": 8929339.0,
      "step": 4300
    },
    {
      "entropy": 5.576473188400269,
      "epoch": 0.39051161103047893,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989848700517849,
      "loss": 5.4732,
      "mean_token_accuracy": 0.16945151686668397,
      "num_tokens": 8939473.0,
      "step": 4305
    },
    {
      "entropy": 5.632296085357666,
      "epoch": 0.39096516690856314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989817976209526,
      "loss": 5.5056,
      "mean_token_accuracy": 0.16843720525503159,
      "num_tokens": 8949173.0,
      "step": 4310
    },
    {
      "entropy": 5.724737358093262,
      "epoch": 0.3914187227866473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989787205581196,
      "loss": 5.4705,
      "mean_token_accuracy": 0.17448844164609909,
      "num_tokens": 8959879.0,
      "step": 4315
    },
    {
      "entropy": 5.661236667633057,
      "epoch": 0.3918722786647315,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989756388633496,
      "loss": 5.5626,
      "mean_token_accuracy": 0.1667757049202919,
      "num_tokens": 8970643.0,
      "step": 4320
    },
    {
      "entropy": 5.635674095153808,
      "epoch": 0.39232583454281567,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989725525367063,
      "loss": 5.4405,
      "mean_token_accuracy": 0.17421406060457229,
      "num_tokens": 8979882.0,
      "step": 4325
    },
    {
      "entropy": 5.694489431381226,
      "epoch": 0.3927793904208999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989694615782534,
      "loss": 5.5696,
      "mean_token_accuracy": 0.1630749970674515,
      "num_tokens": 8990925.0,
      "step": 4330
    },
    {
      "entropy": 5.679839992523194,
      "epoch": 0.39323294629898403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989663659880552,
      "loss": 5.5923,
      "mean_token_accuracy": 0.1655128762125969,
      "num_tokens": 9001365.0,
      "step": 4335
    },
    {
      "entropy": 5.7446623802185055,
      "epoch": 0.3936865021770682,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989632657661754,
      "loss": 5.5877,
      "mean_token_accuracy": 0.16630492359399796,
      "num_tokens": 9011942.0,
      "step": 4340
    },
    {
      "entropy": 5.694921827316284,
      "epoch": 0.3941400580551524,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498960160912678,
      "loss": 5.6212,
      "mean_token_accuracy": 0.16706815212965012,
      "num_tokens": 9022610.0,
      "step": 4345
    },
    {
      "entropy": 5.657366991043091,
      "epoch": 0.39459361393323655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989570514276276,
      "loss": 5.5033,
      "mean_token_accuracy": 0.1658786192536354,
      "num_tokens": 9033034.0,
      "step": 4350
    },
    {
      "entropy": 5.693386363983154,
      "epoch": 0.39504716981132076,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989539373110883,
      "loss": 5.5471,
      "mean_token_accuracy": 0.1648570790886879,
      "num_tokens": 9043418.0,
      "step": 4355
    },
    {
      "entropy": 5.730675506591797,
      "epoch": 0.3955007256894049,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989508185631245,
      "loss": 5.6473,
      "mean_token_accuracy": 0.16303568482398986,
      "num_tokens": 9052708.0,
      "step": 4360
    },
    {
      "entropy": 5.610522842407226,
      "epoch": 0.39595428156748913,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989476951838006,
      "loss": 5.503,
      "mean_token_accuracy": 0.16568560302257537,
      "num_tokens": 9062883.0,
      "step": 4365
    },
    {
      "entropy": 5.690759563446045,
      "epoch": 0.3964078374455733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989445671731814,
      "loss": 5.4655,
      "mean_token_accuracy": 0.1712013304233551,
      "num_tokens": 9072830.0,
      "step": 4370
    },
    {
      "entropy": 5.583675479888916,
      "epoch": 0.3968613933236575,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989414345313315,
      "loss": 5.417,
      "mean_token_accuracy": 0.16587775200605392,
      "num_tokens": 9082749.0,
      "step": 4375
    },
    {
      "entropy": 5.6703362464904785,
      "epoch": 0.39731494920174165,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989382972583156,
      "loss": 5.5146,
      "mean_token_accuracy": 0.1637563332915306,
      "num_tokens": 9094188.0,
      "step": 4380
    },
    {
      "entropy": 5.6142324924469,
      "epoch": 0.3977685050798258,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989351553541986,
      "loss": 5.5483,
      "mean_token_accuracy": 0.1618421718478203,
      "num_tokens": 9105023.0,
      "step": 4385
    },
    {
      "entropy": 5.60534520149231,
      "epoch": 0.39822206095791,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989320088190456,
      "loss": 5.4782,
      "mean_token_accuracy": 0.17100915908813477,
      "num_tokens": 9115563.0,
      "step": 4390
    },
    {
      "entropy": 5.668853235244751,
      "epoch": 0.3986756168359942,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989288576529216,
      "loss": 5.509,
      "mean_token_accuracy": 0.16283219754695893,
      "num_tokens": 9125394.0,
      "step": 4395
    },
    {
      "entropy": 5.6934130668640135,
      "epoch": 0.3991291727140784,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989257018558917,
      "loss": 5.5032,
      "mean_token_accuracy": 0.16816064417362214,
      "num_tokens": 9137061.0,
      "step": 4400
    },
    {
      "entropy": 5.669379186630249,
      "epoch": 0.39958272859216254,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989225414280213,
      "loss": 5.476,
      "mean_token_accuracy": 0.17045701593160628,
      "num_tokens": 9148196.0,
      "step": 4405
    },
    {
      "entropy": 5.5835472583770756,
      "epoch": 0.40003628447024675,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989193763693756,
      "loss": 5.494,
      "mean_token_accuracy": 0.17453179508447647,
      "num_tokens": 9158013.0,
      "step": 4410
    },
    {
      "entropy": 5.637920618057251,
      "epoch": 0.4004898403483309,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989162066800202,
      "loss": 5.4142,
      "mean_token_accuracy": 0.1738430827856064,
      "num_tokens": 9167813.0,
      "step": 4415
    },
    {
      "entropy": 5.620611953735351,
      "epoch": 0.4009433962264151,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989130323600205,
      "loss": 5.4628,
      "mean_token_accuracy": 0.17226067781448365,
      "num_tokens": 9177783.0,
      "step": 4420
    },
    {
      "entropy": 5.668359851837158,
      "epoch": 0.40139695210449927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989098534094424,
      "loss": 5.5088,
      "mean_token_accuracy": 0.17067906260490417,
      "num_tokens": 9187556.0,
      "step": 4425
    },
    {
      "entropy": 5.655431270599365,
      "epoch": 0.4018505079825834,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989066698283512,
      "loss": 5.5866,
      "mean_token_accuracy": 0.16263027787208556,
      "num_tokens": 9198659.0,
      "step": 4430
    },
    {
      "entropy": 5.669175672531128,
      "epoch": 0.40230406386066764,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989034816168132,
      "loss": 5.5561,
      "mean_token_accuracy": 0.16660318672657012,
      "num_tokens": 9209546.0,
      "step": 4435
    },
    {
      "entropy": 5.627757024765015,
      "epoch": 0.4027576197387518,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989002887748941,
      "loss": 5.4648,
      "mean_token_accuracy": 0.1715000495314598,
      "num_tokens": 9219339.0,
      "step": 4440
    },
    {
      "entropy": 5.656140756607056,
      "epoch": 0.403211175616836,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988970913026601,
      "loss": 5.6487,
      "mean_token_accuracy": 0.16045075058937072,
      "num_tokens": 9230364.0,
      "step": 4445
    },
    {
      "entropy": 5.673274517059326,
      "epoch": 0.40366473149492016,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988938892001771,
      "loss": 5.4897,
      "mean_token_accuracy": 0.16654825061559678,
      "num_tokens": 9240732.0,
      "step": 4450
    },
    {
      "entropy": 5.599252319335937,
      "epoch": 0.40411828737300437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988906824675113,
      "loss": 5.4359,
      "mean_token_accuracy": 0.17193734049797058,
      "num_tokens": 9251049.0,
      "step": 4455
    },
    {
      "entropy": 5.626266956329346,
      "epoch": 0.4045718432510885,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988874711047293,
      "loss": 5.4841,
      "mean_token_accuracy": 0.17143381983041764,
      "num_tokens": 9261660.0,
      "step": 4460
    },
    {
      "entropy": 5.650390625,
      "epoch": 0.40502539912917274,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988842551118972,
      "loss": 5.5465,
      "mean_token_accuracy": 0.170607990026474,
      "num_tokens": 9272114.0,
      "step": 4465
    },
    {
      "entropy": 5.645572662353516,
      "epoch": 0.4054789550072569,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988810344890818,
      "loss": 5.5135,
      "mean_token_accuracy": 0.17275023311376572,
      "num_tokens": 9282125.0,
      "step": 4470
    },
    {
      "entropy": 5.7307713508605955,
      "epoch": 0.40593251088534105,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988778092363494,
      "loss": 5.5466,
      "mean_token_accuracy": 0.16705104261636733,
      "num_tokens": 9294078.0,
      "step": 4475
    },
    {
      "entropy": 5.658994054794311,
      "epoch": 0.40638606676342526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988745793537671,
      "loss": 5.5095,
      "mean_token_accuracy": 0.16595903933048248,
      "num_tokens": 9304026.0,
      "step": 4480
    },
    {
      "entropy": 5.674858808517456,
      "epoch": 0.4068396226415094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988713448414011,
      "loss": 5.5545,
      "mean_token_accuracy": 0.1764197275042534,
      "num_tokens": 9313933.0,
      "step": 4485
    },
    {
      "entropy": 5.660690259933472,
      "epoch": 0.4072931785195936,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498868105699319,
      "loss": 5.5765,
      "mean_token_accuracy": 0.16426668763160707,
      "num_tokens": 9323932.0,
      "step": 4490
    },
    {
      "entropy": 5.5835301876068115,
      "epoch": 0.4077467343976778,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988648619275872,
      "loss": 5.4754,
      "mean_token_accuracy": 0.17924543023109435,
      "num_tokens": 9333467.0,
      "step": 4495
    },
    {
      "entropy": 5.666203594207763,
      "epoch": 0.408200290275762,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988616135262729,
      "loss": 5.5319,
      "mean_token_accuracy": 0.16330583691596984,
      "num_tokens": 9344909.0,
      "step": 4500
    },
    {
      "entropy": 5.749293851852417,
      "epoch": 0.40865384615384615,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988583604954436,
      "loss": 5.5827,
      "mean_token_accuracy": 0.1672707676887512,
      "num_tokens": 9355630.0,
      "step": 4505
    },
    {
      "entropy": 5.601924467086792,
      "epoch": 0.40910740203193036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988551028351663,
      "loss": 5.5262,
      "mean_token_accuracy": 0.165559022128582,
      "num_tokens": 9365700.0,
      "step": 4510
    },
    {
      "entropy": 5.630667209625244,
      "epoch": 0.4095609579100145,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988518405455084,
      "loss": 5.6004,
      "mean_token_accuracy": 0.16178341209888458,
      "num_tokens": 9377088.0,
      "step": 4515
    },
    {
      "entropy": 5.690512752532959,
      "epoch": 0.41001451378809867,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988485736265374,
      "loss": 5.3891,
      "mean_token_accuracy": 0.17818945944309234,
      "num_tokens": 9386855.0,
      "step": 4520
    },
    {
      "entropy": 5.622279644012451,
      "epoch": 0.4104680696661829,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498845302078321,
      "loss": 5.5203,
      "mean_token_accuracy": 0.16728461235761644,
      "num_tokens": 9398661.0,
      "step": 4525
    },
    {
      "entropy": 5.662732839584351,
      "epoch": 0.41092162554426703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988420259009266,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17487220466136932,
      "num_tokens": 9408670.0,
      "step": 4530
    },
    {
      "entropy": 5.680340766906738,
      "epoch": 0.41137518142235124,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498838745094422,
      "loss": 5.531,
      "mean_token_accuracy": 0.1707653671503067,
      "num_tokens": 9418740.0,
      "step": 4535
    },
    {
      "entropy": 5.592594003677368,
      "epoch": 0.4118287373004354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988354596588751,
      "loss": 5.5133,
      "mean_token_accuracy": 0.1718428000807762,
      "num_tokens": 9429339.0,
      "step": 4540
    },
    {
      "entropy": 5.652766942977905,
      "epoch": 0.4122822931785196,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988321695943539,
      "loss": 5.4573,
      "mean_token_accuracy": 0.17410281002521516,
      "num_tokens": 9438996.0,
      "step": 4545
    },
    {
      "entropy": 5.635617685317993,
      "epoch": 0.41273584905660377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988288749009264,
      "loss": 5.4764,
      "mean_token_accuracy": 0.17122164964675904,
      "num_tokens": 9448669.0,
      "step": 4550
    },
    {
      "entropy": 5.588901138305664,
      "epoch": 0.413189404934688,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988255755786608,
      "loss": 5.4564,
      "mean_token_accuracy": 0.17379783987998962,
      "num_tokens": 9458857.0,
      "step": 4555
    },
    {
      "entropy": 5.644811010360717,
      "epoch": 0.41364296081277213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988222716276251,
      "loss": 5.4576,
      "mean_token_accuracy": 0.17216896563768386,
      "num_tokens": 9470241.0,
      "step": 4560
    },
    {
      "entropy": 5.688581943511963,
      "epoch": 0.4140965166908563,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498818963047888,
      "loss": 5.6176,
      "mean_token_accuracy": 0.16512774527072907,
      "num_tokens": 9480528.0,
      "step": 4565
    },
    {
      "entropy": 5.707815265655517,
      "epoch": 0.4145500725689405,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988156498395176,
      "loss": 5.5291,
      "mean_token_accuracy": 0.16812336444854736,
      "num_tokens": 9491000.0,
      "step": 4570
    },
    {
      "entropy": 5.693633604049682,
      "epoch": 0.41500362844702465,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988123320025825,
      "loss": 5.5084,
      "mean_token_accuracy": 0.1725713789463043,
      "num_tokens": 9501612.0,
      "step": 4575
    },
    {
      "entropy": 5.645354413986206,
      "epoch": 0.41545718432510886,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988090095371514,
      "loss": 5.6264,
      "mean_token_accuracy": 0.1642218790948391,
      "num_tokens": 9513999.0,
      "step": 4580
    },
    {
      "entropy": 5.672644186019897,
      "epoch": 0.415910740203193,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988056824432928,
      "loss": 5.4907,
      "mean_token_accuracy": 0.16187755018472672,
      "num_tokens": 9524392.0,
      "step": 4585
    },
    {
      "entropy": 5.691357946395874,
      "epoch": 0.41636429608127723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988023507210758,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1723600745201111,
      "num_tokens": 9534227.0,
      "step": 4590
    },
    {
      "entropy": 5.637865591049194,
      "epoch": 0.4168178519593614,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987990143705692,
      "loss": 5.4877,
      "mean_token_accuracy": 0.17282977551221848,
      "num_tokens": 9544242.0,
      "step": 4595
    },
    {
      "entropy": 5.601320028305054,
      "epoch": 0.4172714078374456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987956733918418,
      "loss": 5.4146,
      "mean_token_accuracy": 0.16889237761497497,
      "num_tokens": 9553921.0,
      "step": 4600
    },
    {
      "entropy": 5.529018545150757,
      "epoch": 0.41772496371552975,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987923277849629,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1648541048169136,
      "num_tokens": 9564768.0,
      "step": 4605
    },
    {
      "entropy": 5.6799657344818115,
      "epoch": 0.4181785195936139,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987889775500017,
      "loss": 5.5097,
      "mean_token_accuracy": 0.16558238118886948,
      "num_tokens": 9576360.0,
      "step": 4610
    },
    {
      "entropy": 5.645033693313598,
      "epoch": 0.4186320754716981,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987856226870272,
      "loss": 5.5206,
      "mean_token_accuracy": 0.16887864768505095,
      "num_tokens": 9587623.0,
      "step": 4615
    },
    {
      "entropy": 5.6239402294158936,
      "epoch": 0.4190856313497823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987822631961092,
      "loss": 5.5487,
      "mean_token_accuracy": 0.1658405616879463,
      "num_tokens": 9597837.0,
      "step": 4620
    },
    {
      "entropy": 5.668896579742432,
      "epoch": 0.4195391872278665,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498778899077317,
      "loss": 5.4938,
      "mean_token_accuracy": 0.1671911209821701,
      "num_tokens": 9609509.0,
      "step": 4625
    },
    {
      "entropy": 5.637753105163574,
      "epoch": 0.41999274310595064,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049877553033072,
      "loss": 5.4786,
      "mean_token_accuracy": 0.1714413955807686,
      "num_tokens": 9619954.0,
      "step": 4630
    },
    {
      "entropy": 5.5957275390625,
      "epoch": 0.42044629898403485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987721569563882,
      "loss": 5.4746,
      "mean_token_accuracy": 0.1704351544380188,
      "num_tokens": 9629742.0,
      "step": 4635
    },
    {
      "entropy": 5.6290453433990475,
      "epoch": 0.420899854862119,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498768778954391,
      "loss": 5.4947,
      "mean_token_accuracy": 0.17307782173156738,
      "num_tokens": 9641378.0,
      "step": 4640
    },
    {
      "entropy": 5.7364380836486815,
      "epoch": 0.4213534107402032,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987653963247985,
      "loss": 5.5701,
      "mean_token_accuracy": 0.1703578621149063,
      "num_tokens": 9652279.0,
      "step": 4645
    },
    {
      "entropy": 5.666488122940064,
      "epoch": 0.42180696661828737,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987620090676805,
      "loss": 5.5464,
      "mean_token_accuracy": 0.1691744804382324,
      "num_tokens": 9663416.0,
      "step": 4650
    },
    {
      "entropy": 5.635804986953735,
      "epoch": 0.4222605224963715,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987586171831072,
      "loss": 5.5084,
      "mean_token_accuracy": 0.17069009840488433,
      "num_tokens": 9674647.0,
      "step": 4655
    },
    {
      "entropy": 5.627348804473877,
      "epoch": 0.42271407837445574,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987552206711487,
      "loss": 5.3729,
      "mean_token_accuracy": 0.17568349838256836,
      "num_tokens": 9685084.0,
      "step": 4660
    },
    {
      "entropy": 5.546395874023437,
      "epoch": 0.4231676342525399,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987518195318752,
      "loss": 5.4367,
      "mean_token_accuracy": 0.1702194795012474,
      "num_tokens": 9695661.0,
      "step": 4665
    },
    {
      "entropy": 5.675781869888306,
      "epoch": 0.4236211901306241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498748413765357,
      "loss": 5.4906,
      "mean_token_accuracy": 0.17196833938360215,
      "num_tokens": 9705773.0,
      "step": 4670
    },
    {
      "entropy": 5.642674303054809,
      "epoch": 0.42407474600870826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987450033716646,
      "loss": 5.5318,
      "mean_token_accuracy": 0.16780486702919006,
      "num_tokens": 9715799.0,
      "step": 4675
    },
    {
      "entropy": 5.678176498413086,
      "epoch": 0.42452830188679247,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987415883508686,
      "loss": 5.4922,
      "mean_token_accuracy": 0.1741717353463173,
      "num_tokens": 9727012.0,
      "step": 4680
    },
    {
      "entropy": 5.624595403671265,
      "epoch": 0.4249818577648766,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987381687030394,
      "loss": 5.5047,
      "mean_token_accuracy": 0.1686483457684517,
      "num_tokens": 9737044.0,
      "step": 4685
    },
    {
      "entropy": 5.73467435836792,
      "epoch": 0.42543541364296084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987347444282479,
      "loss": 5.5885,
      "mean_token_accuracy": 0.15938138514757155,
      "num_tokens": 9747648.0,
      "step": 4690
    },
    {
      "entropy": 5.68295259475708,
      "epoch": 0.425888969521045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987313155265649,
      "loss": 5.4984,
      "mean_token_accuracy": 0.16666015088558198,
      "num_tokens": 9757812.0,
      "step": 4695
    },
    {
      "entropy": 5.677083969116211,
      "epoch": 0.42634252539912915,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987278819980613,
      "loss": 5.5077,
      "mean_token_accuracy": 0.175352643430233,
      "num_tokens": 9768337.0,
      "step": 4700
    },
    {
      "entropy": 5.682326030731201,
      "epoch": 0.42679608127721336,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987244438428081,
      "loss": 5.548,
      "mean_token_accuracy": 0.16384970843791963,
      "num_tokens": 9779433.0,
      "step": 4705
    },
    {
      "entropy": 5.598498392105102,
      "epoch": 0.4272496371552975,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987210010608764,
      "loss": 5.4256,
      "mean_token_accuracy": 0.1729959085583687,
      "num_tokens": 9790069.0,
      "step": 4710
    },
    {
      "entropy": 5.612740516662598,
      "epoch": 0.4277031930333817,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987175536523373,
      "loss": 5.413,
      "mean_token_accuracy": 0.17409739047288894,
      "num_tokens": 9799669.0,
      "step": 4715
    },
    {
      "entropy": 5.602735471725464,
      "epoch": 0.4281567489114659,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987141016172622,
      "loss": 5.4565,
      "mean_token_accuracy": 0.1733306437730789,
      "num_tokens": 9809615.0,
      "step": 4720
    },
    {
      "entropy": 5.598521518707275,
      "epoch": 0.4286103047895501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987106449557225,
      "loss": 5.4288,
      "mean_token_accuracy": 0.17376641631126405,
      "num_tokens": 9819666.0,
      "step": 4725
    },
    {
      "entropy": 5.727816772460938,
      "epoch": 0.42906386066763424,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987071836677896,
      "loss": 5.5897,
      "mean_token_accuracy": 0.16895099133253097,
      "num_tokens": 9831015.0,
      "step": 4730
    },
    {
      "entropy": 5.61959285736084,
      "epoch": 0.42951741654571846,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987037177535353,
      "loss": 5.445,
      "mean_token_accuracy": 0.164634008705616,
      "num_tokens": 9841354.0,
      "step": 4735
    },
    {
      "entropy": 5.564237451553344,
      "epoch": 0.4299709724238026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987002472130309,
      "loss": 5.4452,
      "mean_token_accuracy": 0.16783540546894074,
      "num_tokens": 9850467.0,
      "step": 4740
    },
    {
      "entropy": 5.710805034637451,
      "epoch": 0.43042452830188677,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986967720463485,
      "loss": 5.5211,
      "mean_token_accuracy": 0.16725909262895583,
      "num_tokens": 9861115.0,
      "step": 4745
    },
    {
      "entropy": 5.713682174682617,
      "epoch": 0.430878084179971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986932922535598,
      "loss": 5.5218,
      "mean_token_accuracy": 0.168181936442852,
      "num_tokens": 9871172.0,
      "step": 4750
    },
    {
      "entropy": 5.633159351348877,
      "epoch": 0.43133164005805513,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986898078347368,
      "loss": 5.4289,
      "mean_token_accuracy": 0.1761467546224594,
      "num_tokens": 9881997.0,
      "step": 4755
    },
    {
      "entropy": 5.609486722946167,
      "epoch": 0.43178519593613934,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986863187899516,
      "loss": 5.4029,
      "mean_token_accuracy": 0.17335348427295685,
      "num_tokens": 9892310.0,
      "step": 4760
    },
    {
      "entropy": 5.551136064529419,
      "epoch": 0.4322387518142235,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986828251192764,
      "loss": 5.4672,
      "mean_token_accuracy": 0.1719319611787796,
      "num_tokens": 9902664.0,
      "step": 4765
    },
    {
      "entropy": 5.614825916290283,
      "epoch": 0.4326923076923077,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986793268227833,
      "loss": 5.446,
      "mean_token_accuracy": 0.17410656809806824,
      "num_tokens": 9912491.0,
      "step": 4770
    },
    {
      "entropy": 5.6420478343963625,
      "epoch": 0.43314586357039186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986758239005448,
      "loss": 5.4114,
      "mean_token_accuracy": 0.17530306428670883,
      "num_tokens": 9922057.0,
      "step": 4775
    },
    {
      "entropy": 5.603491544723511,
      "epoch": 0.4335994194484761,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986723163526331,
      "loss": 5.4703,
      "mean_token_accuracy": 0.16902925074100494,
      "num_tokens": 9933661.0,
      "step": 4780
    },
    {
      "entropy": 5.545865392684936,
      "epoch": 0.43405297532656023,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986688041791209,
      "loss": 5.4331,
      "mean_token_accuracy": 0.16436014771461488,
      "num_tokens": 9945177.0,
      "step": 4785
    },
    {
      "entropy": 5.630179929733276,
      "epoch": 0.4345065312046444,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498665287380081,
      "loss": 5.4649,
      "mean_token_accuracy": 0.16838951110839845,
      "num_tokens": 9956226.0,
      "step": 4790
    },
    {
      "entropy": 5.670184659957886,
      "epoch": 0.4349600870827286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986617659555859,
      "loss": 5.5281,
      "mean_token_accuracy": 0.1700868308544159,
      "num_tokens": 9965877.0,
      "step": 4795
    },
    {
      "entropy": 5.601760721206665,
      "epoch": 0.43541364296081275,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986582399057085,
      "loss": 5.4742,
      "mean_token_accuracy": 0.17551955133676528,
      "num_tokens": 9975592.0,
      "step": 4800
    },
    {
      "entropy": 5.680169630050659,
      "epoch": 0.43586719883889696,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986547092305216,
      "loss": 5.5056,
      "mean_token_accuracy": 0.16957487165927887,
      "num_tokens": 9986094.0,
      "step": 4805
    },
    {
      "entropy": 5.717319345474243,
      "epoch": 0.4363207547169811,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986511739300984,
      "loss": 5.5406,
      "mean_token_accuracy": 0.16883666813373566,
      "num_tokens": 9997241.0,
      "step": 4810
    },
    {
      "entropy": 5.607332468032837,
      "epoch": 0.43677431059506533,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986476340045119,
      "loss": 5.4581,
      "mean_token_accuracy": 0.17112519294023515,
      "num_tokens": 10008091.0,
      "step": 4815
    },
    {
      "entropy": 5.587670564651489,
      "epoch": 0.4372278664731495,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986440894538354,
      "loss": 5.5149,
      "mean_token_accuracy": 0.16817084699869156,
      "num_tokens": 10018794.0,
      "step": 4820
    },
    {
      "entropy": 5.617847633361817,
      "epoch": 0.4376814223512337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986405402781421,
      "loss": 5.4295,
      "mean_token_accuracy": 0.17280754894018174,
      "num_tokens": 10028414.0,
      "step": 4825
    },
    {
      "entropy": 5.634881925582886,
      "epoch": 0.43813497822931785,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986369864775054,
      "loss": 5.5077,
      "mean_token_accuracy": 0.16782476902008056,
      "num_tokens": 10038749.0,
      "step": 4830
    },
    {
      "entropy": 5.6245917797088625,
      "epoch": 0.438588534107402,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986334280519989,
      "loss": 5.4539,
      "mean_token_accuracy": 0.17029527723789215,
      "num_tokens": 10049866.0,
      "step": 4835
    },
    {
      "entropy": 5.6066179275512695,
      "epoch": 0.4390420899854862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498629865001696,
      "loss": 5.4886,
      "mean_token_accuracy": 0.17579846531152726,
      "num_tokens": 10060680.0,
      "step": 4840
    },
    {
      "entropy": 5.681512403488159,
      "epoch": 0.43949564586357037,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986262973266706,
      "loss": 5.5728,
      "mean_token_accuracy": 0.16835084557533264,
      "num_tokens": 10070931.0,
      "step": 4845
    },
    {
      "entropy": 5.582299041748047,
      "epoch": 0.4399492017416546,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986227250269964,
      "loss": 5.5073,
      "mean_token_accuracy": 0.1723349377512932,
      "num_tokens": 10080189.0,
      "step": 4850
    },
    {
      "entropy": 5.640298318862915,
      "epoch": 0.44040275761973874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986191481027472,
      "loss": 5.4149,
      "mean_token_accuracy": 0.17179947793483735,
      "num_tokens": 10088813.0,
      "step": 4855
    },
    {
      "entropy": 5.5767192363739015,
      "epoch": 0.44085631349782295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986155665539971,
      "loss": 5.4386,
      "mean_token_accuracy": 0.17210467159748077,
      "num_tokens": 10098125.0,
      "step": 4860
    },
    {
      "entropy": 5.666972684860229,
      "epoch": 0.4413098693759071,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049861198038082,
      "loss": 5.4766,
      "mean_token_accuracy": 0.16837365478277205,
      "num_tokens": 10109125.0,
      "step": 4865
    },
    {
      "entropy": 5.607843923568725,
      "epoch": 0.4417634252539913,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986083895832903,
      "loss": 5.4597,
      "mean_token_accuracy": 0.17088931798934937,
      "num_tokens": 10120182.0,
      "step": 4870
    },
    {
      "entropy": 5.563991737365723,
      "epoch": 0.44221698113207547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986047941614821,
      "loss": 5.5187,
      "mean_token_accuracy": 0.1657083287835121,
      "num_tokens": 10130941.0,
      "step": 4875
    },
    {
      "entropy": 5.683834409713745,
      "epoch": 0.4426705370101596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986011941154696,
      "loss": 5.4817,
      "mean_token_accuracy": 0.16850205659866332,
      "num_tokens": 10141708.0,
      "step": 4880
    },
    {
      "entropy": 5.627482271194458,
      "epoch": 0.44312409288824384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985975894453275,
      "loss": 5.4216,
      "mean_token_accuracy": 0.16762909889221192,
      "num_tokens": 10151816.0,
      "step": 4885
    },
    {
      "entropy": 5.666119718551636,
      "epoch": 0.443577648766328,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985939801511304,
      "loss": 5.5404,
      "mean_token_accuracy": 0.16791291683912277,
      "num_tokens": 10162470.0,
      "step": 4890
    },
    {
      "entropy": 5.635271310806274,
      "epoch": 0.4440312046444122,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985903662329527,
      "loss": 5.4278,
      "mean_token_accuracy": 0.17861440777778625,
      "num_tokens": 10173171.0,
      "step": 4895
    },
    {
      "entropy": 5.6425152778625485,
      "epoch": 0.44448476052249636,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985867476908693,
      "loss": 5.4696,
      "mean_token_accuracy": 0.17133731096982957,
      "num_tokens": 10184185.0,
      "step": 4900
    },
    {
      "entropy": 5.674123239517212,
      "epoch": 0.44493831640058057,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985831245249549,
      "loss": 5.5259,
      "mean_token_accuracy": 0.17461849972605706,
      "num_tokens": 10195516.0,
      "step": 4905
    },
    {
      "entropy": 5.618943500518799,
      "epoch": 0.4453918722786647,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985794967352846,
      "loss": 5.4537,
      "mean_token_accuracy": 0.17528860867023469,
      "num_tokens": 10204936.0,
      "step": 4910
    },
    {
      "entropy": 5.646315479278565,
      "epoch": 0.44584542815674894,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985758643219334,
      "loss": 5.4884,
      "mean_token_accuracy": 0.1720237150788307,
      "num_tokens": 10216218.0,
      "step": 4915
    },
    {
      "entropy": 5.66897406578064,
      "epoch": 0.4462989840348331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985722272849762,
      "loss": 5.4856,
      "mean_token_accuracy": 0.16680216640233994,
      "num_tokens": 10226370.0,
      "step": 4920
    },
    {
      "entropy": 5.541078567504883,
      "epoch": 0.44675253991291725,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985685856244884,
      "loss": 5.418,
      "mean_token_accuracy": 0.18198757171630858,
      "num_tokens": 10236170.0,
      "step": 4925
    },
    {
      "entropy": 5.612875270843506,
      "epoch": 0.44720609579100146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985649393405453,
      "loss": 5.4902,
      "mean_token_accuracy": 0.16974830478429795,
      "num_tokens": 10246119.0,
      "step": 4930
    },
    {
      "entropy": 5.630044221878052,
      "epoch": 0.4476596516690856,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985612884332223,
      "loss": 5.5032,
      "mean_token_accuracy": 0.16304934322834014,
      "num_tokens": 10256762.0,
      "step": 4935
    },
    {
      "entropy": 5.5931806564331055,
      "epoch": 0.4481132075471698,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985576329025949,
      "loss": 5.4349,
      "mean_token_accuracy": 0.17295026481151582,
      "num_tokens": 10267317.0,
      "step": 4940
    },
    {
      "entropy": 5.6881492137908936,
      "epoch": 0.448566763425254,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985539727487385,
      "loss": 5.5798,
      "mean_token_accuracy": 0.15689071714878083,
      "num_tokens": 10279111.0,
      "step": 4945
    },
    {
      "entropy": 5.61900143623352,
      "epoch": 0.4490203193033382,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985503079717292,
      "loss": 5.4445,
      "mean_token_accuracy": 0.17076381295919418,
      "num_tokens": 10289383.0,
      "step": 4950
    },
    {
      "entropy": 5.611308670043945,
      "epoch": 0.44947387518142234,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985466385716425,
      "loss": 5.4578,
      "mean_token_accuracy": 0.1737099438905716,
      "num_tokens": 10299806.0,
      "step": 4955
    },
    {
      "entropy": 5.5558473587036135,
      "epoch": 0.44992743105950656,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985429645485543,
      "loss": 5.4864,
      "mean_token_accuracy": 0.16585210114717483,
      "num_tokens": 10310348.0,
      "step": 4960
    },
    {
      "entropy": 5.625494909286499,
      "epoch": 0.4503809869375907,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985392859025407,
      "loss": 5.4737,
      "mean_token_accuracy": 0.17041534036397935,
      "num_tokens": 10320354.0,
      "step": 4965
    },
    {
      "entropy": 5.601232576370239,
      "epoch": 0.45083454281567487,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985356026336776,
      "loss": 5.3997,
      "mean_token_accuracy": 0.17218953967094422,
      "num_tokens": 10330292.0,
      "step": 4970
    },
    {
      "entropy": 5.576334428787232,
      "epoch": 0.4512880986937591,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985319147420414,
      "loss": 5.366,
      "mean_token_accuracy": 0.18071993440389633,
      "num_tokens": 10340449.0,
      "step": 4975
    },
    {
      "entropy": 5.603624248504639,
      "epoch": 0.45174165457184323,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985282222277082,
      "loss": 5.442,
      "mean_token_accuracy": 0.17101425379514695,
      "num_tokens": 10350637.0,
      "step": 4980
    },
    {
      "entropy": 5.54790735244751,
      "epoch": 0.45219521044992744,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985245250907544,
      "loss": 5.3815,
      "mean_token_accuracy": 0.17386381030082704,
      "num_tokens": 10361548.0,
      "step": 4985
    },
    {
      "entropy": 5.566483211517334,
      "epoch": 0.4526487663280116,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985208233312564,
      "loss": 5.3907,
      "mean_token_accuracy": 0.1698409304022789,
      "num_tokens": 10371765.0,
      "step": 4990
    },
    {
      "entropy": 5.661066150665283,
      "epoch": 0.4531023222060958,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985171169492908,
      "loss": 5.4967,
      "mean_token_accuracy": 0.1759715974330902,
      "num_tokens": 10382108.0,
      "step": 4995
    },
    {
      "entropy": 5.606108808517456,
      "epoch": 0.45355587808417996,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985134059449344,
      "loss": 5.4415,
      "mean_token_accuracy": 0.1780141994357109,
      "num_tokens": 10392390.0,
      "step": 5000
    },
    {
      "entropy": 5.593348503112793,
      "epoch": 0.4540094339622642,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985096903182637,
      "loss": 5.5074,
      "mean_token_accuracy": 0.17851048856973648,
      "num_tokens": 10402094.0,
      "step": 5005
    },
    {
      "entropy": 5.660210275650025,
      "epoch": 0.45446298984034833,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985059700693557,
      "loss": 5.4447,
      "mean_token_accuracy": 0.17331443428993226,
      "num_tokens": 10413350.0,
      "step": 5010
    },
    {
      "entropy": 5.674293041229248,
      "epoch": 0.4549165457184325,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985022451982872,
      "loss": 5.5776,
      "mean_token_accuracy": 0.16241093277931212,
      "num_tokens": 10424292.0,
      "step": 5015
    },
    {
      "entropy": 5.69236478805542,
      "epoch": 0.4553701015965167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984985157051354,
      "loss": 5.4254,
      "mean_token_accuracy": 0.17173273414373397,
      "num_tokens": 10433724.0,
      "step": 5020
    },
    {
      "entropy": 5.644928026199341,
      "epoch": 0.45582365747460085,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984947815899774,
      "loss": 5.5155,
      "mean_token_accuracy": 0.16911423206329346,
      "num_tokens": 10444724.0,
      "step": 5025
    },
    {
      "entropy": 5.585797357559204,
      "epoch": 0.45627721335268506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984910428528902,
      "loss": 5.4332,
      "mean_token_accuracy": 0.17657725960016252,
      "num_tokens": 10455086.0,
      "step": 5030
    },
    {
      "entropy": 5.552824258804321,
      "epoch": 0.4567307692307692,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984872994939514,
      "loss": 5.3995,
      "mean_token_accuracy": 0.16780003756284714,
      "num_tokens": 10465076.0,
      "step": 5035
    },
    {
      "entropy": 5.628185558319092,
      "epoch": 0.45718432510885343,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984835515132382,
      "loss": 5.4097,
      "mean_token_accuracy": 0.17855704724788665,
      "num_tokens": 10474817.0,
      "step": 5040
    },
    {
      "entropy": 5.590842294692993,
      "epoch": 0.4576378809869376,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984797989108282,
      "loss": 5.4381,
      "mean_token_accuracy": 0.1680952191352844,
      "num_tokens": 10485735.0,
      "step": 5045
    },
    {
      "entropy": 5.577192163467407,
      "epoch": 0.4580914368650218,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984760416867991,
      "loss": 5.5278,
      "mean_token_accuracy": 0.16953626424074172,
      "num_tokens": 10496731.0,
      "step": 5050
    },
    {
      "entropy": 5.570709562301635,
      "epoch": 0.45854499274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984722798412286,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1745694950222969,
      "num_tokens": 10506787.0,
      "step": 5055
    },
    {
      "entropy": 5.653943300247192,
      "epoch": 0.4589985486211901,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984685133741941,
      "loss": 5.4581,
      "mean_token_accuracy": 0.1692812368273735,
      "num_tokens": 10517170.0,
      "step": 5060
    },
    {
      "entropy": 5.604233932495117,
      "epoch": 0.4594521044992743,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498464742285774,
      "loss": 5.46,
      "mean_token_accuracy": 0.1756867364048958,
      "num_tokens": 10527342.0,
      "step": 5065
    },
    {
      "entropy": 5.5937105178833,
      "epoch": 0.45990566037735847,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498460966576046,
      "loss": 5.433,
      "mean_token_accuracy": 0.1744078904390335,
      "num_tokens": 10537253.0,
      "step": 5070
    },
    {
      "entropy": 5.653906154632568,
      "epoch": 0.4603592162554427,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984571862450884,
      "loss": 5.4459,
      "mean_token_accuracy": 0.16896500289440156,
      "num_tokens": 10547249.0,
      "step": 5075
    },
    {
      "entropy": 5.689875221252441,
      "epoch": 0.46081277213352684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984534012929791,
      "loss": 5.4873,
      "mean_token_accuracy": 0.17034367769956588,
      "num_tokens": 10558174.0,
      "step": 5080
    },
    {
      "entropy": 5.67174391746521,
      "epoch": 0.46126632801161105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984496117197967,
      "loss": 5.4359,
      "mean_token_accuracy": 0.17626606971025466,
      "num_tokens": 10567610.0,
      "step": 5085
    },
    {
      "entropy": 5.512884426116943,
      "epoch": 0.4617198838896952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984458175256193,
      "loss": 5.4515,
      "mean_token_accuracy": 0.17041234970092772,
      "num_tokens": 10578494.0,
      "step": 5090
    },
    {
      "entropy": 5.667339754104614,
      "epoch": 0.4621734397677794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984420187105253,
      "loss": 5.5543,
      "mean_token_accuracy": 0.16734294444322587,
      "num_tokens": 10589482.0,
      "step": 5095
    },
    {
      "entropy": 5.714143896102906,
      "epoch": 0.46262699564586357,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984382152745936,
      "loss": 5.5033,
      "mean_token_accuracy": 0.16372307240962983,
      "num_tokens": 10599965.0,
      "step": 5100
    },
    {
      "entropy": 5.607032728195191,
      "epoch": 0.4630805515239477,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984344072179026,
      "loss": 5.5384,
      "mean_token_accuracy": 0.17114560455083846,
      "num_tokens": 10610500.0,
      "step": 5105
    },
    {
      "entropy": 5.703943061828613,
      "epoch": 0.46353410740203194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984305945405312,
      "loss": 5.4792,
      "mean_token_accuracy": 0.17351720333099366,
      "num_tokens": 10619843.0,
      "step": 5110
    },
    {
      "entropy": 5.620040988922119,
      "epoch": 0.4639876632801161,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498426777242558,
      "loss": 5.4189,
      "mean_token_accuracy": 0.17909690588712693,
      "num_tokens": 10631026.0,
      "step": 5115
    },
    {
      "entropy": 5.440194177627563,
      "epoch": 0.4644412191582003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984229553240623,
      "loss": 5.4071,
      "mean_token_accuracy": 0.17747279554605483,
      "num_tokens": 10641130.0,
      "step": 5120
    },
    {
      "entropy": 5.710953187942505,
      "epoch": 0.46489477503628446,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984191287851228,
      "loss": 5.4911,
      "mean_token_accuracy": 0.17253502756357192,
      "num_tokens": 10652305.0,
      "step": 5125
    },
    {
      "entropy": 5.679422616958618,
      "epoch": 0.46534833091436867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984152976258188,
      "loss": 5.5138,
      "mean_token_accuracy": 0.17212531566619874,
      "num_tokens": 10663083.0,
      "step": 5130
    },
    {
      "entropy": 5.602247858047486,
      "epoch": 0.4658018867924528,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984114618462296,
      "loss": 5.5342,
      "mean_token_accuracy": 0.1666450873017311,
      "num_tokens": 10673602.0,
      "step": 5135
    },
    {
      "entropy": 5.576183080673218,
      "epoch": 0.46625544267053703,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984076214464343,
      "loss": 5.3597,
      "mean_token_accuracy": 0.18366137892007828,
      "num_tokens": 10684622.0,
      "step": 5140
    },
    {
      "entropy": 5.608708000183105,
      "epoch": 0.4667089985486212,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984037764265126,
      "loss": 5.4792,
      "mean_token_accuracy": 0.171570348739624,
      "num_tokens": 10695129.0,
      "step": 5145
    },
    {
      "entropy": 5.566263818740845,
      "epoch": 0.46716255442670535,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983999267865438,
      "loss": 5.3269,
      "mean_token_accuracy": 0.1831385925412178,
      "num_tokens": 10705236.0,
      "step": 5150
    },
    {
      "entropy": 5.559924459457397,
      "epoch": 0.46761611030478956,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983960725266078,
      "loss": 5.4354,
      "mean_token_accuracy": 0.17425220608711242,
      "num_tokens": 10715748.0,
      "step": 5155
    },
    {
      "entropy": 5.570201396942139,
      "epoch": 0.4680696661828737,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983922136467838,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18027273714542388,
      "num_tokens": 10725799.0,
      "step": 5160
    },
    {
      "entropy": 5.638973999023437,
      "epoch": 0.4685232220609579,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498388350147152,
      "loss": 5.4945,
      "mean_token_accuracy": 0.16939412355422973,
      "num_tokens": 10735668.0,
      "step": 5165
    },
    {
      "entropy": 5.662532472610474,
      "epoch": 0.4689767779390421,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983844820277922,
      "loss": 5.5932,
      "mean_token_accuracy": 0.1683347389101982,
      "num_tokens": 10746466.0,
      "step": 5170
    },
    {
      "entropy": 5.7226646900177,
      "epoch": 0.4694303338171263,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983806092887843,
      "loss": 5.4903,
      "mean_token_accuracy": 0.170631568133831,
      "num_tokens": 10756724.0,
      "step": 5175
    },
    {
      "entropy": 5.611787223815918,
      "epoch": 0.46988388969521044,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983767319302086,
      "loss": 5.5413,
      "mean_token_accuracy": 0.1692361503839493,
      "num_tokens": 10766919.0,
      "step": 5180
    },
    {
      "entropy": 5.565976285934449,
      "epoch": 0.47033744557329465,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983728499521451,
      "loss": 5.3111,
      "mean_token_accuracy": 0.1810767963528633,
      "num_tokens": 10777550.0,
      "step": 5185
    },
    {
      "entropy": 5.609437608718872,
      "epoch": 0.4707910014513788,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983689633546741,
      "loss": 5.3816,
      "mean_token_accuracy": 0.17074895054101943,
      "num_tokens": 10788456.0,
      "step": 5190
    },
    {
      "entropy": 5.579924440383911,
      "epoch": 0.47124455732946297,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983650721378761,
      "loss": 5.4118,
      "mean_token_accuracy": 0.1763128623366356,
      "num_tokens": 10800049.0,
      "step": 5195
    },
    {
      "entropy": 5.6112401485443115,
      "epoch": 0.4716981132075472,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983611763018314,
      "loss": 5.4538,
      "mean_token_accuracy": 0.1694786950945854,
      "num_tokens": 10810743.0,
      "step": 5200
    },
    {
      "entropy": 5.722457981109619,
      "epoch": 0.47215166908563133,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983572758466208,
      "loss": 5.6501,
      "mean_token_accuracy": 0.16584568619728088,
      "num_tokens": 10821897.0,
      "step": 5205
    },
    {
      "entropy": 5.692024803161621,
      "epoch": 0.47260522496371554,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983533707723247,
      "loss": 5.5064,
      "mean_token_accuracy": 0.17261601835489274,
      "num_tokens": 10831589.0,
      "step": 5210
    },
    {
      "entropy": 5.609469699859619,
      "epoch": 0.4730587808417997,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983494610790241,
      "loss": 5.4244,
      "mean_token_accuracy": 0.16788049787282944,
      "num_tokens": 10841404.0,
      "step": 5215
    },
    {
      "entropy": 5.577998208999634,
      "epoch": 0.4735123367198839,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983455467667996,
      "loss": 5.4547,
      "mean_token_accuracy": 0.16858301162719727,
      "num_tokens": 10851577.0,
      "step": 5220
    },
    {
      "entropy": 5.585168886184692,
      "epoch": 0.47396589259796806,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983416278357322,
      "loss": 5.3676,
      "mean_token_accuracy": 0.17800675630569457,
      "num_tokens": 10861403.0,
      "step": 5225
    },
    {
      "entropy": 5.595915460586548,
      "epoch": 0.4744194484760523,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983377042859032,
      "loss": 5.416,
      "mean_token_accuracy": 0.16920799762010574,
      "num_tokens": 10871933.0,
      "step": 5230
    },
    {
      "entropy": 5.5529121398925785,
      "epoch": 0.47487300435413643,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983337761173936,
      "loss": 5.4844,
      "mean_token_accuracy": 0.17355937510728836,
      "num_tokens": 10881800.0,
      "step": 5235
    },
    {
      "entropy": 5.67032265663147,
      "epoch": 0.4753265602322206,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983298433302843,
      "loss": 5.4219,
      "mean_token_accuracy": 0.1732437029480934,
      "num_tokens": 10891830.0,
      "step": 5240
    },
    {
      "entropy": 5.5511308193206785,
      "epoch": 0.4757801161103048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983259059246572,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1802822083234787,
      "num_tokens": 10901557.0,
      "step": 5245
    },
    {
      "entropy": 5.5265007495880125,
      "epoch": 0.47623367198838895,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983219639005934,
      "loss": 5.4896,
      "mean_token_accuracy": 0.16791147738695145,
      "num_tokens": 10912155.0,
      "step": 5250
    },
    {
      "entropy": 5.609215116500854,
      "epoch": 0.47668722786647316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983180172581745,
      "loss": 5.3888,
      "mean_token_accuracy": 0.17777832001447677,
      "num_tokens": 10922291.0,
      "step": 5255
    },
    {
      "entropy": 5.660725116729736,
      "epoch": 0.4771407837445573,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983140659974819,
      "loss": 5.4694,
      "mean_token_accuracy": 0.17444853037595748,
      "num_tokens": 10933176.0,
      "step": 5260
    },
    {
      "entropy": 5.595969295501709,
      "epoch": 0.47759433962264153,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983101101185977,
      "loss": 5.5461,
      "mean_token_accuracy": 0.165020552277565,
      "num_tokens": 10944049.0,
      "step": 5265
    },
    {
      "entropy": 5.5567363739013675,
      "epoch": 0.4780478955007257,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983061496216035,
      "loss": 5.3823,
      "mean_token_accuracy": 0.1773837149143219,
      "num_tokens": 10954347.0,
      "step": 5270
    },
    {
      "entropy": 5.603483200073242,
      "epoch": 0.4785014513788099,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983021845065812,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17545117288827897,
      "num_tokens": 10965123.0,
      "step": 5275
    },
    {
      "entropy": 5.5426102638244625,
      "epoch": 0.47895500725689405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982982147736128,
      "loss": 5.429,
      "mean_token_accuracy": 0.17363492995500565,
      "num_tokens": 10976395.0,
      "step": 5280
    },
    {
      "entropy": 5.617046165466308,
      "epoch": 0.4794085631349782,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982942404227805,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17311217486858368,
      "num_tokens": 10986797.0,
      "step": 5285
    },
    {
      "entropy": 5.655211639404297,
      "epoch": 0.4798621190130624,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982902614541662,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1699137955904007,
      "num_tokens": 10997229.0,
      "step": 5290
    },
    {
      "entropy": 5.557010078430176,
      "epoch": 0.48031567489114657,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982862778678526,
      "loss": 5.4061,
      "mean_token_accuracy": 0.18039322644472122,
      "num_tokens": 11008180.0,
      "step": 5295
    },
    {
      "entropy": 5.612529039382935,
      "epoch": 0.4807692307692308,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982822896639218,
      "loss": 5.4279,
      "mean_token_accuracy": 0.179520645737648,
      "num_tokens": 11019248.0,
      "step": 5300
    },
    {
      "entropy": 5.558038425445557,
      "epoch": 0.48122278664731494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982782968424563,
      "loss": 5.352,
      "mean_token_accuracy": 0.1724263459444046,
      "num_tokens": 11029760.0,
      "step": 5305
    },
    {
      "entropy": 5.563699245452881,
      "epoch": 0.48167634252539915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982742994035388,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1763938307762146,
      "num_tokens": 11039211.0,
      "step": 5310
    },
    {
      "entropy": 5.577710294723511,
      "epoch": 0.4821298984034833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982702973472518,
      "loss": 5.4115,
      "mean_token_accuracy": 0.17324711233377457,
      "num_tokens": 11049892.0,
      "step": 5315
    },
    {
      "entropy": 5.665327739715576,
      "epoch": 0.4825834542815675,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982662906736783,
      "loss": 5.4634,
      "mean_token_accuracy": 0.17439174950122832,
      "num_tokens": 11060823.0,
      "step": 5320
    },
    {
      "entropy": 5.543615531921387,
      "epoch": 0.48303701015965167,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982622793829009,
      "loss": 5.4282,
      "mean_token_accuracy": 0.17477934807538986,
      "num_tokens": 11072353.0,
      "step": 5325
    },
    {
      "entropy": 5.606219053268433,
      "epoch": 0.4834905660377358,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982582634750027,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1762908637523651,
      "num_tokens": 11082655.0,
      "step": 5330
    },
    {
      "entropy": 5.619220972061157,
      "epoch": 0.48394412191582004,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982542429500667,
      "loss": 5.4757,
      "mean_token_accuracy": 0.1755259931087494,
      "num_tokens": 11093315.0,
      "step": 5335
    },
    {
      "entropy": 5.478820943832398,
      "epoch": 0.4843976777939042,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982502178081761,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1801849588751793,
      "num_tokens": 11103074.0,
      "step": 5340
    },
    {
      "entropy": 5.587391710281372,
      "epoch": 0.4848512336719884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982461880494141,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1792117804288864,
      "num_tokens": 11113636.0,
      "step": 5345
    },
    {
      "entropy": 5.571300601959228,
      "epoch": 0.48530478955007256,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498242153673864,
      "loss": 5.367,
      "mean_token_accuracy": 0.17792698591947556,
      "num_tokens": 11123509.0,
      "step": 5350
    },
    {
      "entropy": 5.555966091156006,
      "epoch": 0.48575834542815677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982381146816094,
      "loss": 5.413,
      "mean_token_accuracy": 0.17205502539873124,
      "num_tokens": 11133405.0,
      "step": 5355
    },
    {
      "entropy": 5.676434469223023,
      "epoch": 0.4862119013062409,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982340710727336,
      "loss": 5.4994,
      "mean_token_accuracy": 0.1640792816877365,
      "num_tokens": 11144725.0,
      "step": 5360
    },
    {
      "entropy": 5.577190208435058,
      "epoch": 0.48666545718432513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982300228473203,
      "loss": 5.378,
      "mean_token_accuracy": 0.17402091771364211,
      "num_tokens": 11155446.0,
      "step": 5365
    },
    {
      "entropy": 5.566140747070312,
      "epoch": 0.4871190130624093,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982259700054533,
      "loss": 5.4423,
      "mean_token_accuracy": 0.1782988503575325,
      "num_tokens": 11165380.0,
      "step": 5370
    },
    {
      "entropy": 5.599629878997803,
      "epoch": 0.48757256894049344,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982219125472163,
      "loss": 5.4825,
      "mean_token_accuracy": 0.17474941462278365,
      "num_tokens": 11175641.0,
      "step": 5375
    },
    {
      "entropy": 5.750193023681641,
      "epoch": 0.48802612481857766,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982178504726933,
      "loss": 5.6549,
      "mean_token_accuracy": 0.17236959785223008,
      "num_tokens": 11186257.0,
      "step": 5380
    },
    {
      "entropy": 5.677424907684326,
      "epoch": 0.4884796806966618,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982137837819682,
      "loss": 5.4656,
      "mean_token_accuracy": 0.16541945934295654,
      "num_tokens": 11197284.0,
      "step": 5385
    },
    {
      "entropy": 5.629317426681519,
      "epoch": 0.488933236574746,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982097124751252,
      "loss": 5.4406,
      "mean_token_accuracy": 0.17467922419309617,
      "num_tokens": 11208413.0,
      "step": 5390
    },
    {
      "entropy": 5.571043968200684,
      "epoch": 0.4893867924528302,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982056365522486,
      "loss": 5.438,
      "mean_token_accuracy": 0.17819569408893585,
      "num_tokens": 11219016.0,
      "step": 5395
    },
    {
      "entropy": 5.607671308517456,
      "epoch": 0.4898403483309144,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982015560134225,
      "loss": 5.405,
      "mean_token_accuracy": 0.17524216771125795,
      "num_tokens": 11228934.0,
      "step": 5400
    },
    {
      "entropy": 5.603353881835938,
      "epoch": 0.49029390420899854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981974708587312,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1815441757440567,
      "num_tokens": 11239596.0,
      "step": 5405
    },
    {
      "entropy": 5.519843006134034,
      "epoch": 0.49074746008708275,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981933810882594,
      "loss": 5.3232,
      "mean_token_accuracy": 0.17441688776016234,
      "num_tokens": 11249560.0,
      "step": 5410
    },
    {
      "entropy": 5.6501236915588375,
      "epoch": 0.4912010159651669,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981892867020917,
      "loss": 5.5281,
      "mean_token_accuracy": 0.17009060010313987,
      "num_tokens": 11259787.0,
      "step": 5415
    },
    {
      "entropy": 5.669698715209961,
      "epoch": 0.49165457184325106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981851877003125,
      "loss": 5.4462,
      "mean_token_accuracy": 0.17556439489126205,
      "num_tokens": 11269717.0,
      "step": 5420
    },
    {
      "entropy": 5.540651464462281,
      "epoch": 0.4921081277213353,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981810840830068,
      "loss": 5.4329,
      "mean_token_accuracy": 0.18177998811006546,
      "num_tokens": 11279863.0,
      "step": 5425
    },
    {
      "entropy": 5.613092374801636,
      "epoch": 0.49256168359941943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981769758502596,
      "loss": 5.555,
      "mean_token_accuracy": 0.16333883106708527,
      "num_tokens": 11291161.0,
      "step": 5430
    },
    {
      "entropy": 5.5472633838653564,
      "epoch": 0.49301523947750364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981728630021555,
      "loss": 5.3762,
      "mean_token_accuracy": 0.17442505955696105,
      "num_tokens": 11301151.0,
      "step": 5435
    },
    {
      "entropy": 5.524767255783081,
      "epoch": 0.4934687953555878,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981687455387797,
      "loss": 5.3818,
      "mean_token_accuracy": 0.1735568791627884,
      "num_tokens": 11311466.0,
      "step": 5440
    },
    {
      "entropy": 5.597518110275269,
      "epoch": 0.493922351233672,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981646234602173,
      "loss": 5.3424,
      "mean_token_accuracy": 0.17718082964420317,
      "num_tokens": 11321522.0,
      "step": 5445
    },
    {
      "entropy": 5.68077597618103,
      "epoch": 0.49437590711175616,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981604967665539,
      "loss": 5.4617,
      "mean_token_accuracy": 0.17782650887966156,
      "num_tokens": 11332317.0,
      "step": 5450
    },
    {
      "entropy": 5.593668365478516,
      "epoch": 0.4948294629898404,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981563654578743,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1720297485589981,
      "num_tokens": 11343024.0,
      "step": 5455
    },
    {
      "entropy": 5.524274826049805,
      "epoch": 0.49528301886792453,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981522295342643,
      "loss": 5.365,
      "mean_token_accuracy": 0.17542414665222167,
      "num_tokens": 11353114.0,
      "step": 5460
    },
    {
      "entropy": 5.566524648666382,
      "epoch": 0.4957365747460087,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981480889958092,
      "loss": 5.411,
      "mean_token_accuracy": 0.17415190041065215,
      "num_tokens": 11363446.0,
      "step": 5465
    },
    {
      "entropy": 5.521554613113404,
      "epoch": 0.4961901306240929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981439438425949,
      "loss": 5.4142,
      "mean_token_accuracy": 0.18057068884372712,
      "num_tokens": 11373413.0,
      "step": 5470
    },
    {
      "entropy": 5.528734445571899,
      "epoch": 0.49664368650217705,
      "grad_norm": 1.0,
      "learning_rate": 0.000498139794074707,
      "loss": 5.2985,
      "mean_token_accuracy": 0.17956365793943405,
      "num_tokens": 11383142.0,
      "step": 5475
    },
    {
      "entropy": 5.62391300201416,
      "epoch": 0.49709724238026126,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981356396922313,
      "loss": 5.4436,
      "mean_token_accuracy": 0.17436255961656572,
      "num_tokens": 11394635.0,
      "step": 5480
    },
    {
      "entropy": 5.548166561126709,
      "epoch": 0.4975507982583454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004981314806952535,
      "loss": 5.3495,
      "mean_token_accuracy": 0.17908620834350586,
      "num_tokens": 11404442.0,
      "step": 5485
    },
    {
      "entropy": 5.488615989685059,
      "epoch": 0.49800435413642963,
      "grad_norm": 1.0,
      "learning_rate": 0.00049812731708386,
      "loss": 5.3055,
      "mean_token_accuracy": 0.1778433859348297,
      "num_tokens": 11414217.0,
      "step": 5490
    },
    {
      "entropy": 5.632096195220948,
      "epoch": 0.4984579100145138,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981231488581367,
      "loss": 5.435,
      "mean_token_accuracy": 0.1736169531941414,
      "num_tokens": 11423834.0,
      "step": 5495
    },
    {
      "entropy": 5.467576503753662,
      "epoch": 0.498911465892598,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981189760181699,
      "loss": 5.3164,
      "mean_token_accuracy": 0.17622301876544952,
      "num_tokens": 11433613.0,
      "step": 5500
    },
    {
      "entropy": 5.507274103164673,
      "epoch": 0.49936502177068215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981147985640456,
      "loss": 5.3959,
      "mean_token_accuracy": 0.1766986832022667,
      "num_tokens": 11443915.0,
      "step": 5505
    },
    {
      "entropy": 5.668768692016601,
      "epoch": 0.4998185776487663,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981106164958506,
      "loss": 5.4205,
      "mean_token_accuracy": 0.17393814772367477,
      "num_tokens": 11454588.0,
      "step": 5510
    },
    {
      "entropy": 5.569959020614624,
      "epoch": 0.5002721335268505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981064298136712,
      "loss": 5.4766,
      "mean_token_accuracy": 0.17282384485006333,
      "num_tokens": 11465533.0,
      "step": 5515
    },
    {
      "entropy": 5.569711923599243,
      "epoch": 0.5007256894049347,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981022385175939,
      "loss": 5.4259,
      "mean_token_accuracy": 0.1630531370639801,
      "num_tokens": 11475776.0,
      "step": 5520
    },
    {
      "entropy": 5.567735767364502,
      "epoch": 0.5011792452830188,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980980426077054,
      "loss": 5.4036,
      "mean_token_accuracy": 0.18170717507600784,
      "num_tokens": 11486200.0,
      "step": 5525
    },
    {
      "entropy": 5.587890720367431,
      "epoch": 0.5016328011611031,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980938420840926,
      "loss": 5.4315,
      "mean_token_accuracy": 0.1731704905629158,
      "num_tokens": 11496768.0,
      "step": 5530
    },
    {
      "entropy": 5.611096572875977,
      "epoch": 0.5020863570391872,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980896369468423,
      "loss": 5.3881,
      "mean_token_accuracy": 0.17920028865337373,
      "num_tokens": 11507096.0,
      "step": 5535
    },
    {
      "entropy": 5.5796503067016605,
      "epoch": 0.5025399129172714,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980854271960415,
      "loss": 5.4389,
      "mean_token_accuracy": 0.1763727992773056,
      "num_tokens": 11517301.0,
      "step": 5540
    },
    {
      "entropy": 5.62597918510437,
      "epoch": 0.5029934687953556,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980812128317772,
      "loss": 5.5175,
      "mean_token_accuracy": 0.17588469237089158,
      "num_tokens": 11528616.0,
      "step": 5545
    },
    {
      "entropy": 5.653271436691284,
      "epoch": 0.5034470246734397,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980769938541364,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18069574236869812,
      "num_tokens": 11539344.0,
      "step": 5550
    },
    {
      "entropy": 5.5967125415802,
      "epoch": 0.503900580551524,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980727702632066,
      "loss": 5.4819,
      "mean_token_accuracy": 0.17302369624376296,
      "num_tokens": 11549511.0,
      "step": 5555
    },
    {
      "entropy": 5.592402076721191,
      "epoch": 0.5043541364296081,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980685420590752,
      "loss": 5.4184,
      "mean_token_accuracy": 0.17681699395179748,
      "num_tokens": 11559425.0,
      "step": 5560
    },
    {
      "entropy": 5.674917411804199,
      "epoch": 0.5048076923076923,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980643092418294,
      "loss": 5.5278,
      "mean_token_accuracy": 0.16729987114667894,
      "num_tokens": 11570406.0,
      "step": 5565
    },
    {
      "entropy": 5.604274845123291,
      "epoch": 0.5052612481857764,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980600718115568,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17531515061855316,
      "num_tokens": 11580509.0,
      "step": 5570
    },
    {
      "entropy": 5.593520736694336,
      "epoch": 0.5057148040638607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980558297683452,
      "loss": 5.431,
      "mean_token_accuracy": 0.1766576960682869,
      "num_tokens": 11590860.0,
      "step": 5575
    },
    {
      "entropy": 5.602160930633545,
      "epoch": 0.5061683599419449,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980515831122822,
      "loss": 5.4209,
      "mean_token_accuracy": 0.17564485520124434,
      "num_tokens": 11601954.0,
      "step": 5580
    },
    {
      "entropy": 5.609522008895874,
      "epoch": 0.506621915820029,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980473318434555,
      "loss": 5.4391,
      "mean_token_accuracy": 0.17501976639032363,
      "num_tokens": 11612581.0,
      "step": 5585
    },
    {
      "entropy": 5.5128148078918455,
      "epoch": 0.5070754716981132,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004980430759619532,
      "loss": 5.3458,
      "mean_token_accuracy": 0.17789707481861114,
      "num_tokens": 11622599.0,
      "step": 5590
    },
    {
      "entropy": 5.4920727729797365,
      "epoch": 0.5075290275761973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980388154678633,
      "loss": 5.3789,
      "mean_token_accuracy": 0.17600786685943604,
      "num_tokens": 11633611.0,
      "step": 5595
    },
    {
      "entropy": 5.540155076980591,
      "epoch": 0.5079825834542816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004980345503612738,
      "loss": 5.3483,
      "mean_token_accuracy": 0.17380252927541734,
      "num_tokens": 11643159.0,
      "step": 5600
    },
    {
      "entropy": 5.554949855804443,
      "epoch": 0.5084361393323658,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980302806422729,
      "loss": 5.3301,
      "mean_token_accuracy": 0.18010354191064834,
      "num_tokens": 11653174.0,
      "step": 5605
    },
    {
      "entropy": 5.570931339263916,
      "epoch": 0.5088896952104499,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980260063109491,
      "loss": 5.3818,
      "mean_token_accuracy": 0.17837395519018173,
      "num_tokens": 11662621.0,
      "step": 5610
    },
    {
      "entropy": 5.487890195846558,
      "epoch": 0.5093432510885341,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980217273673906,
      "loss": 5.3409,
      "mean_token_accuracy": 0.1755935877561569,
      "num_tokens": 11673538.0,
      "step": 5615
    },
    {
      "entropy": 5.584263563156128,
      "epoch": 0.5097968069666183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980174438116859,
      "loss": 5.355,
      "mean_token_accuracy": 0.18188655823469163,
      "num_tokens": 11683203.0,
      "step": 5620
    },
    {
      "entropy": 5.511223268508911,
      "epoch": 0.5102503628447025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980131556439236,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18352581113576888,
      "num_tokens": 11693313.0,
      "step": 5625
    },
    {
      "entropy": 5.590123558044434,
      "epoch": 0.5107039187227866,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980088628641925,
      "loss": 5.4414,
      "mean_token_accuracy": 0.17572334557771682,
      "num_tokens": 11704194.0,
      "step": 5630
    },
    {
      "entropy": 5.572856807708741,
      "epoch": 0.5111574746008708,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980045654725812,
      "loss": 5.3004,
      "mean_token_accuracy": 0.18134156614542007,
      "num_tokens": 11715109.0,
      "step": 5635
    },
    {
      "entropy": 5.505990886688233,
      "epoch": 0.511611030478955,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980002634691787,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18385408073663712,
      "num_tokens": 11724961.0,
      "step": 5640
    },
    {
      "entropy": 5.480311536788941,
      "epoch": 0.5120645863570392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497995956854074,
      "loss": 5.3565,
      "mean_token_accuracy": 0.1773085355758667,
      "num_tokens": 11735334.0,
      "step": 5645
    },
    {
      "entropy": 5.5139413356781,
      "epoch": 0.5125181422351234,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979916456273561,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17331959158182145,
      "num_tokens": 11746858.0,
      "step": 5650
    },
    {
      "entropy": 5.6394970417022705,
      "epoch": 0.5129716981132075,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979873297891141,
      "loss": 5.4643,
      "mean_token_accuracy": 0.17096892297267913,
      "num_tokens": 11757819.0,
      "step": 5655
    },
    {
      "entropy": 5.578557634353638,
      "epoch": 0.5134252539912917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979830093394372,
      "loss": 5.3963,
      "mean_token_accuracy": 0.17578381448984146,
      "num_tokens": 11767470.0,
      "step": 5660
    },
    {
      "entropy": 5.511473894119263,
      "epoch": 0.513878809869376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979786842784149,
      "loss": 5.3225,
      "mean_token_accuracy": 0.18720675110816956,
      "num_tokens": 11777060.0,
      "step": 5665
    },
    {
      "entropy": 5.539907693862915,
      "epoch": 0.5143323657474601,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979743546061366,
      "loss": 5.3779,
      "mean_token_accuracy": 0.179094859957695,
      "num_tokens": 11787176.0,
      "step": 5670
    },
    {
      "entropy": 5.490522623062134,
      "epoch": 0.5147859216255443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979700203226918,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1881948873400688,
      "num_tokens": 11796419.0,
      "step": 5675
    },
    {
      "entropy": 5.593269300460816,
      "epoch": 0.5152394775036284,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979656814281702,
      "loss": 5.4765,
      "mean_token_accuracy": 0.1741935670375824,
      "num_tokens": 11807279.0,
      "step": 5680
    },
    {
      "entropy": 5.544387626647949,
      "epoch": 0.5156930333817126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979613379226615,
      "loss": 5.3819,
      "mean_token_accuracy": 0.1720898911356926,
      "num_tokens": 11817669.0,
      "step": 5685
    },
    {
      "entropy": 5.501872730255127,
      "epoch": 0.5161465892597968,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979569898062554,
      "loss": 5.3753,
      "mean_token_accuracy": 0.18298592865467073,
      "num_tokens": 11827247.0,
      "step": 5690
    },
    {
      "entropy": 5.57254548072815,
      "epoch": 0.516600145137881,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497952637079042,
      "loss": 5.362,
      "mean_token_accuracy": 0.17560323923826218,
      "num_tokens": 11838201.0,
      "step": 5695
    },
    {
      "entropy": 5.53571834564209,
      "epoch": 0.5170537010159652,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979482797411114,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1851049169898033,
      "num_tokens": 11847561.0,
      "step": 5700
    },
    {
      "entropy": 5.582940626144409,
      "epoch": 0.5175072568940493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979439177925534,
      "loss": 5.4319,
      "mean_token_accuracy": 0.1830618754029274,
      "num_tokens": 11858272.0,
      "step": 5705
    },
    {
      "entropy": 5.607063817977905,
      "epoch": 0.5179608127721336,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979395512334584,
      "loss": 5.4662,
      "mean_token_accuracy": 0.17159156501293182,
      "num_tokens": 11869660.0,
      "step": 5710
    },
    {
      "entropy": 5.618724775314331,
      "epoch": 0.5184143686502177,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979351800639166,
      "loss": 5.4987,
      "mean_token_accuracy": 0.171364863216877,
      "num_tokens": 11880239.0,
      "step": 5715
    },
    {
      "entropy": 5.676260328292846,
      "epoch": 0.5188679245283019,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979308042840186,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1713451027870178,
      "num_tokens": 11890630.0,
      "step": 5720
    },
    {
      "entropy": 5.644466686248779,
      "epoch": 0.519321480406386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979264238938547,
      "loss": 5.4871,
      "mean_token_accuracy": 0.16992147564888,
      "num_tokens": 11900948.0,
      "step": 5725
    },
    {
      "entropy": 5.512283897399902,
      "epoch": 0.5197750362844702,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979220388935155,
      "loss": 5.3824,
      "mean_token_accuracy": 0.17406109869480133,
      "num_tokens": 11912528.0,
      "step": 5730
    },
    {
      "entropy": 5.519187307357788,
      "epoch": 0.5202285921625545,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979176492830918,
      "loss": 5.3426,
      "mean_token_accuracy": 0.17976761609315872,
      "num_tokens": 11922211.0,
      "step": 5735
    },
    {
      "entropy": 5.571788311004639,
      "epoch": 0.5206821480406386,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979132550626742,
      "loss": 5.377,
      "mean_token_accuracy": 0.17560189813375474,
      "num_tokens": 11932520.0,
      "step": 5740
    },
    {
      "entropy": 5.545016241073609,
      "epoch": 0.5211357039187228,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979088562323538,
      "loss": 5.293,
      "mean_token_accuracy": 0.18467068821191787,
      "num_tokens": 11943312.0,
      "step": 5745
    },
    {
      "entropy": 5.4827127933502195,
      "epoch": 0.5215892597968069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979044527922214,
      "loss": 5.4606,
      "mean_token_accuracy": 0.17181172966957092,
      "num_tokens": 11954455.0,
      "step": 5750
    },
    {
      "entropy": 5.57046070098877,
      "epoch": 0.5220428156748912,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979000447423681,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1794617787003517,
      "num_tokens": 11964650.0,
      "step": 5755
    },
    {
      "entropy": 5.524151277542114,
      "epoch": 0.5224963715529753,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978956320828851,
      "loss": 5.3791,
      "mean_token_accuracy": 0.17838129699230193,
      "num_tokens": 11975134.0,
      "step": 5760
    },
    {
      "entropy": 5.5696474552154545,
      "epoch": 0.5229499274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978912148138637,
      "loss": 5.4274,
      "mean_token_accuracy": 0.17480608224868774,
      "num_tokens": 11986486.0,
      "step": 5765
    },
    {
      "entropy": 5.633633422851562,
      "epoch": 0.5234034833091437,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497886792935395,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1734721377491951,
      "num_tokens": 11998045.0,
      "step": 5770
    },
    {
      "entropy": 5.634159660339355,
      "epoch": 0.5238570391872278,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978823664475707,
      "loss": 5.4267,
      "mean_token_accuracy": 0.17873497009277345,
      "num_tokens": 12008130.0,
      "step": 5775
    },
    {
      "entropy": 5.544890308380127,
      "epoch": 0.5243105950653121,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978779353504824,
      "loss": 5.4998,
      "mean_token_accuracy": 0.16745634078979493,
      "num_tokens": 12018753.0,
      "step": 5780
    },
    {
      "entropy": 5.579509639739991,
      "epoch": 0.5247641509433962,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978734996442214,
      "loss": 5.3882,
      "mean_token_accuracy": 0.17942756414413452,
      "num_tokens": 12028202.0,
      "step": 5785
    },
    {
      "entropy": 5.576986360549927,
      "epoch": 0.5252177068214804,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978690593288798,
      "loss": 5.4305,
      "mean_token_accuracy": 0.18245516568422318,
      "num_tokens": 12039652.0,
      "step": 5790
    },
    {
      "entropy": 5.620332813262939,
      "epoch": 0.5256712626995645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978646144045491,
      "loss": 5.3697,
      "mean_token_accuracy": 0.17105551362037658,
      "num_tokens": 12051269.0,
      "step": 5795
    },
    {
      "entropy": 5.445338344573974,
      "epoch": 0.5261248185776488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978601648713215,
      "loss": 5.2544,
      "mean_token_accuracy": 0.17919528782367705,
      "num_tokens": 12060647.0,
      "step": 5800
    },
    {
      "entropy": 5.515891742706299,
      "epoch": 0.526578374455733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978557107292889,
      "loss": 5.403,
      "mean_token_accuracy": 0.1814267262816429,
      "num_tokens": 12071669.0,
      "step": 5805
    },
    {
      "entropy": 5.515309429168701,
      "epoch": 0.5270319303338171,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978512519785433,
      "loss": 5.2696,
      "mean_token_accuracy": 0.18379224985837936,
      "num_tokens": 12081971.0,
      "step": 5810
    },
    {
      "entropy": 5.512398910522461,
      "epoch": 0.5274854862119013,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497846788619177,
      "loss": 5.4025,
      "mean_token_accuracy": 0.18137486279010773,
      "num_tokens": 12092465.0,
      "step": 5815
    },
    {
      "entropy": 5.5811525821685795,
      "epoch": 0.5279390420899854,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978423206512825,
      "loss": 5.351,
      "mean_token_accuracy": 0.1765703558921814,
      "num_tokens": 12101752.0,
      "step": 5820
    },
    {
      "entropy": 5.611680221557617,
      "epoch": 0.5283925979680697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497837848074952,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1774984195828438,
      "num_tokens": 12111805.0,
      "step": 5825
    },
    {
      "entropy": 5.605977153778076,
      "epoch": 0.5288461538461539,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497833370890278,
      "loss": 5.4652,
      "mean_token_accuracy": 0.17292639762163162,
      "num_tokens": 12122511.0,
      "step": 5830
    },
    {
      "entropy": 5.5387050151824955,
      "epoch": 0.529299709724238,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978288890973529,
      "loss": 5.3391,
      "mean_token_accuracy": 0.17948324829339982,
      "num_tokens": 12132510.0,
      "step": 5835
    },
    {
      "entropy": 5.529839754104614,
      "epoch": 0.5297532656023222,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978244026962698,
      "loss": 5.3596,
      "mean_token_accuracy": 0.17620361894369124,
      "num_tokens": 12142419.0,
      "step": 5840
    },
    {
      "entropy": 5.548328828811646,
      "epoch": 0.5302068214804064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978199116871212,
      "loss": 5.3942,
      "mean_token_accuracy": 0.17589981257915496,
      "num_tokens": 12152759.0,
      "step": 5845
    },
    {
      "entropy": 5.519029521942139,
      "epoch": 0.5306603773584906,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049781541607,
      "loss": 5.352,
      "mean_token_accuracy": 0.17770048826932908,
      "num_tokens": 12162758.0,
      "step": 5850
    },
    {
      "entropy": 5.528452920913696,
      "epoch": 0.5311139332365747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978109158449993,
      "loss": 5.4023,
      "mean_token_accuracy": 0.17557057440280915,
      "num_tokens": 12172664.0,
      "step": 5855
    },
    {
      "entropy": 5.534512901306153,
      "epoch": 0.5315674891146589,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978064110122121,
      "loss": 5.3733,
      "mean_token_accuracy": 0.17872561067342757,
      "num_tokens": 12182028.0,
      "step": 5860
    },
    {
      "entropy": 5.5767758846282955,
      "epoch": 0.532021044992743,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978019015717314,
      "loss": 5.4254,
      "mean_token_accuracy": 0.17671894729137422,
      "num_tokens": 12192168.0,
      "step": 5865
    },
    {
      "entropy": 5.480035495758057,
      "epoch": 0.5324746008708273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977973875236508,
      "loss": 5.3214,
      "mean_token_accuracy": 0.17603713423013687,
      "num_tokens": 12202889.0,
      "step": 5870
    },
    {
      "entropy": 5.484272623062134,
      "epoch": 0.5329281567489115,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977928688680634,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1813318744301796,
      "num_tokens": 12212514.0,
      "step": 5875
    },
    {
      "entropy": 5.513952398300171,
      "epoch": 0.5333817126269956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977883456050626,
      "loss": 5.2224,
      "mean_token_accuracy": 0.18456854969263076,
      "num_tokens": 12222353.0,
      "step": 5880
    },
    {
      "entropy": 5.614924669265747,
      "epoch": 0.5338352685050798,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497783817734742,
      "loss": 5.5128,
      "mean_token_accuracy": 0.17102179080247878,
      "num_tokens": 12232845.0,
      "step": 5885
    },
    {
      "entropy": 5.568334293365479,
      "epoch": 0.534288824383164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977792852571954,
      "loss": 5.4261,
      "mean_token_accuracy": 0.17077252715826036,
      "num_tokens": 12242950.0,
      "step": 5890
    },
    {
      "entropy": 5.563507461547852,
      "epoch": 0.5347423802612482,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977747481725164,
      "loss": 5.406,
      "mean_token_accuracy": 0.17852654159069062,
      "num_tokens": 12253132.0,
      "step": 5895
    },
    {
      "entropy": 5.545976257324218,
      "epoch": 0.5351959361393324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977702064807987,
      "loss": 5.3649,
      "mean_token_accuracy": 0.17776333838701247,
      "num_tokens": 12263214.0,
      "step": 5900
    },
    {
      "entropy": 5.4558085918426515,
      "epoch": 0.5356494920174165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977656601821366,
      "loss": 5.2902,
      "mean_token_accuracy": 0.18908421099185943,
      "num_tokens": 12273250.0,
      "step": 5905
    },
    {
      "entropy": 5.684365606307983,
      "epoch": 0.5361030478955007,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977611092766238,
      "loss": 5.4839,
      "mean_token_accuracy": 0.18275137096643448,
      "num_tokens": 12283767.0,
      "step": 5910
    },
    {
      "entropy": 5.496643781661987,
      "epoch": 0.5365566037735849,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977565537643544,
      "loss": 5.3221,
      "mean_token_accuracy": 0.17724623829126357,
      "num_tokens": 12294612.0,
      "step": 5915
    },
    {
      "entropy": 5.497534513473511,
      "epoch": 0.5370101596516691,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977519936454227,
      "loss": 5.4078,
      "mean_token_accuracy": 0.17884652316570282,
      "num_tokens": 12305614.0,
      "step": 5920
    },
    {
      "entropy": 5.499866771697998,
      "epoch": 0.5374637155297532,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977474289199231,
      "loss": 5.2892,
      "mean_token_accuracy": 0.17646802216768265,
      "num_tokens": 12316152.0,
      "step": 5925
    },
    {
      "entropy": 5.460436773300171,
      "epoch": 0.5379172714078374,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977428595879499,
      "loss": 5.36,
      "mean_token_accuracy": 0.18156999796628953,
      "num_tokens": 12325940.0,
      "step": 5930
    },
    {
      "entropy": 5.518941020965576,
      "epoch": 0.5383708272859217,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004977382856495975,
      "loss": 5.349,
      "mean_token_accuracy": 0.18231007754802703,
      "num_tokens": 12335482.0,
      "step": 5935
    },
    {
      "entropy": 5.581254148483277,
      "epoch": 0.5388243831640058,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977337071049606,
      "loss": 5.436,
      "mean_token_accuracy": 0.17495866268873214,
      "num_tokens": 12346386.0,
      "step": 5940
    },
    {
      "entropy": 5.62946310043335,
      "epoch": 0.53927793904209,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497729123954134,
      "loss": 5.4003,
      "mean_token_accuracy": 0.17766263484954833,
      "num_tokens": 12356282.0,
      "step": 5945
    },
    {
      "entropy": 5.5623939037323,
      "epoch": 0.5397314949201741,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977245361972123,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1799081191420555,
      "num_tokens": 12366994.0,
      "step": 5950
    },
    {
      "entropy": 5.515581130981445,
      "epoch": 0.5401850507982583,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977199438342904,
      "loss": 5.2576,
      "mean_token_accuracy": 0.18349101692438125,
      "num_tokens": 12377524.0,
      "step": 5955
    },
    {
      "entropy": 5.505928182601929,
      "epoch": 0.5406386066763426,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977153468654634,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1794543370604515,
      "num_tokens": 12387593.0,
      "step": 5960
    },
    {
      "entropy": 5.605174350738525,
      "epoch": 0.5410921625544267,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977107452908262,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1782233878970146,
      "num_tokens": 12398732.0,
      "step": 5965
    },
    {
      "entropy": 5.553425455093384,
      "epoch": 0.5415457184325109,
      "grad_norm": 1.0,
      "learning_rate": 0.000497706139110474,
      "loss": 5.393,
      "mean_token_accuracy": 0.18314189165830613,
      "num_tokens": 12409821.0,
      "step": 5970
    },
    {
      "entropy": 5.556648111343383,
      "epoch": 0.541999274310595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977015283245021,
      "loss": 5.4116,
      "mean_token_accuracy": 0.17372148483991623,
      "num_tokens": 12420044.0,
      "step": 5975
    },
    {
      "entropy": 5.4501725196838375,
      "epoch": 0.5424528301886793,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976969129330059,
      "loss": 5.32,
      "mean_token_accuracy": 0.1746181458234787,
      "num_tokens": 12430980.0,
      "step": 5980
    },
    {
      "entropy": 5.516680097579956,
      "epoch": 0.5429063860667634,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976922929360808,
      "loss": 5.3777,
      "mean_token_accuracy": 0.1816636934876442,
      "num_tokens": 12441999.0,
      "step": 5985
    },
    {
      "entropy": 5.596748208999633,
      "epoch": 0.5433599419448476,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976876683338222,
      "loss": 5.4189,
      "mean_token_accuracy": 0.17574200332164763,
      "num_tokens": 12452837.0,
      "step": 5990
    },
    {
      "entropy": 5.547367906570434,
      "epoch": 0.5438134978229318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497683039126326,
      "loss": 5.3353,
      "mean_token_accuracy": 0.18248194307088852,
      "num_tokens": 12462142.0,
      "step": 5995
    },
    {
      "entropy": 5.537907600402832,
      "epoch": 0.5442670537010159,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976784053136878,
      "loss": 5.3846,
      "mean_token_accuracy": 0.17923251092433928,
      "num_tokens": 12472756.0,
      "step": 6000
    },
    {
      "epoch": 0.5442670537010159,
      "eval_entropy": 5.300756282288564,
      "eval_loss": 5.398468971252441,
      "eval_mean_token_accuracy": 0.1842530439409735,
      "eval_num_tokens": 12472756.0,
      "eval_runtime": 20.7729,
      "eval_samples_per_second": 1702.214,
      "eval_steps_per_second": 212.777,
      "step": 6000
    },
    {
      "entropy": 5.589378356933594,
      "epoch": 0.5447206095791002,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976737668960034,
      "loss": 5.5023,
      "mean_token_accuracy": 0.16921191662549973,
      "num_tokens": 12483115.0,
      "step": 6005
    },
    {
      "entropy": 5.606891059875489,
      "epoch": 0.5451741654571843,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976691238733688,
      "loss": 5.3434,
      "mean_token_accuracy": 0.17560307085514068,
      "num_tokens": 12493280.0,
      "step": 6010
    },
    {
      "entropy": 5.555288028717041,
      "epoch": 0.5456277213352685,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049766447624588,
      "loss": 5.4899,
      "mean_token_accuracy": 0.1748093470931053,
      "num_tokens": 12503665.0,
      "step": 6015
    },
    {
      "entropy": 5.522700214385987,
      "epoch": 0.5460812772133526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497659824013633,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17242820858955382,
      "num_tokens": 12513381.0,
      "step": 6020
    },
    {
      "entropy": 5.633601522445678,
      "epoch": 0.5465348330914369,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976551671767241,
      "loss": 5.4278,
      "mean_token_accuracy": 0.1727713316679001,
      "num_tokens": 12524121.0,
      "step": 6025
    },
    {
      "entropy": 5.546627807617187,
      "epoch": 0.5469883889695211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976505057352496,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17922226190567017,
      "num_tokens": 12533972.0,
      "step": 6030
    },
    {
      "entropy": 5.598764562606812,
      "epoch": 0.5474419448476052,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976458396893059,
      "loss": 5.5091,
      "mean_token_accuracy": 0.16582815796136857,
      "num_tokens": 12544110.0,
      "step": 6035
    },
    {
      "entropy": 5.587916040420533,
      "epoch": 0.5478955007256894,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976411690389896,
      "loss": 5.3301,
      "mean_token_accuracy": 0.17198276072740554,
      "num_tokens": 12554273.0,
      "step": 6040
    },
    {
      "entropy": 5.588861322402954,
      "epoch": 0.5483490566037735,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497636493784397,
      "loss": 5.4118,
      "mean_token_accuracy": 0.1747738778591156,
      "num_tokens": 12564633.0,
      "step": 6045
    },
    {
      "entropy": 5.5353899002075195,
      "epoch": 0.5488026124818578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976318139256251,
      "loss": 5.3269,
      "mean_token_accuracy": 0.18243416100740434,
      "num_tokens": 12573790.0,
      "step": 6050
    },
    {
      "entropy": 5.546478748321533,
      "epoch": 0.549256168359942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976271294627706,
      "loss": 5.4166,
      "mean_token_accuracy": 0.17173689603805542,
      "num_tokens": 12584265.0,
      "step": 6055
    },
    {
      "entropy": 5.532885789871216,
      "epoch": 0.5497097242380261,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976224403959303,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17905397415161134,
      "num_tokens": 12594413.0,
      "step": 6060
    },
    {
      "entropy": 5.52966341972351,
      "epoch": 0.5501632801161103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976177467252012,
      "loss": 5.3433,
      "mean_token_accuracy": 0.18766953349113463,
      "num_tokens": 12604637.0,
      "step": 6065
    },
    {
      "entropy": 5.580493021011352,
      "epoch": 0.5506168359941945,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976130484506804,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18023614436388016,
      "num_tokens": 12614873.0,
      "step": 6070
    },
    {
      "entropy": 5.555962562561035,
      "epoch": 0.5510703918722787,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976083455724649,
      "loss": 5.3565,
      "mean_token_accuracy": 0.17727595716714858,
      "num_tokens": 12626182.0,
      "step": 6075
    },
    {
      "entropy": 5.501303195953369,
      "epoch": 0.5515239477503628,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976036380906523,
      "loss": 5.2818,
      "mean_token_accuracy": 0.18620312511920928,
      "num_tokens": 12635418.0,
      "step": 6080
    },
    {
      "entropy": 5.5395348072052,
      "epoch": 0.551977503628447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975989260053395,
      "loss": 5.4414,
      "mean_token_accuracy": 0.16959528923034667,
      "num_tokens": 12645942.0,
      "step": 6085
    },
    {
      "entropy": 5.557878732681274,
      "epoch": 0.5524310595065312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975942093166242,
      "loss": 5.3681,
      "mean_token_accuracy": 0.17604986876249312,
      "num_tokens": 12656460.0,
      "step": 6090
    },
    {
      "entropy": 5.562954711914062,
      "epoch": 0.5528846153846154,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975894880246041,
      "loss": 5.3042,
      "mean_token_accuracy": 0.17875290811061859,
      "num_tokens": 12666916.0,
      "step": 6095
    },
    {
      "entropy": 5.491450262069702,
      "epoch": 0.5533381712626996,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975847621293766,
      "loss": 5.3514,
      "mean_token_accuracy": 0.18175258487462997,
      "num_tokens": 12677562.0,
      "step": 6100
    },
    {
      "entropy": 5.489293241500855,
      "epoch": 0.5537917271407837,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975800316310394,
      "loss": 5.3207,
      "mean_token_accuracy": 0.18005092144012452,
      "num_tokens": 12687349.0,
      "step": 6105
    },
    {
      "entropy": 5.5710242748260494,
      "epoch": 0.5542452830188679,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975752965296906,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17738821506500244,
      "num_tokens": 12697850.0,
      "step": 6110
    },
    {
      "entropy": 5.493455505371093,
      "epoch": 0.5546988388969522,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975705568254278,
      "loss": 5.2428,
      "mean_token_accuracy": 0.17960745394229888,
      "num_tokens": 12708283.0,
      "step": 6115
    },
    {
      "entropy": 5.540375185012818,
      "epoch": 0.5551523947750363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975658125183493,
      "loss": 5.3476,
      "mean_token_accuracy": 0.1787947803735733,
      "num_tokens": 12718201.0,
      "step": 6120
    },
    {
      "entropy": 5.479549789428711,
      "epoch": 0.5556059506531205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497561063608553,
      "loss": 5.292,
      "mean_token_accuracy": 0.18358754068613053,
      "num_tokens": 12728860.0,
      "step": 6125
    },
    {
      "entropy": 5.527003002166748,
      "epoch": 0.5560595065312046,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975563100961372,
      "loss": 5.354,
      "mean_token_accuracy": 0.17611497938632964,
      "num_tokens": 12739007.0,
      "step": 6130
    },
    {
      "entropy": 5.524245548248291,
      "epoch": 0.5565130624092888,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975515519812002,
      "loss": 5.2078,
      "mean_token_accuracy": 0.18419712632894517,
      "num_tokens": 12749004.0,
      "step": 6135
    },
    {
      "entropy": 5.496869516372681,
      "epoch": 0.556966618287373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975467892638402,
      "loss": 5.3626,
      "mean_token_accuracy": 0.18162387162446975,
      "num_tokens": 12760085.0,
      "step": 6140
    },
    {
      "entropy": 5.532464933395386,
      "epoch": 0.5574201741654572,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975420219441561,
      "loss": 5.3144,
      "mean_token_accuracy": 0.18599002957344055,
      "num_tokens": 12770303.0,
      "step": 6145
    },
    {
      "entropy": 5.5668566703796385,
      "epoch": 0.5578737300435413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975372500222462,
      "loss": 5.3268,
      "mean_token_accuracy": 0.18526234328746796,
      "num_tokens": 12780334.0,
      "step": 6150
    },
    {
      "entropy": 5.521327352523803,
      "epoch": 0.5583272859216255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975324734982094,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17213321775197982,
      "num_tokens": 12790980.0,
      "step": 6155
    },
    {
      "entropy": 5.559855127334595,
      "epoch": 0.5587808417997098,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975276923721442,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17847093641757966,
      "num_tokens": 12801713.0,
      "step": 6160
    },
    {
      "entropy": 5.581979513168335,
      "epoch": 0.5592343976777939,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975229066441496,
      "loss": 5.4391,
      "mean_token_accuracy": 0.17114087343215942,
      "num_tokens": 12813040.0,
      "step": 6165
    },
    {
      "entropy": 5.521907711029053,
      "epoch": 0.5596879535558781,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975181163143247,
      "loss": 5.373,
      "mean_token_accuracy": 0.1722074955701828,
      "num_tokens": 12824048.0,
      "step": 6170
    },
    {
      "entropy": 5.556381750106811,
      "epoch": 0.5601415094339622,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975133213827684,
      "loss": 5.3349,
      "mean_token_accuracy": 0.1731349363923073,
      "num_tokens": 12834297.0,
      "step": 6175
    },
    {
      "entropy": 5.628904151916504,
      "epoch": 0.5605950653120464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975085218495798,
      "loss": 5.4009,
      "mean_token_accuracy": 0.17205490320920944,
      "num_tokens": 12844574.0,
      "step": 6180
    },
    {
      "entropy": 5.476094388961792,
      "epoch": 0.5610486211901307,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975037177148585,
      "loss": 5.3343,
      "mean_token_accuracy": 0.1756583869457245,
      "num_tokens": 12854670.0,
      "step": 6185
    },
    {
      "entropy": 5.502668714523315,
      "epoch": 0.5615021770682148,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974989089787035,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18209386765956878,
      "num_tokens": 12865281.0,
      "step": 6190
    },
    {
      "entropy": 5.6341406345367435,
      "epoch": 0.561955732946299,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974940956412143,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1820847824215889,
      "num_tokens": 12875222.0,
      "step": 6195
    },
    {
      "entropy": 5.65907883644104,
      "epoch": 0.5624092888243831,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974892777024905,
      "loss": 5.4245,
      "mean_token_accuracy": 0.16868771463632584,
      "num_tokens": 12885079.0,
      "step": 6200
    },
    {
      "entropy": 5.475528192520142,
      "epoch": 0.5628628447024674,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004974844551626318,
      "loss": 5.3626,
      "mean_token_accuracy": 0.18181288093328477,
      "num_tokens": 12896402.0,
      "step": 6205
    },
    {
      "entropy": 5.545027017593384,
      "epoch": 0.5633164005805515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974796280217379,
      "loss": 5.3348,
      "mean_token_accuracy": 0.1814814731478691,
      "num_tokens": 12906403.0,
      "step": 6210
    },
    {
      "entropy": 5.571792650222778,
      "epoch": 0.5637699564586357,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974747962799086,
      "loss": 5.3618,
      "mean_token_accuracy": 0.17638490051031114,
      "num_tokens": 12917061.0,
      "step": 6215
    },
    {
      "entropy": 5.523290824890137,
      "epoch": 0.5642235123367199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974699599372437,
      "loss": 5.378,
      "mean_token_accuracy": 0.18156666308641434,
      "num_tokens": 12927182.0,
      "step": 6220
    },
    {
      "entropy": 5.494165849685669,
      "epoch": 0.564677068214804,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974651189938434,
      "loss": 5.3496,
      "mean_token_accuracy": 0.1800533875823021,
      "num_tokens": 12937371.0,
      "step": 6225
    },
    {
      "entropy": 5.538199424743652,
      "epoch": 0.5651306240928883,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974602734498079,
      "loss": 5.4027,
      "mean_token_accuracy": 0.17287339568138121,
      "num_tokens": 12948569.0,
      "step": 6230
    },
    {
      "entropy": 5.569438743591308,
      "epoch": 0.5655841799709724,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497455423305237,
      "loss": 5.3436,
      "mean_token_accuracy": 0.17129313349723815,
      "num_tokens": 12958503.0,
      "step": 6235
    },
    {
      "entropy": 5.575509071350098,
      "epoch": 0.5660377358490566,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974505685602314,
      "loss": 5.368,
      "mean_token_accuracy": 0.1803085759282112,
      "num_tokens": 12968714.0,
      "step": 6240
    },
    {
      "entropy": 5.497352027893067,
      "epoch": 0.5664912917271407,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974457092148912,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17698344886302947,
      "num_tokens": 12978857.0,
      "step": 6245
    },
    {
      "entropy": 5.623220825195313,
      "epoch": 0.566944847605225,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974408452693172,
      "loss": 5.4974,
      "mean_token_accuracy": 0.17146317958831786,
      "num_tokens": 12989863.0,
      "step": 6250
    },
    {
      "entropy": 5.5716962814331055,
      "epoch": 0.5673984034833092,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974359767236099,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1790088012814522,
      "num_tokens": 13000307.0,
      "step": 6255
    },
    {
      "entropy": 5.550614261627198,
      "epoch": 0.5678519593613933,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974311035778698,
      "loss": 5.3787,
      "mean_token_accuracy": 0.1838524878025055,
      "num_tokens": 13009053.0,
      "step": 6260
    },
    {
      "entropy": 5.540517282485962,
      "epoch": 0.5683055152394775,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974262258321979,
      "loss": 5.406,
      "mean_token_accuracy": 0.17350224703550338,
      "num_tokens": 13019136.0,
      "step": 6265
    },
    {
      "entropy": 5.594241189956665,
      "epoch": 0.5687590711175616,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974213434866949,
      "loss": 5.446,
      "mean_token_accuracy": 0.1712695136666298,
      "num_tokens": 13030356.0,
      "step": 6270
    },
    {
      "entropy": 5.602076768875122,
      "epoch": 0.5692126269956459,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974164565414617,
      "loss": 5.4559,
      "mean_token_accuracy": 0.17462238520383835,
      "num_tokens": 13040477.0,
      "step": 6275
    },
    {
      "entropy": 5.55633134841919,
      "epoch": 0.56966618287373,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974115649965998,
      "loss": 5.3222,
      "mean_token_accuracy": 0.18193186223506927,
      "num_tokens": 13051755.0,
      "step": 6280
    },
    {
      "entropy": 5.541445636749268,
      "epoch": 0.5701197387518142,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974066688522098,
      "loss": 5.3493,
      "mean_token_accuracy": 0.18550432324409485,
      "num_tokens": 13061670.0,
      "step": 6285
    },
    {
      "entropy": 5.4989865779876705,
      "epoch": 0.5705732946298984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974017681083934,
      "loss": 5.3297,
      "mean_token_accuracy": 0.1777109295129776,
      "num_tokens": 13070917.0,
      "step": 6290
    },
    {
      "entropy": 5.532304143905639,
      "epoch": 0.5710268505079826,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004973968627652516,
      "loss": 5.3885,
      "mean_token_accuracy": 0.17264553159475327,
      "num_tokens": 13080577.0,
      "step": 6295
    },
    {
      "entropy": 5.580759334564209,
      "epoch": 0.5714804063860668,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497391952822886,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17470873892307281,
      "num_tokens": 13091446.0,
      "step": 6300
    },
    {
      "entropy": 5.586512184143066,
      "epoch": 0.5719339622641509,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973870382813983,
      "loss": 5.4488,
      "mean_token_accuracy": 0.17591034173965453,
      "num_tokens": 13102752.0,
      "step": 6305
    },
    {
      "entropy": 5.6238028049469,
      "epoch": 0.5723875181422351,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973821191408899,
      "loss": 5.4457,
      "mean_token_accuracy": 0.17629251033067703,
      "num_tokens": 13113944.0,
      "step": 6310
    },
    {
      "entropy": 5.575624752044678,
      "epoch": 0.5728410740203193,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973771954014626,
      "loss": 5.3887,
      "mean_token_accuracy": 0.17858594805002212,
      "num_tokens": 13124316.0,
      "step": 6315
    },
    {
      "entropy": 5.624091053009034,
      "epoch": 0.5732946298984035,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004973722670632182,
      "loss": 5.4551,
      "mean_token_accuracy": 0.1725321814417839,
      "num_tokens": 13133567.0,
      "step": 6320
    },
    {
      "entropy": 5.511906528472901,
      "epoch": 0.5737481857764877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973673341262587,
      "loss": 5.3886,
      "mean_token_accuracy": 0.17628957331180573,
      "num_tokens": 13144279.0,
      "step": 6325
    },
    {
      "entropy": 5.572101020812989,
      "epoch": 0.5742017416545718,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497362396590686,
      "loss": 5.4501,
      "mean_token_accuracy": 0.17332976907491685,
      "num_tokens": 13154463.0,
      "step": 6330
    },
    {
      "entropy": 5.607514715194702,
      "epoch": 0.574655297532656,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973574544566024,
      "loss": 5.4049,
      "mean_token_accuracy": 0.16702012121677398,
      "num_tokens": 13165565.0,
      "step": 6335
    },
    {
      "entropy": 5.5187538146972654,
      "epoch": 0.5751088534107403,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973525077241098,
      "loss": 5.3295,
      "mean_token_accuracy": 0.18137111365795136,
      "num_tokens": 13175839.0,
      "step": 6340
    },
    {
      "entropy": 5.462864255905151,
      "epoch": 0.5755624092888244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973475563933109,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1800212725996971,
      "num_tokens": 13186372.0,
      "step": 6345
    },
    {
      "entropy": 5.509666633605957,
      "epoch": 0.5760159651669086,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973426004643077,
      "loss": 5.37,
      "mean_token_accuracy": 0.18036460280418395,
      "num_tokens": 13196259.0,
      "step": 6350
    },
    {
      "entropy": 5.572441911697387,
      "epoch": 0.5764695210449927,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973376399372029,
      "loss": 5.344,
      "mean_token_accuracy": 0.1819022089242935,
      "num_tokens": 13206497.0,
      "step": 6355
    },
    {
      "entropy": 5.549651670455932,
      "epoch": 0.5769230769230769,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973326748120991,
      "loss": 5.3967,
      "mean_token_accuracy": 0.17948508709669114,
      "num_tokens": 13217188.0,
      "step": 6360
    },
    {
      "entropy": 5.524637079238891,
      "epoch": 0.5773766328011611,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973277050890988,
      "loss": 5.3873,
      "mean_token_accuracy": 0.17857213616371154,
      "num_tokens": 13226544.0,
      "step": 6365
    },
    {
      "entropy": 5.600632667541504,
      "epoch": 0.5778301886792453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973227307683049,
      "loss": 5.4246,
      "mean_token_accuracy": 0.17547977715730667,
      "num_tokens": 13236572.0,
      "step": 6370
    },
    {
      "entropy": 5.669041633605957,
      "epoch": 0.5782837445573294,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973177518498204,
      "loss": 5.4281,
      "mean_token_accuracy": 0.17712328135967254,
      "num_tokens": 13247958.0,
      "step": 6375
    },
    {
      "entropy": 5.547800731658936,
      "epoch": 0.5787373004354136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973127683337482,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18092712610960007,
      "num_tokens": 13258255.0,
      "step": 6380
    },
    {
      "entropy": 5.465287494659424,
      "epoch": 0.5791908563134979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973077802201911,
      "loss": 5.3557,
      "mean_token_accuracy": 0.17363798320293428,
      "num_tokens": 13268416.0,
      "step": 6385
    },
    {
      "entropy": 5.503123950958252,
      "epoch": 0.579644412191582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973027875092526,
      "loss": 5.3293,
      "mean_token_accuracy": 0.17840101420879365,
      "num_tokens": 13278087.0,
      "step": 6390
    },
    {
      "entropy": 5.524763393402099,
      "epoch": 0.5800979680696662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972977902010357,
      "loss": 5.2447,
      "mean_token_accuracy": 0.1844359055161476,
      "num_tokens": 13288121.0,
      "step": 6395
    },
    {
      "entropy": 5.6309037685394285,
      "epoch": 0.5805515239477503,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972927882956439,
      "loss": 5.4657,
      "mean_token_accuracy": 0.16494472771883012,
      "num_tokens": 13298620.0,
      "step": 6400
    },
    {
      "entropy": 5.517560768127441,
      "epoch": 0.5810050798258345,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972877817931807,
      "loss": 5.2753,
      "mean_token_accuracy": 0.17818261831998825,
      "num_tokens": 13308413.0,
      "step": 6405
    },
    {
      "entropy": 5.477993535995483,
      "epoch": 0.5814586357039188,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972827706937493,
      "loss": 5.3562,
      "mean_token_accuracy": 0.18004698753356935,
      "num_tokens": 13319387.0,
      "step": 6410
    },
    {
      "entropy": 5.563785362243652,
      "epoch": 0.5819121915820029,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972777549974537,
      "loss": 5.3143,
      "mean_token_accuracy": 0.17967743724584578,
      "num_tokens": 13330494.0,
      "step": 6415
    },
    {
      "entropy": 5.541955375671387,
      "epoch": 0.5823657474600871,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972727347043975,
      "loss": 5.3155,
      "mean_token_accuracy": 0.1786987692117691,
      "num_tokens": 13341009.0,
      "step": 6420
    },
    {
      "entropy": 5.490311861038208,
      "epoch": 0.5828193033381712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972677098146845,
      "loss": 5.2766,
      "mean_token_accuracy": 0.1829589769244194,
      "num_tokens": 13351269.0,
      "step": 6425
    },
    {
      "entropy": 5.4687377452850345,
      "epoch": 0.5832728592162555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972626803284185,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1788855239748955,
      "num_tokens": 13362160.0,
      "step": 6430
    },
    {
      "entropy": 5.5932941913604735,
      "epoch": 0.5837264150943396,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972576462457038,
      "loss": 5.4421,
      "mean_token_accuracy": 0.18139465898275375,
      "num_tokens": 13372497.0,
      "step": 6435
    },
    {
      "entropy": 5.603217458724975,
      "epoch": 0.5841799709724238,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972526075666442,
      "loss": 5.3849,
      "mean_token_accuracy": 0.17963762879371642,
      "num_tokens": 13381665.0,
      "step": 6440
    },
    {
      "entropy": 5.481933689117431,
      "epoch": 0.584633526850508,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972475642913442,
      "loss": 5.2862,
      "mean_token_accuracy": 0.18237100541591644,
      "num_tokens": 13392507.0,
      "step": 6445
    },
    {
      "entropy": 5.484732341766358,
      "epoch": 0.5850870827285921,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972425164199078,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18799172341823578,
      "num_tokens": 13402171.0,
      "step": 6450
    },
    {
      "entropy": 5.608668804168701,
      "epoch": 0.5855406386066764,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972374639524396,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1756341800093651,
      "num_tokens": 13412933.0,
      "step": 6455
    },
    {
      "entropy": 5.547793292999268,
      "epoch": 0.5859941944847605,
      "grad_norm": 0.859375,
      "learning_rate": 0.000497232406889044,
      "loss": 5.3367,
      "mean_token_accuracy": 0.17190575897693633,
      "num_tokens": 13423563.0,
      "step": 6460
    },
    {
      "entropy": 5.600666761398315,
      "epoch": 0.5864477503628447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972273452298255,
      "loss": 5.3459,
      "mean_token_accuracy": 0.18437232375144957,
      "num_tokens": 13433998.0,
      "step": 6465
    },
    {
      "entropy": 5.516443872451783,
      "epoch": 0.5869013062409288,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972222789748888,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18250495493412017,
      "num_tokens": 13443878.0,
      "step": 6470
    },
    {
      "entropy": 5.495645189285279,
      "epoch": 0.5873548621190131,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972172081243389,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17166997939348222,
      "num_tokens": 13456051.0,
      "step": 6475
    },
    {
      "entropy": 5.5309515476226805,
      "epoch": 0.5878084179970973,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972121326782806,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18319807052612305,
      "num_tokens": 13467276.0,
      "step": 6480
    },
    {
      "entropy": 5.55036153793335,
      "epoch": 0.5882619738751814,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972070526368186,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17851588428020476,
      "num_tokens": 13478331.0,
      "step": 6485
    },
    {
      "entropy": 5.58794755935669,
      "epoch": 0.5887155297532656,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972019680000582,
      "loss": 5.2955,
      "mean_token_accuracy": 0.17752917855978012,
      "num_tokens": 13489879.0,
      "step": 6490
    },
    {
      "entropy": 5.493243551254272,
      "epoch": 0.5891690856313497,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971968787681044,
      "loss": 5.4016,
      "mean_token_accuracy": 0.181935915350914,
      "num_tokens": 13500879.0,
      "step": 6495
    },
    {
      "entropy": 5.616210842132569,
      "epoch": 0.589622641509434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971917849410627,
      "loss": 5.3875,
      "mean_token_accuracy": 0.17739370465278625,
      "num_tokens": 13511719.0,
      "step": 6500
    },
    {
      "entropy": 5.50895209312439,
      "epoch": 0.5900761973875182,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971866865190381,
      "loss": 5.318,
      "mean_token_accuracy": 0.17822761982679367,
      "num_tokens": 13521850.0,
      "step": 6505
    },
    {
      "entropy": 5.563834190368652,
      "epoch": 0.5905297532656023,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971815835021362,
      "loss": 5.3519,
      "mean_token_accuracy": 0.1825216919183731,
      "num_tokens": 13532095.0,
      "step": 6510
    },
    {
      "entropy": 5.515305042266846,
      "epoch": 0.5909833091436865,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971764758904626,
      "loss": 5.3413,
      "mean_token_accuracy": 0.17511919140815735,
      "num_tokens": 13543133.0,
      "step": 6515
    },
    {
      "entropy": 5.525648593902588,
      "epoch": 0.5914368650217707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971713636841229,
      "loss": 5.402,
      "mean_token_accuracy": 0.17349975258111955,
      "num_tokens": 13553771.0,
      "step": 6520
    },
    {
      "entropy": 5.501309537887574,
      "epoch": 0.5918904208998549,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971662468832228,
      "loss": 5.3479,
      "mean_token_accuracy": 0.18141622096300125,
      "num_tokens": 13564548.0,
      "step": 6525
    },
    {
      "entropy": 5.475589799880981,
      "epoch": 0.592343976777939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497161125487868,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1807534858584404,
      "num_tokens": 13574567.0,
      "step": 6530
    },
    {
      "entropy": 5.496382856369019,
      "epoch": 0.5927975326560232,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971559994981645,
      "loss": 5.3686,
      "mean_token_accuracy": 0.1793726459145546,
      "num_tokens": 13585302.0,
      "step": 6535
    },
    {
      "entropy": 5.572533655166626,
      "epoch": 0.5932510885341074,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971508689142184,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17545087039470672,
      "num_tokens": 13594754.0,
      "step": 6540
    },
    {
      "entropy": 5.486157846450806,
      "epoch": 0.5937046444121916,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971457337361356,
      "loss": 5.3268,
      "mean_token_accuracy": 0.177515971660614,
      "num_tokens": 13604691.0,
      "step": 6545
    },
    {
      "entropy": 5.4501392364501955,
      "epoch": 0.5941582002902758,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971405939640225,
      "loss": 5.2214,
      "mean_token_accuracy": 0.18570286631584168,
      "num_tokens": 13613911.0,
      "step": 6550
    },
    {
      "entropy": 5.636072492599487,
      "epoch": 0.5946117561683599,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971354495979854,
      "loss": 5.4398,
      "mean_token_accuracy": 0.17701357752084732,
      "num_tokens": 13624285.0,
      "step": 6555
    },
    {
      "entropy": 5.521263599395752,
      "epoch": 0.5950653120464441,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971303006381305,
      "loss": 5.2916,
      "mean_token_accuracy": 0.18147354274988176,
      "num_tokens": 13635434.0,
      "step": 6560
    },
    {
      "entropy": 5.4838104248046875,
      "epoch": 0.5955188679245284,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971251470845643,
      "loss": 5.345,
      "mean_token_accuracy": 0.1768563508987427,
      "num_tokens": 13645582.0,
      "step": 6565
    },
    {
      "entropy": 5.5116071701049805,
      "epoch": 0.5959724238026125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971199889373935,
      "loss": 5.3055,
      "mean_token_accuracy": 0.18206693828105927,
      "num_tokens": 13655029.0,
      "step": 6570
    },
    {
      "entropy": 5.480277967453003,
      "epoch": 0.5964259796806967,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971148261967249,
      "loss": 5.294,
      "mean_token_accuracy": 0.18128707706928254,
      "num_tokens": 13665204.0,
      "step": 6575
    },
    {
      "entropy": 5.5318831443786625,
      "epoch": 0.5968795355587808,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971096588626649,
      "loss": 5.3953,
      "mean_token_accuracy": 0.17587229311466218,
      "num_tokens": 13676281.0,
      "step": 6580
    },
    {
      "entropy": 5.528276824951172,
      "epoch": 0.597333091436865,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971044869353206,
      "loss": 5.353,
      "mean_token_accuracy": 0.17965753376483917,
      "num_tokens": 13687144.0,
      "step": 6585
    },
    {
      "entropy": 5.529869556427002,
      "epoch": 0.5977866473149492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497099310414799,
      "loss": 5.3345,
      "mean_token_accuracy": 0.18199270814657212,
      "num_tokens": 13696371.0,
      "step": 6590
    },
    {
      "entropy": 5.573552417755127,
      "epoch": 0.5982402031930334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970941293012069,
      "loss": 5.2835,
      "mean_token_accuracy": 0.17838052809238433,
      "num_tokens": 13706900.0,
      "step": 6595
    },
    {
      "entropy": 5.5088542938232425,
      "epoch": 0.5986937590711175,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970889435946518,
      "loss": 5.3598,
      "mean_token_accuracy": 0.1811433807015419,
      "num_tokens": 13717874.0,
      "step": 6600
    },
    {
      "entropy": 5.550598001480102,
      "epoch": 0.5991473149492017,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970837532952406,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1788433775305748,
      "num_tokens": 13728020.0,
      "step": 6605
    },
    {
      "entropy": 5.562851047515869,
      "epoch": 0.599600870827286,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970785584030808,
      "loss": 5.372,
      "mean_token_accuracy": 0.1790483370423317,
      "num_tokens": 13738007.0,
      "step": 6610
    },
    {
      "entropy": 5.53517484664917,
      "epoch": 0.6000544267053701,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970733589182799,
      "loss": 5.3681,
      "mean_token_accuracy": 0.17741776257753372,
      "num_tokens": 13749429.0,
      "step": 6615
    },
    {
      "entropy": 5.50953049659729,
      "epoch": 0.6005079825834543,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970681548409452,
      "loss": 5.2357,
      "mean_token_accuracy": 0.18484993875026703,
      "num_tokens": 13760279.0,
      "step": 6620
    },
    {
      "entropy": 5.569452714920044,
      "epoch": 0.6009615384615384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970629461711846,
      "loss": 5.4547,
      "mean_token_accuracy": 0.1746961295604706,
      "num_tokens": 13770575.0,
      "step": 6625
    },
    {
      "entropy": 5.4972024917602536,
      "epoch": 0.6014150943396226,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970577329091057,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18609136939048768,
      "num_tokens": 13781068.0,
      "step": 6630
    },
    {
      "entropy": 5.549247598648071,
      "epoch": 0.6018686502177069,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970525150548162,
      "loss": 5.3836,
      "mean_token_accuracy": 0.18321933150291442,
      "num_tokens": 13790745.0,
      "step": 6635
    },
    {
      "entropy": 5.521773195266723,
      "epoch": 0.602322206095791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970472926084242,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1791524738073349,
      "num_tokens": 13800196.0,
      "step": 6640
    },
    {
      "entropy": 5.490709257125855,
      "epoch": 0.6027757619738752,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970420655700375,
      "loss": 5.2291,
      "mean_token_accuracy": 0.18693480491638184,
      "num_tokens": 13809954.0,
      "step": 6645
    },
    {
      "entropy": 5.512080097198487,
      "epoch": 0.6032293178519593,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004970368339397643,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1746376097202301,
      "num_tokens": 13820281.0,
      "step": 6650
    },
    {
      "entropy": 5.586239385604858,
      "epoch": 0.6036828737300436,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004970315977177129,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18440013825893403,
      "num_tokens": 13830614.0,
      "step": 6655
    },
    {
      "entropy": 5.500546884536743,
      "epoch": 0.6041364296081277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970263569039914,
      "loss": 5.2936,
      "mean_token_accuracy": 0.181098048388958,
      "num_tokens": 13840551.0,
      "step": 6660
    },
    {
      "entropy": 5.625216722488403,
      "epoch": 0.6045899854862119,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970211114987083,
      "loss": 5.478,
      "mean_token_accuracy": 0.17690049409866332,
      "num_tokens": 13851830.0,
      "step": 6665
    },
    {
      "entropy": 5.556969070434571,
      "epoch": 0.6050435413642961,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970158615019719,
      "loss": 5.286,
      "mean_token_accuracy": 0.18250710368156434,
      "num_tokens": 13862299.0,
      "step": 6670
    },
    {
      "entropy": 5.575377559661865,
      "epoch": 0.6054970972423802,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970106069138912,
      "loss": 5.4059,
      "mean_token_accuracy": 0.1716219589114189,
      "num_tokens": 13872852.0,
      "step": 6675
    },
    {
      "entropy": 5.499694538116455,
      "epoch": 0.6059506531204645,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004970053477345743,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18448759168386458,
      "num_tokens": 13883601.0,
      "step": 6680
    },
    {
      "entropy": 5.550814008712768,
      "epoch": 0.6064042089985486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970000839641304,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18434916138648988,
      "num_tokens": 13894246.0,
      "step": 6685
    },
    {
      "entropy": 5.4789330005645756,
      "epoch": 0.6068577648766328,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969948156026681,
      "loss": 5.3844,
      "mean_token_accuracy": 0.17830681800842285,
      "num_tokens": 13904823.0,
      "step": 6690
    },
    {
      "entropy": 5.598898124694824,
      "epoch": 0.6073113207547169,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969895426502965,
      "loss": 5.3862,
      "mean_token_accuracy": 0.17962592840194702,
      "num_tokens": 13915338.0,
      "step": 6695
    },
    {
      "entropy": 5.507212829589844,
      "epoch": 0.6077648766328012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969842651071246,
      "loss": 5.3413,
      "mean_token_accuracy": 0.17749970853328706,
      "num_tokens": 13925838.0,
      "step": 6700
    },
    {
      "entropy": 5.5650975704193115,
      "epoch": 0.6082184325108854,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969789829732615,
      "loss": 5.3388,
      "mean_token_accuracy": 0.17786233574151994,
      "num_tokens": 13936341.0,
      "step": 6705
    },
    {
      "entropy": 5.559619951248169,
      "epoch": 0.6086719883889695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969736962488164,
      "loss": 5.3525,
      "mean_token_accuracy": 0.1801693171262741,
      "num_tokens": 13945438.0,
      "step": 6710
    },
    {
      "entropy": 5.509589719772339,
      "epoch": 0.6091255442670537,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969684049338987,
      "loss": 5.2386,
      "mean_token_accuracy": 0.180333910882473,
      "num_tokens": 13956029.0,
      "step": 6715
    },
    {
      "entropy": 5.5770862102508545,
      "epoch": 0.6095791001451378,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496963109028618,
      "loss": 5.3515,
      "mean_token_accuracy": 0.17293535470962523,
      "num_tokens": 13966498.0,
      "step": 6720
    },
    {
      "entropy": 5.5024824142456055,
      "epoch": 0.6100326560232221,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969578085330834,
      "loss": 5.2984,
      "mean_token_accuracy": 0.1841035008430481,
      "num_tokens": 13976997.0,
      "step": 6725
    },
    {
      "entropy": 5.555294322967529,
      "epoch": 0.6104862119013063,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969525034474049,
      "loss": 5.3116,
      "mean_token_accuracy": 0.18184979408979415,
      "num_tokens": 13986664.0,
      "step": 6730
    },
    {
      "entropy": 5.456352567672729,
      "epoch": 0.6109397677793904,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496947193771692,
      "loss": 5.3061,
      "mean_token_accuracy": 0.18491313904523848,
      "num_tokens": 13997828.0,
      "step": 6735
    },
    {
      "entropy": 5.480406856536865,
      "epoch": 0.6113933236574746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969418795060548,
      "loss": 5.2586,
      "mean_token_accuracy": 0.18284006267786027,
      "num_tokens": 14008293.0,
      "step": 6740
    },
    {
      "entropy": 5.478730154037476,
      "epoch": 0.6118468795355588,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969365606506027,
      "loss": 5.2436,
      "mean_token_accuracy": 0.186459219455719,
      "num_tokens": 14018031.0,
      "step": 6745
    },
    {
      "entropy": 5.46801495552063,
      "epoch": 0.612300435413643,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969312372054461,
      "loss": 5.2332,
      "mean_token_accuracy": 0.18784987032413483,
      "num_tokens": 14028013.0,
      "step": 6750
    },
    {
      "entropy": 5.491974496841431,
      "epoch": 0.6127539912917271,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969259091706948,
      "loss": 5.3637,
      "mean_token_accuracy": 0.1775025248527527,
      "num_tokens": 14039787.0,
      "step": 6755
    },
    {
      "entropy": 5.5373763084411625,
      "epoch": 0.6132075471698113,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969205765464593,
      "loss": 5.2606,
      "mean_token_accuracy": 0.18175734281539918,
      "num_tokens": 14049223.0,
      "step": 6760
    },
    {
      "entropy": 5.5544661521911625,
      "epoch": 0.6136611030478955,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969152393328497,
      "loss": 5.2985,
      "mean_token_accuracy": 0.18559947311878205,
      "num_tokens": 14059004.0,
      "step": 6765
    },
    {
      "entropy": 5.536505842208863,
      "epoch": 0.6141146589259797,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969098975299764,
      "loss": 5.4062,
      "mean_token_accuracy": 0.17162830382585526,
      "num_tokens": 14069615.0,
      "step": 6770
    },
    {
      "entropy": 5.525615072250366,
      "epoch": 0.6145682148040639,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969045511379499,
      "loss": 5.3252,
      "mean_token_accuracy": 0.18242162317037583,
      "num_tokens": 14078886.0,
      "step": 6775
    },
    {
      "entropy": 5.610492515563965,
      "epoch": 0.615021770682148,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968992001568806,
      "loss": 5.4129,
      "mean_token_accuracy": 0.17199260890483856,
      "num_tokens": 14089021.0,
      "step": 6780
    },
    {
      "entropy": 5.4925415992736815,
      "epoch": 0.6154753265602322,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968938445868795,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1890949234366417,
      "num_tokens": 14098316.0,
      "step": 6785
    },
    {
      "entropy": 5.463628149032592,
      "epoch": 0.6159288824383164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968884844280569,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1768708735704422,
      "num_tokens": 14108874.0,
      "step": 6790
    },
    {
      "entropy": 5.520146799087525,
      "epoch": 0.6163824383164006,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968831196805242,
      "loss": 5.3576,
      "mean_token_accuracy": 0.1788746789097786,
      "num_tokens": 14119613.0,
      "step": 6795
    },
    {
      "entropy": 5.559129047393799,
      "epoch": 0.6168359941944848,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968777503443918,
      "loss": 5.3215,
      "mean_token_accuracy": 0.17656567692756653,
      "num_tokens": 14129085.0,
      "step": 6800
    },
    {
      "entropy": 5.440955114364624,
      "epoch": 0.6172895500725689,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968723764197711,
      "loss": 5.2268,
      "mean_token_accuracy": 0.1918613076210022,
      "num_tokens": 14138461.0,
      "step": 6805
    },
    {
      "entropy": 5.47119174003601,
      "epoch": 0.6177431059506531,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496866997906773,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18615336567163468,
      "num_tokens": 14149126.0,
      "step": 6810
    },
    {
      "entropy": 5.4483057975769045,
      "epoch": 0.6181966618287373,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004968616148055089,
      "loss": 5.2881,
      "mean_token_accuracy": 0.18620866686105728,
      "num_tokens": 14160068.0,
      "step": 6815
    },
    {
      "entropy": 5.509595775604248,
      "epoch": 0.6186502177068215,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968562271160901,
      "loss": 5.3862,
      "mean_token_accuracy": 0.18133724480867386,
      "num_tokens": 14170288.0,
      "step": 6820
    },
    {
      "entropy": 5.591276216506958,
      "epoch": 0.6191037735849056,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496850834838628,
      "loss": 5.3755,
      "mean_token_accuracy": 0.17940264344215393,
      "num_tokens": 14181231.0,
      "step": 6825
    },
    {
      "entropy": 5.559400033950806,
      "epoch": 0.6195573294629898,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968454379732342,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1832197815179825,
      "num_tokens": 14191715.0,
      "step": 6830
    },
    {
      "entropy": 5.463568258285522,
      "epoch": 0.6200108853410741,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049684003652002,
      "loss": 5.2953,
      "mean_token_accuracy": 0.17444696128368378,
      "num_tokens": 14203416.0,
      "step": 6835
    },
    {
      "entropy": 5.4512610912323,
      "epoch": 0.6204644412191582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968346304790973,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19206695556640624,
      "num_tokens": 14213658.0,
      "step": 6840
    },
    {
      "entropy": 5.544649982452393,
      "epoch": 0.6209179970972424,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968292198505781,
      "loss": 5.4084,
      "mean_token_accuracy": 0.17560362964868545,
      "num_tokens": 14223956.0,
      "step": 6845
    },
    {
      "entropy": 5.562683057785034,
      "epoch": 0.6213715529753265,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496823804634574,
      "loss": 5.3982,
      "mean_token_accuracy": 0.17950718104839325,
      "num_tokens": 14236801.0,
      "step": 6850
    },
    {
      "entropy": 5.515814352035522,
      "epoch": 0.6218251088534107,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968183848311972,
      "loss": 5.2099,
      "mean_token_accuracy": 0.18720452338457108,
      "num_tokens": 14247382.0,
      "step": 6855
    },
    {
      "entropy": 5.555305004119873,
      "epoch": 0.622278664731495,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968129604405596,
      "loss": 5.2996,
      "mean_token_accuracy": 0.1855861246585846,
      "num_tokens": 14258627.0,
      "step": 6860
    },
    {
      "entropy": 5.4478870868682865,
      "epoch": 0.6227322206095791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968075314627734,
      "loss": 5.3,
      "mean_token_accuracy": 0.18473073691129685,
      "num_tokens": 14269013.0,
      "step": 6865
    },
    {
      "entropy": 5.474765825271606,
      "epoch": 0.6231857764876633,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496802097897951,
      "loss": 5.2832,
      "mean_token_accuracy": 0.18246868699789048,
      "num_tokens": 14279620.0,
      "step": 6870
    },
    {
      "entropy": 5.5203248977661135,
      "epoch": 0.6236393323657474,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967966597462046,
      "loss": 5.2874,
      "mean_token_accuracy": 0.1900581583380699,
      "num_tokens": 14289448.0,
      "step": 6875
    },
    {
      "entropy": 5.53533444404602,
      "epoch": 0.6240928882438317,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967912170076469,
      "loss": 5.291,
      "mean_token_accuracy": 0.17730221003293992,
      "num_tokens": 14299999.0,
      "step": 6880
    },
    {
      "entropy": 5.460947799682617,
      "epoch": 0.6245464441219158,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967857696823901,
      "loss": 5.2673,
      "mean_token_accuracy": 0.1848415583372116,
      "num_tokens": 14310384.0,
      "step": 6885
    },
    {
      "entropy": 5.509242057800293,
      "epoch": 0.625,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967803177705472,
      "loss": 5.3818,
      "mean_token_accuracy": 0.17576736062765122,
      "num_tokens": 14320644.0,
      "step": 6890
    },
    {
      "entropy": 5.552798414230347,
      "epoch": 0.6254535558780842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967748612722308,
      "loss": 5.3416,
      "mean_token_accuracy": 0.1810254707932472,
      "num_tokens": 14330077.0,
      "step": 6895
    },
    {
      "entropy": 5.5750954151153564,
      "epoch": 0.6259071117561683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967694001875538,
      "loss": 5.4386,
      "mean_token_accuracy": 0.17333186119794847,
      "num_tokens": 14341549.0,
      "step": 6900
    },
    {
      "entropy": 5.564305830001831,
      "epoch": 0.6263606676342526,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496763934516629,
      "loss": 5.2874,
      "mean_token_accuracy": 0.18876885771751403,
      "num_tokens": 14352146.0,
      "step": 6905
    },
    {
      "entropy": 5.471659088134766,
      "epoch": 0.6268142235123367,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967584642595696,
      "loss": 5.325,
      "mean_token_accuracy": 0.17631787657737732,
      "num_tokens": 14362521.0,
      "step": 6910
    },
    {
      "entropy": 5.541382551193237,
      "epoch": 0.6272677793904209,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967529894164887,
      "loss": 5.3537,
      "mean_token_accuracy": 0.17927271872758865,
      "num_tokens": 14373216.0,
      "step": 6915
    },
    {
      "entropy": 5.488159418106079,
      "epoch": 0.627721335268505,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967475099874992,
      "loss": 5.2567,
      "mean_token_accuracy": 0.1907469481229782,
      "num_tokens": 14383104.0,
      "step": 6920
    },
    {
      "entropy": 5.4569638729095455,
      "epoch": 0.6281748911465893,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496742025972715,
      "loss": 5.2899,
      "mean_token_accuracy": 0.19110392779111862,
      "num_tokens": 14393854.0,
      "step": 6925
    },
    {
      "entropy": 5.516732025146484,
      "epoch": 0.6286284470246735,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496736537372249,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1899970456957817,
      "num_tokens": 14403213.0,
      "step": 6930
    },
    {
      "entropy": 5.603625202178955,
      "epoch": 0.6290820029027576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967310441862151,
      "loss": 5.4528,
      "mean_token_accuracy": 0.1735685408115387,
      "num_tokens": 14412792.0,
      "step": 6935
    },
    {
      "entropy": 5.546102237701416,
      "epoch": 0.6295355587808418,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004967255464147265,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18926624953746796,
      "num_tokens": 14423061.0,
      "step": 6940
    },
    {
      "entropy": 5.512146854400635,
      "epoch": 0.6299891146589259,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967200440578972,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19065099507570266,
      "num_tokens": 14433338.0,
      "step": 6945
    },
    {
      "entropy": 5.45038628578186,
      "epoch": 0.6304426705370102,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967145371158408,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19122012555599213,
      "num_tokens": 14444124.0,
      "step": 6950
    },
    {
      "entropy": 5.501307201385498,
      "epoch": 0.6308962264150944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967090255886714,
      "loss": 5.3908,
      "mean_token_accuracy": 0.17991872131824493,
      "num_tokens": 14453868.0,
      "step": 6955
    },
    {
      "entropy": 5.517397832870484,
      "epoch": 0.6313497822931785,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967035094765028,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18014791905879973,
      "num_tokens": 14465015.0,
      "step": 6960
    },
    {
      "entropy": 5.467123174667359,
      "epoch": 0.6318033381712627,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004966979887794491,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19137182533740998,
      "num_tokens": 14475523.0,
      "step": 6965
    },
    {
      "entropy": 5.466613149642944,
      "epoch": 0.6322568940493469,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966924634976245,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1851485162973404,
      "num_tokens": 14485523.0,
      "step": 6970
    },
    {
      "entropy": 5.538769245147705,
      "epoch": 0.6327104499274311,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966869336311433,
      "loss": 5.3623,
      "mean_token_accuracy": 0.18265851140022277,
      "num_tokens": 14495237.0,
      "step": 6975
    },
    {
      "entropy": 5.564757013320923,
      "epoch": 0.6331640058055152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966813991801197,
      "loss": 5.4848,
      "mean_token_accuracy": 0.1746712803840637,
      "num_tokens": 14506431.0,
      "step": 6980
    },
    {
      "entropy": 5.602010774612427,
      "epoch": 0.6336175616835994,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966758601446684,
      "loss": 5.4194,
      "mean_token_accuracy": 0.17271297425031662,
      "num_tokens": 14517445.0,
      "step": 6985
    },
    {
      "entropy": 5.554762315750122,
      "epoch": 0.6340711175616836,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966703165249038,
      "loss": 5.3482,
      "mean_token_accuracy": 0.17258450388908386,
      "num_tokens": 14527818.0,
      "step": 6990
    },
    {
      "entropy": 5.549514579772949,
      "epoch": 0.6345246734397678,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966647683209406,
      "loss": 5.4332,
      "mean_token_accuracy": 0.16915443241596223,
      "num_tokens": 14538235.0,
      "step": 6995
    },
    {
      "entropy": 5.6472475051879885,
      "epoch": 0.634978229317852,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966592155328935,
      "loss": 5.4463,
      "mean_token_accuracy": 0.1713977947831154,
      "num_tokens": 14549784.0,
      "step": 7000
    },
    {
      "entropy": 5.499601411819458,
      "epoch": 0.6354317851959361,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966536581608771,
      "loss": 5.4273,
      "mean_token_accuracy": 0.17051296085119247,
      "num_tokens": 14561057.0,
      "step": 7005
    },
    {
      "entropy": 5.546831130981445,
      "epoch": 0.6358853410740203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966480962050069,
      "loss": 5.3432,
      "mean_token_accuracy": 0.1802305594086647,
      "num_tokens": 14570859.0,
      "step": 7010
    },
    {
      "entropy": 5.5499834537506105,
      "epoch": 0.6363388969521045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966425296653974,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18582831472158431,
      "num_tokens": 14581066.0,
      "step": 7015
    },
    {
      "entropy": 5.500417184829712,
      "epoch": 0.6367924528301887,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496636958542164,
      "loss": 5.3133,
      "mean_token_accuracy": 0.18046621680259706,
      "num_tokens": 14590709.0,
      "step": 7020
    },
    {
      "entropy": 5.511330270767212,
      "epoch": 0.6372460087082729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966313828354218,
      "loss": 5.2974,
      "mean_token_accuracy": 0.18662593364715577,
      "num_tokens": 14600849.0,
      "step": 7025
    },
    {
      "entropy": 5.455791187286377,
      "epoch": 0.637699564586357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496625802545286,
      "loss": 5.1919,
      "mean_token_accuracy": 0.18537684082984923,
      "num_tokens": 14611427.0,
      "step": 7030
    },
    {
      "entropy": 5.508231830596924,
      "epoch": 0.6381531204644412,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966202176718723,
      "loss": 5.3486,
      "mean_token_accuracy": 0.18688730001449586,
      "num_tokens": 14623860.0,
      "step": 7035
    },
    {
      "entropy": 5.537319087982178,
      "epoch": 0.6386066763425254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966146282152959,
      "loss": 5.2776,
      "mean_token_accuracy": 0.18067865520715715,
      "num_tokens": 14633481.0,
      "step": 7040
    },
    {
      "entropy": 5.535480594635009,
      "epoch": 0.6390602322206096,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966090341756726,
      "loss": 5.2586,
      "mean_token_accuracy": 0.19394663125276565,
      "num_tokens": 14643519.0,
      "step": 7045
    },
    {
      "entropy": 5.390807437896728,
      "epoch": 0.6395137880986937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496603435553118,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19258788973093033,
      "num_tokens": 14654570.0,
      "step": 7050
    },
    {
      "entropy": 5.439500379562378,
      "epoch": 0.6399673439767779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965978323477478,
      "loss": 5.2894,
      "mean_token_accuracy": 0.17954834699630737,
      "num_tokens": 14665596.0,
      "step": 7055
    },
    {
      "entropy": 5.537118530273437,
      "epoch": 0.6404208998548622,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496592224559678,
      "loss": 5.2384,
      "mean_token_accuracy": 0.18761214017868041,
      "num_tokens": 14675660.0,
      "step": 7060
    },
    {
      "entropy": 5.550399494171143,
      "epoch": 0.6408744557329463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965866121890245,
      "loss": 5.3923,
      "mean_token_accuracy": 0.1781720206141472,
      "num_tokens": 14685229.0,
      "step": 7065
    },
    {
      "entropy": 5.52332239151001,
      "epoch": 0.6413280116110305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965809952359034,
      "loss": 5.2985,
      "mean_token_accuracy": 0.17515805810689927,
      "num_tokens": 14695672.0,
      "step": 7070
    },
    {
      "entropy": 5.501980543136597,
      "epoch": 0.6417815674891146,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004965753737004309,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1830699175596237,
      "num_tokens": 14705160.0,
      "step": 7075
    },
    {
      "entropy": 5.48081636428833,
      "epoch": 0.6422351233671988,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965697475827231,
      "loss": 5.2957,
      "mean_token_accuracy": 0.18334358930587769,
      "num_tokens": 14715487.0,
      "step": 7080
    },
    {
      "entropy": 5.545005559921265,
      "epoch": 0.6426886792452831,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965641168828965,
      "loss": 5.3329,
      "mean_token_accuracy": 0.18440192192792892,
      "num_tokens": 14726987.0,
      "step": 7085
    },
    {
      "entropy": 5.4650185108184814,
      "epoch": 0.6431422351233672,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965584816010676,
      "loss": 5.2829,
      "mean_token_accuracy": 0.1867426112294197,
      "num_tokens": 14737810.0,
      "step": 7090
    },
    {
      "entropy": 5.546742534637451,
      "epoch": 0.6435957910014514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965528417373528,
      "loss": 5.3085,
      "mean_token_accuracy": 0.18062587231397628,
      "num_tokens": 14748997.0,
      "step": 7095
    },
    {
      "entropy": 5.515628051757813,
      "epoch": 0.6440493468795355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965471972918688,
      "loss": 5.3136,
      "mean_token_accuracy": 0.17992234230041504,
      "num_tokens": 14760439.0,
      "step": 7100
    },
    {
      "entropy": 5.42392373085022,
      "epoch": 0.6445029027576198,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965415482647322,
      "loss": 5.2268,
      "mean_token_accuracy": 0.1856522426009178,
      "num_tokens": 14770828.0,
      "step": 7105
    },
    {
      "entropy": 5.4523519515991214,
      "epoch": 0.6449564586357039,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049653589465606,
      "loss": 5.3692,
      "mean_token_accuracy": 0.17964595556259155,
      "num_tokens": 14781753.0,
      "step": 7110
    },
    {
      "entropy": 5.59224910736084,
      "epoch": 0.6454100145137881,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496530236465969,
      "loss": 5.3884,
      "mean_token_accuracy": 0.17452527433633805,
      "num_tokens": 14792241.0,
      "step": 7115
    },
    {
      "entropy": 5.466744709014892,
      "epoch": 0.6458635703918723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965245736945763,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19304679036140443,
      "num_tokens": 14802580.0,
      "step": 7120
    },
    {
      "entropy": 5.487527847290039,
      "epoch": 0.6463171262699564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965189063419989,
      "loss": 5.4117,
      "mean_token_accuracy": 0.17699665278196336,
      "num_tokens": 14813103.0,
      "step": 7125
    },
    {
      "entropy": 5.527288436889648,
      "epoch": 0.6467706821480407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965132344083541,
      "loss": 5.2994,
      "mean_token_accuracy": 0.17992949336767197,
      "num_tokens": 14823334.0,
      "step": 7130
    },
    {
      "entropy": 5.533066177368164,
      "epoch": 0.6472242380261248,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965075578937593,
      "loss": 5.3173,
      "mean_token_accuracy": 0.17671651989221573,
      "num_tokens": 14834850.0,
      "step": 7135
    },
    {
      "entropy": 5.469491338729858,
      "epoch": 0.647677793904209,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965018767983316,
      "loss": 5.3381,
      "mean_token_accuracy": 0.1797954857349396,
      "num_tokens": 14844276.0,
      "step": 7140
    },
    {
      "entropy": 5.385136699676513,
      "epoch": 0.6481313497822931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964961911221886,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19689569771289825,
      "num_tokens": 14854352.0,
      "step": 7145
    },
    {
      "entropy": 5.52975492477417,
      "epoch": 0.6485849056603774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496490500865448,
      "loss": 5.3462,
      "mean_token_accuracy": 0.1827696070075035,
      "num_tokens": 14864395.0,
      "step": 7150
    },
    {
      "entropy": 5.498007678985596,
      "epoch": 0.6490384615384616,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964848060282275,
      "loss": 5.2283,
      "mean_token_accuracy": 0.18727268576622008,
      "num_tokens": 14874373.0,
      "step": 7155
    },
    {
      "entropy": 5.491959953308106,
      "epoch": 0.6494920174165457,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964791066106447,
      "loss": 5.2253,
      "mean_token_accuracy": 0.18243176639080047,
      "num_tokens": 14884932.0,
      "step": 7160
    },
    {
      "entropy": 5.5285765647888185,
      "epoch": 0.6499455732946299,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964734026128175,
      "loss": 5.3897,
      "mean_token_accuracy": 0.17832595705986024,
      "num_tokens": 14896254.0,
      "step": 7165
    },
    {
      "entropy": 5.572258043289184,
      "epoch": 0.650399129172714,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964676940348638,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1801915466785431,
      "num_tokens": 14907392.0,
      "step": 7170
    },
    {
      "entropy": 5.564219951629639,
      "epoch": 0.6508526850507983,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496461980876902,
      "loss": 5.4305,
      "mean_token_accuracy": 0.17841965556144715,
      "num_tokens": 14917676.0,
      "step": 7175
    },
    {
      "entropy": 5.515146255493164,
      "epoch": 0.6513062409288825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964562631390498,
      "loss": 5.2854,
      "mean_token_accuracy": 0.1819108560681343,
      "num_tokens": 14928500.0,
      "step": 7180
    },
    {
      "entropy": 5.483141279220581,
      "epoch": 0.6517597968069666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964505408214257,
      "loss": 5.243,
      "mean_token_accuracy": 0.18780999034643173,
      "num_tokens": 14938970.0,
      "step": 7185
    },
    {
      "entropy": 5.482848167419434,
      "epoch": 0.6522133526850508,
      "grad_norm": 1.0,
      "learning_rate": 0.000496444813924148,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1825691819190979,
      "num_tokens": 14949017.0,
      "step": 7190
    },
    {
      "entropy": 5.521585893630982,
      "epoch": 0.652666908563135,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964390824473351,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18902090042829514,
      "num_tokens": 14958529.0,
      "step": 7195
    },
    {
      "entropy": 5.471276807785034,
      "epoch": 0.6531204644412192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964333463911056,
      "loss": 5.3325,
      "mean_token_accuracy": 0.18068593144416809,
      "num_tokens": 14968044.0,
      "step": 7200
    },
    {
      "entropy": 5.519154214859009,
      "epoch": 0.6535740203193033,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964276057555779,
      "loss": 5.3461,
      "mean_token_accuracy": 0.18163236528635024,
      "num_tokens": 14977867.0,
      "step": 7205
    },
    {
      "entropy": 5.576827430725098,
      "epoch": 0.6540275761973875,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496421860540871,
      "loss": 5.3499,
      "mean_token_accuracy": 0.18708072751760482,
      "num_tokens": 14986936.0,
      "step": 7210
    },
    {
      "entropy": 5.499208927154541,
      "epoch": 0.6544811320754716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964161107471036,
      "loss": 5.3146,
      "mean_token_accuracy": 0.1859233096241951,
      "num_tokens": 14996681.0,
      "step": 7215
    },
    {
      "entropy": 5.50371241569519,
      "epoch": 0.6549346879535559,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964103563743946,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18214258253574372,
      "num_tokens": 15007100.0,
      "step": 7220
    },
    {
      "entropy": 5.530852222442627,
      "epoch": 0.6553882438316401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496404597422863,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1881464421749115,
      "num_tokens": 15017452.0,
      "step": 7225
    },
    {
      "entropy": 5.495759153366089,
      "epoch": 0.6558417997097242,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496398833892628,
      "loss": 5.2317,
      "mean_token_accuracy": 0.18454673141241074,
      "num_tokens": 15028362.0,
      "step": 7230
    },
    {
      "entropy": 5.53149824142456,
      "epoch": 0.6562953555878084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963930657838086,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1792814999818802,
      "num_tokens": 15038691.0,
      "step": 7235
    },
    {
      "entropy": 5.503436279296875,
      "epoch": 0.6567489114658926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963872930965242,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18664567470550536,
      "num_tokens": 15050037.0,
      "step": 7240
    },
    {
      "entropy": 5.5270062446594235,
      "epoch": 0.6572024673439768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963815158308941,
      "loss": 5.2687,
      "mean_token_accuracy": 0.18515150398015975,
      "num_tokens": 15059793.0,
      "step": 7245
    },
    {
      "entropy": 5.48774676322937,
      "epoch": 0.657656023222061,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963757339870379,
      "loss": 5.3611,
      "mean_token_accuracy": 0.1739514410495758,
      "num_tokens": 15070500.0,
      "step": 7250
    },
    {
      "entropy": 5.557533264160156,
      "epoch": 0.6581095791001451,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004963699475650751,
      "loss": 5.4125,
      "mean_token_accuracy": 0.17817330807447435,
      "num_tokens": 15081109.0,
      "step": 7255
    },
    {
      "entropy": 5.464095830917358,
      "epoch": 0.6585631349782293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963641565651255,
      "loss": 5.295,
      "mean_token_accuracy": 0.18745609223842621,
      "num_tokens": 15092315.0,
      "step": 7260
    },
    {
      "entropy": 5.526384782791138,
      "epoch": 0.6590166908563135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963583609873084,
      "loss": 5.3606,
      "mean_token_accuracy": 0.18208035826683044,
      "num_tokens": 15103957.0,
      "step": 7265
    },
    {
      "entropy": 5.556434249877929,
      "epoch": 0.6594702467343977,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963525608317443,
      "loss": 5.3472,
      "mean_token_accuracy": 0.17901953905820847,
      "num_tokens": 15114996.0,
      "step": 7270
    },
    {
      "entropy": 5.518425130844117,
      "epoch": 0.6599238026124818,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963467560985527,
      "loss": 5.2853,
      "mean_token_accuracy": 0.17808068990707399,
      "num_tokens": 15125098.0,
      "step": 7275
    },
    {
      "entropy": 5.448391628265381,
      "epoch": 0.660377358490566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963409467878538,
      "loss": 5.2763,
      "mean_token_accuracy": 0.1811485469341278,
      "num_tokens": 15134746.0,
      "step": 7280
    },
    {
      "entropy": 5.566460371017456,
      "epoch": 0.6608309143686503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963351328997678,
      "loss": 5.3115,
      "mean_token_accuracy": 0.17681878209114074,
      "num_tokens": 15145698.0,
      "step": 7285
    },
    {
      "entropy": 5.540733051300049,
      "epoch": 0.6612844702467344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963293144344148,
      "loss": 5.3639,
      "mean_token_accuracy": 0.1819513663649559,
      "num_tokens": 15156026.0,
      "step": 7290
    },
    {
      "entropy": 5.5315940380096436,
      "epoch": 0.6617380261248186,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004963234913919151,
      "loss": 5.2934,
      "mean_token_accuracy": 0.17740343809127807,
      "num_tokens": 15165900.0,
      "step": 7295
    },
    {
      "entropy": 5.499019193649292,
      "epoch": 0.6621915820029027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963176637723893,
      "loss": 5.3654,
      "mean_token_accuracy": 0.177557073533535,
      "num_tokens": 15176693.0,
      "step": 7300
    },
    {
      "entropy": 5.536378002166748,
      "epoch": 0.6626451378809869,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004963118315759578,
      "loss": 5.3817,
      "mean_token_accuracy": 0.17143151313066482,
      "num_tokens": 15186535.0,
      "step": 7305
    },
    {
      "entropy": 5.5319818496704105,
      "epoch": 0.6630986937590712,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963059948027411,
      "loss": 5.2914,
      "mean_token_accuracy": 0.17779401242733,
      "num_tokens": 15196019.0,
      "step": 7310
    },
    {
      "entropy": 5.423022127151489,
      "epoch": 0.6635522496371553,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963001534528601,
      "loss": 5.1571,
      "mean_token_accuracy": 0.19192803800106048,
      "num_tokens": 15205273.0,
      "step": 7315
    },
    {
      "entropy": 5.527570486068726,
      "epoch": 0.6640058055152395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962943075264355,
      "loss": 5.3163,
      "mean_token_accuracy": 0.17965196520090104,
      "num_tokens": 15216716.0,
      "step": 7320
    },
    {
      "entropy": 5.532290601730347,
      "epoch": 0.6644593613933236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962884570235884,
      "loss": 5.3649,
      "mean_token_accuracy": 0.1758708655834198,
      "num_tokens": 15227306.0,
      "step": 7325
    },
    {
      "entropy": 5.456675958633423,
      "epoch": 0.6649129172714079,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962826019444394,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18692494183778763,
      "num_tokens": 15237302.0,
      "step": 7330
    },
    {
      "entropy": 5.492032623291015,
      "epoch": 0.665366473149492,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962767422891099,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1889828249812126,
      "num_tokens": 15247064.0,
      "step": 7335
    },
    {
      "entropy": 5.5133216857910154,
      "epoch": 0.6658200290275762,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496270878057721,
      "loss": 5.3161,
      "mean_token_accuracy": 0.18626102805137634,
      "num_tokens": 15257006.0,
      "step": 7340
    },
    {
      "entropy": 5.5310060501098635,
      "epoch": 0.6662735849056604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962650092503939,
      "loss": 5.3115,
      "mean_token_accuracy": 0.19045017808675765,
      "num_tokens": 15266720.0,
      "step": 7345
    },
    {
      "entropy": 5.508117914199829,
      "epoch": 0.6667271407837445,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962591358672502,
      "loss": 5.3201,
      "mean_token_accuracy": 0.18087735772132874,
      "num_tokens": 15277330.0,
      "step": 7350
    },
    {
      "entropy": 5.490307807922363,
      "epoch": 0.6671806966618288,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496253257908411,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18703862726688386,
      "num_tokens": 15287401.0,
      "step": 7355
    },
    {
      "entropy": 5.460469102859497,
      "epoch": 0.6676342525399129,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004962473753739981,
      "loss": 5.2144,
      "mean_token_accuracy": 0.18987584263086318,
      "num_tokens": 15297426.0,
      "step": 7360
    },
    {
      "entropy": 5.512601900100708,
      "epoch": 0.6680878084179971,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496241488264133,
      "loss": 5.2122,
      "mean_token_accuracy": 0.17890442311763763,
      "num_tokens": 15308412.0,
      "step": 7365
    },
    {
      "entropy": 5.494791793823242,
      "epoch": 0.6685413642960812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962355965789377,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18206070810556413,
      "num_tokens": 15318821.0,
      "step": 7370
    },
    {
      "entropy": 5.428279018402099,
      "epoch": 0.6689949201741655,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962297003185338,
      "loss": 5.284,
      "mean_token_accuracy": 0.18319466859102249,
      "num_tokens": 15328737.0,
      "step": 7375
    },
    {
      "entropy": 5.52919020652771,
      "epoch": 0.6694484760522497,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004962237994830433,
      "loss": 5.3792,
      "mean_token_accuracy": 0.17462361603975296,
      "num_tokens": 15339695.0,
      "step": 7380
    },
    {
      "entropy": 5.563119554519654,
      "epoch": 0.6699020319303338,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962178940725883,
      "loss": 5.3496,
      "mean_token_accuracy": 0.18139780610799788,
      "num_tokens": 15350270.0,
      "step": 7385
    },
    {
      "entropy": 5.520648050308227,
      "epoch": 0.670355587808418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962119840872908,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1830776020884514,
      "num_tokens": 15360986.0,
      "step": 7390
    },
    {
      "entropy": 5.482888650894165,
      "epoch": 0.6708091436865021,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962060695272733,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18322502970695495,
      "num_tokens": 15371786.0,
      "step": 7395
    },
    {
      "entropy": 5.507193279266358,
      "epoch": 0.6712626995645864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962001503926577,
      "loss": 5.2871,
      "mean_token_accuracy": 0.19009895771741867,
      "num_tokens": 15382960.0,
      "step": 7400
    },
    {
      "entropy": 5.521780300140381,
      "epoch": 0.6717162554426706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961942266835667,
      "loss": 5.2713,
      "mean_token_accuracy": 0.18810440748929977,
      "num_tokens": 15392807.0,
      "step": 7405
    },
    {
      "entropy": 5.498525714874267,
      "epoch": 0.6721698113207547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961882984001227,
      "loss": 5.3172,
      "mean_token_accuracy": 0.17630916088819504,
      "num_tokens": 15403712.0,
      "step": 7410
    },
    {
      "entropy": 5.498222351074219,
      "epoch": 0.6726233671988389,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961823655424484,
      "loss": 5.3845,
      "mean_token_accuracy": 0.17348767071962357,
      "num_tokens": 15414839.0,
      "step": 7415
    },
    {
      "entropy": 5.530002164840698,
      "epoch": 0.6730769230769231,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961764281106664,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1857857435941696,
      "num_tokens": 15425320.0,
      "step": 7420
    },
    {
      "entropy": 5.50027289390564,
      "epoch": 0.6735304789550073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961704861048995,
      "loss": 5.263,
      "mean_token_accuracy": 0.18382281810045242,
      "num_tokens": 15435476.0,
      "step": 7425
    },
    {
      "entropy": 5.478959846496582,
      "epoch": 0.6739840348330914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961645395252705,
      "loss": 5.3773,
      "mean_token_accuracy": 0.1767744243144989,
      "num_tokens": 15445786.0,
      "step": 7430
    },
    {
      "entropy": 5.524657535552978,
      "epoch": 0.6744375907111756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961585883719025,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18489955067634584,
      "num_tokens": 15455915.0,
      "step": 7435
    },
    {
      "entropy": 5.54436674118042,
      "epoch": 0.6748911465892597,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961526326449186,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1796664759516716,
      "num_tokens": 15466523.0,
      "step": 7440
    },
    {
      "entropy": 5.51652889251709,
      "epoch": 0.675344702467344,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961466723444419,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1811014547944069,
      "num_tokens": 15476821.0,
      "step": 7445
    },
    {
      "entropy": 5.483820390701294,
      "epoch": 0.6757982583454282,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961407074705955,
      "loss": 5.2628,
      "mean_token_accuracy": 0.18104134052991866,
      "num_tokens": 15487182.0,
      "step": 7450
    },
    {
      "entropy": 5.501073026657105,
      "epoch": 0.6762518142235123,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961347380235031,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18460666239261628,
      "num_tokens": 15497600.0,
      "step": 7455
    },
    {
      "entropy": 5.524596357345581,
      "epoch": 0.6767053701015965,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961287640032879,
      "loss": 5.3944,
      "mean_token_accuracy": 0.17238223403692246,
      "num_tokens": 15508278.0,
      "step": 7460
    },
    {
      "entropy": 5.57806396484375,
      "epoch": 0.6771589259796807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004961227854100734,
      "loss": 5.4086,
      "mean_token_accuracy": 0.17855887562036515,
      "num_tokens": 15518458.0,
      "step": 7465
    },
    {
      "entropy": 5.475199174880982,
      "epoch": 0.6776124818577649,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961168022439835,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18490078896284104,
      "num_tokens": 15528738.0,
      "step": 7470
    },
    {
      "entropy": 5.493577146530152,
      "epoch": 0.6780660377358491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961108145051417,
      "loss": 5.2666,
      "mean_token_accuracy": 0.18824383467435837,
      "num_tokens": 15538656.0,
      "step": 7475
    },
    {
      "entropy": 5.446999168395996,
      "epoch": 0.6785195936139332,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496104822193672,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18458060622215272,
      "num_tokens": 15549285.0,
      "step": 7480
    },
    {
      "entropy": 5.559575128555298,
      "epoch": 0.6789731494920174,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960988253096982,
      "loss": 5.3514,
      "mean_token_accuracy": 0.18502087444067,
      "num_tokens": 15559766.0,
      "step": 7485
    },
    {
      "entropy": 5.494141435623169,
      "epoch": 0.6794267053701016,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960928238533443,
      "loss": 5.218,
      "mean_token_accuracy": 0.19351499229669572,
      "num_tokens": 15570873.0,
      "step": 7490
    },
    {
      "entropy": 5.517694282531738,
      "epoch": 0.6798802612481858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004960868178247344,
      "loss": 5.4704,
      "mean_token_accuracy": 0.17666775584220887,
      "num_tokens": 15581229.0,
      "step": 7495
    },
    {
      "entropy": 5.497984743118286,
      "epoch": 0.68033381712627,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960808072239929,
      "loss": 5.3298,
      "mean_token_accuracy": 0.1761513829231262,
      "num_tokens": 15591686.0,
      "step": 7500
    },
    {
      "entropy": 5.607373332977295,
      "epoch": 0.6807873730043541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004960747920512439,
      "loss": 5.3475,
      "mean_token_accuracy": 0.18001337945461274,
      "num_tokens": 15601869.0,
      "step": 7505
    },
    {
      "entropy": 5.4866539478302006,
      "epoch": 0.6812409288824384,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496068772306612,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1745358318090439,
      "num_tokens": 15612946.0,
      "step": 7510
    },
    {
      "entropy": 5.529070234298706,
      "epoch": 0.6816944847605225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960627479902214,
      "loss": 5.3561,
      "mean_token_accuracy": 0.17982869893312453,
      "num_tokens": 15623761.0,
      "step": 7515
    },
    {
      "entropy": 5.523828363418579,
      "epoch": 0.6821480406386067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496056719102197,
      "loss": 5.3549,
      "mean_token_accuracy": 0.17677418738603592,
      "num_tokens": 15634500.0,
      "step": 7520
    },
    {
      "entropy": 5.451393985748291,
      "epoch": 0.6826015965166908,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960506856426631,
      "loss": 5.2098,
      "mean_token_accuracy": 0.1831522673368454,
      "num_tokens": 15644896.0,
      "step": 7525
    },
    {
      "entropy": 5.528468227386474,
      "epoch": 0.683055152394775,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960446476117448,
      "loss": 5.3356,
      "mean_token_accuracy": 0.1845097690820694,
      "num_tokens": 15655008.0,
      "step": 7530
    },
    {
      "entropy": 5.501651287078857,
      "epoch": 0.6835087082728593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960386050095668,
      "loss": 5.2435,
      "mean_token_accuracy": 0.18057304620742798,
      "num_tokens": 15665944.0,
      "step": 7535
    },
    {
      "entropy": 5.510901641845703,
      "epoch": 0.6839622641509434,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960325578362543,
      "loss": 5.3456,
      "mean_token_accuracy": 0.18406951576471328,
      "num_tokens": 15677099.0,
      "step": 7540
    },
    {
      "entropy": 5.578004693984985,
      "epoch": 0.6844158200290276,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496026506091932,
      "loss": 5.3118,
      "mean_token_accuracy": 0.18596543669700621,
      "num_tokens": 15687531.0,
      "step": 7545
    },
    {
      "entropy": 5.537898349761963,
      "epoch": 0.6848693759071117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960204497767253,
      "loss": 5.323,
      "mean_token_accuracy": 0.18122205436229705,
      "num_tokens": 15697002.0,
      "step": 7550
    },
    {
      "entropy": 5.39183406829834,
      "epoch": 0.685322931785196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960143888907594,
      "loss": 5.1951,
      "mean_token_accuracy": 0.1852844089269638,
      "num_tokens": 15707249.0,
      "step": 7555
    },
    {
      "entropy": 5.48224287033081,
      "epoch": 0.6857764876632801,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960083234341595,
      "loss": 5.275,
      "mean_token_accuracy": 0.18008209019899368,
      "num_tokens": 15718764.0,
      "step": 7560
    },
    {
      "entropy": 5.485287809371949,
      "epoch": 0.6862300435413643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960022534070513,
      "loss": 5.223,
      "mean_token_accuracy": 0.18764107674360275,
      "num_tokens": 15728854.0,
      "step": 7565
    },
    {
      "entropy": 5.415549468994141,
      "epoch": 0.6866835994194485,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049599617880956,
      "loss": 5.2349,
      "mean_token_accuracy": 0.18654119819402695,
      "num_tokens": 15739682.0,
      "step": 7570
    },
    {
      "entropy": 5.4649327278137205,
      "epoch": 0.6871371552975326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959900996418115,
      "loss": 5.2324,
      "mean_token_accuracy": 0.18843563199043273,
      "num_tokens": 15749628.0,
      "step": 7575
    },
    {
      "entropy": 5.513627624511718,
      "epoch": 0.6875907111756169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959840159039315,
      "loss": 5.3056,
      "mean_token_accuracy": 0.18155467957258226,
      "num_tokens": 15760137.0,
      "step": 7580
    },
    {
      "entropy": 5.538849496841431,
      "epoch": 0.688044267053701,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959779275960456,
      "loss": 5.2556,
      "mean_token_accuracy": 0.1825544536113739,
      "num_tokens": 15769889.0,
      "step": 7585
    },
    {
      "entropy": 5.480565595626831,
      "epoch": 0.6884978229317852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049597183471828,
      "loss": 5.3771,
      "mean_token_accuracy": 0.18086028099060059,
      "num_tokens": 15780912.0,
      "step": 7590
    },
    {
      "entropy": 5.501943922042846,
      "epoch": 0.6889513788098693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959657372707604,
      "loss": 5.2532,
      "mean_token_accuracy": 0.189553838968277,
      "num_tokens": 15791719.0,
      "step": 7595
    },
    {
      "entropy": 5.527008438110352,
      "epoch": 0.6894049346879536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959596352536132,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18491054624319075,
      "num_tokens": 15802551.0,
      "step": 7600
    },
    {
      "entropy": 5.546923732757568,
      "epoch": 0.6898584905660378,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959535286669643,
      "loss": 5.3388,
      "mean_token_accuracy": 0.17994246035814285,
      "num_tokens": 15813551.0,
      "step": 7605
    },
    {
      "entropy": 5.498889207839966,
      "epoch": 0.6903120464441219,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959474175109402,
      "loss": 5.3034,
      "mean_token_accuracy": 0.18564930260181428,
      "num_tokens": 15824262.0,
      "step": 7610
    },
    {
      "entropy": 5.550105953216553,
      "epoch": 0.6907656023222061,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959413017856672,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1801539644598961,
      "num_tokens": 15835493.0,
      "step": 7615
    },
    {
      "entropy": 5.419754505157471,
      "epoch": 0.6912191582002902,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959351814912716,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1831692114472389,
      "num_tokens": 15846224.0,
      "step": 7620
    },
    {
      "entropy": 5.5195647239685055,
      "epoch": 0.6916727140783745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959290566278803,
      "loss": 5.2845,
      "mean_token_accuracy": 0.18517156839370727,
      "num_tokens": 15855622.0,
      "step": 7625
    },
    {
      "entropy": 5.537323617935181,
      "epoch": 0.6921262699564587,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959229271956199,
      "loss": 5.344,
      "mean_token_accuracy": 0.18175625056028366,
      "num_tokens": 15865835.0,
      "step": 7630
    },
    {
      "entropy": 5.5008909702301025,
      "epoch": 0.6925798258345428,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959167931946169,
      "loss": 5.2823,
      "mean_token_accuracy": 0.18331976979970932,
      "num_tokens": 15875915.0,
      "step": 7635
    },
    {
      "entropy": 5.514472818374633,
      "epoch": 0.693033381712627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959106546249984,
      "loss": 5.2829,
      "mean_token_accuracy": 0.1838786020874977,
      "num_tokens": 15887111.0,
      "step": 7640
    },
    {
      "entropy": 5.530413293838501,
      "epoch": 0.6934869375907112,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959045114868913,
      "loss": 5.39,
      "mean_token_accuracy": 0.1729227066040039,
      "num_tokens": 15899528.0,
      "step": 7645
    },
    {
      "entropy": 5.5232024669647215,
      "epoch": 0.6939404934687954,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958983637804225,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19027511328458785,
      "num_tokens": 15910881.0,
      "step": 7650
    },
    {
      "entropy": 5.569855737686157,
      "epoch": 0.6943940493468795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958922115057194,
      "loss": 5.3411,
      "mean_token_accuracy": 0.17807182669639587,
      "num_tokens": 15921189.0,
      "step": 7655
    },
    {
      "entropy": 5.563821744918823,
      "epoch": 0.6948476052249637,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958860546629089,
      "loss": 5.3297,
      "mean_token_accuracy": 0.18766747564077377,
      "num_tokens": 15931276.0,
      "step": 7660
    },
    {
      "entropy": 5.550361442565918,
      "epoch": 0.6953011611030478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958798932521186,
      "loss": 5.4139,
      "mean_token_accuracy": 0.17329273372888565,
      "num_tokens": 15941506.0,
      "step": 7665
    },
    {
      "entropy": 5.4555493831634525,
      "epoch": 0.6957547169811321,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958737272734759,
      "loss": 5.245,
      "mean_token_accuracy": 0.18558478802442552,
      "num_tokens": 15951158.0,
      "step": 7670
    },
    {
      "entropy": 5.522123765945435,
      "epoch": 0.6962082728592163,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958675567271081,
      "loss": 5.2866,
      "mean_token_accuracy": 0.18499088436365127,
      "num_tokens": 15960852.0,
      "step": 7675
    },
    {
      "entropy": 5.5311461925506595,
      "epoch": 0.6966618287373004,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495861381613143,
      "loss": 5.2857,
      "mean_token_accuracy": 0.179577599465847,
      "num_tokens": 15971291.0,
      "step": 7680
    },
    {
      "entropy": 5.507929944992066,
      "epoch": 0.6971153846153846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958552019317082,
      "loss": 5.3468,
      "mean_token_accuracy": 0.18295310884714128,
      "num_tokens": 15982180.0,
      "step": 7685
    },
    {
      "entropy": 5.420698976516723,
      "epoch": 0.6975689404934688,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958490176829316,
      "loss": 5.2349,
      "mean_token_accuracy": 0.18435447812080383,
      "num_tokens": 15992363.0,
      "step": 7690
    },
    {
      "entropy": 5.496499824523926,
      "epoch": 0.698022496371553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495842828866941,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17525662034749984,
      "num_tokens": 16003477.0,
      "step": 7695
    },
    {
      "entropy": 5.558476448059082,
      "epoch": 0.6984760522496372,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958366354838645,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18492270410060882,
      "num_tokens": 16013747.0,
      "step": 7700
    },
    {
      "entropy": 5.521188449859619,
      "epoch": 0.6989296081277213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049583043753383,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18768803924322128,
      "num_tokens": 16025662.0,
      "step": 7705
    },
    {
      "entropy": 5.532258319854736,
      "epoch": 0.6993831640058055,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958242350169658,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18370112180709838,
      "num_tokens": 16037315.0,
      "step": 7710
    },
    {
      "entropy": 5.446314811706543,
      "epoch": 0.6998367198838897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004958180279334001,
      "loss": 5.2128,
      "mean_token_accuracy": 0.18639328330755234,
      "num_tokens": 16047297.0,
      "step": 7715
    },
    {
      "entropy": 5.467506217956543,
      "epoch": 0.7002902757619739,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958118162832614,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19042861312627793,
      "num_tokens": 16057573.0,
      "step": 7720
    },
    {
      "entropy": 5.496659231185913,
      "epoch": 0.700743831640058,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958056000666781,
      "loss": 5.3569,
      "mean_token_accuracy": 0.18165837675333024,
      "num_tokens": 16067926.0,
      "step": 7725
    },
    {
      "entropy": 5.4943725109100345,
      "epoch": 0.7011973875181422,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004957993792837786,
      "loss": 5.2781,
      "mean_token_accuracy": 0.18150821179151536,
      "num_tokens": 16077938.0,
      "step": 7730
    },
    {
      "entropy": 5.42755537033081,
      "epoch": 0.7016509433962265,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004957931539346917,
      "loss": 5.175,
      "mean_token_accuracy": 0.18916916996240615,
      "num_tokens": 16088163.0,
      "step": 7735
    },
    {
      "entropy": 5.556448745727539,
      "epoch": 0.7021044992743106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957869240195462,
      "loss": 5.394,
      "mean_token_accuracy": 0.17612502872943878,
      "num_tokens": 16100640.0,
      "step": 7740
    },
    {
      "entropy": 5.486884880065918,
      "epoch": 0.7025580551523948,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957806895384707,
      "loss": 5.2049,
      "mean_token_accuracy": 0.1893024906516075,
      "num_tokens": 16111018.0,
      "step": 7745
    },
    {
      "entropy": 5.448366641998291,
      "epoch": 0.7030116110304789,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957744504915944,
      "loss": 5.2455,
      "mean_token_accuracy": 0.1839633345603943,
      "num_tokens": 16120873.0,
      "step": 7750
    },
    {
      "entropy": 5.529433631896973,
      "epoch": 0.7034651669085631,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957682068790463,
      "loss": 5.2736,
      "mean_token_accuracy": 0.18772852271795273,
      "num_tokens": 16131178.0,
      "step": 7755
    },
    {
      "entropy": 5.466021537780762,
      "epoch": 0.7039187227866474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957619587009553,
      "loss": 5.278,
      "mean_token_accuracy": 0.1836032971739769,
      "num_tokens": 16141156.0,
      "step": 7760
    },
    {
      "entropy": 5.52126636505127,
      "epoch": 0.7043722786647315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957557059574507,
      "loss": 5.3781,
      "mean_token_accuracy": 0.17983128130435944,
      "num_tokens": 16150077.0,
      "step": 7765
    },
    {
      "entropy": 5.486139154434204,
      "epoch": 0.7048258345428157,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957494486486618,
      "loss": 5.296,
      "mean_token_accuracy": 0.1811152681708336,
      "num_tokens": 16160514.0,
      "step": 7770
    },
    {
      "entropy": 5.528773260116577,
      "epoch": 0.7052793904208998,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957431867747182,
      "loss": 5.3239,
      "mean_token_accuracy": 0.18625716120004654,
      "num_tokens": 16170811.0,
      "step": 7775
    },
    {
      "entropy": 5.452904558181762,
      "epoch": 0.7057329462989841,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957369203357492,
      "loss": 5.2389,
      "mean_token_accuracy": 0.18305610865354538,
      "num_tokens": 16181553.0,
      "step": 7780
    },
    {
      "entropy": 5.484501028060913,
      "epoch": 0.7061865021770682,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957306493318844,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19118969291448593,
      "num_tokens": 16191308.0,
      "step": 7785
    },
    {
      "entropy": 5.503818416595459,
      "epoch": 0.7066400580551524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957243737632535,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19155516624450683,
      "num_tokens": 16201311.0,
      "step": 7790
    },
    {
      "entropy": 5.545148134231567,
      "epoch": 0.7070936139332366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004957180936299863,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1893450289964676,
      "num_tokens": 16211105.0,
      "step": 7795
    },
    {
      "entropy": 5.520404529571533,
      "epoch": 0.7075471698113207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004957118089322127,
      "loss": 5.2874,
      "mean_token_accuracy": 0.17717190086841583,
      "num_tokens": 16220836.0,
      "step": 7800
    },
    {
      "entropy": 5.462588691711426,
      "epoch": 0.708000725689405,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957055196700628,
      "loss": 5.1898,
      "mean_token_accuracy": 0.18572719693183898,
      "num_tokens": 16231630.0,
      "step": 7805
    },
    {
      "entropy": 5.494060945510864,
      "epoch": 0.7084542815674891,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956992258436663,
      "loss": 5.2909,
      "mean_token_accuracy": 0.18368653506040572,
      "num_tokens": 16242272.0,
      "step": 7810
    },
    {
      "entropy": 5.399591398239136,
      "epoch": 0.7089078374455733,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004956929274531536,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1873309940099716,
      "num_tokens": 16252555.0,
      "step": 7815
    },
    {
      "entropy": 5.486877584457398,
      "epoch": 0.7093613933236574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495686624498655,
      "loss": 5.3216,
      "mean_token_accuracy": 0.18683241903781891,
      "num_tokens": 16261964.0,
      "step": 7820
    },
    {
      "entropy": 5.574159717559814,
      "epoch": 0.7098149492017417,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956803169803007,
      "loss": 5.3752,
      "mean_token_accuracy": 0.175976400077343,
      "num_tokens": 16274543.0,
      "step": 7825
    },
    {
      "entropy": 5.467151403427124,
      "epoch": 0.7102685050798259,
      "grad_norm": 0.875,
      "learning_rate": 0.0004956740048982212,
      "loss": 5.28,
      "mean_token_accuracy": 0.17519069761037825,
      "num_tokens": 16285124.0,
      "step": 7830
    },
    {
      "entropy": 5.484114265441894,
      "epoch": 0.71072206095791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956676882525472,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18310529738664627,
      "num_tokens": 16295727.0,
      "step": 7835
    },
    {
      "entropy": 5.574627113342285,
      "epoch": 0.7111756168359942,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495661367043409,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19142441600561141,
      "num_tokens": 16305496.0,
      "step": 7840
    },
    {
      "entropy": 5.482682752609253,
      "epoch": 0.7116291727140783,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956550412709376,
      "loss": 5.222,
      "mean_token_accuracy": 0.19398910850286483,
      "num_tokens": 16315112.0,
      "step": 7845
    },
    {
      "entropy": 5.463079977035522,
      "epoch": 0.7120827285921626,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956487109352637,
      "loss": 5.2869,
      "mean_token_accuracy": 0.18460012674331666,
      "num_tokens": 16325613.0,
      "step": 7850
    },
    {
      "entropy": 5.49309515953064,
      "epoch": 0.7125362844702468,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956423760365181,
      "loss": 5.2163,
      "mean_token_accuracy": 0.18853310793638228,
      "num_tokens": 16335317.0,
      "step": 7855
    },
    {
      "entropy": 5.528402090072632,
      "epoch": 0.7129898403483309,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956360365748321,
      "loss": 5.3547,
      "mean_token_accuracy": 0.17705050855875015,
      "num_tokens": 16345003.0,
      "step": 7860
    },
    {
      "entropy": 5.511237478256225,
      "epoch": 0.7134433962264151,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956296925503367,
      "loss": 5.3451,
      "mean_token_accuracy": 0.18143388777971267,
      "num_tokens": 16356108.0,
      "step": 7865
    },
    {
      "entropy": 5.448657751083374,
      "epoch": 0.7138969521044993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956233439631628,
      "loss": 5.1452,
      "mean_token_accuracy": 0.18960337191820145,
      "num_tokens": 16364733.0,
      "step": 7870
    },
    {
      "entropy": 5.506351900100708,
      "epoch": 0.7143505079825835,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004956169908134421,
      "loss": 5.2525,
      "mean_token_accuracy": 0.18633940815925598,
      "num_tokens": 16375229.0,
      "step": 7875
    },
    {
      "entropy": 5.485099172592163,
      "epoch": 0.7148040638606676,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956106331013057,
      "loss": 5.339,
      "mean_token_accuracy": 0.18616413027048112,
      "num_tokens": 16386366.0,
      "step": 7880
    },
    {
      "entropy": 5.479721736907959,
      "epoch": 0.7152576197387518,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004956042708268853,
      "loss": 5.2205,
      "mean_token_accuracy": 0.19334444254636765,
      "num_tokens": 16397067.0,
      "step": 7885
    },
    {
      "entropy": 5.424285650253296,
      "epoch": 0.715711175616836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955979039903123,
      "loss": 5.187,
      "mean_token_accuracy": 0.18900911509990692,
      "num_tokens": 16407256.0,
      "step": 7890
    },
    {
      "entropy": 5.455164909362793,
      "epoch": 0.7161647314949202,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004955915325917183,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1819542557001114,
      "num_tokens": 16418701.0,
      "step": 7895
    },
    {
      "entropy": 5.432131624221801,
      "epoch": 0.7166182873730044,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955851566312354,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19276488721370696,
      "num_tokens": 16428019.0,
      "step": 7900
    },
    {
      "entropy": 5.4328320026397705,
      "epoch": 0.7170718432510885,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004955787761089952,
      "loss": 5.2523,
      "mean_token_accuracy": 0.1807563617825508,
      "num_tokens": 16440122.0,
      "step": 7905
    },
    {
      "entropy": 5.56105408668518,
      "epoch": 0.7175253991291727,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955723910251296,
      "loss": 5.338,
      "mean_token_accuracy": 0.1794456034898758,
      "num_tokens": 16449801.0,
      "step": 7910
    },
    {
      "entropy": 5.44252028465271,
      "epoch": 0.717978955007257,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004955660013797709,
      "loss": 5.1941,
      "mean_token_accuracy": 0.189080049097538,
      "num_tokens": 16460155.0,
      "step": 7915
    },
    {
      "entropy": 5.547311067581177,
      "epoch": 0.7184325108853411,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495559607173051,
      "loss": 5.3922,
      "mean_token_accuracy": 0.17179529070854188,
      "num_tokens": 16470656.0,
      "step": 7920
    },
    {
      "entropy": 5.511463022232055,
      "epoch": 0.7188860667634253,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955532084051022,
      "loss": 5.2813,
      "mean_token_accuracy": 0.18540270030498504,
      "num_tokens": 16481449.0,
      "step": 7925
    },
    {
      "entropy": 5.474109745025634,
      "epoch": 0.7193396226415094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955468050760568,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1859189748764038,
      "num_tokens": 16492032.0,
      "step": 7930
    },
    {
      "entropy": 5.499914312362671,
      "epoch": 0.7197931785195936,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955403971860474,
      "loss": 5.2276,
      "mean_token_accuracy": 0.18490062803030013,
      "num_tokens": 16502221.0,
      "step": 7935
    },
    {
      "entropy": 5.50188307762146,
      "epoch": 0.7202467343976778,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955339847352063,
      "loss": 5.2127,
      "mean_token_accuracy": 0.18936268985271454,
      "num_tokens": 16512021.0,
      "step": 7940
    },
    {
      "entropy": 5.486470651626587,
      "epoch": 0.720700290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004955275677236662,
      "loss": 5.27,
      "mean_token_accuracy": 0.1855945870280266,
      "num_tokens": 16522060.0,
      "step": 7945
    },
    {
      "entropy": 5.504107522964477,
      "epoch": 0.7211538461538461,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955211461515599,
      "loss": 5.2715,
      "mean_token_accuracy": 0.18720357716083527,
      "num_tokens": 16532827.0,
      "step": 7950
    },
    {
      "entropy": 5.490920114517212,
      "epoch": 0.7216074020319303,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049551472001902,
      "loss": 5.3106,
      "mean_token_accuracy": 0.18881746232509614,
      "num_tokens": 16542838.0,
      "step": 7955
    },
    {
      "entropy": 5.46091480255127,
      "epoch": 0.7220609579100146,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955082893261796,
      "loss": 5.252,
      "mean_token_accuracy": 0.18916592001914978,
      "num_tokens": 16552615.0,
      "step": 7960
    },
    {
      "entropy": 5.521874952316284,
      "epoch": 0.7225145137880987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955018540731714,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1830597385764122,
      "num_tokens": 16561777.0,
      "step": 7965
    },
    {
      "entropy": 5.519312000274658,
      "epoch": 0.7229680696661829,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954954142601288,
      "loss": 5.2888,
      "mean_token_accuracy": 0.182749043405056,
      "num_tokens": 16572111.0,
      "step": 7970
    },
    {
      "entropy": 5.5174407958984375,
      "epoch": 0.723421625544267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954889698871849,
      "loss": 5.3275,
      "mean_token_accuracy": 0.18790917694568635,
      "num_tokens": 16581674.0,
      "step": 7975
    },
    {
      "entropy": 5.516995191574097,
      "epoch": 0.7238751814223512,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954825209544728,
      "loss": 5.3103,
      "mean_token_accuracy": 0.1911083474755287,
      "num_tokens": 16591802.0,
      "step": 7980
    },
    {
      "entropy": 5.4708818912506105,
      "epoch": 0.7243287373004355,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495476067462126,
      "loss": 5.202,
      "mean_token_accuracy": 0.1951405331492424,
      "num_tokens": 16602772.0,
      "step": 7985
    },
    {
      "entropy": 5.5349139213562015,
      "epoch": 0.7247822931785196,
      "grad_norm": 0.96875,
      "learning_rate": 0.000495469609410278,
      "loss": 5.2517,
      "mean_token_accuracy": 0.18888139277696608,
      "num_tokens": 16613526.0,
      "step": 7990
    },
    {
      "entropy": 5.402984237670898,
      "epoch": 0.7252358490566038,
      "grad_norm": 1.125,
      "learning_rate": 0.0004954631467990622,
      "loss": 5.26,
      "mean_token_accuracy": 0.19152738451957702,
      "num_tokens": 16623583.0,
      "step": 7995
    },
    {
      "entropy": 5.496296977996826,
      "epoch": 0.7256894049346879,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954566796286124,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18329164236783982,
      "num_tokens": 16635068.0,
      "step": 8000
    },
    {
      "entropy": 5.568232727050781,
      "epoch": 0.7261429608127722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954502078990623,
      "loss": 5.2167,
      "mean_token_accuracy": 0.18837719559669494,
      "num_tokens": 16645021.0,
      "step": 8005
    },
    {
      "entropy": 5.470982456207276,
      "epoch": 0.7265965166908563,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954437316105457,
      "loss": 5.3,
      "mean_token_accuracy": 0.17855734527111053,
      "num_tokens": 16656021.0,
      "step": 8010
    },
    {
      "entropy": 5.408538675308227,
      "epoch": 0.7270500725689405,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954372507631965,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19202053397893906,
      "num_tokens": 16665236.0,
      "step": 8015
    },
    {
      "entropy": 5.490168190002441,
      "epoch": 0.7275036284470247,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954307653571489,
      "loss": 5.396,
      "mean_token_accuracy": 0.1767436221241951,
      "num_tokens": 16676690.0,
      "step": 8020
    },
    {
      "entropy": 5.514575815200805,
      "epoch": 0.7279571843251088,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954242753925367,
      "loss": 5.2762,
      "mean_token_accuracy": 0.17500143349170685,
      "num_tokens": 16687288.0,
      "step": 8025
    },
    {
      "entropy": 5.487605571746826,
      "epoch": 0.7284107402031931,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954177808694946,
      "loss": 5.2849,
      "mean_token_accuracy": 0.18090034574270247,
      "num_tokens": 16697873.0,
      "step": 8030
    },
    {
      "entropy": 5.501708555221557,
      "epoch": 0.7288642960812772,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954112817881565,
      "loss": 5.3056,
      "mean_token_accuracy": 0.18123792260885238,
      "num_tokens": 16709002.0,
      "step": 8035
    },
    {
      "entropy": 5.468709087371826,
      "epoch": 0.7293178519593614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954047781486568,
      "loss": 5.2637,
      "mean_token_accuracy": 0.18713146299123765,
      "num_tokens": 16718989.0,
      "step": 8040
    },
    {
      "entropy": 5.572985076904297,
      "epoch": 0.7297714078374455,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004953982699511303,
      "loss": 5.355,
      "mean_token_accuracy": 0.18397848755121232,
      "num_tokens": 16730425.0,
      "step": 8045
    },
    {
      "entropy": 5.519139575958252,
      "epoch": 0.7302249637155298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953917571957114,
      "loss": 5.2586,
      "mean_token_accuracy": 0.18431825190782547,
      "num_tokens": 16740216.0,
      "step": 8050
    },
    {
      "entropy": 5.4953625202178955,
      "epoch": 0.730678519593614,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953852398825346,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1867049127817154,
      "num_tokens": 16752436.0,
      "step": 8055
    },
    {
      "entropy": 5.4429257869720455,
      "epoch": 0.7311320754716981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495378718011735,
      "loss": 5.2692,
      "mean_token_accuracy": 0.1828521177172661,
      "num_tokens": 16762631.0,
      "step": 8060
    },
    {
      "entropy": 5.514633131027222,
      "epoch": 0.7315856313497823,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953721915834473,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18744696229696273,
      "num_tokens": 16772613.0,
      "step": 8065
    },
    {
      "entropy": 5.547576761245727,
      "epoch": 0.7320391872278664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953656605978067,
      "loss": 5.2805,
      "mean_token_accuracy": 0.17784342020750046,
      "num_tokens": 16782963.0,
      "step": 8070
    },
    {
      "entropy": 5.477340126037598,
      "epoch": 0.7324927431059507,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000495359125054948,
      "loss": 5.205,
      "mean_token_accuracy": 0.1900916039943695,
      "num_tokens": 16793892.0,
      "step": 8075
    },
    {
      "entropy": 5.3859124183654785,
      "epoch": 0.7329462989840348,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004953525849550063,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18191331773996353,
      "num_tokens": 16803388.0,
      "step": 8080
    },
    {
      "entropy": 5.457746791839599,
      "epoch": 0.733399854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004953460402981172,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19219024181365968,
      "num_tokens": 16813201.0,
      "step": 8085
    },
    {
      "entropy": 5.542439699172974,
      "epoch": 0.7338534107402032,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004953394910844157,
      "loss": 5.4383,
      "mean_token_accuracy": 0.178664830327034,
      "num_tokens": 16823265.0,
      "step": 8090
    },
    {
      "entropy": 5.487712240219116,
      "epoch": 0.7343069666182874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004953329373140374,
      "loss": 5.2248,
      "mean_token_accuracy": 0.18390753567218782,
      "num_tokens": 16833057.0,
      "step": 8095
    },
    {
      "entropy": 5.401983737945557,
      "epoch": 0.7347605224963716,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953263789871179,
      "loss": 5.1818,
      "mean_token_accuracy": 0.18794455230236054,
      "num_tokens": 16843488.0,
      "step": 8100
    },
    {
      "entropy": 5.413467597961426,
      "epoch": 0.7352140783744557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953198161037925,
      "loss": 5.1973,
      "mean_token_accuracy": 0.19814343005418777,
      "num_tokens": 16852448.0,
      "step": 8105
    },
    {
      "entropy": 5.396668195724487,
      "epoch": 0.7356676342525399,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953132486641975,
      "loss": 5.2457,
      "mean_token_accuracy": 0.18623815029859542,
      "num_tokens": 16863944.0,
      "step": 8110
    },
    {
      "entropy": 5.5121378898620605,
      "epoch": 0.736121190130624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495306676668468,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19186616986989974,
      "num_tokens": 16873923.0,
      "step": 8115
    },
    {
      "entropy": 5.5024964809417725,
      "epoch": 0.7365747460087083,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953001001167406,
      "loss": 5.2731,
      "mean_token_accuracy": 0.18318339437246323,
      "num_tokens": 16884373.0,
      "step": 8120
    },
    {
      "entropy": 5.522818374633789,
      "epoch": 0.7370283018867925,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952935190091508,
      "loss": 5.4047,
      "mean_token_accuracy": 0.17949366867542266,
      "num_tokens": 16895869.0,
      "step": 8125
    },
    {
      "entropy": 5.411563777923584,
      "epoch": 0.7374818577648766,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952869333458349,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18748896867036818,
      "num_tokens": 16906737.0,
      "step": 8130
    },
    {
      "entropy": 5.4336510181427,
      "epoch": 0.7379354136429608,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952803431269292,
      "loss": 5.1823,
      "mean_token_accuracy": 0.1944849282503128,
      "num_tokens": 16916083.0,
      "step": 8135
    },
    {
      "entropy": 5.546496820449829,
      "epoch": 0.738388969521045,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952737483525699,
      "loss": 5.346,
      "mean_token_accuracy": 0.1862415462732315,
      "num_tokens": 16926419.0,
      "step": 8140
    },
    {
      "entropy": 5.537788152694702,
      "epoch": 0.7388425253991292,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952671490228932,
      "loss": 5.2821,
      "mean_token_accuracy": 0.18204317539930343,
      "num_tokens": 16936067.0,
      "step": 8145
    },
    {
      "entropy": 5.551399993896484,
      "epoch": 0.7392960812772134,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952605451380357,
      "loss": 5.2271,
      "mean_token_accuracy": 0.18934836983680725,
      "num_tokens": 16946571.0,
      "step": 8150
    },
    {
      "entropy": 5.41324372291565,
      "epoch": 0.7397496371552975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952539366981342,
      "loss": 5.2941,
      "mean_token_accuracy": 0.18353919684886932,
      "num_tokens": 16956764.0,
      "step": 8155
    },
    {
      "entropy": 5.457695579528808,
      "epoch": 0.7402031930333817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495247323703325,
      "loss": 5.2137,
      "mean_token_accuracy": 0.18735239505767823,
      "num_tokens": 16966002.0,
      "step": 8160
    },
    {
      "entropy": 5.569545698165894,
      "epoch": 0.7406567489114659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495240706153745,
      "loss": 5.3036,
      "mean_token_accuracy": 0.1865418493747711,
      "num_tokens": 16975705.0,
      "step": 8165
    },
    {
      "entropy": 5.4822554111480715,
      "epoch": 0.7411103047895501,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952340840495312,
      "loss": 5.2087,
      "mean_token_accuracy": 0.18633690029382705,
      "num_tokens": 16985945.0,
      "step": 8170
    },
    {
      "entropy": 5.4184746742248535,
      "epoch": 0.7415638606676342,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952274573908203,
      "loss": 5.189,
      "mean_token_accuracy": 0.18653265982866288,
      "num_tokens": 16995475.0,
      "step": 8175
    },
    {
      "entropy": 5.464861154556274,
      "epoch": 0.7420174165457184,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952208261777496,
      "loss": 5.2251,
      "mean_token_accuracy": 0.18869319707155227,
      "num_tokens": 17005028.0,
      "step": 8180
    },
    {
      "entropy": 5.505099534988403,
      "epoch": 0.7424709724238027,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495214190410456,
      "loss": 5.2747,
      "mean_token_accuracy": 0.18964312076568604,
      "num_tokens": 17014786.0,
      "step": 8185
    },
    {
      "entropy": 5.522277116775513,
      "epoch": 0.7429245283018868,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952075500890768,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1866888329386711,
      "num_tokens": 17024431.0,
      "step": 8190
    },
    {
      "entropy": 5.457682704925537,
      "epoch": 0.743378084179971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952009052137493,
      "loss": 5.259,
      "mean_token_accuracy": 0.18620658963918685,
      "num_tokens": 17034368.0,
      "step": 8195
    },
    {
      "entropy": 5.472687244415283,
      "epoch": 0.7438316400580551,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951942557846112,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1834196001291275,
      "num_tokens": 17043818.0,
      "step": 8200
    },
    {
      "entropy": 5.5276368141174315,
      "epoch": 0.7442851959361393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951876018017997,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18539572805166243,
      "num_tokens": 17054843.0,
      "step": 8205
    },
    {
      "entropy": 5.494559049606323,
      "epoch": 0.7447387518142236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951809432654524,
      "loss": 5.3072,
      "mean_token_accuracy": 0.1883242279291153,
      "num_tokens": 17065539.0,
      "step": 8210
    },
    {
      "entropy": 5.442562866210937,
      "epoch": 0.7451923076923077,
      "grad_norm": 0.984375,
      "learning_rate": 0.000495174280175707,
      "loss": 5.2166,
      "mean_token_accuracy": 0.1864542081952095,
      "num_tokens": 17075443.0,
      "step": 8215
    },
    {
      "entropy": 5.560985994338989,
      "epoch": 0.7456458635703919,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951676125327015,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19198401570320128,
      "num_tokens": 17085490.0,
      "step": 8220
    },
    {
      "entropy": 5.551355648040771,
      "epoch": 0.746099419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951609403365737,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18634077459573745,
      "num_tokens": 17094803.0,
      "step": 8225
    },
    {
      "entropy": 5.491112995147705,
      "epoch": 0.7465529753265603,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951542635874615,
      "loss": 5.18,
      "mean_token_accuracy": 0.18782074451446534,
      "num_tokens": 17104373.0,
      "step": 8230
    },
    {
      "entropy": 5.529582977294922,
      "epoch": 0.7470065312046444,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004951475822855029,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1870327353477478,
      "num_tokens": 17115469.0,
      "step": 8235
    },
    {
      "entropy": 5.464807987213135,
      "epoch": 0.7474600870827286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951408964308364,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1856093615293503,
      "num_tokens": 17125443.0,
      "step": 8240
    },
    {
      "entropy": 5.451996564865112,
      "epoch": 0.7479136429608128,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951342060236,
      "loss": 5.2143,
      "mean_token_accuracy": 0.18845360726118088,
      "num_tokens": 17136253.0,
      "step": 8245
    },
    {
      "entropy": 5.5428102016448975,
      "epoch": 0.7483671988388969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951275110639321,
      "loss": 5.275,
      "mean_token_accuracy": 0.18393643498420714,
      "num_tokens": 17145544.0,
      "step": 8250
    },
    {
      "entropy": 5.495501470565796,
      "epoch": 0.7488207547169812,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951208115519712,
      "loss": 5.2632,
      "mean_token_accuracy": 0.18630698323249817,
      "num_tokens": 17155043.0,
      "step": 8255
    },
    {
      "entropy": 5.464247703552246,
      "epoch": 0.7492743105950653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004951141074878558,
      "loss": 5.308,
      "mean_token_accuracy": 0.1831713780760765,
      "num_tokens": 17164885.0,
      "step": 8260
    },
    {
      "entropy": 5.5879473209381105,
      "epoch": 0.7497278664731495,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951073988717246,
      "loss": 5.3179,
      "mean_token_accuracy": 0.18246327489614486,
      "num_tokens": 17175681.0,
      "step": 8265
    },
    {
      "entropy": 5.449235534667968,
      "epoch": 0.7501814223512336,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951006857037163,
      "loss": 5.184,
      "mean_token_accuracy": 0.18728435188531875,
      "num_tokens": 17186002.0,
      "step": 8270
    },
    {
      "entropy": 5.469281768798828,
      "epoch": 0.7506349782293179,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950939679839696,
      "loss": 5.2305,
      "mean_token_accuracy": 0.1872097671031952,
      "num_tokens": 17195499.0,
      "step": 8275
    },
    {
      "entropy": 5.444141530990601,
      "epoch": 0.7510885341074021,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950872457126237,
      "loss": 5.24,
      "mean_token_accuracy": 0.18864094465970993,
      "num_tokens": 17204684.0,
      "step": 8280
    },
    {
      "entropy": 5.476453733444214,
      "epoch": 0.7515420899854862,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950805188898175,
      "loss": 5.2717,
      "mean_token_accuracy": 0.1793087586760521,
      "num_tokens": 17213987.0,
      "step": 8285
    },
    {
      "entropy": 5.512142181396484,
      "epoch": 0.7519956458635704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049507378751569,
      "loss": 5.2851,
      "mean_token_accuracy": 0.18152845948934554,
      "num_tokens": 17224512.0,
      "step": 8290
    },
    {
      "entropy": 5.550980758666992,
      "epoch": 0.7524492017416545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950670515903805,
      "loss": 5.3058,
      "mean_token_accuracy": 0.18777280449867248,
      "num_tokens": 17234362.0,
      "step": 8295
    },
    {
      "entropy": 5.5337158203125,
      "epoch": 0.7529027576197388,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950603111140283,
      "loss": 5.299,
      "mean_token_accuracy": 0.1910579890012741,
      "num_tokens": 17244241.0,
      "step": 8300
    },
    {
      "entropy": 5.466256761550904,
      "epoch": 0.753356313497823,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950535660867728,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19583994895219803,
      "num_tokens": 17254689.0,
      "step": 8305
    },
    {
      "entropy": 5.5242081642150875,
      "epoch": 0.7538098693759071,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950468165087535,
      "loss": 5.3076,
      "mean_token_accuracy": 0.18249109089374543,
      "num_tokens": 17265506.0,
      "step": 8310
    },
    {
      "entropy": 5.509704494476319,
      "epoch": 0.7542634252539913,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004950400623801098,
      "loss": 5.3849,
      "mean_token_accuracy": 0.18589098006486893,
      "num_tokens": 17276376.0,
      "step": 8315
    },
    {
      "entropy": 5.559766817092895,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950333037009818,
      "loss": 5.2798,
      "mean_token_accuracy": 0.175719191133976,
      "num_tokens": 17287005.0,
      "step": 8320
    },
    {
      "entropy": 5.537489891052246,
      "epoch": 0.7551705370101597,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950265404715089,
      "loss": 5.2791,
      "mean_token_accuracy": 0.18351641595363616,
      "num_tokens": 17297595.0,
      "step": 8325
    },
    {
      "entropy": 5.4321057319641115,
      "epoch": 0.7556240928882438,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495019772691831,
      "loss": 5.1553,
      "mean_token_accuracy": 0.18927870392799379,
      "num_tokens": 17307319.0,
      "step": 8330
    },
    {
      "entropy": 5.5272448539733885,
      "epoch": 0.756077648766328,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950130003620882,
      "loss": 5.3748,
      "mean_token_accuracy": 0.18703329414129258,
      "num_tokens": 17317978.0,
      "step": 8335
    },
    {
      "entropy": 5.455478620529175,
      "epoch": 0.7565312046444121,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950062234824204,
      "loss": 5.2836,
      "mean_token_accuracy": 0.1799110844731331,
      "num_tokens": 17328339.0,
      "step": 8340
    },
    {
      "entropy": 5.499893665313721,
      "epoch": 0.7569847605224964,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004949994420529679,
      "loss": 5.317,
      "mean_token_accuracy": 0.17965888679027558,
      "num_tokens": 17338668.0,
      "step": 8345
    },
    {
      "entropy": 5.474080753326416,
      "epoch": 0.7574383164005806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004949926560738709,
      "loss": 5.2619,
      "mean_token_accuracy": 0.17932692617177964,
      "num_tokens": 17348609.0,
      "step": 8350
    },
    {
      "entropy": 5.548439264297485,
      "epoch": 0.7578918722786647,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949858655452697,
      "loss": 5.2713,
      "mean_token_accuracy": 0.18396163880825042,
      "num_tokens": 17358014.0,
      "step": 8355
    },
    {
      "entropy": 5.541851425170899,
      "epoch": 0.7583454281567489,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949790704673048,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18673214316368103,
      "num_tokens": 17367975.0,
      "step": 8360
    },
    {
      "entropy": 5.51186900138855,
      "epoch": 0.7587989840348331,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949722708401166,
      "loss": 5.2276,
      "mean_token_accuracy": 0.18283698856830596,
      "num_tokens": 17377960.0,
      "step": 8365
    },
    {
      "entropy": 5.555224323272705,
      "epoch": 0.7592525399129173,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949654666638457,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1900601267814636,
      "num_tokens": 17387334.0,
      "step": 8370
    },
    {
      "entropy": 5.493747091293335,
      "epoch": 0.7597060957910015,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494958657938633,
      "loss": 5.2611,
      "mean_token_accuracy": 0.18564350605010987,
      "num_tokens": 17396521.0,
      "step": 8375
    },
    {
      "entropy": 5.420101881027222,
      "epoch": 0.7601596516690856,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949518446646192,
      "loss": 5.2373,
      "mean_token_accuracy": 0.18725302666425706,
      "num_tokens": 17407217.0,
      "step": 8380
    },
    {
      "entropy": 5.617413759231567,
      "epoch": 0.7606132075471698,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949450268419451,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17854755967855454,
      "num_tokens": 17416982.0,
      "step": 8385
    },
    {
      "entropy": 5.44600477218628,
      "epoch": 0.761066763425254,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494938204470752,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19048412889242172,
      "num_tokens": 17427636.0,
      "step": 8390
    },
    {
      "entropy": 5.491252088546753,
      "epoch": 0.7615203193033382,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949313775511806,
      "loss": 5.3197,
      "mean_token_accuracy": 0.17598724365234375,
      "num_tokens": 17437488.0,
      "step": 8395
    },
    {
      "entropy": 5.467138290405273,
      "epoch": 0.7619738751814223,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949245460833723,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19133297652006148,
      "num_tokens": 17447838.0,
      "step": 8400
    },
    {
      "entropy": 5.547179698944092,
      "epoch": 0.7624274310595065,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949177100674684,
      "loss": 5.3324,
      "mean_token_accuracy": 0.1857564628124237,
      "num_tokens": 17457257.0,
      "step": 8405
    },
    {
      "entropy": 5.489846420288086,
      "epoch": 0.7628809869375908,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004949108695036102,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18621718734502793,
      "num_tokens": 17467729.0,
      "step": 8410
    },
    {
      "entropy": 5.480379867553711,
      "epoch": 0.7633345428156749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949040243919392,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18046046197414398,
      "num_tokens": 17477692.0,
      "step": 8415
    },
    {
      "entropy": 5.470694160461425,
      "epoch": 0.7637880986937591,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948971747325968,
      "loss": 5.2666,
      "mean_token_accuracy": 0.18498124331235885,
      "num_tokens": 17488730.0,
      "step": 8420
    },
    {
      "entropy": 5.538025569915772,
      "epoch": 0.7642416545718432,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494890320525725,
      "loss": 5.3334,
      "mean_token_accuracy": 0.1852163478732109,
      "num_tokens": 17498374.0,
      "step": 8425
    },
    {
      "entropy": 5.477238607406616,
      "epoch": 0.7646952104499274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948834617714652,
      "loss": 5.192,
      "mean_token_accuracy": 0.18794307559728624,
      "num_tokens": 17508601.0,
      "step": 8430
    },
    {
      "entropy": 5.487441873550415,
      "epoch": 0.7651487663280117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948765984699595,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18388939797878265,
      "num_tokens": 17518164.0,
      "step": 8435
    },
    {
      "entropy": 5.520015287399292,
      "epoch": 0.7656023222060958,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948697306213494,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19726908057928086,
      "num_tokens": 17527244.0,
      "step": 8440
    },
    {
      "entropy": 5.4783806800842285,
      "epoch": 0.76605587808418,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948628582257775,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18394037932157517,
      "num_tokens": 17538045.0,
      "step": 8445
    },
    {
      "entropy": 5.4761889457702635,
      "epoch": 0.7665094339622641,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948559812833857,
      "loss": 5.2373,
      "mean_token_accuracy": 0.18717231154441832,
      "num_tokens": 17547584.0,
      "step": 8450
    },
    {
      "entropy": 5.484969186782837,
      "epoch": 0.7669629898403484,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004948490997943162,
      "loss": 5.2864,
      "mean_token_accuracy": 0.18479120582342148,
      "num_tokens": 17558925.0,
      "step": 8455
    },
    {
      "entropy": 5.506250715255737,
      "epoch": 0.7674165457184325,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948422137587111,
      "loss": 5.3651,
      "mean_token_accuracy": 0.17608188539743425,
      "num_tokens": 17569355.0,
      "step": 8460
    },
    {
      "entropy": 5.5164354801177975,
      "epoch": 0.7678701015965167,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948353231767131,
      "loss": 5.2246,
      "mean_token_accuracy": 0.18756223171949388,
      "num_tokens": 17579782.0,
      "step": 8465
    },
    {
      "entropy": 5.50532283782959,
      "epoch": 0.7683236574746009,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948284280484645,
      "loss": 5.2839,
      "mean_token_accuracy": 0.19143763184547424,
      "num_tokens": 17589772.0,
      "step": 8470
    },
    {
      "entropy": 5.406950950622559,
      "epoch": 0.768777213352685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004948215283741081,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1878790646791458,
      "num_tokens": 17599613.0,
      "step": 8475
    },
    {
      "entropy": 5.485948133468628,
      "epoch": 0.7692307692307693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948146241537866,
      "loss": 5.2951,
      "mean_token_accuracy": 0.1790087714791298,
      "num_tokens": 17609056.0,
      "step": 8480
    },
    {
      "entropy": 5.497835874557495,
      "epoch": 0.7696843251088534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948077153876425,
      "loss": 5.214,
      "mean_token_accuracy": 0.18795309066772461,
      "num_tokens": 17618620.0,
      "step": 8485
    },
    {
      "entropy": 5.446607255935669,
      "epoch": 0.7701378809869376,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948008020758188,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19390195310115815,
      "num_tokens": 17627630.0,
      "step": 8490
    },
    {
      "entropy": 5.36888632774353,
      "epoch": 0.7705914368650217,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004947938842184585,
      "loss": 5.1691,
      "mean_token_accuracy": 0.1886688217520714,
      "num_tokens": 17638903.0,
      "step": 8495
    },
    {
      "entropy": 5.444310665130615,
      "epoch": 0.771044992743106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947869618157048,
      "loss": 5.2229,
      "mean_token_accuracy": 0.18465732783079147,
      "num_tokens": 17649257.0,
      "step": 8500
    },
    {
      "entropy": 5.456356811523437,
      "epoch": 0.7714985486211902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947800348677006,
      "loss": 5.1987,
      "mean_token_accuracy": 0.19576434046030045,
      "num_tokens": 17659495.0,
      "step": 8505
    },
    {
      "entropy": 5.399709033966064,
      "epoch": 0.7719521044992743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947731033745893,
      "loss": 5.1733,
      "mean_token_accuracy": 0.18452270925045014,
      "num_tokens": 17669671.0,
      "step": 8510
    },
    {
      "entropy": 5.529618835449218,
      "epoch": 0.7724056603773585,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947661673365142,
      "loss": 5.3506,
      "mean_token_accuracy": 0.18083200752735137,
      "num_tokens": 17680927.0,
      "step": 8515
    },
    {
      "entropy": 5.483820056915283,
      "epoch": 0.7728592162554426,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947592267536188,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18810635805130005,
      "num_tokens": 17691378.0,
      "step": 8520
    },
    {
      "entropy": 5.477559185028076,
      "epoch": 0.7733127721335269,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947522816260466,
      "loss": 5.2962,
      "mean_token_accuracy": 0.1867695137858391,
      "num_tokens": 17700942.0,
      "step": 8525
    },
    {
      "entropy": 5.467412757873535,
      "epoch": 0.773766328011611,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004947453319539411,
      "loss": 5.2025,
      "mean_token_accuracy": 0.18547020852565765,
      "num_tokens": 17710076.0,
      "step": 8530
    },
    {
      "entropy": 5.474762058258056,
      "epoch": 0.7742198838896952,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947383777374462,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1894417494535446,
      "num_tokens": 17721009.0,
      "step": 8535
    },
    {
      "entropy": 5.449787902832031,
      "epoch": 0.7746734397677794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947314189767058,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18772920221090317,
      "num_tokens": 17731821.0,
      "step": 8540
    },
    {
      "entropy": 5.430331468582153,
      "epoch": 0.7751269956458636,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947244556718635,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19465826600790023,
      "num_tokens": 17741425.0,
      "step": 8545
    },
    {
      "entropy": 5.462887001037598,
      "epoch": 0.7755805515239478,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947174878230635,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1927179977297783,
      "num_tokens": 17752381.0,
      "step": 8550
    },
    {
      "entropy": 5.409137153625489,
      "epoch": 0.7760341074020319,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947105154304498,
      "loss": 5.1963,
      "mean_token_accuracy": 0.187833933532238,
      "num_tokens": 17762768.0,
      "step": 8555
    },
    {
      "entropy": 5.462284898757934,
      "epoch": 0.7764876632801161,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947035384941667,
      "loss": 5.2253,
      "mean_token_accuracy": 0.18217650353908538,
      "num_tokens": 17774032.0,
      "step": 8560
    },
    {
      "entropy": 5.554944324493408,
      "epoch": 0.7769412191582002,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946965570143584,
      "loss": 5.3676,
      "mean_token_accuracy": 0.1783722385764122,
      "num_tokens": 17785098.0,
      "step": 8565
    },
    {
      "entropy": 5.524931526184082,
      "epoch": 0.7773947750362845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946895709911695,
      "loss": 5.2255,
      "mean_token_accuracy": 0.18829830288887023,
      "num_tokens": 17795487.0,
      "step": 8570
    },
    {
      "entropy": 5.490832090377808,
      "epoch": 0.7778483309143687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946825804247442,
      "loss": 5.345,
      "mean_token_accuracy": 0.17945600003004075,
      "num_tokens": 17806045.0,
      "step": 8575
    },
    {
      "entropy": 5.543007564544678,
      "epoch": 0.7783018867924528,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494675585315227,
      "loss": 5.2967,
      "mean_token_accuracy": 0.17957210540771484,
      "num_tokens": 17816948.0,
      "step": 8580
    },
    {
      "entropy": 5.540530586242676,
      "epoch": 0.778755442670537,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004946685856627628,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18615331947803498,
      "num_tokens": 17826929.0,
      "step": 8585
    },
    {
      "entropy": 5.514763975143433,
      "epoch": 0.7792089985486212,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946615814674963,
      "loss": 5.3319,
      "mean_token_accuracy": 0.18369301408529282,
      "num_tokens": 17837780.0,
      "step": 8590
    },
    {
      "entropy": 5.387348318099976,
      "epoch": 0.7796625544267054,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946545727295724,
      "loss": 5.1051,
      "mean_token_accuracy": 0.19867068976163865,
      "num_tokens": 17847835.0,
      "step": 8595
    },
    {
      "entropy": 5.452921438217163,
      "epoch": 0.7801161103047896,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946475594491359,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19558655917644502,
      "num_tokens": 17858324.0,
      "step": 8600
    },
    {
      "entropy": 5.481823539733886,
      "epoch": 0.7805696661828737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946405416263319,
      "loss": 5.2549,
      "mean_token_accuracy": 0.18132758736610413,
      "num_tokens": 17868455.0,
      "step": 8605
    },
    {
      "entropy": 5.487615919113159,
      "epoch": 0.7810232220609579,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946335192613054,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18736030757427216,
      "num_tokens": 17879038.0,
      "step": 8610
    },
    {
      "entropy": 5.404131507873535,
      "epoch": 0.7814767779390421,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946264923542019,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1924211636185646,
      "num_tokens": 17888153.0,
      "step": 8615
    },
    {
      "entropy": 5.440420913696289,
      "epoch": 0.7819303338171263,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946194609051664,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1910945162177086,
      "num_tokens": 17898289.0,
      "step": 8620
    },
    {
      "entropy": 5.474812173843384,
      "epoch": 0.7823838896952104,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946124249143447,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1902774378657341,
      "num_tokens": 17909192.0,
      "step": 8625
    },
    {
      "entropy": 5.445071649551392,
      "epoch": 0.7828374455732946,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494605384381882,
      "loss": 5.1791,
      "mean_token_accuracy": 0.1815720483660698,
      "num_tokens": 17918937.0,
      "step": 8630
    },
    {
      "entropy": 5.420489835739136,
      "epoch": 0.7832910014513788,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004945983393079239,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19512285739183427,
      "num_tokens": 17929346.0,
      "step": 8635
    },
    {
      "entropy": 5.47286205291748,
      "epoch": 0.783744557329463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945912896926163,
      "loss": 5.2335,
      "mean_token_accuracy": 0.18775832206010817,
      "num_tokens": 17939479.0,
      "step": 8640
    },
    {
      "entropy": 5.573257970809936,
      "epoch": 0.7841981132075472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945842355361049,
      "loss": 5.3215,
      "mean_token_accuracy": 0.1810726061463356,
      "num_tokens": 17949841.0,
      "step": 8645
    },
    {
      "entropy": 5.429520606994629,
      "epoch": 0.7846516690856313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945771768385355,
      "loss": 5.2357,
      "mean_token_accuracy": 0.19100900441408158,
      "num_tokens": 17960334.0,
      "step": 8650
    },
    {
      "entropy": 5.42313871383667,
      "epoch": 0.7851052249637155,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000494570113600054,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1927325651049614,
      "num_tokens": 17970595.0,
      "step": 8655
    },
    {
      "entropy": 5.488664007186889,
      "epoch": 0.7855587808417998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945630458208068,
      "loss": 5.3267,
      "mean_token_accuracy": 0.17650770395994186,
      "num_tokens": 17981301.0,
      "step": 8660
    },
    {
      "entropy": 5.435780143737793,
      "epoch": 0.7860123367198839,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945559735009397,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1920144185423851,
      "num_tokens": 17991262.0,
      "step": 8665
    },
    {
      "entropy": 5.522634553909302,
      "epoch": 0.7864658925979681,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945488966405993,
      "loss": 5.1962,
      "mean_token_accuracy": 0.1927980050444603,
      "num_tokens": 18001212.0,
      "step": 8670
    },
    {
      "entropy": 5.468816947937012,
      "epoch": 0.7869194484760522,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945418152399317,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19336934387683868,
      "num_tokens": 18010432.0,
      "step": 8675
    },
    {
      "entropy": 5.452950716018677,
      "epoch": 0.7873730043541364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004945347292990833,
      "loss": 5.2545,
      "mean_token_accuracy": 0.17933656126260758,
      "num_tokens": 18020902.0,
      "step": 8680
    },
    {
      "entropy": 5.4519189357757565,
      "epoch": 0.7878265602322206,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945276388182008,
      "loss": 5.1303,
      "mean_token_accuracy": 0.1907524511218071,
      "num_tokens": 18031133.0,
      "step": 8685
    },
    {
      "entropy": 5.5043336868286135,
      "epoch": 0.7882801161103048,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945205437974309,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19138137102127076,
      "num_tokens": 18042050.0,
      "step": 8690
    },
    {
      "entropy": 5.376919841766357,
      "epoch": 0.788733671988389,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945134442369201,
      "loss": 5.201,
      "mean_token_accuracy": 0.18819751739501953,
      "num_tokens": 18051642.0,
      "step": 8695
    },
    {
      "entropy": 5.506156253814697,
      "epoch": 0.7891872278664731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004945063401368154,
      "loss": 5.2433,
      "mean_token_accuracy": 0.18314868658781053,
      "num_tokens": 18062180.0,
      "step": 8700
    },
    {
      "entropy": 5.473495769500732,
      "epoch": 0.7896407837445574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004944992314972637,
      "loss": 5.2047,
      "mean_token_accuracy": 0.18621009290218354,
      "num_tokens": 18072302.0,
      "step": 8705
    },
    {
      "entropy": 5.4587818622589115,
      "epoch": 0.7900943396226415,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944921183184118,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1914062112569809,
      "num_tokens": 18081824.0,
      "step": 8710
    },
    {
      "entropy": 5.457731771469116,
      "epoch": 0.7905478955007257,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000494485000600407,
      "loss": 5.2583,
      "mean_token_accuracy": 0.19050151854753494,
      "num_tokens": 18092696.0,
      "step": 8715
    },
    {
      "entropy": 5.413540267944336,
      "epoch": 0.7910014513788098,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004944778783433965,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19154656678438187,
      "num_tokens": 18101790.0,
      "step": 8720
    },
    {
      "entropy": 5.525717163085938,
      "epoch": 0.791455007256894,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944707515475276,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1849859058856964,
      "num_tokens": 18111572.0,
      "step": 8725
    },
    {
      "entropy": 5.505932140350342,
      "epoch": 0.7919085631349783,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944636202129475,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1949717327952385,
      "num_tokens": 18120650.0,
      "step": 8730
    },
    {
      "entropy": 5.435335779190064,
      "epoch": 0.7923621190130624,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944564843398039,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1990776017308235,
      "num_tokens": 18131055.0,
      "step": 8735
    },
    {
      "entropy": 5.546191930770874,
      "epoch": 0.7928156748911466,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944493439282443,
      "loss": 5.2667,
      "mean_token_accuracy": 0.18769678622484207,
      "num_tokens": 18141423.0,
      "step": 8740
    },
    {
      "entropy": 5.409640216827393,
      "epoch": 0.7932692307692307,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004944421989784162,
      "loss": 5.1811,
      "mean_token_accuracy": 0.18719988465309143,
      "num_tokens": 18152015.0,
      "step": 8745
    },
    {
      "entropy": 5.4817914962768555,
      "epoch": 0.793722786647315,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944350494904676,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18807127922773362,
      "num_tokens": 18161934.0,
      "step": 8750
    },
    {
      "entropy": 5.461260223388672,
      "epoch": 0.7941763425253991,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004944278954645463,
      "loss": 5.1952,
      "mean_token_accuracy": 0.18708185702562333,
      "num_tokens": 18172287.0,
      "step": 8755
    },
    {
      "entropy": 5.47312331199646,
      "epoch": 0.7946298984034833,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004944207369008002,
      "loss": 5.2517,
      "mean_token_accuracy": 0.19569893479347228,
      "num_tokens": 18181015.0,
      "step": 8760
    },
    {
      "entropy": 5.543885898590088,
      "epoch": 0.7950834542815675,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004944135737993773,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18109244257211685,
      "num_tokens": 18192382.0,
      "step": 8765
    },
    {
      "entropy": 5.452446174621582,
      "epoch": 0.7955370101596516,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944064061604258,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19768597930669785,
      "num_tokens": 18202597.0,
      "step": 8770
    },
    {
      "entropy": 5.4939878463745115,
      "epoch": 0.7959905660377359,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943992339840939,
      "loss": 5.264,
      "mean_token_accuracy": 0.18397243022918702,
      "num_tokens": 18213756.0,
      "step": 8775
    },
    {
      "entropy": 5.4102050304412845,
      "epoch": 0.79644412191582,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049439205727053,
      "loss": 5.1772,
      "mean_token_accuracy": 0.18465391248464585,
      "num_tokens": 18223669.0,
      "step": 8780
    },
    {
      "entropy": 5.444017505645752,
      "epoch": 0.7968976777939042,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943848760198823,
      "loss": 5.202,
      "mean_token_accuracy": 0.19344062507152557,
      "num_tokens": 18234884.0,
      "step": 8785
    },
    {
      "entropy": 5.4118711948394775,
      "epoch": 0.7973512336719883,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943776902322995,
      "loss": 5.1329,
      "mean_token_accuracy": 0.19479400366544725,
      "num_tokens": 18245155.0,
      "step": 8790
    },
    {
      "entropy": 5.349363708496094,
      "epoch": 0.7978047895500726,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943704999079302,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19781140685081483,
      "num_tokens": 18254686.0,
      "step": 8795
    },
    {
      "entropy": 5.433163785934449,
      "epoch": 0.7982583454281568,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943633050469231,
      "loss": 5.1091,
      "mean_token_accuracy": 0.19448212683200836,
      "num_tokens": 18265473.0,
      "step": 8800
    },
    {
      "entropy": 5.5096200466156,
      "epoch": 0.7987119013062409,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943561056494269,
      "loss": 5.306,
      "mean_token_accuracy": 0.18551455140113832,
      "num_tokens": 18276024.0,
      "step": 8805
    },
    {
      "entropy": 5.466209077835083,
      "epoch": 0.7991654571843251,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004943489017155906,
      "loss": 5.2774,
      "mean_token_accuracy": 0.17907433956861496,
      "num_tokens": 18287261.0,
      "step": 8810
    },
    {
      "entropy": 5.505891227722168,
      "epoch": 0.7996190130624092,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943416932455631,
      "loss": 5.295,
      "mean_token_accuracy": 0.18324223309755325,
      "num_tokens": 18299374.0,
      "step": 8815
    },
    {
      "entropy": 5.438243436813354,
      "epoch": 0.8000725689404935,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943344802394935,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18090363442897797,
      "num_tokens": 18309344.0,
      "step": 8820
    },
    {
      "entropy": 5.50924162864685,
      "epoch": 0.8005261248185777,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943272626975309,
      "loss": 5.2282,
      "mean_token_accuracy": 0.1878451332449913,
      "num_tokens": 18319778.0,
      "step": 8825
    },
    {
      "entropy": 5.483279848098755,
      "epoch": 0.8009796806966618,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004943200406198246,
      "loss": 5.2424,
      "mean_token_accuracy": 0.18929843008518218,
      "num_tokens": 18329787.0,
      "step": 8830
    },
    {
      "entropy": 5.42649998664856,
      "epoch": 0.801433236574746,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494312814006524,
      "loss": 5.2021,
      "mean_token_accuracy": 0.1958215445280075,
      "num_tokens": 18340670.0,
      "step": 8835
    },
    {
      "entropy": 5.4468505859375,
      "epoch": 0.8018867924528302,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004943055828577784,
      "loss": 5.1937,
      "mean_token_accuracy": 0.1845944657921791,
      "num_tokens": 18352484.0,
      "step": 8840
    },
    {
      "entropy": 5.418262958526611,
      "epoch": 0.8023403483309144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942983471737376,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19169240295886994,
      "num_tokens": 18363103.0,
      "step": 8845
    },
    {
      "entropy": 5.37873272895813,
      "epoch": 0.8027939042089985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942911069545511,
      "loss": 5.2267,
      "mean_token_accuracy": 0.1946313798427582,
      "num_tokens": 18373925.0,
      "step": 8850
    },
    {
      "entropy": 5.49118709564209,
      "epoch": 0.8032474600870827,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942838622003686,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18206191956996917,
      "num_tokens": 18384081.0,
      "step": 8855
    },
    {
      "entropy": 5.541353750228882,
      "epoch": 0.8037010159651669,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942766129113399,
      "loss": 5.1539,
      "mean_token_accuracy": 0.19194665402173997,
      "num_tokens": 18392510.0,
      "step": 8860
    },
    {
      "entropy": 5.532504987716675,
      "epoch": 0.8041545718432511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942693590876151,
      "loss": 5.2968,
      "mean_token_accuracy": 0.18387772738933564,
      "num_tokens": 18402809.0,
      "step": 8865
    },
    {
      "entropy": 5.542223501205444,
      "epoch": 0.8046081277213353,
      "grad_norm": 1.09375,
      "learning_rate": 0.000494262100729344,
      "loss": 5.3171,
      "mean_token_accuracy": 0.18584613353013993,
      "num_tokens": 18413342.0,
      "step": 8870
    },
    {
      "entropy": 5.513447427749634,
      "epoch": 0.8050616835994194,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942548378366768,
      "loss": 5.3013,
      "mean_token_accuracy": 0.18480869382619858,
      "num_tokens": 18424735.0,
      "step": 8875
    },
    {
      "entropy": 5.482922840118408,
      "epoch": 0.8055152394775036,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942475704097635,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19012280106544494,
      "num_tokens": 18434797.0,
      "step": 8880
    },
    {
      "entropy": 5.44505763053894,
      "epoch": 0.8059687953555879,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942402984487548,
      "loss": 5.2439,
      "mean_token_accuracy": 0.18273085951805115,
      "num_tokens": 18445456.0,
      "step": 8885
    },
    {
      "entropy": 5.477434682846069,
      "epoch": 0.806422351233672,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942330219538008,
      "loss": 5.2316,
      "mean_token_accuracy": 0.1858235090970993,
      "num_tokens": 18456728.0,
      "step": 8890
    },
    {
      "entropy": 5.699557685852051,
      "epoch": 0.8068759071117562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000494225740925052,
      "loss": 5.3628,
      "mean_token_accuracy": 0.18159916400909423,
      "num_tokens": 18468813.0,
      "step": 8895
    },
    {
      "entropy": 5.512146091461181,
      "epoch": 0.8073294629898403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942184553626591,
      "loss": 5.2507,
      "mean_token_accuracy": 0.18834080696105956,
      "num_tokens": 18477912.0,
      "step": 8900
    },
    {
      "entropy": 5.40484938621521,
      "epoch": 0.8077830188679245,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942111652667726,
      "loss": 5.2008,
      "mean_token_accuracy": 0.1940469965338707,
      "num_tokens": 18488115.0,
      "step": 8905
    },
    {
      "entropy": 5.466488409042358,
      "epoch": 0.8082365747460087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942038706375433,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18689305633306502,
      "num_tokens": 18498853.0,
      "step": 8910
    },
    {
      "entropy": 5.460456275939942,
      "epoch": 0.8086901306240929,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941965714751222,
      "loss": 5.2812,
      "mean_token_accuracy": 0.18107305467128754,
      "num_tokens": 18510047.0,
      "step": 8915
    },
    {
      "entropy": 5.472559976577759,
      "epoch": 0.809143686502177,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941892677796601,
      "loss": 5.258,
      "mean_token_accuracy": 0.19089266210794448,
      "num_tokens": 18519673.0,
      "step": 8920
    },
    {
      "entropy": 5.475661754608154,
      "epoch": 0.8095972423802612,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004941819595513081,
      "loss": 5.2702,
      "mean_token_accuracy": 0.18570117205381392,
      "num_tokens": 18530090.0,
      "step": 8925
    },
    {
      "entropy": 5.567102146148682,
      "epoch": 0.8100507982583455,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941746467902173,
      "loss": 5.3165,
      "mean_token_accuracy": 0.17997350692749023,
      "num_tokens": 18541862.0,
      "step": 8930
    },
    {
      "entropy": 5.540829753875732,
      "epoch": 0.8105043541364296,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494167329496539,
      "loss": 5.3158,
      "mean_token_accuracy": 0.17742486000061036,
      "num_tokens": 18551624.0,
      "step": 8935
    },
    {
      "entropy": 5.449011516571045,
      "epoch": 0.8109579100145138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004941600076704244,
      "loss": 5.2422,
      "mean_token_accuracy": 0.18720260560512542,
      "num_tokens": 18561784.0,
      "step": 8940
    },
    {
      "entropy": 5.501886558532715,
      "epoch": 0.8114114658925979,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941526813120251,
      "loss": 5.3173,
      "mean_token_accuracy": 0.17972312569618226,
      "num_tokens": 18572110.0,
      "step": 8945
    },
    {
      "entropy": 5.539577198028565,
      "epoch": 0.8118650217706821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941453504214924,
      "loss": 5.2618,
      "mean_token_accuracy": 0.18143393248319625,
      "num_tokens": 18582012.0,
      "step": 8950
    },
    {
      "entropy": 5.399498987197876,
      "epoch": 0.8123185776487664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941380149989781,
      "loss": 5.0988,
      "mean_token_accuracy": 0.198584945499897,
      "num_tokens": 18592850.0,
      "step": 8955
    },
    {
      "entropy": 5.40629301071167,
      "epoch": 0.8127721335268505,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941306750446336,
      "loss": 5.2769,
      "mean_token_accuracy": 0.18415168970823287,
      "num_tokens": 18602590.0,
      "step": 8960
    },
    {
      "entropy": 5.431480407714844,
      "epoch": 0.8132256894049347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494123330558611,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19291519969701768,
      "num_tokens": 18612607.0,
      "step": 8965
    },
    {
      "entropy": 5.474500274658203,
      "epoch": 0.8136792452830188,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941159815410622,
      "loss": 5.166,
      "mean_token_accuracy": 0.19350909292697907,
      "num_tokens": 18622936.0,
      "step": 8970
    },
    {
      "entropy": 5.4926918029785154,
      "epoch": 0.8141328011611031,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494108627992139,
      "loss": 5.3095,
      "mean_token_accuracy": 0.17892413288354875,
      "num_tokens": 18633997.0,
      "step": 8975
    },
    {
      "entropy": 5.414634323120117,
      "epoch": 0.8145863570391872,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941012699119936,
      "loss": 5.1478,
      "mean_token_accuracy": 0.1955949380993843,
      "num_tokens": 18644347.0,
      "step": 8980
    },
    {
      "entropy": 5.4861256122589115,
      "epoch": 0.8150399129172714,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940939073007779,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1909121960401535,
      "num_tokens": 18654989.0,
      "step": 8985
    },
    {
      "entropy": 5.464990901947021,
      "epoch": 0.8154934687953556,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940865401586446,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19030434787273406,
      "num_tokens": 18665526.0,
      "step": 8990
    },
    {
      "entropy": 5.438522624969482,
      "epoch": 0.8159470246734397,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940791684857458,
      "loss": 5.2008,
      "mean_token_accuracy": 0.19021283239126205,
      "num_tokens": 18676226.0,
      "step": 8995
    },
    {
      "entropy": 5.461506366729736,
      "epoch": 0.816400580551524,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940717922822339,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1934579148888588,
      "num_tokens": 18686499.0,
      "step": 9000
    },
    {
      "epoch": 0.816400580551524,
      "eval_entropy": 5.295148694784933,
      "eval_loss": 5.239758491516113,
      "eval_mean_token_accuracy": 0.1953426175653867,
      "eval_num_tokens": 18686499.0,
      "eval_runtime": 20.5731,
      "eval_samples_per_second": 1718.747,
      "eval_steps_per_second": 214.843,
      "step": 9000
    },
    {
      "entropy": 5.512953996658325,
      "epoch": 0.8168541364296081,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004940644115482615,
      "loss": 5.2113,
      "mean_token_accuracy": 0.18784044086933135,
      "num_tokens": 18696543.0,
      "step": 9005
    },
    {
      "entropy": 5.411309862136841,
      "epoch": 0.8173076923076923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940570262839812,
      "loss": 5.2609,
      "mean_token_accuracy": 0.18269165456295014,
      "num_tokens": 18707029.0,
      "step": 9010
    },
    {
      "entropy": 5.458065605163574,
      "epoch": 0.8177612481857764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940496364895457,
      "loss": 5.2382,
      "mean_token_accuracy": 0.1901217520236969,
      "num_tokens": 18716937.0,
      "step": 9015
    },
    {
      "entropy": 5.589706897735596,
      "epoch": 0.8182148040638607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940422421651082,
      "loss": 5.2344,
      "mean_token_accuracy": 0.18793250024318695,
      "num_tokens": 18727478.0,
      "step": 9020
    },
    {
      "entropy": 5.449050188064575,
      "epoch": 0.8186683599419449,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940348433108211,
      "loss": 5.1626,
      "mean_token_accuracy": 0.18938371986150743,
      "num_tokens": 18737339.0,
      "step": 9025
    },
    {
      "entropy": 5.455372714996338,
      "epoch": 0.819121915820029,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940274399268377,
      "loss": 5.278,
      "mean_token_accuracy": 0.1877202123403549,
      "num_tokens": 18747810.0,
      "step": 9030
    },
    {
      "entropy": 5.471349382400513,
      "epoch": 0.8195754716981132,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940200320133109,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1876297190785408,
      "num_tokens": 18757564.0,
      "step": 9035
    },
    {
      "entropy": 5.529532814025879,
      "epoch": 0.8200290275761973,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940126195703941,
      "loss": 5.2642,
      "mean_token_accuracy": 0.18516496270895005,
      "num_tokens": 18768196.0,
      "step": 9040
    },
    {
      "entropy": 5.503971004486084,
      "epoch": 0.8204825834542816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940052025982405,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19783169627189637,
      "num_tokens": 18778350.0,
      "step": 9045
    },
    {
      "entropy": 5.457395792007446,
      "epoch": 0.8209361393323658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939977810970035,
      "loss": 5.1951,
      "mean_token_accuracy": 0.18221480399370193,
      "num_tokens": 18789825.0,
      "step": 9050
    },
    {
      "entropy": 5.469322443008423,
      "epoch": 0.8213896952104499,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939903550668366,
      "loss": 5.1901,
      "mean_token_accuracy": 0.189435413479805,
      "num_tokens": 18799793.0,
      "step": 9055
    },
    {
      "entropy": 5.452561235427856,
      "epoch": 0.8218432510885341,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939829245078934,
      "loss": 5.236,
      "mean_token_accuracy": 0.19126893430948258,
      "num_tokens": 18809771.0,
      "step": 9060
    },
    {
      "entropy": 5.465332746505737,
      "epoch": 0.8222968069666183,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939754894203275,
      "loss": 5.2524,
      "mean_token_accuracy": 0.1888252705335617,
      "num_tokens": 18820529.0,
      "step": 9065
    },
    {
      "entropy": 5.393265581130981,
      "epoch": 0.8227503628447025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939680498042926,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20150044560432434,
      "num_tokens": 18830797.0,
      "step": 9070
    },
    {
      "entropy": 5.421929931640625,
      "epoch": 0.8232039187227866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004939606056599428,
      "loss": 5.1794,
      "mean_token_accuracy": 0.19247416406869888,
      "num_tokens": 18840526.0,
      "step": 9075
    },
    {
      "entropy": 5.450492906570434,
      "epoch": 0.8236574746008708,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939531569874316,
      "loss": 5.2073,
      "mean_token_accuracy": 0.18442903757095336,
      "num_tokens": 18851201.0,
      "step": 9080
    },
    {
      "entropy": 5.366197156906128,
      "epoch": 0.824111030478955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939457037869135,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21173077523708345,
      "num_tokens": 18861676.0,
      "step": 9085
    },
    {
      "entropy": 5.387629175186158,
      "epoch": 0.8245645863570392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939382460585424,
      "loss": 5.1667,
      "mean_token_accuracy": 0.1845756560564041,
      "num_tokens": 18872669.0,
      "step": 9090
    },
    {
      "entropy": 5.379079055786133,
      "epoch": 0.8250181422351234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939307838024726,
      "loss": 5.2288,
      "mean_token_accuracy": 0.1850636750459671,
      "num_tokens": 18882130.0,
      "step": 9095
    },
    {
      "entropy": 5.423450231552124,
      "epoch": 0.8254716981132075,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939233170188585,
      "loss": 5.1872,
      "mean_token_accuracy": 0.19012628942728044,
      "num_tokens": 18892457.0,
      "step": 9100
    },
    {
      "entropy": 5.511678886413574,
      "epoch": 0.8259252539912917,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939158457078543,
      "loss": 5.2185,
      "mean_token_accuracy": 0.18623092919588088,
      "num_tokens": 18903152.0,
      "step": 9105
    },
    {
      "entropy": 5.443718433380127,
      "epoch": 0.826378809869376,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939083698696146,
      "loss": 5.1775,
      "mean_token_accuracy": 0.18756801933050155,
      "num_tokens": 18913740.0,
      "step": 9110
    },
    {
      "entropy": 5.431342744827271,
      "epoch": 0.8268323657474601,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493900889504294,
      "loss": 5.2354,
      "mean_token_accuracy": 0.18877937346696855,
      "num_tokens": 18923501.0,
      "step": 9115
    },
    {
      "entropy": 5.4776379585266115,
      "epoch": 0.8272859216255443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938934046120473,
      "loss": 5.223,
      "mean_token_accuracy": 0.1891995906829834,
      "num_tokens": 18933255.0,
      "step": 9120
    },
    {
      "entropy": 5.438175106048584,
      "epoch": 0.8277394775036284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938859151930293,
      "loss": 5.1909,
      "mean_token_accuracy": 0.18860452771186828,
      "num_tokens": 18943649.0,
      "step": 9125
    },
    {
      "entropy": 5.4302948951721195,
      "epoch": 0.8281930333817126,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938784212473947,
      "loss": 5.093,
      "mean_token_accuracy": 0.198411026597023,
      "num_tokens": 18953187.0,
      "step": 9130
    },
    {
      "entropy": 5.519911003112793,
      "epoch": 0.8286465892597968,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938709227752985,
      "loss": 5.1945,
      "mean_token_accuracy": 0.19375702738761902,
      "num_tokens": 18963015.0,
      "step": 9135
    },
    {
      "entropy": 5.448123931884766,
      "epoch": 0.829100145137881,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493863419776896,
      "loss": 5.2526,
      "mean_token_accuracy": 0.18133300095796584,
      "num_tokens": 18974184.0,
      "step": 9140
    },
    {
      "entropy": 5.43465895652771,
      "epoch": 0.8295537010159652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004938559122523421,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18903249651193618,
      "num_tokens": 18985036.0,
      "step": 9145
    },
    {
      "entropy": 5.530106878280639,
      "epoch": 0.8300072568940493,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938484002017921,
      "loss": 5.3025,
      "mean_token_accuracy": 0.18530908972024918,
      "num_tokens": 18995237.0,
      "step": 9150
    },
    {
      "entropy": 5.515827941894531,
      "epoch": 0.8304608127721336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938408836254015,
      "loss": 5.3055,
      "mean_token_accuracy": 0.17978544384241105,
      "num_tokens": 19006272.0,
      "step": 9155
    },
    {
      "entropy": 5.436789894104004,
      "epoch": 0.8309143686502177,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938333625233258,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19587559551000594,
      "num_tokens": 19016728.0,
      "step": 9160
    },
    {
      "entropy": 5.479961681365967,
      "epoch": 0.8313679245283019,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938258368957202,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1806613489985466,
      "num_tokens": 19028024.0,
      "step": 9165
    },
    {
      "entropy": 5.475996828079223,
      "epoch": 0.831821480406386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938183067427406,
      "loss": 5.206,
      "mean_token_accuracy": 0.19370831400156022,
      "num_tokens": 19038218.0,
      "step": 9170
    },
    {
      "entropy": 5.427654123306274,
      "epoch": 0.8322750362844702,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938107720645426,
      "loss": 5.246,
      "mean_token_accuracy": 0.18549474328756332,
      "num_tokens": 19048574.0,
      "step": 9175
    },
    {
      "entropy": 5.439662885665894,
      "epoch": 0.8327285921625545,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938032328612821,
      "loss": 5.2156,
      "mean_token_accuracy": 0.187961608171463,
      "num_tokens": 19057721.0,
      "step": 9180
    },
    {
      "entropy": 5.447835779190063,
      "epoch": 0.8331821480406386,
      "grad_norm": 0.96875,
      "learning_rate": 0.000493795689133115,
      "loss": 5.1821,
      "mean_token_accuracy": 0.18198906183242797,
      "num_tokens": 19067582.0,
      "step": 9185
    },
    {
      "entropy": 5.447179794311523,
      "epoch": 0.8336357039187228,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004937881408801974,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19615552127361296,
      "num_tokens": 19077896.0,
      "step": 9190
    },
    {
      "entropy": 5.443184280395508,
      "epoch": 0.8340892597968069,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493780588102685,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19482969790697097,
      "num_tokens": 19088082.0,
      "step": 9195
    },
    {
      "entropy": 5.4979020118713375,
      "epoch": 0.8345428156748912,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937730308007346,
      "loss": 5.2671,
      "mean_token_accuracy": 0.18952690958976745,
      "num_tokens": 19097176.0,
      "step": 9200
    },
    {
      "entropy": 5.460204553604126,
      "epoch": 0.8349963715529753,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004937654689745021,
      "loss": 5.1457,
      "mean_token_accuracy": 0.19767686128616332,
      "num_tokens": 19107835.0,
      "step": 9205
    },
    {
      "entropy": 5.40765962600708,
      "epoch": 0.8354499274310595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004937579026241439,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19384574443101882,
      "num_tokens": 19118028.0,
      "step": 9210
    },
    {
      "entropy": 5.465951061248779,
      "epoch": 0.8359034833091437,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937503317498165,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19141697138547897,
      "num_tokens": 19129223.0,
      "step": 9215
    },
    {
      "entropy": 5.4654120922088625,
      "epoch": 0.8363570391872278,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004937427563516765,
      "loss": 5.3005,
      "mean_token_accuracy": 0.17993318885564805,
      "num_tokens": 19140048.0,
      "step": 9220
    },
    {
      "entropy": 5.52028489112854,
      "epoch": 0.8368105950653121,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937351764298805,
      "loss": 5.279,
      "mean_token_accuracy": 0.17879910469055177,
      "num_tokens": 19150853.0,
      "step": 9225
    },
    {
      "entropy": 5.499497604370117,
      "epoch": 0.8372641509433962,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937275919845855,
      "loss": 5.2588,
      "mean_token_accuracy": 0.18826455175876616,
      "num_tokens": 19161645.0,
      "step": 9230
    },
    {
      "entropy": 5.476880359649658,
      "epoch": 0.8377177068214804,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937200030159481,
      "loss": 5.3458,
      "mean_token_accuracy": 0.17791823595762252,
      "num_tokens": 19173276.0,
      "step": 9235
    },
    {
      "entropy": 5.52899203300476,
      "epoch": 0.8381712626995645,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937124095241252,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19551825225353242,
      "num_tokens": 19182949.0,
      "step": 9240
    },
    {
      "entropy": 5.539991760253907,
      "epoch": 0.8386248185776488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937048115092739,
      "loss": 5.2211,
      "mean_token_accuracy": 0.18341253250837325,
      "num_tokens": 19193504.0,
      "step": 9245
    },
    {
      "entropy": 5.45371150970459,
      "epoch": 0.839078374455733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936972089715515,
      "loss": 5.2937,
      "mean_token_accuracy": 0.18242430686950684,
      "num_tokens": 19203406.0,
      "step": 9250
    },
    {
      "entropy": 5.4398195266723635,
      "epoch": 0.8395319303338171,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493689601911115,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18913086950778962,
      "num_tokens": 19214325.0,
      "step": 9255
    },
    {
      "entropy": 5.470502233505249,
      "epoch": 0.8399854862119013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936819903281219,
      "loss": 5.2876,
      "mean_token_accuracy": 0.17488500624895095,
      "num_tokens": 19226466.0,
      "step": 9260
    },
    {
      "entropy": 5.565874528884888,
      "epoch": 0.8404390420899854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936743742227295,
      "loss": 5.3116,
      "mean_token_accuracy": 0.18711801916360854,
      "num_tokens": 19237032.0,
      "step": 9265
    },
    {
      "entropy": 5.536182641983032,
      "epoch": 0.8408925979680697,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936667535950952,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1812775194644928,
      "num_tokens": 19248009.0,
      "step": 9270
    },
    {
      "entropy": 5.523895597457885,
      "epoch": 0.8413461538461539,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936591284453768,
      "loss": 5.3465,
      "mean_token_accuracy": 0.1860367998480797,
      "num_tokens": 19259363.0,
      "step": 9275
    },
    {
      "entropy": 5.519369649887085,
      "epoch": 0.841799709724238,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936514987737319,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1884769007563591,
      "num_tokens": 19269191.0,
      "step": 9280
    },
    {
      "entropy": 5.515477180480957,
      "epoch": 0.8422532656023222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004936438645803182,
      "loss": 5.2103,
      "mean_token_accuracy": 0.18393854945898055,
      "num_tokens": 19280278.0,
      "step": 9285
    },
    {
      "entropy": 5.453181600570678,
      "epoch": 0.8427068214804064,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936362258652938,
      "loss": 5.2026,
      "mean_token_accuracy": 0.19249465465545654,
      "num_tokens": 19290130.0,
      "step": 9290
    },
    {
      "entropy": 5.559951686859131,
      "epoch": 0.8431603773584906,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936285826288165,
      "loss": 5.3212,
      "mean_token_accuracy": 0.19275434464216232,
      "num_tokens": 19301669.0,
      "step": 9295
    },
    {
      "entropy": 5.448169231414795,
      "epoch": 0.8436139332365747,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004936209348710444,
      "loss": 5.1782,
      "mean_token_accuracy": 0.19250596314668655,
      "num_tokens": 19312490.0,
      "step": 9300
    },
    {
      "entropy": 5.4438097953796385,
      "epoch": 0.8440674891146589,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004936132825921355,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18785629719495772,
      "num_tokens": 19321743.0,
      "step": 9305
    },
    {
      "entropy": 5.431347608566284,
      "epoch": 0.844521044992743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936056257922484,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1909039482474327,
      "num_tokens": 19332149.0,
      "step": 9310
    },
    {
      "entropy": 5.466032552719116,
      "epoch": 0.8449746008708273,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004935979644715412,
      "loss": 5.2232,
      "mean_token_accuracy": 0.1855064868927002,
      "num_tokens": 19341655.0,
      "step": 9315
    },
    {
      "entropy": 5.519804620742798,
      "epoch": 0.8454281567489115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004935902986301724,
      "loss": 5.2669,
      "mean_token_accuracy": 0.18666884005069734,
      "num_tokens": 19351404.0,
      "step": 9320
    },
    {
      "entropy": 5.51761212348938,
      "epoch": 0.8458817126269956,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935826282683005,
      "loss": 5.306,
      "mean_token_accuracy": 0.18108808994293213,
      "num_tokens": 19361788.0,
      "step": 9325
    },
    {
      "entropy": 5.4535401344299315,
      "epoch": 0.8463352685050798,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935749533860842,
      "loss": 5.2601,
      "mean_token_accuracy": 0.1881358101963997,
      "num_tokens": 19372902.0,
      "step": 9330
    },
    {
      "entropy": 5.440549039840699,
      "epoch": 0.846788824383164,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493567273983682,
      "loss": 5.1421,
      "mean_token_accuracy": 0.18869245946407318,
      "num_tokens": 19382181.0,
      "step": 9335
    },
    {
      "entropy": 5.489796781539917,
      "epoch": 0.8472423802612482,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000493559590061253,
      "loss": 5.2085,
      "mean_token_accuracy": 0.18647148460149765,
      "num_tokens": 19393692.0,
      "step": 9340
    },
    {
      "entropy": 5.487021207809448,
      "epoch": 0.8476959361393324,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935519016189559,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19533468037843704,
      "num_tokens": 19404587.0,
      "step": 9345
    },
    {
      "entropy": 5.476869916915893,
      "epoch": 0.8481494920174165,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935442086569498,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1869882181286812,
      "num_tokens": 19415296.0,
      "step": 9350
    },
    {
      "entropy": 5.532775974273681,
      "epoch": 0.8486030478955007,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935365111753939,
      "loss": 5.2712,
      "mean_token_accuracy": 0.18344380110502242,
      "num_tokens": 19426119.0,
      "step": 9355
    },
    {
      "entropy": 5.515724086761475,
      "epoch": 0.8490566037735849,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935288091744472,
      "loss": 5.2665,
      "mean_token_accuracy": 0.18467478454113007,
      "num_tokens": 19436828.0,
      "step": 9360
    },
    {
      "entropy": 5.534423828125,
      "epoch": 0.8495101596516691,
      "grad_norm": 1.0,
      "learning_rate": 0.000493521102654269,
      "loss": 5.2324,
      "mean_token_accuracy": 0.188790924847126,
      "num_tokens": 19446583.0,
      "step": 9365
    },
    {
      "entropy": 5.423384094238282,
      "epoch": 0.8499637155297532,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935133916150187,
      "loss": 5.1168,
      "mean_token_accuracy": 0.19754848927259444,
      "num_tokens": 19456757.0,
      "step": 9370
    },
    {
      "entropy": 5.427696084976196,
      "epoch": 0.8504172714078374,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004935056760568558,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18450350314378738,
      "num_tokens": 19467624.0,
      "step": 9375
    },
    {
      "entropy": 5.441474723815918,
      "epoch": 0.8508708272859217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934979559799399,
      "loss": 5.157,
      "mean_token_accuracy": 0.18935077488422394,
      "num_tokens": 19477759.0,
      "step": 9380
    },
    {
      "entropy": 5.460384941101074,
      "epoch": 0.8513243831640058,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934902313844306,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1911305770277977,
      "num_tokens": 19488877.0,
      "step": 9385
    },
    {
      "entropy": 5.406986904144287,
      "epoch": 0.85177793904209,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934825022704875,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1952897772192955,
      "num_tokens": 19499170.0,
      "step": 9390
    },
    {
      "entropy": 5.454417419433594,
      "epoch": 0.8522314949201741,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934747686382707,
      "loss": 5.2065,
      "mean_token_accuracy": 0.1808039888739586,
      "num_tokens": 19509944.0,
      "step": 9395
    },
    {
      "entropy": 5.47445969581604,
      "epoch": 0.8526850507982583,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004934670304879398,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19298242628574372,
      "num_tokens": 19519014.0,
      "step": 9400
    },
    {
      "entropy": 5.4161628723144535,
      "epoch": 0.8531386066763426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934592878196553,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1934538498520851,
      "num_tokens": 19528586.0,
      "step": 9405
    },
    {
      "entropy": 5.484393167495727,
      "epoch": 0.8535921625544267,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493451540633577,
      "loss": 5.3113,
      "mean_token_accuracy": 0.18901212215423585,
      "num_tokens": 19539912.0,
      "step": 9410
    },
    {
      "entropy": 5.476560163497925,
      "epoch": 0.8540457184325109,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934437889298651,
      "loss": 5.2646,
      "mean_token_accuracy": 0.17929179668426515,
      "num_tokens": 19549860.0,
      "step": 9415
    },
    {
      "entropy": 5.498154592514038,
      "epoch": 0.854499274310595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934360327086799,
      "loss": 5.2267,
      "mean_token_accuracy": 0.184198497235775,
      "num_tokens": 19559446.0,
      "step": 9420
    },
    {
      "entropy": 5.525316429138184,
      "epoch": 0.8549528301886793,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493428271970182,
      "loss": 5.2297,
      "mean_token_accuracy": 0.18811480402946473,
      "num_tokens": 19570282.0,
      "step": 9425
    },
    {
      "entropy": 5.475820112228393,
      "epoch": 0.8554063860667634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934205067145318,
      "loss": 5.1505,
      "mean_token_accuracy": 0.18823657035827637,
      "num_tokens": 19581087.0,
      "step": 9430
    },
    {
      "entropy": 5.425998735427856,
      "epoch": 0.8558599419448476,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934127369418898,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18897894769906998,
      "num_tokens": 19592100.0,
      "step": 9435
    },
    {
      "entropy": 5.4478099822998045,
      "epoch": 0.8563134978229318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934049626524168,
      "loss": 5.233,
      "mean_token_accuracy": 0.18636782914400102,
      "num_tokens": 19602557.0,
      "step": 9440
    },
    {
      "entropy": 5.522075128555298,
      "epoch": 0.8567670537010159,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933971838462734,
      "loss": 5.2775,
      "mean_token_accuracy": 0.18067894577980043,
      "num_tokens": 19613643.0,
      "step": 9445
    },
    {
      "entropy": 5.467205333709717,
      "epoch": 0.8572206095791002,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004933894005236207,
      "loss": 5.1977,
      "mean_token_accuracy": 0.18942511677742005,
      "num_tokens": 19624180.0,
      "step": 9450
    },
    {
      "entropy": 5.476292467117309,
      "epoch": 0.8576741654571843,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004933816126846195,
      "loss": 5.2077,
      "mean_token_accuracy": 0.189624784886837,
      "num_tokens": 19634799.0,
      "step": 9455
    },
    {
      "entropy": 5.411914920806884,
      "epoch": 0.8581277213352685,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933738203294309,
      "loss": 5.1713,
      "mean_token_accuracy": 0.200113882124424,
      "num_tokens": 19644618.0,
      "step": 9460
    },
    {
      "entropy": 5.411866044998169,
      "epoch": 0.8585812772133526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493366023458216,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18326595425605774,
      "num_tokens": 19655550.0,
      "step": 9465
    },
    {
      "entropy": 5.4921434879302975,
      "epoch": 0.8590348330914369,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933582220711362,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19442976713180543,
      "num_tokens": 19665734.0,
      "step": 9470
    },
    {
      "entropy": 5.509519481658936,
      "epoch": 0.8594883889695211,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004933504161683526,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18399560004472731,
      "num_tokens": 19677130.0,
      "step": 9475
    },
    {
      "entropy": 5.414972305297852,
      "epoch": 0.8599419448476052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933426057500269,
      "loss": 5.1227,
      "mean_token_accuracy": 0.19844300150871277,
      "num_tokens": 19687374.0,
      "step": 9480
    },
    {
      "entropy": 5.498849821090698,
      "epoch": 0.8603955007256894,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933347908163203,
      "loss": 5.2317,
      "mean_token_accuracy": 0.18636648654937743,
      "num_tokens": 19698182.0,
      "step": 9485
    },
    {
      "entropy": 5.457022476196289,
      "epoch": 0.8608490566037735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933269713673947,
      "loss": 5.1897,
      "mean_token_accuracy": 0.18997897803783417,
      "num_tokens": 19707664.0,
      "step": 9490
    },
    {
      "entropy": 5.358859634399414,
      "epoch": 0.8613026124818578,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933191474034116,
      "loss": 5.2329,
      "mean_token_accuracy": 0.18852289021015167,
      "num_tokens": 19717647.0,
      "step": 9495
    },
    {
      "entropy": 5.440193605422974,
      "epoch": 0.861756168359942,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493311318924533,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18900999724864959,
      "num_tokens": 19727444.0,
      "step": 9500
    },
    {
      "entropy": 5.504938125610352,
      "epoch": 0.8622097242380261,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933034859309208,
      "loss": 5.2762,
      "mean_token_accuracy": 0.17947544306516647,
      "num_tokens": 19738719.0,
      "step": 9505
    },
    {
      "entropy": 5.543647336959839,
      "epoch": 0.8626632801161103,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004932956484227367,
      "loss": 5.2594,
      "mean_token_accuracy": 0.1917922765016556,
      "num_tokens": 19749786.0,
      "step": 9510
    },
    {
      "entropy": 5.467603588104248,
      "epoch": 0.8631168359941945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493287806400143,
      "loss": 5.2005,
      "mean_token_accuracy": 0.19913371056318283,
      "num_tokens": 19759623.0,
      "step": 9515
    },
    {
      "entropy": 5.470474147796631,
      "epoch": 0.8635703918722787,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004932799598633018,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18518185019493102,
      "num_tokens": 19770117.0,
      "step": 9520
    },
    {
      "entropy": 5.495015096664429,
      "epoch": 0.8640239477503628,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932721088123753,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1860126957297325,
      "num_tokens": 19780748.0,
      "step": 9525
    },
    {
      "entropy": 5.509607982635498,
      "epoch": 0.864477503628447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004932642532475262,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19054854065179824,
      "num_tokens": 19791816.0,
      "step": 9530
    },
    {
      "entropy": 5.537548017501831,
      "epoch": 0.8649310595065312,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004932563931689168,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18602988123893738,
      "num_tokens": 19801096.0,
      "step": 9535
    },
    {
      "entropy": 5.464199876785278,
      "epoch": 0.8653846153846154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004932485285767093,
      "loss": 5.1812,
      "mean_token_accuracy": 0.19209689795970916,
      "num_tokens": 19811258.0,
      "step": 9540
    },
    {
      "entropy": 5.499051570892334,
      "epoch": 0.8658381712626996,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932406594710667,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1938839554786682,
      "num_tokens": 19821765.0,
      "step": 9545
    },
    {
      "entropy": 5.470533132553101,
      "epoch": 0.8662917271407837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932327858521517,
      "loss": 5.114,
      "mean_token_accuracy": 0.19665188640356063,
      "num_tokens": 19831537.0,
      "step": 9550
    },
    {
      "entropy": 5.406108570098877,
      "epoch": 0.8667452830188679,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493224907720127,
      "loss": 5.1387,
      "mean_token_accuracy": 0.1899405762553215,
      "num_tokens": 19841830.0,
      "step": 9555
    },
    {
      "entropy": 5.479475736618042,
      "epoch": 0.8671988388969522,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932170250751556,
      "loss": 5.2427,
      "mean_token_accuracy": 0.18139030337333678,
      "num_tokens": 19851896.0,
      "step": 9560
    },
    {
      "entropy": 5.500799608230591,
      "epoch": 0.8676523947750363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932091379174006,
      "loss": 5.2659,
      "mean_token_accuracy": 0.1821811392903328,
      "num_tokens": 19861735.0,
      "step": 9565
    },
    {
      "entropy": 5.479330682754517,
      "epoch": 0.8681059506531205,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493201246247025,
      "loss": 5.1533,
      "mean_token_accuracy": 0.19605274200439454,
      "num_tokens": 19872333.0,
      "step": 9570
    },
    {
      "entropy": 5.469655227661133,
      "epoch": 0.8685595065312046,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493193350064192,
      "loss": 5.189,
      "mean_token_accuracy": 0.18709672391414642,
      "num_tokens": 19882896.0,
      "step": 9575
    },
    {
      "entropy": 5.439773464202881,
      "epoch": 0.8690130624092888,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931854493690649,
      "loss": 5.1423,
      "mean_token_accuracy": 0.19140439331531525,
      "num_tokens": 19893767.0,
      "step": 9580
    },
    {
      "entropy": 5.377742004394531,
      "epoch": 0.869466618287373,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931775441618071,
      "loss": 5.1093,
      "mean_token_accuracy": 0.19865830689668657,
      "num_tokens": 19904242.0,
      "step": 9585
    },
    {
      "entropy": 5.438934183120727,
      "epoch": 0.8699201741654572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931696344425821,
      "loss": 5.25,
      "mean_token_accuracy": 0.19252071380615235,
      "num_tokens": 19913898.0,
      "step": 9590
    },
    {
      "entropy": 5.5521914005279545,
      "epoch": 0.8703737300435413,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004931617202115535,
      "loss": 5.2507,
      "mean_token_accuracy": 0.1872472956776619,
      "num_tokens": 19923756.0,
      "step": 9595
    },
    {
      "entropy": 5.4616528987884525,
      "epoch": 0.8708272859216255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931538014688849,
      "loss": 5.1822,
      "mean_token_accuracy": 0.1898686945438385,
      "num_tokens": 19932720.0,
      "step": 9600
    },
    {
      "entropy": 5.530621290206909,
      "epoch": 0.8712808417997098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004931458782147401,
      "loss": 5.3842,
      "mean_token_accuracy": 0.17831998318433762,
      "num_tokens": 19942962.0,
      "step": 9605
    },
    {
      "entropy": 5.552232170104981,
      "epoch": 0.8717343976777939,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493137950449283,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1898573726415634,
      "num_tokens": 19953735.0,
      "step": 9610
    },
    {
      "entropy": 5.449032592773437,
      "epoch": 0.8721879535558781,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931300181726775,
      "loss": 5.2334,
      "mean_token_accuracy": 0.18857302963733674,
      "num_tokens": 19964416.0,
      "step": 9615
    },
    {
      "entropy": 5.468363428115845,
      "epoch": 0.8726415094339622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931220813850875,
      "loss": 5.2418,
      "mean_token_accuracy": 0.18374694138765335,
      "num_tokens": 19974531.0,
      "step": 9620
    },
    {
      "entropy": 5.505137968063354,
      "epoch": 0.8730950653120464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931141400866773,
      "loss": 5.2263,
      "mean_token_accuracy": 0.19138203859329223,
      "num_tokens": 19984964.0,
      "step": 9625
    },
    {
      "entropy": 5.444048309326172,
      "epoch": 0.8735486211901307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931061942776112,
      "loss": 5.1759,
      "mean_token_accuracy": 0.19382957071065904,
      "num_tokens": 19995568.0,
      "step": 9630
    },
    {
      "entropy": 5.571946382522583,
      "epoch": 0.8740021770682148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930982439580534,
      "loss": 5.3621,
      "mean_token_accuracy": 0.17532415837049484,
      "num_tokens": 20005856.0,
      "step": 9635
    },
    {
      "entropy": 5.51176905632019,
      "epoch": 0.874455732946299,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004930902891281683,
      "loss": 5.2643,
      "mean_token_accuracy": 0.18537067621946335,
      "num_tokens": 20015683.0,
      "step": 9640
    },
    {
      "entropy": 5.3699976921081545,
      "epoch": 0.8749092888243831,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004930823297881205,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1967955693602562,
      "num_tokens": 20025998.0,
      "step": 9645
    },
    {
      "entropy": 5.497171115875244,
      "epoch": 0.8753628447024674,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930743659380745,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18628917187452315,
      "num_tokens": 20036486.0,
      "step": 9650
    },
    {
      "entropy": 5.4253710269927975,
      "epoch": 0.8758164005805515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493066397578195,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19823774695396423,
      "num_tokens": 20046784.0,
      "step": 9655
    },
    {
      "entropy": 5.450145435333252,
      "epoch": 0.8762699564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004930584247086468,
      "loss": 5.2119,
      "mean_token_accuracy": 0.1873118743300438,
      "num_tokens": 20056999.0,
      "step": 9660
    },
    {
      "entropy": 5.436940431594849,
      "epoch": 0.8767235123367199,
      "grad_norm": 0.96875,
      "learning_rate": 0.000493050447329595,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19763690680265428,
      "num_tokens": 20067612.0,
      "step": 9665
    },
    {
      "entropy": 5.494093751907348,
      "epoch": 0.877177068214804,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930424654412043,
      "loss": 5.2578,
      "mean_token_accuracy": 0.19012722074985505,
      "num_tokens": 20077330.0,
      "step": 9670
    },
    {
      "entropy": 5.48297963142395,
      "epoch": 0.8776306240928883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930344790436398,
      "loss": 5.256,
      "mean_token_accuracy": 0.18199552744626998,
      "num_tokens": 20086801.0,
      "step": 9675
    },
    {
      "entropy": 5.499620962142944,
      "epoch": 0.8780841799709724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930264881370666,
      "loss": 5.1915,
      "mean_token_accuracy": 0.18674761801958084,
      "num_tokens": 20097047.0,
      "step": 9680
    },
    {
      "entropy": 5.465418910980224,
      "epoch": 0.8785377358490566,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930184927216504,
      "loss": 5.2343,
      "mean_token_accuracy": 0.191551773250103,
      "num_tokens": 20107185.0,
      "step": 9685
    },
    {
      "entropy": 5.489413642883301,
      "epoch": 0.8789912917271407,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930104927975559,
      "loss": 5.1704,
      "mean_token_accuracy": 0.19116564840078354,
      "num_tokens": 20118493.0,
      "step": 9690
    },
    {
      "entropy": 5.467175674438477,
      "epoch": 0.879444847605225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004930024883649489,
      "loss": 5.2275,
      "mean_token_accuracy": 0.18803865611553192,
      "num_tokens": 20129294.0,
      "step": 9695
    },
    {
      "entropy": 5.431062269210815,
      "epoch": 0.8798984034833092,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492994479423995,
      "loss": 5.1918,
      "mean_token_accuracy": 0.19232005923986434,
      "num_tokens": 20139194.0,
      "step": 9700
    },
    {
      "entropy": 5.45842924118042,
      "epoch": 0.8803519593613933,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929864659748597,
      "loss": 5.2081,
      "mean_token_accuracy": 0.18481118232011795,
      "num_tokens": 20149681.0,
      "step": 9705
    },
    {
      "entropy": 5.508456516265869,
      "epoch": 0.8808055152394775,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929784480177087,
      "loss": 5.2977,
      "mean_token_accuracy": 0.1847716510295868,
      "num_tokens": 20160557.0,
      "step": 9710
    },
    {
      "entropy": 5.457950210571289,
      "epoch": 0.8812590711175616,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929704255527078,
      "loss": 5.169,
      "mean_token_accuracy": 0.1884682521224022,
      "num_tokens": 20171047.0,
      "step": 9715
    },
    {
      "entropy": 5.476606464385986,
      "epoch": 0.8817126269956459,
      "grad_norm": 0.890625,
      "learning_rate": 0.000492962398580023,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1803833708167076,
      "num_tokens": 20182148.0,
      "step": 9720
    },
    {
      "entropy": 5.409409523010254,
      "epoch": 0.88216618287373,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004929543670998203,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19479717761278154,
      "num_tokens": 20192131.0,
      "step": 9725
    },
    {
      "entropy": 5.4837233543396,
      "epoch": 0.8826197387518142,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004929463311122657,
      "loss": 5.219,
      "mean_token_accuracy": 0.1879323437809944,
      "num_tokens": 20202805.0,
      "step": 9730
    },
    {
      "entropy": 5.467478561401367,
      "epoch": 0.8830732946298984,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004929382906175253,
      "loss": 5.1584,
      "mean_token_accuracy": 0.193536177277565,
      "num_tokens": 20212702.0,
      "step": 9735
    },
    {
      "entropy": 5.4351178169250485,
      "epoch": 0.8835268505079826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929302456157657,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1951988160610199,
      "num_tokens": 20222661.0,
      "step": 9740
    },
    {
      "entropy": 5.474312114715576,
      "epoch": 0.8839804063860668,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929221961071531,
      "loss": 5.2279,
      "mean_token_accuracy": 0.18725454956293106,
      "num_tokens": 20233012.0,
      "step": 9745
    },
    {
      "entropy": 5.498449611663818,
      "epoch": 0.8844339622641509,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929141420918538,
      "loss": 5.2431,
      "mean_token_accuracy": 0.19418645799160003,
      "num_tokens": 20242918.0,
      "step": 9750
    },
    {
      "entropy": 5.457444906234741,
      "epoch": 0.8848875181422351,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929060835700347,
      "loss": 5.1631,
      "mean_token_accuracy": 0.19698089808225633,
      "num_tokens": 20253254.0,
      "step": 9755
    },
    {
      "entropy": 5.45939040184021,
      "epoch": 0.8853410740203193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492898020541862,
      "loss": 5.2168,
      "mean_token_accuracy": 0.18872032910585404,
      "num_tokens": 20263355.0,
      "step": 9760
    },
    {
      "entropy": 5.419777965545654,
      "epoch": 0.8857946298984035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928899530075029,
      "loss": 5.1452,
      "mean_token_accuracy": 0.19159601032733917,
      "num_tokens": 20274716.0,
      "step": 9765
    },
    {
      "entropy": 5.386063766479492,
      "epoch": 0.8862481857764877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004928818809671241,
      "loss": 5.1268,
      "mean_token_accuracy": 0.19315028339624404,
      "num_tokens": 20285631.0,
      "step": 9770
    },
    {
      "entropy": 5.49028263092041,
      "epoch": 0.8867017416545718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928738044208923,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19549182206392288,
      "num_tokens": 20295717.0,
      "step": 9775
    },
    {
      "entropy": 5.394902086257934,
      "epoch": 0.887155297532656,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004928657233689748,
      "loss": 5.1927,
      "mean_token_accuracy": 0.19078177809715272,
      "num_tokens": 20306637.0,
      "step": 9780
    },
    {
      "entropy": 5.46283483505249,
      "epoch": 0.8876088534107403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928576378115387,
      "loss": 5.1645,
      "mean_token_accuracy": 0.18774652034044265,
      "num_tokens": 20316629.0,
      "step": 9785
    },
    {
      "entropy": 5.560376405715942,
      "epoch": 0.8880624092888244,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928495477487509,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1772662043571472,
      "num_tokens": 20327935.0,
      "step": 9790
    },
    {
      "entropy": 5.461290884017944,
      "epoch": 0.8885159651669086,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004928414531807792,
      "loss": 5.1561,
      "mean_token_accuracy": 0.18659840822219848,
      "num_tokens": 20339142.0,
      "step": 9795
    },
    {
      "entropy": 5.415150117874146,
      "epoch": 0.8889695210449927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928333541077906,
      "loss": 5.1661,
      "mean_token_accuracy": 0.20060190111398696,
      "num_tokens": 20348474.0,
      "step": 9800
    },
    {
      "entropy": 5.486493635177612,
      "epoch": 0.8894230769230769,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928252505299527,
      "loss": 5.2328,
      "mean_token_accuracy": 0.18882507979869842,
      "num_tokens": 20359440.0,
      "step": 9805
    },
    {
      "entropy": 5.55666823387146,
      "epoch": 0.8898766328011611,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004928171424474332,
      "loss": 5.2732,
      "mean_token_accuracy": 0.18582544177770616,
      "num_tokens": 20370355.0,
      "step": 9810
    },
    {
      "entropy": 5.410488796234131,
      "epoch": 0.8903301886792453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928090298603996,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19132491052150727,
      "num_tokens": 20380214.0,
      "step": 9815
    },
    {
      "entropy": 5.48265962600708,
      "epoch": 0.8907837445573294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928009127690199,
      "loss": 5.1669,
      "mean_token_accuracy": 0.18830980807542802,
      "num_tokens": 20389190.0,
      "step": 9820
    },
    {
      "entropy": 5.463851690292358,
      "epoch": 0.8912373004354136,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004927927911734617,
      "loss": 5.2319,
      "mean_token_accuracy": 0.18881546407938005,
      "num_tokens": 20399531.0,
      "step": 9825
    },
    {
      "entropy": 5.42667145729065,
      "epoch": 0.8916908563134979,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004927846650738931,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1890368089079857,
      "num_tokens": 20410201.0,
      "step": 9830
    },
    {
      "entropy": 5.436289834976196,
      "epoch": 0.892144412191582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927765344704821,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1938535362482071,
      "num_tokens": 20421579.0,
      "step": 9835
    },
    {
      "entropy": 5.458463096618653,
      "epoch": 0.8925979680696662,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927683993633969,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1929639533162117,
      "num_tokens": 20432048.0,
      "step": 9840
    },
    {
      "entropy": 5.534075403213501,
      "epoch": 0.8930515239477503,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004927602597528057,
      "loss": 5.2518,
      "mean_token_accuracy": 0.18371923565864562,
      "num_tokens": 20441752.0,
      "step": 9845
    },
    {
      "entropy": 5.4015154361724855,
      "epoch": 0.8935050798258345,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927521156388769,
      "loss": 5.1202,
      "mean_token_accuracy": 0.19143102169036866,
      "num_tokens": 20451693.0,
      "step": 9850
    },
    {
      "entropy": 5.4664863586425785,
      "epoch": 0.8939586357039188,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004927439670217788,
      "loss": 5.2549,
      "mean_token_accuracy": 0.18985022753477096,
      "num_tokens": 20462312.0,
      "step": 9855
    },
    {
      "entropy": 5.4695860862731935,
      "epoch": 0.8944121915820029,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049273581390168,
      "loss": 5.1755,
      "mean_token_accuracy": 0.18485859036445618,
      "num_tokens": 20472546.0,
      "step": 9860
    },
    {
      "entropy": 5.461061286926269,
      "epoch": 0.8948657474600871,
      "grad_norm": 0.875,
      "learning_rate": 0.000492727656278749,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19223169088363648,
      "num_tokens": 20483266.0,
      "step": 9865
    },
    {
      "entropy": 5.432494115829468,
      "epoch": 0.8953193033381712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927194941531547,
      "loss": 5.1281,
      "mean_token_accuracy": 0.18978212475776673,
      "num_tokens": 20492490.0,
      "step": 9870
    },
    {
      "entropy": 5.45404806137085,
      "epoch": 0.8957728592162555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927113275250657,
      "loss": 5.1775,
      "mean_token_accuracy": 0.19612860679626465,
      "num_tokens": 20503077.0,
      "step": 9875
    },
    {
      "entropy": 5.425263690948486,
      "epoch": 0.8962264150943396,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927031563946511,
      "loss": 5.2018,
      "mean_token_accuracy": 0.18478720486164094,
      "num_tokens": 20513771.0,
      "step": 9880
    },
    {
      "entropy": 5.487907648086548,
      "epoch": 0.8966799709724238,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926949807620796,
      "loss": 5.2973,
      "mean_token_accuracy": 0.18408113718032837,
      "num_tokens": 20523616.0,
      "step": 9885
    },
    {
      "entropy": 5.4724946975708,
      "epoch": 0.897133526850508,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004926868006275204,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19698286801576614,
      "num_tokens": 20533613.0,
      "step": 9890
    },
    {
      "entropy": 5.463048982620239,
      "epoch": 0.8975870827285921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926786159911427,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1881907179951668,
      "num_tokens": 20543812.0,
      "step": 9895
    },
    {
      "entropy": 5.4218991756439205,
      "epoch": 0.8980406386066764,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492670426853116,
      "loss": 5.1287,
      "mean_token_accuracy": 0.1955796554684639,
      "num_tokens": 20553785.0,
      "step": 9900
    },
    {
      "entropy": 5.5498435497283936,
      "epoch": 0.8984941944847605,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926622332136092,
      "loss": 5.2431,
      "mean_token_accuracy": 0.190711410343647,
      "num_tokens": 20563326.0,
      "step": 9905
    },
    {
      "entropy": 5.443296480178833,
      "epoch": 0.8989477503628447,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926540350727918,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19946378320455552,
      "num_tokens": 20573643.0,
      "step": 9910
    },
    {
      "entropy": 5.396259069442749,
      "epoch": 0.8994013062409288,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926458324308337,
      "loss": 5.1156,
      "mean_token_accuracy": 0.1931847482919693,
      "num_tokens": 20583916.0,
      "step": 9915
    },
    {
      "entropy": 5.465553617477417,
      "epoch": 0.8998548621190131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926376252879043,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1940779507160187,
      "num_tokens": 20593984.0,
      "step": 9920
    },
    {
      "entropy": 5.466677331924439,
      "epoch": 0.9003084179970973,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926294136441733,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19847935438156128,
      "num_tokens": 20603927.0,
      "step": 9925
    },
    {
      "entropy": 5.486162567138672,
      "epoch": 0.9007619738751814,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926211974998106,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19301336407661437,
      "num_tokens": 20614147.0,
      "step": 9930
    },
    {
      "entropy": 5.497865104675293,
      "epoch": 0.9012155297532656,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492612976854986,
      "loss": 5.2085,
      "mean_token_accuracy": 0.19398757964372634,
      "num_tokens": 20623631.0,
      "step": 9935
    },
    {
      "entropy": 5.412996292114258,
      "epoch": 0.9016690856313497,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926047517098697,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19451865702867507,
      "num_tokens": 20632566.0,
      "step": 9940
    },
    {
      "entropy": 5.490465974807739,
      "epoch": 0.902122641509434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925965220646315,
      "loss": 5.2074,
      "mean_token_accuracy": 0.1901567503809929,
      "num_tokens": 20643246.0,
      "step": 9945
    },
    {
      "entropy": 5.451488637924195,
      "epoch": 0.9025761973875182,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925882879194419,
      "loss": 5.2783,
      "mean_token_accuracy": 0.18662670105695725,
      "num_tokens": 20653372.0,
      "step": 9950
    },
    {
      "entropy": 5.446879196166992,
      "epoch": 0.9030297532656023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925800492744712,
      "loss": 5.2239,
      "mean_token_accuracy": 0.18523058593273162,
      "num_tokens": 20662687.0,
      "step": 9955
    },
    {
      "entropy": 5.436415195465088,
      "epoch": 0.9034833091436865,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925718061298895,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19940126687288284,
      "num_tokens": 20673168.0,
      "step": 9960
    },
    {
      "entropy": 5.574391031265259,
      "epoch": 0.9039368650217707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925635584858675,
      "loss": 5.2616,
      "mean_token_accuracy": 0.1847979247570038,
      "num_tokens": 20683100.0,
      "step": 9965
    },
    {
      "entropy": 5.563992643356324,
      "epoch": 0.9043904208998549,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925553063425755,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18730091899633408,
      "num_tokens": 20694097.0,
      "step": 9970
    },
    {
      "entropy": 5.4666440963745115,
      "epoch": 0.904843976777939,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925470497001844,
      "loss": 5.2329,
      "mean_token_accuracy": 0.19150329530239105,
      "num_tokens": 20704640.0,
      "step": 9975
    },
    {
      "entropy": 5.427973079681396,
      "epoch": 0.9052975326560232,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925387885588648,
      "loss": 5.179,
      "mean_token_accuracy": 0.1923139587044716,
      "num_tokens": 20714419.0,
      "step": 9980
    },
    {
      "entropy": 5.5230906963348385,
      "epoch": 0.9057510885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925305229187877,
      "loss": 5.2626,
      "mean_token_accuracy": 0.1851169213652611,
      "num_tokens": 20724705.0,
      "step": 9985
    },
    {
      "entropy": 5.553748750686646,
      "epoch": 0.9062046444121916,
      "grad_norm": 1.09375,
      "learning_rate": 0.000492522252780124,
      "loss": 5.3253,
      "mean_token_accuracy": 0.18897947520017624,
      "num_tokens": 20734753.0,
      "step": 9990
    },
    {
      "entropy": 5.435711288452149,
      "epoch": 0.9066582002902758,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004925139781430446,
      "loss": 5.1086,
      "mean_token_accuracy": 0.19855519235134125,
      "num_tokens": 20745389.0,
      "step": 9995
    },
    {
      "entropy": 5.46137261390686,
      "epoch": 0.9071117561683599,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004925056990077207,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1954839289188385,
      "num_tokens": 20755702.0,
      "step": 10000
    },
    {
      "entropy": 5.471224355697632,
      "epoch": 0.9075653120464441,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924974153743235,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2000062495470047,
      "num_tokens": 20766046.0,
      "step": 10005
    },
    {
      "entropy": 5.457051086425781,
      "epoch": 0.9080188679245284,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924891272430244,
      "loss": 5.1963,
      "mean_token_accuracy": 0.19161363542079926,
      "num_tokens": 20776654.0,
      "step": 10010
    },
    {
      "entropy": 5.489421749114991,
      "epoch": 0.9084724238026125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924808346139948,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19583165645599365,
      "num_tokens": 20787066.0,
      "step": 10015
    },
    {
      "entropy": 5.47434401512146,
      "epoch": 0.9089259796806967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924725374874059,
      "loss": 5.1592,
      "mean_token_accuracy": 0.1912713587284088,
      "num_tokens": 20796822.0,
      "step": 10020
    },
    {
      "entropy": 5.468144750595092,
      "epoch": 0.9093795355587808,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924642358634297,
      "loss": 5.2155,
      "mean_token_accuracy": 0.1913888320326805,
      "num_tokens": 20807602.0,
      "step": 10025
    },
    {
      "entropy": 5.5108407497406,
      "epoch": 0.909833091436865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924559297422376,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19738072901964188,
      "num_tokens": 20819046.0,
      "step": 10030
    },
    {
      "entropy": 5.51841721534729,
      "epoch": 0.9102866473149492,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924476191240014,
      "loss": 5.1916,
      "mean_token_accuracy": 0.18738922476768494,
      "num_tokens": 20828989.0,
      "step": 10035
    },
    {
      "entropy": 5.4257941246032715,
      "epoch": 0.9107402031930334,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492439304008893,
      "loss": 5.2032,
      "mean_token_accuracy": 0.18497975170612335,
      "num_tokens": 20839366.0,
      "step": 10040
    },
    {
      "entropy": 5.4617719650268555,
      "epoch": 0.9111937590711175,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004924309843970847,
      "loss": 5.1952,
      "mean_token_accuracy": 0.19048805832862853,
      "num_tokens": 20849513.0,
      "step": 10045
    },
    {
      "entropy": 5.54318790435791,
      "epoch": 0.9116473149492017,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924226602887479,
      "loss": 5.2508,
      "mean_token_accuracy": 0.18627595603466035,
      "num_tokens": 20861406.0,
      "step": 10050
    },
    {
      "entropy": 5.462168550491333,
      "epoch": 0.912100870827286,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924143316840552,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1994623914361,
      "num_tokens": 20871190.0,
      "step": 10055
    },
    {
      "entropy": 5.488011312484741,
      "epoch": 0.9125544267053701,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004924059985831787,
      "loss": 5.3135,
      "mean_token_accuracy": 0.18306103050708772,
      "num_tokens": 20881589.0,
      "step": 10060
    },
    {
      "entropy": 5.480007028579712,
      "epoch": 0.9130079825834543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004923976609862908,
      "loss": 5.2285,
      "mean_token_accuracy": 0.18074655085802077,
      "num_tokens": 20892072.0,
      "step": 10065
    },
    {
      "entropy": 5.519109153747559,
      "epoch": 0.9134615384615384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923893188935638,
      "loss": 5.2174,
      "mean_token_accuracy": 0.18683876395225524,
      "num_tokens": 20902616.0,
      "step": 10070
    },
    {
      "entropy": 5.5475568771362305,
      "epoch": 0.9139150943396226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004923809723051702,
      "loss": 5.218,
      "mean_token_accuracy": 0.1869152545928955,
      "num_tokens": 20913832.0,
      "step": 10075
    },
    {
      "entropy": 5.40401291847229,
      "epoch": 0.9143686502177069,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923726212212829,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1958935961127281,
      "num_tokens": 20923322.0,
      "step": 10080
    },
    {
      "entropy": 5.413718605041504,
      "epoch": 0.914822206095791,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923642656420742,
      "loss": 5.1963,
      "mean_token_accuracy": 0.1901403397321701,
      "num_tokens": 20933334.0,
      "step": 10085
    },
    {
      "entropy": 5.449005460739135,
      "epoch": 0.9152757619738752,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923559055677173,
      "loss": 5.1714,
      "mean_token_accuracy": 0.18672747910022736,
      "num_tokens": 20944171.0,
      "step": 10090
    },
    {
      "entropy": 5.463436222076416,
      "epoch": 0.9157293178519593,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923475409983848,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19056689441204072,
      "num_tokens": 20954200.0,
      "step": 10095
    },
    {
      "entropy": 5.472059440612793,
      "epoch": 0.9161828737300436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923391719342497,
      "loss": 5.2217,
      "mean_token_accuracy": 0.186924010515213,
      "num_tokens": 20964935.0,
      "step": 10100
    },
    {
      "entropy": 5.444699954986572,
      "epoch": 0.9166364296081277,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923307983754852,
      "loss": 5.2157,
      "mean_token_accuracy": 0.19322005957365035,
      "num_tokens": 20975499.0,
      "step": 10105
    },
    {
      "entropy": 5.40584716796875,
      "epoch": 0.9170899854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004923224203222645,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19008391797542573,
      "num_tokens": 20986199.0,
      "step": 10110
    },
    {
      "entropy": 5.468102693557739,
      "epoch": 0.9175435413642961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923140377747606,
      "loss": 5.2175,
      "mean_token_accuracy": 0.19170612692832947,
      "num_tokens": 20996313.0,
      "step": 10115
    },
    {
      "entropy": 5.4784221172332765,
      "epoch": 0.9179970972423802,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923056507331471,
      "loss": 5.198,
      "mean_token_accuracy": 0.18825269341468812,
      "num_tokens": 21006133.0,
      "step": 10120
    },
    {
      "entropy": 5.489863491058349,
      "epoch": 0.9184506531204645,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922972591975974,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19215502738952636,
      "num_tokens": 21016448.0,
      "step": 10125
    },
    {
      "entropy": 5.407008695602417,
      "epoch": 0.9189042089985486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922888631682849,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19257669001817704,
      "num_tokens": 21026458.0,
      "step": 10130
    },
    {
      "entropy": 5.399579048156738,
      "epoch": 0.9193577648766328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922804626453834,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19592053592205047,
      "num_tokens": 21037111.0,
      "step": 10135
    },
    {
      "entropy": 5.428416681289673,
      "epoch": 0.9198113207547169,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922720576290665,
      "loss": 5.202,
      "mean_token_accuracy": 0.19176458269357682,
      "num_tokens": 21048041.0,
      "step": 10140
    },
    {
      "entropy": 5.43221607208252,
      "epoch": 0.9202648766328012,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922636481195082,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2073404759168625,
      "num_tokens": 21058099.0,
      "step": 10145
    },
    {
      "entropy": 5.4595221996307375,
      "epoch": 0.9207184325108854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004922552341168823,
      "loss": 5.2142,
      "mean_token_accuracy": 0.18525523990392684,
      "num_tokens": 21068137.0,
      "step": 10150
    },
    {
      "entropy": 5.39358959197998,
      "epoch": 0.9211719883889695,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922468156213627,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19592654556035996,
      "num_tokens": 21077556.0,
      "step": 10155
    },
    {
      "entropy": 5.444814157485962,
      "epoch": 0.9216255442670537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922383926331236,
      "loss": 5.1259,
      "mean_token_accuracy": 0.1955080047249794,
      "num_tokens": 21087524.0,
      "step": 10160
    },
    {
      "entropy": 5.4274458408355715,
      "epoch": 0.9220791001451378,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922299651523392,
      "loss": 5.2001,
      "mean_token_accuracy": 0.18721571266651155,
      "num_tokens": 21098446.0,
      "step": 10165
    },
    {
      "entropy": 5.437089157104492,
      "epoch": 0.9225326560232221,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922215331791837,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1849062278866768,
      "num_tokens": 21109291.0,
      "step": 10170
    },
    {
      "entropy": 5.444094038009643,
      "epoch": 0.9229862119013063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922130967138316,
      "loss": 5.089,
      "mean_token_accuracy": 0.19493570625782014,
      "num_tokens": 21120043.0,
      "step": 10175
    },
    {
      "entropy": 5.47716703414917,
      "epoch": 0.9234397677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922046557564574,
      "loss": 5.2076,
      "mean_token_accuracy": 0.1816290646791458,
      "num_tokens": 21130339.0,
      "step": 10180
    },
    {
      "entropy": 5.493131637573242,
      "epoch": 0.9238933236574746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921962103072355,
      "loss": 5.2752,
      "mean_token_accuracy": 0.18547841012477875,
      "num_tokens": 21142307.0,
      "step": 10185
    },
    {
      "entropy": 5.484469032287597,
      "epoch": 0.9243468795355588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921877603663406,
      "loss": 5.1808,
      "mean_token_accuracy": 0.19219895750284194,
      "num_tokens": 21151843.0,
      "step": 10190
    },
    {
      "entropy": 5.4950824737548825,
      "epoch": 0.924800435413643,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004921793059339474,
      "loss": 5.1816,
      "mean_token_accuracy": 0.19396258443593978,
      "num_tokens": 21162574.0,
      "step": 10195
    },
    {
      "entropy": 5.452712297439575,
      "epoch": 0.9252539912917271,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921708470102309,
      "loss": 5.1382,
      "mean_token_accuracy": 0.19424404352903366,
      "num_tokens": 21172637.0,
      "step": 10200
    },
    {
      "entropy": 5.42418475151062,
      "epoch": 0.9257075471698113,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492162383595366,
      "loss": 5.1475,
      "mean_token_accuracy": 0.1976383075118065,
      "num_tokens": 21182272.0,
      "step": 10205
    },
    {
      "entropy": 5.523120403289795,
      "epoch": 0.9261611030478955,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004921539156895275,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18169392496347428,
      "num_tokens": 21193712.0,
      "step": 10210
    },
    {
      "entropy": 5.541693687438965,
      "epoch": 0.9266146589259797,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004921454432928908,
      "loss": 5.3006,
      "mean_token_accuracy": 0.1801042675971985,
      "num_tokens": 21205104.0,
      "step": 10215
    },
    {
      "entropy": 5.4941614151000975,
      "epoch": 0.9270682148040639,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492136966405631,
      "loss": 5.1576,
      "mean_token_accuracy": 0.19228855818510054,
      "num_tokens": 21215495.0,
      "step": 10220
    },
    {
      "entropy": 5.437613391876221,
      "epoch": 0.927521770682148,
      "grad_norm": 1.125,
      "learning_rate": 0.0004921284850279233,
      "loss": 5.1911,
      "mean_token_accuracy": 0.18829042315483094,
      "num_tokens": 21226047.0,
      "step": 10225
    },
    {
      "entropy": 5.4525219917297365,
      "epoch": 0.9279753265602322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921199991599433,
      "loss": 5.2463,
      "mean_token_accuracy": 0.18756283074617386,
      "num_tokens": 21235784.0,
      "step": 10230
    },
    {
      "entropy": 5.526726865768433,
      "epoch": 0.9284288824383164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004921115088018665,
      "loss": 5.1999,
      "mean_token_accuracy": 0.18950969576835633,
      "num_tokens": 21245493.0,
      "step": 10235
    },
    {
      "entropy": 5.584573888778687,
      "epoch": 0.9288824383164006,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492103013953868,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1842719316482544,
      "num_tokens": 21255541.0,
      "step": 10240
    },
    {
      "entropy": 5.423892831802368,
      "epoch": 0.9293359941944848,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920945146161242,
      "loss": 5.1304,
      "mean_token_accuracy": 0.196283920109272,
      "num_tokens": 21265427.0,
      "step": 10245
    },
    {
      "entropy": 5.45904951095581,
      "epoch": 0.9297895500725689,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920860107888105,
      "loss": 5.1504,
      "mean_token_accuracy": 0.19452783316373826,
      "num_tokens": 21276597.0,
      "step": 10250
    },
    {
      "entropy": 5.411054039001465,
      "epoch": 0.9302431059506531,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004920775024721026,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19338110834360123,
      "num_tokens": 21286416.0,
      "step": 10255
    },
    {
      "entropy": 5.446629858016967,
      "epoch": 0.9306966618287373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920689896661767,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1925641641020775,
      "num_tokens": 21296064.0,
      "step": 10260
    },
    {
      "entropy": 5.426272392272949,
      "epoch": 0.9311502177068215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920604723712088,
      "loss": 5.1953,
      "mean_token_accuracy": 0.1939927965402603,
      "num_tokens": 21306836.0,
      "step": 10265
    },
    {
      "entropy": 5.4682684421539305,
      "epoch": 0.9316037735849056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920519505873751,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19016034603118898,
      "num_tokens": 21316936.0,
      "step": 10270
    },
    {
      "entropy": 5.444418096542359,
      "epoch": 0.9320573294629898,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920434243148515,
      "loss": 5.1067,
      "mean_token_accuracy": 0.1991023138165474,
      "num_tokens": 21326892.0,
      "step": 10275
    },
    {
      "entropy": 5.4291871070861815,
      "epoch": 0.9325108853410741,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920348935538147,
      "loss": 5.1313,
      "mean_token_accuracy": 0.18786720633506776,
      "num_tokens": 21337270.0,
      "step": 10280
    },
    {
      "entropy": 5.484059762954712,
      "epoch": 0.9329644412191582,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492026358304441,
      "loss": 5.2205,
      "mean_token_accuracy": 0.18756954073905946,
      "num_tokens": 21347968.0,
      "step": 10285
    },
    {
      "entropy": 5.4217315196990965,
      "epoch": 0.9334179970972424,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492017818566907,
      "loss": 5.1521,
      "mean_token_accuracy": 0.19510474503040315,
      "num_tokens": 21358312.0,
      "step": 10290
    },
    {
      "entropy": 5.430413627624512,
      "epoch": 0.9338715529753265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920092743413892,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19392232745885848,
      "num_tokens": 21368794.0,
      "step": 10295
    },
    {
      "entropy": 5.517674350738526,
      "epoch": 0.9343251088534107,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920007256280642,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19540511071681976,
      "num_tokens": 21380080.0,
      "step": 10300
    },
    {
      "entropy": 5.465408277511597,
      "epoch": 0.934778664731495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491992172427109,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1919153705239296,
      "num_tokens": 21389961.0,
      "step": 10305
    },
    {
      "entropy": 5.423232889175415,
      "epoch": 0.9352322206095791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004919836147387002,
      "loss": 5.2048,
      "mean_token_accuracy": 0.19409374743700028,
      "num_tokens": 21399869.0,
      "step": 10310
    },
    {
      "entropy": 5.502434301376343,
      "epoch": 0.9356857764876633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919750525630151,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19113861620426179,
      "num_tokens": 21410343.0,
      "step": 10315
    },
    {
      "entropy": 5.52030725479126,
      "epoch": 0.9361393323657474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919664859002306,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19153531044721603,
      "num_tokens": 21420915.0,
      "step": 10320
    },
    {
      "entropy": 5.468138742446899,
      "epoch": 0.9365928882438317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004919579147505239,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19598333686590194,
      "num_tokens": 21430879.0,
      "step": 10325
    },
    {
      "entropy": 5.482166004180908,
      "epoch": 0.9370464441219158,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919493391140721,
      "loss": 5.2311,
      "mean_token_accuracy": 0.19225272685289382,
      "num_tokens": 21440843.0,
      "step": 10330
    },
    {
      "entropy": 5.5462892055511475,
      "epoch": 0.9375,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004919407589910529,
      "loss": 5.3852,
      "mean_token_accuracy": 0.18182432949542998,
      "num_tokens": 21452132.0,
      "step": 10335
    },
    {
      "entropy": 5.4647728443145756,
      "epoch": 0.9379535558780842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004919321743816434,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1902553766965866,
      "num_tokens": 21461861.0,
      "step": 10340
    },
    {
      "entropy": 5.461749696731568,
      "epoch": 0.9384071117561683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004919235852860213,
      "loss": 5.1619,
      "mean_token_accuracy": 0.18584073036909105,
      "num_tokens": 21472244.0,
      "step": 10345
    },
    {
      "entropy": 5.505767917633056,
      "epoch": 0.9388606676342526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919149917043641,
      "loss": 5.2325,
      "mean_token_accuracy": 0.18217187970876694,
      "num_tokens": 21482394.0,
      "step": 10350
    },
    {
      "entropy": 5.476301479339599,
      "epoch": 0.9393142235123367,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919063936368496,
      "loss": 5.2623,
      "mean_token_accuracy": 0.1858245924115181,
      "num_tokens": 21493334.0,
      "step": 10355
    },
    {
      "entropy": 5.526564693450927,
      "epoch": 0.9397677793904209,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918977910836557,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1870238721370697,
      "num_tokens": 21504443.0,
      "step": 10360
    },
    {
      "entropy": 5.524574422836304,
      "epoch": 0.940221335268505,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918891840449601,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1889067441225052,
      "num_tokens": 21513908.0,
      "step": 10365
    },
    {
      "entropy": 5.509088134765625,
      "epoch": 0.9406748911465893,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000491880572520941,
      "loss": 5.2706,
      "mean_token_accuracy": 0.18197775781154632,
      "num_tokens": 21524349.0,
      "step": 10370
    },
    {
      "entropy": 5.601503610610962,
      "epoch": 0.9411284470246735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918719565117763,
      "loss": 5.3169,
      "mean_token_accuracy": 0.19112387597560881,
      "num_tokens": 21535003.0,
      "step": 10375
    },
    {
      "entropy": 5.616733551025391,
      "epoch": 0.9415820029027576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918633360176444,
      "loss": 5.3396,
      "mean_token_accuracy": 0.17721167653799058,
      "num_tokens": 21545045.0,
      "step": 10380
    },
    {
      "entropy": 5.420244455337524,
      "epoch": 0.9420355587808418,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918547110387234,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19451433718204497,
      "num_tokens": 21554830.0,
      "step": 10385
    },
    {
      "entropy": 5.470901536941528,
      "epoch": 0.9424891146589259,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918460815751916,
      "loss": 5.147,
      "mean_token_accuracy": 0.1933375909924507,
      "num_tokens": 21564693.0,
      "step": 10390
    },
    {
      "entropy": 5.455016374588013,
      "epoch": 0.9429426705370102,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918374476272277,
      "loss": 5.151,
      "mean_token_accuracy": 0.18868277966976166,
      "num_tokens": 21575475.0,
      "step": 10395
    },
    {
      "entropy": 5.444744110107422,
      "epoch": 0.9433962264150944,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004918288091950101,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19193150252103805,
      "num_tokens": 21586944.0,
      "step": 10400
    },
    {
      "entropy": 5.508098602294922,
      "epoch": 0.9438497822931785,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004918201662787174,
      "loss": 5.1872,
      "mean_token_accuracy": 0.188102026283741,
      "num_tokens": 21598170.0,
      "step": 10405
    },
    {
      "entropy": 5.457108020782471,
      "epoch": 0.9443033381712627,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004918115188785285,
      "loss": 5.2038,
      "mean_token_accuracy": 0.1924883246421814,
      "num_tokens": 21609170.0,
      "step": 10410
    },
    {
      "entropy": 5.5134035587310795,
      "epoch": 0.9447568940493469,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004918028669946221,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1856304958462715,
      "num_tokens": 21620171.0,
      "step": 10415
    },
    {
      "entropy": 5.444842720031739,
      "epoch": 0.9452104499274311,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004917942106271771,
      "loss": 5.204,
      "mean_token_accuracy": 0.18975074589252472,
      "num_tokens": 21631072.0,
      "step": 10420
    },
    {
      "entropy": 5.446480655670166,
      "epoch": 0.9456640058055152,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917855497763727,
      "loss": 5.0733,
      "mean_token_accuracy": 0.19692519158124924,
      "num_tokens": 21641543.0,
      "step": 10425
    },
    {
      "entropy": 5.450199317932129,
      "epoch": 0.9461175616835994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917768844423878,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19204503446817398,
      "num_tokens": 21651342.0,
      "step": 10430
    },
    {
      "entropy": 5.479935264587402,
      "epoch": 0.9465711175616836,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004917682146254017,
      "loss": 5.158,
      "mean_token_accuracy": 0.19622128307819367,
      "num_tokens": 21661043.0,
      "step": 10435
    },
    {
      "entropy": 5.386088800430298,
      "epoch": 0.9470246734397678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004917595403255938,
      "loss": 5.04,
      "mean_token_accuracy": 0.20307991057634353,
      "num_tokens": 21671197.0,
      "step": 10440
    },
    {
      "entropy": 5.388993835449218,
      "epoch": 0.947478229317852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004917508615431431,
      "loss": 5.1074,
      "mean_token_accuracy": 0.19641754031181335,
      "num_tokens": 21682066.0,
      "step": 10445
    },
    {
      "entropy": 5.509283256530762,
      "epoch": 0.9479317851959361,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004917421782782296,
      "loss": 5.1637,
      "mean_token_accuracy": 0.1928715631365776,
      "num_tokens": 21692117.0,
      "step": 10450
    },
    {
      "entropy": 5.497783470153808,
      "epoch": 0.9483853410740203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004917334905310325,
      "loss": 5.2899,
      "mean_token_accuracy": 0.18302855789661407,
      "num_tokens": 21703544.0,
      "step": 10455
    },
    {
      "entropy": 5.399119567871094,
      "epoch": 0.9488388969521045,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004917247983017317,
      "loss": 5.1063,
      "mean_token_accuracy": 0.19429357051849366,
      "num_tokens": 21712850.0,
      "step": 10460
    },
    {
      "entropy": 5.465300464630127,
      "epoch": 0.9492924528301887,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917161015905067,
      "loss": 5.2568,
      "mean_token_accuracy": 0.19032784849405288,
      "num_tokens": 21723015.0,
      "step": 10465
    },
    {
      "entropy": 5.50637149810791,
      "epoch": 0.9497460087082729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917074003975376,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1906416416168213,
      "num_tokens": 21732453.0,
      "step": 10470
    },
    {
      "entropy": 5.502040147781372,
      "epoch": 0.950199564586357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004916986947230042,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1872340992093086,
      "num_tokens": 21743251.0,
      "step": 10475
    },
    {
      "entropy": 5.475094842910766,
      "epoch": 0.9506531204644412,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004916899845670866,
      "loss": 5.1852,
      "mean_token_accuracy": 0.19052052050828933,
      "num_tokens": 21752891.0,
      "step": 10480
    },
    {
      "entropy": 5.50803575515747,
      "epoch": 0.9511066763425254,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004916812699299649,
      "loss": 5.1881,
      "mean_token_accuracy": 0.19168722480535508,
      "num_tokens": 21763230.0,
      "step": 10485
    },
    {
      "entropy": 5.468595266342163,
      "epoch": 0.9515602322206096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916725508118193,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19173318296670913,
      "num_tokens": 21773663.0,
      "step": 10490
    },
    {
      "entropy": 5.454308080673218,
      "epoch": 0.9520137880986937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916638272128301,
      "loss": 5.2005,
      "mean_token_accuracy": 0.19195641875267028,
      "num_tokens": 21784265.0,
      "step": 10495
    },
    {
      "entropy": 5.421849203109741,
      "epoch": 0.9524673439767779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916550991331777,
      "loss": 5.1688,
      "mean_token_accuracy": 0.19283314794301987,
      "num_tokens": 21795056.0,
      "step": 10500
    },
    {
      "entropy": 5.451410961151123,
      "epoch": 0.9529208998548622,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916463665730427,
      "loss": 5.1743,
      "mean_token_accuracy": 0.19450462907552718,
      "num_tokens": 21806204.0,
      "step": 10505
    },
    {
      "entropy": 5.452659845352173,
      "epoch": 0.9533744557329463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916376295326057,
      "loss": 5.1935,
      "mean_token_accuracy": 0.19136037677526474,
      "num_tokens": 21816993.0,
      "step": 10510
    },
    {
      "entropy": 5.424757862091065,
      "epoch": 0.9538280116110305,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916288880120472,
      "loss": 5.0334,
      "mean_token_accuracy": 0.206775464117527,
      "num_tokens": 21827457.0,
      "step": 10515
    },
    {
      "entropy": 5.468567562103272,
      "epoch": 0.9542815674891146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916201420115482,
      "loss": 5.1734,
      "mean_token_accuracy": 0.18725205212831497,
      "num_tokens": 21837555.0,
      "step": 10520
    },
    {
      "entropy": 5.469878101348877,
      "epoch": 0.9547351233671988,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004916113915312895,
      "loss": 5.2008,
      "mean_token_accuracy": 0.19283771514892578,
      "num_tokens": 21847074.0,
      "step": 10525
    },
    {
      "entropy": 5.503349781036377,
      "epoch": 0.9551886792452831,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916026365714519,
      "loss": 5.1922,
      "mean_token_accuracy": 0.18674940168857573,
      "num_tokens": 21858333.0,
      "step": 10530
    },
    {
      "entropy": 5.4334405899047855,
      "epoch": 0.9556422351233672,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004915938771322165,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19297861605882644,
      "num_tokens": 21869132.0,
      "step": 10535
    },
    {
      "entropy": 5.437691497802734,
      "epoch": 0.9560957910014514,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915851132137646,
      "loss": 5.2101,
      "mean_token_accuracy": 0.18776627331972123,
      "num_tokens": 21879153.0,
      "step": 10540
    },
    {
      "entropy": 5.471747541427613,
      "epoch": 0.9565493468795355,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004915763448162774,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1977652981877327,
      "num_tokens": 21889063.0,
      "step": 10545
    },
    {
      "entropy": 5.468826198577881,
      "epoch": 0.9570029027576198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004915675719399362,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18763369768857957,
      "num_tokens": 21899005.0,
      "step": 10550
    },
    {
      "entropy": 5.450291156768799,
      "epoch": 0.9574564586357039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004915587945849225,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20017891079187394,
      "num_tokens": 21908391.0,
      "step": 10555
    },
    {
      "entropy": 5.456761217117309,
      "epoch": 0.9579100145137881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915500127514176,
      "loss": 5.1713,
      "mean_token_accuracy": 0.19450551569461821,
      "num_tokens": 21917655.0,
      "step": 10560
    },
    {
      "entropy": 5.545481443405151,
      "epoch": 0.9583635703918723,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004915412264396034,
      "loss": 5.232,
      "mean_token_accuracy": 0.19058077186346054,
      "num_tokens": 21928038.0,
      "step": 10565
    },
    {
      "entropy": 5.486158227920532,
      "epoch": 0.9588171262699564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004915324356496613,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19642450511455536,
      "num_tokens": 21938106.0,
      "step": 10570
    },
    {
      "entropy": 5.421289777755737,
      "epoch": 0.9592706821480407,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915236403817733,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19307896196842195,
      "num_tokens": 21947484.0,
      "step": 10575
    },
    {
      "entropy": 5.461407136917114,
      "epoch": 0.9597242380261248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915148406361213,
      "loss": 5.21,
      "mean_token_accuracy": 0.19016327410936357,
      "num_tokens": 21957139.0,
      "step": 10580
    },
    {
      "entropy": 5.465814828872681,
      "epoch": 0.960177793904209,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915060364128872,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18789542019367217,
      "num_tokens": 21967030.0,
      "step": 10585
    },
    {
      "entropy": 5.424587106704712,
      "epoch": 0.9606313497822931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914972277122532,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1932281047105789,
      "num_tokens": 21976418.0,
      "step": 10590
    },
    {
      "entropy": 5.436279344558716,
      "epoch": 0.9610849056603774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004914884145344012,
      "loss": 5.15,
      "mean_token_accuracy": 0.19504899382591248,
      "num_tokens": 21987284.0,
      "step": 10595
    },
    {
      "entropy": 5.402978086471558,
      "epoch": 0.9615384615384616,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914795968795137,
      "loss": 5.0709,
      "mean_token_accuracy": 0.20060221701860428,
      "num_tokens": 21997106.0,
      "step": 10600
    },
    {
      "entropy": 5.435127687454224,
      "epoch": 0.9619920174165457,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914707747477728,
      "loss": 5.0591,
      "mean_token_accuracy": 0.19728294610977173,
      "num_tokens": 22007589.0,
      "step": 10605
    },
    {
      "entropy": 5.447371244430542,
      "epoch": 0.9624455732946299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004914619481393613,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1917535647749901,
      "num_tokens": 22017669.0,
      "step": 10610
    },
    {
      "entropy": 5.585604906082153,
      "epoch": 0.962899129172714,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914531170544615,
      "loss": 5.3528,
      "mean_token_accuracy": 0.182457534968853,
      "num_tokens": 22029330.0,
      "step": 10615
    },
    {
      "entropy": 5.460028839111328,
      "epoch": 0.9633526850507983,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004914442814932561,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19587075859308242,
      "num_tokens": 22039877.0,
      "step": 10620
    },
    {
      "entropy": 5.451121139526367,
      "epoch": 0.9638062409288825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914354414559278,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19791433960199356,
      "num_tokens": 22049579.0,
      "step": 10625
    },
    {
      "entropy": 5.337468099594116,
      "epoch": 0.9642597968069666,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004914265969426594,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20568925887346268,
      "num_tokens": 22059608.0,
      "step": 10630
    },
    {
      "entropy": 5.454187726974487,
      "epoch": 0.9647133526850508,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004914177479536338,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19067026823759078,
      "num_tokens": 22070036.0,
      "step": 10635
    },
    {
      "entropy": 5.490912675857544,
      "epoch": 0.965166908563135,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491408894489034,
      "loss": 5.2014,
      "mean_token_accuracy": 0.18741769790649415,
      "num_tokens": 22080679.0,
      "step": 10640
    },
    {
      "entropy": 5.437967729568482,
      "epoch": 0.9656204644412192,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004914000365490432,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20011549443006516,
      "num_tokens": 22092192.0,
      "step": 10645
    },
    {
      "entropy": 5.446996021270752,
      "epoch": 0.9660740203193033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913911741338445,
      "loss": 5.2316,
      "mean_token_accuracy": 0.18467727899551392,
      "num_tokens": 22102722.0,
      "step": 10650
    },
    {
      "entropy": 5.398031520843506,
      "epoch": 0.9665275761973875,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004913823072436211,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20055763721466063,
      "num_tokens": 22113521.0,
      "step": 10655
    },
    {
      "entropy": 5.489092683792114,
      "epoch": 0.9669811320754716,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004913734358785565,
      "loss": 5.1951,
      "mean_token_accuracy": 0.191675665974617,
      "num_tokens": 22124891.0,
      "step": 10660
    },
    {
      "entropy": 5.45971531867981,
      "epoch": 0.9674346879535559,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913645600388341,
      "loss": 5.1955,
      "mean_token_accuracy": 0.1867385983467102,
      "num_tokens": 22135972.0,
      "step": 10665
    },
    {
      "entropy": 5.504754066467285,
      "epoch": 0.9678882438316401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913556797246374,
      "loss": 5.2719,
      "mean_token_accuracy": 0.17639271914958954,
      "num_tokens": 22146623.0,
      "step": 10670
    },
    {
      "entropy": 5.51257095336914,
      "epoch": 0.9683417997097242,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913467949361503,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1888551414012909,
      "num_tokens": 22157619.0,
      "step": 10675
    },
    {
      "entropy": 5.4886709690094,
      "epoch": 0.9687953555878084,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913379056735561,
      "loss": 5.1681,
      "mean_token_accuracy": 0.1849115401506424,
      "num_tokens": 22168655.0,
      "step": 10680
    },
    {
      "entropy": 5.448235177993775,
      "epoch": 0.9692489114658926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491329011937039,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1861863613128662,
      "num_tokens": 22178500.0,
      "step": 10685
    },
    {
      "entropy": 5.498290205001831,
      "epoch": 0.9697024673439768,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004913201137267828,
      "loss": 5.209,
      "mean_token_accuracy": 0.18552537113428116,
      "num_tokens": 22188873.0,
      "step": 10690
    },
    {
      "entropy": 5.423058652877808,
      "epoch": 0.970156023222061,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913112110429715,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1923189640045166,
      "num_tokens": 22199743.0,
      "step": 10695
    },
    {
      "entropy": 5.42342176437378,
      "epoch": 0.9706095791001451,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913023038857892,
      "loss": 5.0774,
      "mean_token_accuracy": 0.2012026712298393,
      "num_tokens": 22210301.0,
      "step": 10700
    },
    {
      "entropy": 5.42089877128601,
      "epoch": 0.9710631349782293,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912933922554203,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19924921542406082,
      "num_tokens": 22220771.0,
      "step": 10705
    },
    {
      "entropy": 5.436170768737793,
      "epoch": 0.9715166908563135,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912844761520487,
      "loss": 5.189,
      "mean_token_accuracy": 0.1894242450594902,
      "num_tokens": 22231100.0,
      "step": 10710
    },
    {
      "entropy": 5.427358102798462,
      "epoch": 0.9719702467343977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912755555758592,
      "loss": 5.1704,
      "mean_token_accuracy": 0.1919427692890167,
      "num_tokens": 22240997.0,
      "step": 10715
    },
    {
      "entropy": 5.469578981399536,
      "epoch": 0.9724238026124818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491266630527036,
      "loss": 5.2051,
      "mean_token_accuracy": 0.18859542161226273,
      "num_tokens": 22251795.0,
      "step": 10720
    },
    {
      "entropy": 5.541397523880005,
      "epoch": 0.972877358490566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004912577010057638,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18489183038473128,
      "num_tokens": 22262465.0,
      "step": 10725
    },
    {
      "entropy": 5.585089445114136,
      "epoch": 0.9733309143686503,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004912487670122273,
      "loss": 5.3521,
      "mean_token_accuracy": 0.17896877229213715,
      "num_tokens": 22272399.0,
      "step": 10730
    },
    {
      "entropy": 5.510864639282227,
      "epoch": 0.9737844702467344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912398285466111,
      "loss": 5.2631,
      "mean_token_accuracy": 0.18806591629981995,
      "num_tokens": 22282545.0,
      "step": 10735
    },
    {
      "entropy": 5.4886369705200195,
      "epoch": 0.9742380261248186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912308856091002,
      "loss": 5.211,
      "mean_token_accuracy": 0.19261860400438308,
      "num_tokens": 22293154.0,
      "step": 10740
    },
    {
      "entropy": 5.491091871261597,
      "epoch": 0.9746915820029027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912219381998795,
      "loss": 5.1295,
      "mean_token_accuracy": 0.2008547693490982,
      "num_tokens": 22302309.0,
      "step": 10745
    },
    {
      "entropy": 5.527703189849854,
      "epoch": 0.9751451378809869,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004912129863191339,
      "loss": 5.241,
      "mean_token_accuracy": 0.17985348105430604,
      "num_tokens": 22312316.0,
      "step": 10750
    },
    {
      "entropy": 5.4772176265716555,
      "epoch": 0.9755986937590712,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004912040299670487,
      "loss": 5.2116,
      "mean_token_accuracy": 0.19279262274503708,
      "num_tokens": 22322387.0,
      "step": 10755
    },
    {
      "entropy": 5.465480613708496,
      "epoch": 0.9760522496371553,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004911950691438091,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1968354642391205,
      "num_tokens": 22332759.0,
      "step": 10760
    },
    {
      "entropy": 5.4949132919311525,
      "epoch": 0.9765058055152395,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004911861038496004,
      "loss": 5.245,
      "mean_token_accuracy": 0.19074769914150239,
      "num_tokens": 22344056.0,
      "step": 10765
    },
    {
      "entropy": 5.44970850944519,
      "epoch": 0.9769593613933236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911771340846079,
      "loss": 5.194,
      "mean_token_accuracy": 0.18383531719446183,
      "num_tokens": 22353877.0,
      "step": 10770
    },
    {
      "entropy": 5.5462226390838625,
      "epoch": 0.9774129172714079,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911681598490172,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1833869755268097,
      "num_tokens": 22363549.0,
      "step": 10775
    },
    {
      "entropy": 5.534874534606933,
      "epoch": 0.977866473149492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911591811430139,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18992500454187394,
      "num_tokens": 22374040.0,
      "step": 10780
    },
    {
      "entropy": 5.518569135665894,
      "epoch": 0.9783200290275762,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911501979667836,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18968116790056228,
      "num_tokens": 22384819.0,
      "step": 10785
    },
    {
      "entropy": 5.458084917068481,
      "epoch": 0.9787735849056604,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004911412103205122,
      "loss": 5.1802,
      "mean_token_accuracy": 0.1892141282558441,
      "num_tokens": 22395130.0,
      "step": 10790
    },
    {
      "entropy": 5.475216674804687,
      "epoch": 0.9792271407837445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004911322182043855,
      "loss": 5.1936,
      "mean_token_accuracy": 0.18980710357427596,
      "num_tokens": 22405093.0,
      "step": 10795
    },
    {
      "entropy": 5.490931606292724,
      "epoch": 0.9796806966618288,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004911232216185893,
      "loss": 5.2105,
      "mean_token_accuracy": 0.18691559731960297,
      "num_tokens": 22415008.0,
      "step": 10800
    },
    {
      "entropy": 5.51223521232605,
      "epoch": 0.9801342525399129,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911142205633099,
      "loss": 5.2572,
      "mean_token_accuracy": 0.18417985886335372,
      "num_tokens": 22426486.0,
      "step": 10805
    },
    {
      "entropy": 5.445620203018189,
      "epoch": 0.9805878084179971,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911052150387334,
      "loss": 5.1643,
      "mean_token_accuracy": 0.1907837927341461,
      "num_tokens": 22436844.0,
      "step": 10810
    },
    {
      "entropy": 5.5398115634918215,
      "epoch": 0.9810413642960812,
      "grad_norm": 1.125,
      "learning_rate": 0.0004910962050450458,
      "loss": 5.2869,
      "mean_token_accuracy": 0.18634007424116134,
      "num_tokens": 22446630.0,
      "step": 10815
    },
    {
      "entropy": 5.5155932903289795,
      "epoch": 0.9814949201741655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910871905824338,
      "loss": 5.1246,
      "mean_token_accuracy": 0.1972308024764061,
      "num_tokens": 22456406.0,
      "step": 10820
    },
    {
      "entropy": 5.421662712097168,
      "epoch": 0.9819484760522497,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910781716510837,
      "loss": 5.1097,
      "mean_token_accuracy": 0.19659640938043593,
      "num_tokens": 22466210.0,
      "step": 10825
    },
    {
      "entropy": 5.473699998855591,
      "epoch": 0.9824020319303338,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910691482511817,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19028413742780687,
      "num_tokens": 22476978.0,
      "step": 10830
    },
    {
      "entropy": 5.563931512832641,
      "epoch": 0.982855587808418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910601203829148,
      "loss": 5.1888,
      "mean_token_accuracy": 0.18743186444044113,
      "num_tokens": 22488613.0,
      "step": 10835
    },
    {
      "entropy": 5.440819263458252,
      "epoch": 0.9833091436865021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910510880464696,
      "loss": 5.0876,
      "mean_token_accuracy": 0.2051035374403,
      "num_tokens": 22498312.0,
      "step": 10840
    },
    {
      "entropy": 5.437257242202759,
      "epoch": 0.9837626995645864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910420512420327,
      "loss": 5.2154,
      "mean_token_accuracy": 0.19098950028419495,
      "num_tokens": 22508131.0,
      "step": 10845
    },
    {
      "entropy": 5.461184930801392,
      "epoch": 0.9842162554426706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910330099697913,
      "loss": 5.1836,
      "mean_token_accuracy": 0.19517870992422104,
      "num_tokens": 22517842.0,
      "step": 10850
    },
    {
      "entropy": 5.508009815216065,
      "epoch": 0.9846698113207547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910239642299321,
      "loss": 5.1597,
      "mean_token_accuracy": 0.19345295876264573,
      "num_tokens": 22528239.0,
      "step": 10855
    },
    {
      "entropy": 5.465402603149414,
      "epoch": 0.9851233671988389,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004910149140226422,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19278796762228012,
      "num_tokens": 22539734.0,
      "step": 10860
    },
    {
      "entropy": 5.471485900878906,
      "epoch": 0.9855769230769231,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910058593481089,
      "loss": 5.14,
      "mean_token_accuracy": 0.19574241936206818,
      "num_tokens": 22549796.0,
      "step": 10865
    },
    {
      "entropy": 5.4115073680877686,
      "epoch": 0.9860304789550073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909968002065195,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19815587401390075,
      "num_tokens": 22559934.0,
      "step": 10870
    },
    {
      "entropy": 5.553992795944214,
      "epoch": 0.9864840348330914,
      "grad_norm": 1.078125,
      "learning_rate": 0.000490987736598061,
      "loss": 5.2505,
      "mean_token_accuracy": 0.17646464705467224,
      "num_tokens": 22569864.0,
      "step": 10875
    },
    {
      "entropy": 5.48905930519104,
      "epoch": 0.9869375907111756,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004909786685229211,
      "loss": 5.1793,
      "mean_token_accuracy": 0.1971484124660492,
      "num_tokens": 22580687.0,
      "step": 10880
    },
    {
      "entropy": 5.515979480743408,
      "epoch": 0.9873911465892597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909695959812875,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19113008826971054,
      "num_tokens": 22590507.0,
      "step": 10885
    },
    {
      "entropy": 5.498454666137695,
      "epoch": 0.987844702467344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909605189733474,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19267081320285798,
      "num_tokens": 22601901.0,
      "step": 10890
    },
    {
      "entropy": 5.446087169647217,
      "epoch": 0.9882982583454282,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909514374992889,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19195072799921037,
      "num_tokens": 22613003.0,
      "step": 10895
    },
    {
      "entropy": 5.4279900074005125,
      "epoch": 0.9887518142235123,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004909423515592996,
      "loss": 5.0441,
      "mean_token_accuracy": 0.19864177107810974,
      "num_tokens": 22623401.0,
      "step": 10900
    },
    {
      "entropy": 5.49190239906311,
      "epoch": 0.9892053701015965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909332611535675,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18905673325061798,
      "num_tokens": 22633147.0,
      "step": 10905
    },
    {
      "entropy": 5.422115516662598,
      "epoch": 0.9896589259796807,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909241662822804,
      "loss": 5.043,
      "mean_token_accuracy": 0.20871215909719468,
      "num_tokens": 22642990.0,
      "step": 10910
    },
    {
      "entropy": 5.443415689468384,
      "epoch": 0.9901124818577649,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909150669456265,
      "loss": 5.2646,
      "mean_token_accuracy": 0.18424686342477797,
      "num_tokens": 22652734.0,
      "step": 10915
    },
    {
      "entropy": 5.483167934417724,
      "epoch": 0.9905660377358491,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000490905963143794,
      "loss": 5.2804,
      "mean_token_accuracy": 0.18312346786260605,
      "num_tokens": 22662968.0,
      "step": 10920
    },
    {
      "entropy": 5.407788610458374,
      "epoch": 0.9910195936139332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908968548769713,
      "loss": 5.102,
      "mean_token_accuracy": 0.2041473627090454,
      "num_tokens": 22673157.0,
      "step": 10925
    },
    {
      "entropy": 5.471360683441162,
      "epoch": 0.9914731494920174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908877421453466,
      "loss": 5.2468,
      "mean_token_accuracy": 0.18830858021974564,
      "num_tokens": 22683105.0,
      "step": 10930
    },
    {
      "entropy": 5.453107118606567,
      "epoch": 0.9919267053701016,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004908786249491082,
      "loss": 5.1757,
      "mean_token_accuracy": 0.20476199686527252,
      "num_tokens": 22693793.0,
      "step": 10935
    },
    {
      "entropy": 5.566500282287597,
      "epoch": 0.9923802612481858,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490869503288445,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1911487653851509,
      "num_tokens": 22703884.0,
      "step": 10940
    },
    {
      "entropy": 5.522388029098511,
      "epoch": 0.99283381712627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908603771635454,
      "loss": 5.2657,
      "mean_token_accuracy": 0.18617092519998552,
      "num_tokens": 22713699.0,
      "step": 10945
    },
    {
      "entropy": 5.451653718948364,
      "epoch": 0.9932873730043541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908512465745982,
      "loss": 5.1013,
      "mean_token_accuracy": 0.1970734789967537,
      "num_tokens": 22722871.0,
      "step": 10950
    },
    {
      "entropy": 5.477705287933349,
      "epoch": 0.9937409288824384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908421115217922,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19224777370691298,
      "num_tokens": 22733314.0,
      "step": 10955
    },
    {
      "entropy": 5.404214572906494,
      "epoch": 0.9941944847605225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004908329720053164,
      "loss": 5.1123,
      "mean_token_accuracy": 0.2003941997885704,
      "num_tokens": 22743255.0,
      "step": 10960
    },
    {
      "entropy": 5.388226270675659,
      "epoch": 0.9946480406386067,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908238280253597,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20025001019239425,
      "num_tokens": 22753857.0,
      "step": 10965
    },
    {
      "entropy": 5.4696001529693605,
      "epoch": 0.9951015965166908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004908146795821112,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1949297696352005,
      "num_tokens": 22764326.0,
      "step": 10970
    },
    {
      "entropy": 5.480047273635864,
      "epoch": 0.995555152394775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908055266757602,
      "loss": 5.1715,
      "mean_token_accuracy": 0.19572446793317794,
      "num_tokens": 22774958.0,
      "step": 10975
    },
    {
      "entropy": 5.403993844985962,
      "epoch": 0.9960087082728593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907963693064958,
      "loss": 5.0864,
      "mean_token_accuracy": 0.19230775833129882,
      "num_tokens": 22785616.0,
      "step": 10980
    },
    {
      "entropy": 5.529326391220093,
      "epoch": 0.9964622641509434,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004907872074745075,
      "loss": 5.2676,
      "mean_token_accuracy": 0.18333887457847595,
      "num_tokens": 22795967.0,
      "step": 10985
    },
    {
      "entropy": 5.450800800323487,
      "epoch": 0.9969158200290276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907780411799849,
      "loss": 5.267,
      "mean_token_accuracy": 0.18713125139474868,
      "num_tokens": 22806335.0,
      "step": 10990
    },
    {
      "entropy": 5.5267292022705075,
      "epoch": 0.9973693759071117,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004907688704231173,
      "loss": 5.1945,
      "mean_token_accuracy": 0.19679880142211914,
      "num_tokens": 22816390.0,
      "step": 10995
    },
    {
      "entropy": 5.578009223937988,
      "epoch": 0.997822931785196,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004907596952040945,
      "loss": 5.2575,
      "mean_token_accuracy": 0.19109320491552353,
      "num_tokens": 22826395.0,
      "step": 11000
    },
    {
      "entropy": 5.526041650772095,
      "epoch": 0.9982764876632801,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907505155231063,
      "loss": 5.2196,
      "mean_token_accuracy": 0.19320557415485382,
      "num_tokens": 22836790.0,
      "step": 11005
    },
    {
      "entropy": 5.450096607208252,
      "epoch": 0.9987300435413643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907413313803424,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19249580502510072,
      "num_tokens": 22846687.0,
      "step": 11010
    },
    {
      "entropy": 5.496992063522339,
      "epoch": 0.9991835994194485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907321427759928,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19380525201559068,
      "num_tokens": 22857208.0,
      "step": 11015
    },
    {
      "entropy": 5.507448673248291,
      "epoch": 0.9996371552975326,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004907229497102475,
      "loss": 5.2677,
      "mean_token_accuracy": 0.18637495636940002,
      "num_tokens": 22868556.0,
      "step": 11020
    },
    {
      "entropy": 5.520956563949585,
      "epoch": 1.0000907111756168,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004907137521832966,
      "loss": 5.2354,
      "mean_token_accuracy": 0.19049685299396515,
      "num_tokens": 22880081.0,
      "step": 11025
    },
    {
      "entropy": 5.454957485198975,
      "epoch": 1.000544267053701,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004907045501953304,
      "loss": 5.0733,
      "mean_token_accuracy": 0.19731134325265884,
      "num_tokens": 22892141.0,
      "step": 11030
    },
    {
      "entropy": 5.452963924407959,
      "epoch": 1.0009978229317853,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490695343746539,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19878874123096466,
      "num_tokens": 22902688.0,
      "step": 11035
    },
    {
      "entropy": 5.38298077583313,
      "epoch": 1.0014513788098693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906861328371131,
      "loss": 5.0207,
      "mean_token_accuracy": 0.19931923151016234,
      "num_tokens": 22912453.0,
      "step": 11040
    },
    {
      "entropy": 5.4921776294708256,
      "epoch": 1.0019049346879536,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906769174672429,
      "loss": 5.182,
      "mean_token_accuracy": 0.18716212958097458,
      "num_tokens": 22923167.0,
      "step": 11045
    },
    {
      "entropy": 5.5100603103637695,
      "epoch": 1.0023584905660377,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000490667697637119,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19183895885944366,
      "num_tokens": 22934753.0,
      "step": 11050
    },
    {
      "entropy": 5.430588102340698,
      "epoch": 1.002812046444122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906584733469324,
      "loss": 5.0642,
      "mean_token_accuracy": 0.20371449440717698,
      "num_tokens": 22945277.0,
      "step": 11055
    },
    {
      "entropy": 5.500294542312622,
      "epoch": 1.0032656023222062,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004906492445968735,
      "loss": 5.1419,
      "mean_token_accuracy": 0.18651341646909714,
      "num_tokens": 22955797.0,
      "step": 11060
    },
    {
      "entropy": 5.483213472366333,
      "epoch": 1.0037191582002902,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906400113871332,
      "loss": 5.109,
      "mean_token_accuracy": 0.19797017574310302,
      "num_tokens": 22965969.0,
      "step": 11065
    },
    {
      "entropy": 5.449386882781982,
      "epoch": 1.0041727140783745,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004906307737179026,
      "loss": 5.094,
      "mean_token_accuracy": 0.20706432759761811,
      "num_tokens": 22975323.0,
      "step": 11070
    },
    {
      "entropy": 5.4158118724822994,
      "epoch": 1.0046262699564585,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906215315893726,
      "loss": 5.0854,
      "mean_token_accuracy": 0.20272112637758255,
      "num_tokens": 22985063.0,
      "step": 11075
    },
    {
      "entropy": 5.557787370681763,
      "epoch": 1.0050798258345428,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004906122850017344,
      "loss": 5.2222,
      "mean_token_accuracy": 0.18601493388414383,
      "num_tokens": 22996001.0,
      "step": 11080
    },
    {
      "entropy": 5.504201650619507,
      "epoch": 1.005533381712627,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906030339551792,
      "loss": 5.132,
      "mean_token_accuracy": 0.19287194013595582,
      "num_tokens": 23006896.0,
      "step": 11085
    },
    {
      "entropy": 5.462795066833496,
      "epoch": 1.0059869375907111,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905937784498983,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19932139068841934,
      "num_tokens": 23017080.0,
      "step": 11090
    },
    {
      "entropy": 5.472736549377442,
      "epoch": 1.0064404934687954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004905845184860831,
      "loss": 5.0633,
      "mean_token_accuracy": 0.1989378586411476,
      "num_tokens": 23027168.0,
      "step": 11095
    },
    {
      "entropy": 5.474420785903931,
      "epoch": 1.0068940493468794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905752540639252,
      "loss": 5.1048,
      "mean_token_accuracy": 0.19014271944761277,
      "num_tokens": 23037364.0,
      "step": 11100
    },
    {
      "entropy": 5.379934692382813,
      "epoch": 1.0073476052249637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905659851836161,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20249858796596526,
      "num_tokens": 23047609.0,
      "step": 11105
    },
    {
      "entropy": 5.465517282485962,
      "epoch": 1.007801161103048,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905567118453475,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19380562752485275,
      "num_tokens": 23057942.0,
      "step": 11110
    },
    {
      "entropy": 5.455103206634521,
      "epoch": 1.008254716981132,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004905474340493111,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19476013481616974,
      "num_tokens": 23069119.0,
      "step": 11115
    },
    {
      "entropy": 5.412275934219361,
      "epoch": 1.0087082728592163,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004905381517956988,
      "loss": 5.0439,
      "mean_token_accuracy": 0.19710911959409713,
      "num_tokens": 23080060.0,
      "step": 11120
    },
    {
      "entropy": 5.417080545425415,
      "epoch": 1.0091618287373005,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905288650847027,
      "loss": 5.0792,
      "mean_token_accuracy": 0.1906353548169136,
      "num_tokens": 23090645.0,
      "step": 11125
    },
    {
      "entropy": 5.480456209182739,
      "epoch": 1.0096153846153846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905195739165146,
      "loss": 5.0965,
      "mean_token_accuracy": 0.19660176038742067,
      "num_tokens": 23100449.0,
      "step": 11130
    },
    {
      "entropy": 5.476227140426635,
      "epoch": 1.0100689404934688,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490510278291327,
      "loss": 5.0782,
      "mean_token_accuracy": 0.19399780929088592,
      "num_tokens": 23110415.0,
      "step": 11135
    },
    {
      "entropy": 5.424685478210449,
      "epoch": 1.010522496371553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905009782093318,
      "loss": 5.07,
      "mean_token_accuracy": 0.19324882477521896,
      "num_tokens": 23120336.0,
      "step": 11140
    },
    {
      "entropy": 5.4333597183227536,
      "epoch": 1.0109760522496372,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904916736707214,
      "loss": 5.1458,
      "mean_token_accuracy": 0.18603336960077285,
      "num_tokens": 23130568.0,
      "step": 11145
    },
    {
      "entropy": 5.512174463272094,
      "epoch": 1.0114296081277214,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904823646756885,
      "loss": 5.1057,
      "mean_token_accuracy": 0.19878295361995696,
      "num_tokens": 23140080.0,
      "step": 11150
    },
    {
      "entropy": 5.445178365707397,
      "epoch": 1.0118831640058055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904730512244252,
      "loss": 5.031,
      "mean_token_accuracy": 0.19515729397535325,
      "num_tokens": 23150360.0,
      "step": 11155
    },
    {
      "entropy": 5.403223466873169,
      "epoch": 1.0123367198838897,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904637333171243,
      "loss": 5.0357,
      "mean_token_accuracy": 0.19459904134273528,
      "num_tokens": 23160620.0,
      "step": 11160
    },
    {
      "entropy": 5.518907165527343,
      "epoch": 1.0127902757619738,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904544109539785,
      "loss": 5.1258,
      "mean_token_accuracy": 0.19950608909130096,
      "num_tokens": 23171920.0,
      "step": 11165
    },
    {
      "entropy": 5.475254821777344,
      "epoch": 1.013243831640058,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904450841351805,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20122201591730118,
      "num_tokens": 23182929.0,
      "step": 11170
    },
    {
      "entropy": 5.488019275665283,
      "epoch": 1.0136973875181423,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904357528609236,
      "loss": 5.0553,
      "mean_token_accuracy": 0.18875425010919572,
      "num_tokens": 23193020.0,
      "step": 11175
    },
    {
      "entropy": 5.40572509765625,
      "epoch": 1.0141509433962264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004904264171314,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1987457439303398,
      "num_tokens": 23204073.0,
      "step": 11180
    },
    {
      "entropy": 5.39017915725708,
      "epoch": 1.0146044992743106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904170769468035,
      "loss": 4.9797,
      "mean_token_accuracy": 0.19448689222335816,
      "num_tokens": 23214376.0,
      "step": 11185
    },
    {
      "entropy": 5.484990119934082,
      "epoch": 1.0150580551523947,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904077323073268,
      "loss": 5.1377,
      "mean_token_accuracy": 0.19592861980199813,
      "num_tokens": 23224304.0,
      "step": 11190
    },
    {
      "entropy": 5.456968784332275,
      "epoch": 1.015511611030479,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903983832131633,
      "loss": 5.1484,
      "mean_token_accuracy": 0.1907399609684944,
      "num_tokens": 23235324.0,
      "step": 11195
    },
    {
      "entropy": 5.451511812210083,
      "epoch": 1.0159651669085632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903890296645063,
      "loss": 5.1226,
      "mean_token_accuracy": 0.19383998960256577,
      "num_tokens": 23245774.0,
      "step": 11200
    },
    {
      "entropy": 5.491146993637085,
      "epoch": 1.0164187227866472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903796716615495,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19635346233844758,
      "num_tokens": 23256274.0,
      "step": 11205
    },
    {
      "entropy": 5.450030946731568,
      "epoch": 1.0168722786647315,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903703092044861,
      "loss": 5.081,
      "mean_token_accuracy": 0.19349735528230666,
      "num_tokens": 23266304.0,
      "step": 11210
    },
    {
      "entropy": 5.372950410842895,
      "epoch": 1.0173258345428158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903609422935097,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20075019150972367,
      "num_tokens": 23276552.0,
      "step": 11215
    },
    {
      "entropy": 5.465046977996826,
      "epoch": 1.0177793904208998,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903515709288143,
      "loss": 5.1085,
      "mean_token_accuracy": 0.18825433999300004,
      "num_tokens": 23287799.0,
      "step": 11220
    },
    {
      "entropy": 5.4047778129577635,
      "epoch": 1.018232946298984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903421951105933,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20188473016023636,
      "num_tokens": 23297958.0,
      "step": 11225
    },
    {
      "entropy": 5.429022884368896,
      "epoch": 1.0186865021770681,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903328148390411,
      "loss": 5.158,
      "mean_token_accuracy": 0.18890633732080458,
      "num_tokens": 23307523.0,
      "step": 11230
    },
    {
      "entropy": 5.496990394592285,
      "epoch": 1.0191400580551524,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004903234301143513,
      "loss": 5.1286,
      "mean_token_accuracy": 0.19909049868583678,
      "num_tokens": 23317970.0,
      "step": 11235
    },
    {
      "entropy": 5.454283809661865,
      "epoch": 1.0195936139332367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004903140409367181,
      "loss": 5.1051,
      "mean_token_accuracy": 0.19640181660652162,
      "num_tokens": 23328612.0,
      "step": 11240
    },
    {
      "entropy": 5.433233976364136,
      "epoch": 1.0200471698113207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004903046473063357,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19017990976572036,
      "num_tokens": 23338104.0,
      "step": 11245
    },
    {
      "entropy": 5.4058473110198975,
      "epoch": 1.020500725689405,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004902952492233983,
      "loss": 4.987,
      "mean_token_accuracy": 0.20831747949123383,
      "num_tokens": 23348105.0,
      "step": 11250
    },
    {
      "entropy": 5.413752603530884,
      "epoch": 1.020954281567489,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004902858466881002,
      "loss": 4.9502,
      "mean_token_accuracy": 0.20240961015224457,
      "num_tokens": 23357150.0,
      "step": 11255
    },
    {
      "entropy": 5.453649282455444,
      "epoch": 1.0214078374455733,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490276439700636,
      "loss": 5.1546,
      "mean_token_accuracy": 0.19585868269205092,
      "num_tokens": 23367223.0,
      "step": 11260
    },
    {
      "entropy": 5.470883369445801,
      "epoch": 1.0218613933236576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004902670282612002,
      "loss": 5.0934,
      "mean_token_accuracy": 0.1947690561413765,
      "num_tokens": 23377664.0,
      "step": 11265
    },
    {
      "entropy": 5.406785202026367,
      "epoch": 1.0223149492017416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902576123699873,
      "loss": 4.9868,
      "mean_token_accuracy": 0.19861319065093994,
      "num_tokens": 23388057.0,
      "step": 11270
    },
    {
      "entropy": 5.334575366973877,
      "epoch": 1.0227685050798259,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004902481920271922,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2003134861588478,
      "num_tokens": 23397808.0,
      "step": 11275
    },
    {
      "entropy": 5.366316890716552,
      "epoch": 1.02322206095791,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004902387672330097,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19368760138750077,
      "num_tokens": 23406979.0,
      "step": 11280
    },
    {
      "entropy": 5.426238918304444,
      "epoch": 1.0236756168359942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902293379876345,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19691926538944243,
      "num_tokens": 23416545.0,
      "step": 11285
    },
    {
      "entropy": 5.450316762924194,
      "epoch": 1.0241291727140784,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004902199042912618,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19223443567752838,
      "num_tokens": 23427464.0,
      "step": 11290
    },
    {
      "entropy": 5.466953468322754,
      "epoch": 1.0245827285921625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902104661440868,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1991693541407585,
      "num_tokens": 23436690.0,
      "step": 11295
    },
    {
      "entropy": 5.4285167217254635,
      "epoch": 1.0250362844702468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902010235463042,
      "loss": 5.0272,
      "mean_token_accuracy": 0.19876836538314818,
      "num_tokens": 23446701.0,
      "step": 11300
    },
    {
      "entropy": 5.519870901107788,
      "epoch": 1.025489840348331,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901915764981098,
      "loss": 5.131,
      "mean_token_accuracy": 0.189468115568161,
      "num_tokens": 23456966.0,
      "step": 11305
    },
    {
      "entropy": 5.4783522605896,
      "epoch": 1.025943396226415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004901821249996988,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2010599821805954,
      "num_tokens": 23467300.0,
      "step": 11310
    },
    {
      "entropy": 5.496423721313477,
      "epoch": 1.0263969521044993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901726690512665,
      "loss": 5.1383,
      "mean_token_accuracy": 0.19087745994329453,
      "num_tokens": 23478457.0,
      "step": 11315
    },
    {
      "entropy": 5.416741752624512,
      "epoch": 1.0268505079825834,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901632086530087,
      "loss": 5.0217,
      "mean_token_accuracy": 0.19998096525669098,
      "num_tokens": 23488066.0,
      "step": 11320
    },
    {
      "entropy": 5.435812044143677,
      "epoch": 1.0273040638606676,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004901537438051207,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19100387990474701,
      "num_tokens": 23498856.0,
      "step": 11325
    },
    {
      "entropy": 5.46122555732727,
      "epoch": 1.027757619738752,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901442745077986,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19592872262001038,
      "num_tokens": 23509845.0,
      "step": 11330
    },
    {
      "entropy": 5.523153638839721,
      "epoch": 1.028211175616836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901348007612381,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19549354612827302,
      "num_tokens": 23520322.0,
      "step": 11335
    },
    {
      "entropy": 5.4768232822418215,
      "epoch": 1.0286647314949202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901253225656351,
      "loss": 5.1097,
      "mean_token_accuracy": 0.19431154578924179,
      "num_tokens": 23530030.0,
      "step": 11340
    },
    {
      "entropy": 5.5160911083221436,
      "epoch": 1.0291182873730043,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901158399211855,
      "loss": 5.1606,
      "mean_token_accuracy": 0.19508347511291504,
      "num_tokens": 23541063.0,
      "step": 11345
    },
    {
      "entropy": 5.4110572814941404,
      "epoch": 1.0295718432510885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901063528280856,
      "loss": 5.0839,
      "mean_token_accuracy": 0.1935800477862358,
      "num_tokens": 23551693.0,
      "step": 11350
    },
    {
      "entropy": 5.448520803451538,
      "epoch": 1.0300253991291728,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900968612865314,
      "loss": 5.11,
      "mean_token_accuracy": 0.20178464502096177,
      "num_tokens": 23561758.0,
      "step": 11355
    },
    {
      "entropy": 5.408980846405029,
      "epoch": 1.0304789550072568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900873652967195,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19118053913116456,
      "num_tokens": 23571912.0,
      "step": 11360
    },
    {
      "entropy": 5.424785518646241,
      "epoch": 1.030932510885341,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004900778648588459,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20222554206848145,
      "num_tokens": 23582600.0,
      "step": 11365
    },
    {
      "entropy": 5.40379490852356,
      "epoch": 1.0313860667634251,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900683599731073,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2006585642695427,
      "num_tokens": 23592761.0,
      "step": 11370
    },
    {
      "entropy": 5.512581634521484,
      "epoch": 1.0318396226415094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900588506397002,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19584969729185103,
      "num_tokens": 23603119.0,
      "step": 11375
    },
    {
      "entropy": 5.387804555892944,
      "epoch": 1.0322931785195937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004900493368588213,
      "loss": 5.048,
      "mean_token_accuracy": 0.20134272575378417,
      "num_tokens": 23613916.0,
      "step": 11380
    },
    {
      "entropy": 5.507238960266113,
      "epoch": 1.0327467343976777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900398186306672,
      "loss": 5.1862,
      "mean_token_accuracy": 0.19214550256729127,
      "num_tokens": 23623956.0,
      "step": 11385
    },
    {
      "entropy": 5.428924179077148,
      "epoch": 1.033200290275762,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000490030295955435,
      "loss": 5.0863,
      "mean_token_accuracy": 0.1945862203836441,
      "num_tokens": 23634666.0,
      "step": 11390
    },
    {
      "entropy": 5.396199321746826,
      "epoch": 1.0336538461538463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900207688333215,
      "loss": 5.0342,
      "mean_token_accuracy": 0.19014710038900376,
      "num_tokens": 23645601.0,
      "step": 11395
    },
    {
      "entropy": 5.490629100799561,
      "epoch": 1.0341074020319303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900112372645236,
      "loss": 5.1063,
      "mean_token_accuracy": 0.19617590457201003,
      "num_tokens": 23656339.0,
      "step": 11400
    },
    {
      "entropy": 5.4447340965271,
      "epoch": 1.0345609579100146,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900017012492386,
      "loss": 5.0659,
      "mean_token_accuracy": 0.1966501459479332,
      "num_tokens": 23667433.0,
      "step": 11405
    },
    {
      "entropy": 5.4997495174407955,
      "epoch": 1.0350145137880986,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004899921607876634,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19640792310237884,
      "num_tokens": 23678186.0,
      "step": 11410
    },
    {
      "entropy": 5.467590475082398,
      "epoch": 1.0354680696661829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899826158799958,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1908168226480484,
      "num_tokens": 23688317.0,
      "step": 11415
    },
    {
      "entropy": 5.428335046768188,
      "epoch": 1.0359216255442671,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899730665264328,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19815969318151475,
      "num_tokens": 23697752.0,
      "step": 11420
    },
    {
      "entropy": 5.3815539360046385,
      "epoch": 1.0363751814223512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899635127271721,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20386362820863724,
      "num_tokens": 23707803.0,
      "step": 11425
    },
    {
      "entropy": 5.378336381912232,
      "epoch": 1.0368287373004355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899539544824111,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20478526055812835,
      "num_tokens": 23717630.0,
      "step": 11430
    },
    {
      "entropy": 5.455833387374878,
      "epoch": 1.0372822931785195,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899443917923476,
      "loss": 5.1178,
      "mean_token_accuracy": 0.19574906826019287,
      "num_tokens": 23728836.0,
      "step": 11435
    },
    {
      "entropy": 5.418558263778687,
      "epoch": 1.0377358490566038,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899348246571792,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19457397907972335,
      "num_tokens": 23740145.0,
      "step": 11440
    },
    {
      "entropy": 5.45527195930481,
      "epoch": 1.038189404934688,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489925253077104,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19532546550035476,
      "num_tokens": 23751162.0,
      "step": 11445
    },
    {
      "entropy": 5.454708051681519,
      "epoch": 1.038642960812772,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004899156770523196,
      "loss": 5.0814,
      "mean_token_accuracy": 0.1966884195804596,
      "num_tokens": 23761634.0,
      "step": 11450
    },
    {
      "entropy": 5.4408070087432865,
      "epoch": 1.0390965166908563,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004899060965830244,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20172560662031175,
      "num_tokens": 23772483.0,
      "step": 11455
    },
    {
      "entropy": 5.417911100387573,
      "epoch": 1.0395500725689404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898965116694162,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1978759214282036,
      "num_tokens": 23782298.0,
      "step": 11460
    },
    {
      "entropy": 5.4759174346923825,
      "epoch": 1.0400036284470247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898869223116935,
      "loss": 5.1069,
      "mean_token_accuracy": 0.19060955047607422,
      "num_tokens": 23792492.0,
      "step": 11465
    },
    {
      "entropy": 5.512869071960449,
      "epoch": 1.040457184325109,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004898773285100546,
      "loss": 5.2549,
      "mean_token_accuracy": 0.1875505417585373,
      "num_tokens": 23802999.0,
      "step": 11470
    },
    {
      "entropy": 5.4411492347717285,
      "epoch": 1.040910740203193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004898677302646975,
      "loss": 5.0535,
      "mean_token_accuracy": 0.19900441765785218,
      "num_tokens": 23812541.0,
      "step": 11475
    },
    {
      "entropy": 5.45083065032959,
      "epoch": 1.0413642960812772,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898581275758211,
      "loss": 5.1469,
      "mean_token_accuracy": 0.18596832305192948,
      "num_tokens": 23823244.0,
      "step": 11480
    },
    {
      "entropy": 5.434124565124511,
      "epoch": 1.0418178519593615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898485204436239,
      "loss": 4.9841,
      "mean_token_accuracy": 0.202264666557312,
      "num_tokens": 23832346.0,
      "step": 11485
    },
    {
      "entropy": 5.484871912002563,
      "epoch": 1.0422714078374455,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898389088683047,
      "loss": 5.117,
      "mean_token_accuracy": 0.19769508838653566,
      "num_tokens": 23842554.0,
      "step": 11490
    },
    {
      "entropy": 5.402813959121704,
      "epoch": 1.0427249637155298,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004898292928500619,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20696790814399718,
      "num_tokens": 23852938.0,
      "step": 11495
    },
    {
      "entropy": 5.412786197662354,
      "epoch": 1.0431785195936139,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004898196723890947,
      "loss": 4.9841,
      "mean_token_accuracy": 0.20367277711629866,
      "num_tokens": 23862367.0,
      "step": 11500
    },
    {
      "entropy": 5.41459174156189,
      "epoch": 1.0436320754716981,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004898100474856019,
      "loss": 5.1084,
      "mean_token_accuracy": 0.19728990495204926,
      "num_tokens": 23873249.0,
      "step": 11505
    },
    {
      "entropy": 5.454684019088745,
      "epoch": 1.0440856313497824,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004898004181397824,
      "loss": 5.1245,
      "mean_token_accuracy": 0.18645377606153488,
      "num_tokens": 23883830.0,
      "step": 11510
    },
    {
      "entropy": 5.490788984298706,
      "epoch": 1.0445391872278664,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897907843518358,
      "loss": 5.033,
      "mean_token_accuracy": 0.1962950274348259,
      "num_tokens": 23894359.0,
      "step": 11515
    },
    {
      "entropy": 5.436550092697144,
      "epoch": 1.0449927431059507,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897811461219609,
      "loss": 5.042,
      "mean_token_accuracy": 0.19545938819646835,
      "num_tokens": 23903779.0,
      "step": 11520
    },
    {
      "entropy": 5.438144254684448,
      "epoch": 1.0454462989840347,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897715034503571,
      "loss": 5.0932,
      "mean_token_accuracy": 0.19297188222408296,
      "num_tokens": 23914523.0,
      "step": 11525
    },
    {
      "entropy": 5.4399497509002686,
      "epoch": 1.045899854862119,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489761856337224,
      "loss": 5.08,
      "mean_token_accuracy": 0.19704519659280778,
      "num_tokens": 23924996.0,
      "step": 11530
    },
    {
      "entropy": 5.451943254470825,
      "epoch": 1.0463534107402033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004897522047827609,
      "loss": 5.0481,
      "mean_token_accuracy": 0.2018464282155037,
      "num_tokens": 23935843.0,
      "step": 11535
    },
    {
      "entropy": 5.422876787185669,
      "epoch": 1.0468069666182873,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004897425487871676,
      "loss": 5.009,
      "mean_token_accuracy": 0.1973980635404587,
      "num_tokens": 23945044.0,
      "step": 11540
    },
    {
      "entropy": 5.415144586563111,
      "epoch": 1.0472605224963716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897328883506436,
      "loss": 5.0648,
      "mean_token_accuracy": 0.19418827295303345,
      "num_tokens": 23955933.0,
      "step": 11545
    },
    {
      "entropy": 5.455891418457031,
      "epoch": 1.0477140783744556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897232234733887,
      "loss": 5.1506,
      "mean_token_accuracy": 0.18856869488954545,
      "num_tokens": 23966320.0,
      "step": 11550
    },
    {
      "entropy": 5.4573126316070555,
      "epoch": 1.04816763425254,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897135541556028,
      "loss": 5.0344,
      "mean_token_accuracy": 0.19937969446182252,
      "num_tokens": 23976233.0,
      "step": 11555
    },
    {
      "entropy": 5.468496084213257,
      "epoch": 1.0486211901306242,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000489703880397486,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19293475151062012,
      "num_tokens": 23987854.0,
      "step": 11560
    },
    {
      "entropy": 5.446594524383545,
      "epoch": 1.0490747460087082,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004896942021992382,
      "loss": 5.1068,
      "mean_token_accuracy": 0.1926787257194519,
      "num_tokens": 23997973.0,
      "step": 11565
    },
    {
      "entropy": 5.448502826690674,
      "epoch": 1.0495283018867925,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004896845195610595,
      "loss": 5.0942,
      "mean_token_accuracy": 0.19503124207258224,
      "num_tokens": 24007313.0,
      "step": 11570
    },
    {
      "entropy": 5.430074739456177,
      "epoch": 1.0499818577648767,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896748324831504,
      "loss": 5.1129,
      "mean_token_accuracy": 0.1907646909356117,
      "num_tokens": 24018254.0,
      "step": 11575
    },
    {
      "entropy": 5.468389320373535,
      "epoch": 1.0504354136429608,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004896651409657109,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1924986645579338,
      "num_tokens": 24029799.0,
      "step": 11580
    },
    {
      "entropy": 5.381548690795898,
      "epoch": 1.050888969521045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004896554450089417,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20224028080701828,
      "num_tokens": 24040972.0,
      "step": 11585
    },
    {
      "entropy": 5.401288747787476,
      "epoch": 1.051342525399129,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004896457446130431,
      "loss": 4.9488,
      "mean_token_accuracy": 0.2015111193060875,
      "num_tokens": 24050781.0,
      "step": 11590
    },
    {
      "entropy": 5.395908451080322,
      "epoch": 1.0517960812772134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896360397782159,
      "loss": 5.0786,
      "mean_token_accuracy": 0.18981412649154664,
      "num_tokens": 24060184.0,
      "step": 11595
    },
    {
      "entropy": 5.503989028930664,
      "epoch": 1.0522496371552976,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004896263305046607,
      "loss": 5.1437,
      "mean_token_accuracy": 0.19646002352237701,
      "num_tokens": 24070777.0,
      "step": 11600
    },
    {
      "entropy": 5.475661325454712,
      "epoch": 1.0527031930333817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896166167925782,
      "loss": 5.169,
      "mean_token_accuracy": 0.19187893271446227,
      "num_tokens": 24080878.0,
      "step": 11605
    },
    {
      "entropy": 5.473572683334351,
      "epoch": 1.053156748911466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896068986421693,
      "loss": 5.1247,
      "mean_token_accuracy": 0.2013508141040802,
      "num_tokens": 24090038.0,
      "step": 11610
    },
    {
      "entropy": 5.542226314544678,
      "epoch": 1.05361030478955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004895971760536351,
      "loss": 5.148,
      "mean_token_accuracy": 0.1930503010749817,
      "num_tokens": 24100425.0,
      "step": 11615
    },
    {
      "entropy": 5.510891008377075,
      "epoch": 1.0540638606676342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004895874490271768,
      "loss": 5.2155,
      "mean_token_accuracy": 0.18790028393268585,
      "num_tokens": 24112182.0,
      "step": 11620
    },
    {
      "entropy": 5.470790004730224,
      "epoch": 1.0545174165457185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895777175629952,
      "loss": 5.0586,
      "mean_token_accuracy": 0.19576068520545958,
      "num_tokens": 24122182.0,
      "step": 11625
    },
    {
      "entropy": 5.5000087261199955,
      "epoch": 1.0549709724238026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895679816612917,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2022800475358963,
      "num_tokens": 24131930.0,
      "step": 11630
    },
    {
      "entropy": 5.490646553039551,
      "epoch": 1.0554245283018868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895582413222678,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20672089010477065,
      "num_tokens": 24141575.0,
      "step": 11635
    },
    {
      "entropy": 5.49963755607605,
      "epoch": 1.0558780841799709,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004895484965461247,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1874408856034279,
      "num_tokens": 24151720.0,
      "step": 11640
    },
    {
      "entropy": 5.440736627578735,
      "epoch": 1.0563316400580551,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489538747333064,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19491846710443497,
      "num_tokens": 24161881.0,
      "step": 11645
    },
    {
      "entropy": 5.440998458862305,
      "epoch": 1.0567851959361394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895289936832874,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2015311449766159,
      "num_tokens": 24171474.0,
      "step": 11650
    },
    {
      "entropy": 5.508123254776001,
      "epoch": 1.0572387518142234,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004895192355969966,
      "loss": 5.1487,
      "mean_token_accuracy": 0.19003770053386687,
      "num_tokens": 24182186.0,
      "step": 11655
    },
    {
      "entropy": 5.491469430923462,
      "epoch": 1.0576923076923077,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004895094730743932,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1989141210913658,
      "num_tokens": 24192318.0,
      "step": 11660
    },
    {
      "entropy": 5.486510848999023,
      "epoch": 1.058145863570392,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004894997061156793,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19250682592391968,
      "num_tokens": 24202918.0,
      "step": 11665
    },
    {
      "entropy": 5.416793775558472,
      "epoch": 1.058599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894899347210567,
      "loss": 5.0251,
      "mean_token_accuracy": 0.19752568751573563,
      "num_tokens": 24213218.0,
      "step": 11670
    },
    {
      "entropy": 5.421510839462281,
      "epoch": 1.0590529753265603,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894801588907278,
      "loss": 5.0285,
      "mean_token_accuracy": 0.19644063711166382,
      "num_tokens": 24223229.0,
      "step": 11675
    },
    {
      "entropy": 5.5234157085418705,
      "epoch": 1.0595065312046443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894703786248944,
      "loss": 5.0986,
      "mean_token_accuracy": 0.1940212830901146,
      "num_tokens": 24233637.0,
      "step": 11680
    },
    {
      "entropy": 5.477429342269898,
      "epoch": 1.0599600870827286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000489460593923759,
      "loss": 5.1354,
      "mean_token_accuracy": 0.19680089503526688,
      "num_tokens": 24244251.0,
      "step": 11685
    },
    {
      "entropy": 5.442391777038575,
      "epoch": 1.0604136429608129,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004894508047875237,
      "loss": 5.0383,
      "mean_token_accuracy": 0.19454250782728194,
      "num_tokens": 24254380.0,
      "step": 11690
    },
    {
      "entropy": 5.318802404403686,
      "epoch": 1.060867198838897,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004894410112163913,
      "loss": 4.9794,
      "mean_token_accuracy": 0.19969397485256196,
      "num_tokens": 24264841.0,
      "step": 11695
    },
    {
      "entropy": 5.373764562606811,
      "epoch": 1.0613207547169812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004894312132105641,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20056951344013213,
      "num_tokens": 24274383.0,
      "step": 11700
    },
    {
      "entropy": 5.514595413208008,
      "epoch": 1.0617743105950652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894214107702447,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19603312909603118,
      "num_tokens": 24284140.0,
      "step": 11705
    },
    {
      "entropy": 5.563332653045654,
      "epoch": 1.0622278664731495,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000489411603895636,
      "loss": 5.1639,
      "mean_token_accuracy": 0.19430688321590422,
      "num_tokens": 24295044.0,
      "step": 11710
    },
    {
      "entropy": 5.4479235172271725,
      "epoch": 1.0626814223512338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894017925869406,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19257737845182418,
      "num_tokens": 24305080.0,
      "step": 11715
    },
    {
      "entropy": 5.455956172943115,
      "epoch": 1.0631349782293178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893919768443615,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19278850853443147,
      "num_tokens": 24316306.0,
      "step": 11720
    },
    {
      "entropy": 5.403377342224121,
      "epoch": 1.063588534107402,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893821566681017,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19761644750833512,
      "num_tokens": 24326740.0,
      "step": 11725
    },
    {
      "entropy": 5.464867639541626,
      "epoch": 1.0640420899854863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893723320583644,
      "loss": 5.1241,
      "mean_token_accuracy": 0.18957178741693498,
      "num_tokens": 24337621.0,
      "step": 11730
    },
    {
      "entropy": 5.38532977104187,
      "epoch": 1.0644956458635704,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893625030153525,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20226975977420808,
      "num_tokens": 24347732.0,
      "step": 11735
    },
    {
      "entropy": 5.4007179737091064,
      "epoch": 1.0649492017416546,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004893526695392694,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2070948839187622,
      "num_tokens": 24359173.0,
      "step": 11740
    },
    {
      "entropy": 5.41707181930542,
      "epoch": 1.0654027576197387,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004893428316303185,
      "loss": 5.0104,
      "mean_token_accuracy": 0.19994874447584152,
      "num_tokens": 24369465.0,
      "step": 11745
    },
    {
      "entropy": 5.376338243484497,
      "epoch": 1.065856313497823,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004893329892887031,
      "loss": 5.0463,
      "mean_token_accuracy": 0.19964333474636078,
      "num_tokens": 24378899.0,
      "step": 11750
    },
    {
      "entropy": 5.45647497177124,
      "epoch": 1.0663098693759072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893231425146271,
      "loss": 5.1578,
      "mean_token_accuracy": 0.19202471226453782,
      "num_tokens": 24388919.0,
      "step": 11755
    },
    {
      "entropy": 5.399653387069702,
      "epoch": 1.0667634252539913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004893132913082937,
      "loss": 5.0326,
      "mean_token_accuracy": 0.19771426767110825,
      "num_tokens": 24399632.0,
      "step": 11760
    },
    {
      "entropy": 5.473707294464111,
      "epoch": 1.0672169811320755,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004893034356699069,
      "loss": 5.071,
      "mean_token_accuracy": 0.19982456862926484,
      "num_tokens": 24410817.0,
      "step": 11765
    },
    {
      "entropy": 5.410086917877197,
      "epoch": 1.0676705370101596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004892935755996705,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1969650834798813,
      "num_tokens": 24421058.0,
      "step": 11770
    },
    {
      "entropy": 5.388855934143066,
      "epoch": 1.0681240928882438,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892837110977883,
      "loss": 4.9924,
      "mean_token_accuracy": 0.19559596329927445,
      "num_tokens": 24431522.0,
      "step": 11775
    },
    {
      "entropy": 5.419882726669312,
      "epoch": 1.068577648766328,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004892738421644644,
      "loss": 5.0341,
      "mean_token_accuracy": 0.19629716575145723,
      "num_tokens": 24441746.0,
      "step": 11780
    },
    {
      "entropy": 5.366162490844727,
      "epoch": 1.0690312046444121,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892639687999027,
      "loss": 4.95,
      "mean_token_accuracy": 0.20535225570201873,
      "num_tokens": 24451916.0,
      "step": 11785
    },
    {
      "entropy": 5.41597089767456,
      "epoch": 1.0694847605224964,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892540910043076,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20218056589365005,
      "num_tokens": 24461935.0,
      "step": 11790
    },
    {
      "entropy": 5.413613605499267,
      "epoch": 1.0699383164005805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892442087778833,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20131217688322067,
      "num_tokens": 24472437.0,
      "step": 11795
    },
    {
      "entropy": 5.4258464813232425,
      "epoch": 1.0703918722786647,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004892343221208342,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1990344926714897,
      "num_tokens": 24481726.0,
      "step": 11800
    },
    {
      "entropy": 5.487134218215942,
      "epoch": 1.070845428156749,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004892244310333646,
      "loss": 5.1581,
      "mean_token_accuracy": 0.19333946853876113,
      "num_tokens": 24492340.0,
      "step": 11805
    },
    {
      "entropy": 5.4427143096923825,
      "epoch": 1.071298984034833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892145355156793,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20001528561115264,
      "num_tokens": 24502138.0,
      "step": 11810
    },
    {
      "entropy": 5.435838603973389,
      "epoch": 1.0717525399129173,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004892046355679829,
      "loss": 5.0742,
      "mean_token_accuracy": 0.19548026472330093,
      "num_tokens": 24512723.0,
      "step": 11815
    },
    {
      "entropy": 5.462838745117187,
      "epoch": 1.0722060957910013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891947311904799,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19110726565122604,
      "num_tokens": 24523766.0,
      "step": 11820
    },
    {
      "entropy": 5.464926099777221,
      "epoch": 1.0726596516690856,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004891848223833753,
      "loss": 5.0878,
      "mean_token_accuracy": 0.1985626518726349,
      "num_tokens": 24534571.0,
      "step": 11825
    },
    {
      "entropy": 5.450968503952026,
      "epoch": 1.0731132075471699,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489174909146874,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20022153854370117,
      "num_tokens": 24545265.0,
      "step": 11830
    },
    {
      "entropy": 5.505740213394165,
      "epoch": 1.073566763425254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004891649914811811,
      "loss": 5.1771,
      "mean_token_accuracy": 0.1910492166876793,
      "num_tokens": 24555524.0,
      "step": 11835
    },
    {
      "entropy": 5.40087103843689,
      "epoch": 1.0740203193033382,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004891550693865014,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2117801010608673,
      "num_tokens": 24565420.0,
      "step": 11840
    },
    {
      "entropy": 5.4563820362091064,
      "epoch": 1.0744738751814222,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004891451428630405,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19624711722135543,
      "num_tokens": 24576794.0,
      "step": 11845
    },
    {
      "entropy": 5.445266962051392,
      "epoch": 1.0749274310595065,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891352119110033,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1940919429063797,
      "num_tokens": 24586320.0,
      "step": 11850
    },
    {
      "entropy": 5.3934401512146,
      "epoch": 1.0753809869375908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891252765305954,
      "loss": 5.0651,
      "mean_token_accuracy": 0.20423324257135392,
      "num_tokens": 24596524.0,
      "step": 11855
    },
    {
      "entropy": 5.46302227973938,
      "epoch": 1.0758345428156748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891153367220222,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20729490667581557,
      "num_tokens": 24605877.0,
      "step": 11860
    },
    {
      "entropy": 5.5180091381073,
      "epoch": 1.076288098693759,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004891053924854894,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1921609804034233,
      "num_tokens": 24616719.0,
      "step": 11865
    },
    {
      "entropy": 5.402059555053711,
      "epoch": 1.0767416545718433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890954438212024,
      "loss": 4.96,
      "mean_token_accuracy": 0.20680669397115709,
      "num_tokens": 24626882.0,
      "step": 11870
    },
    {
      "entropy": 5.457244205474853,
      "epoch": 1.0771952104499274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890854907293671,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19212204962968826,
      "num_tokens": 24637884.0,
      "step": 11875
    },
    {
      "entropy": 5.399580287933349,
      "epoch": 1.0776487663280117,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890755332101893,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19744679927825928,
      "num_tokens": 24647389.0,
      "step": 11880
    },
    {
      "entropy": 5.430721712112427,
      "epoch": 1.0781023222060957,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004890655712638749,
      "loss": 5.0856,
      "mean_token_accuracy": 0.1923668220639229,
      "num_tokens": 24657595.0,
      "step": 11885
    },
    {
      "entropy": 5.4359862327575685,
      "epoch": 1.07855587808418,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048905560489063,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20309585630893706,
      "num_tokens": 24667683.0,
      "step": 11890
    },
    {
      "entropy": 5.470280981063842,
      "epoch": 1.0790094339622642,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004890456340906605,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1932810515165329,
      "num_tokens": 24678041.0,
      "step": 11895
    },
    {
      "entropy": 5.421089029312133,
      "epoch": 1.0794629898403483,
      "grad_norm": 1.0,
      "learning_rate": 0.0004890356588641729,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20076433718204498,
      "num_tokens": 24688569.0,
      "step": 11900
    },
    {
      "entropy": 5.457909011840821,
      "epoch": 1.0799165457184325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890256792113733,
      "loss": 5.0963,
      "mean_token_accuracy": 0.19976598918437957,
      "num_tokens": 24699913.0,
      "step": 11905
    },
    {
      "entropy": 5.5236265659332275,
      "epoch": 1.0803701015965168,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004890156951324681,
      "loss": 5.1152,
      "mean_token_accuracy": 0.1960355222225189,
      "num_tokens": 24710682.0,
      "step": 11910
    },
    {
      "entropy": 5.541093349456787,
      "epoch": 1.0808236574746009,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004890057066276636,
      "loss": 5.1933,
      "mean_token_accuracy": 0.18897853046655655,
      "num_tokens": 24721087.0,
      "step": 11915
    },
    {
      "entropy": 5.404382562637329,
      "epoch": 1.0812772133526851,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004889957136971668,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2035218745470047,
      "num_tokens": 24731323.0,
      "step": 11920
    },
    {
      "entropy": 5.405501651763916,
      "epoch": 1.0817307692307692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488985716341184,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19516118317842485,
      "num_tokens": 24742559.0,
      "step": 11925
    },
    {
      "entropy": 5.49205756187439,
      "epoch": 1.0821843251088534,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004889757145599222,
      "loss": 5.0874,
      "mean_token_accuracy": 0.19950402975082399,
      "num_tokens": 24752307.0,
      "step": 11930
    },
    {
      "entropy": 5.3561821460723875,
      "epoch": 1.0826378809869377,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488965708353588,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19089826941490173,
      "num_tokens": 24762792.0,
      "step": 11935
    },
    {
      "entropy": 5.476434326171875,
      "epoch": 1.0830914368650217,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889556977223883,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19093414396047592,
      "num_tokens": 24774377.0,
      "step": 11940
    },
    {
      "entropy": 5.4761061668396,
      "epoch": 1.083544992743106,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889456826665304,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19885472059249878,
      "num_tokens": 24785919.0,
      "step": 11945
    },
    {
      "entropy": 5.484697580337524,
      "epoch": 1.08399854862119,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004889356631862211,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1904277101159096,
      "num_tokens": 24796594.0,
      "step": 11950
    },
    {
      "entropy": 5.427489614486694,
      "epoch": 1.0844521044992743,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488925639281668,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1992164745926857,
      "num_tokens": 24807322.0,
      "step": 11955
    },
    {
      "entropy": 5.385541248321533,
      "epoch": 1.0849056603773586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004889156109530781,
      "loss": 4.985,
      "mean_token_accuracy": 0.19826211035251617,
      "num_tokens": 24817593.0,
      "step": 11960
    },
    {
      "entropy": 5.427872514724731,
      "epoch": 1.0853592162554426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889055782006588,
      "loss": 4.9904,
      "mean_token_accuracy": 0.205525766313076,
      "num_tokens": 24827577.0,
      "step": 11965
    },
    {
      "entropy": 5.401228189468384,
      "epoch": 1.085812772133527,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888955410246176,
      "loss": 5.0438,
      "mean_token_accuracy": 0.19906434863805772,
      "num_tokens": 24838148.0,
      "step": 11970
    },
    {
      "entropy": 5.435687494277954,
      "epoch": 1.086266328011611,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888854994251621,
      "loss": 4.995,
      "mean_token_accuracy": 0.2041756436228752,
      "num_tokens": 24848451.0,
      "step": 11975
    },
    {
      "entropy": 5.477515411376953,
      "epoch": 1.0867198838896952,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004888754534025,
      "loss": 5.1223,
      "mean_token_accuracy": 0.18711033910512925,
      "num_tokens": 24859619.0,
      "step": 11980
    },
    {
      "entropy": 5.460151338577271,
      "epoch": 1.0871734397677795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004888654029568391,
      "loss": 5.1545,
      "mean_token_accuracy": 0.1858536958694458,
      "num_tokens": 24869863.0,
      "step": 11985
    },
    {
      "entropy": 5.4626185417175295,
      "epoch": 1.0876269956458635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004888553480883871,
      "loss": 5.1,
      "mean_token_accuracy": 0.18833720684051514,
      "num_tokens": 24880524.0,
      "step": 11990
    },
    {
      "entropy": 5.428664207458496,
      "epoch": 1.0880805515239478,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488845288797352,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19847525656223297,
      "num_tokens": 24891514.0,
      "step": 11995
    },
    {
      "entropy": 5.419693374633789,
      "epoch": 1.0885341074020318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888352250839418,
      "loss": 5.012,
      "mean_token_accuracy": 0.200258307158947,
      "num_tokens": 24900842.0,
      "step": 12000
    },
    {
      "epoch": 1.0885341074020318,
      "eval_entropy": 5.213240118803482,
      "eval_loss": 5.159305095672607,
      "eval_mean_token_accuracy": 0.20191680841923298,
      "eval_num_tokens": 24900842.0,
      "eval_runtime": 20.5739,
      "eval_samples_per_second": 1718.686,
      "eval_steps_per_second": 214.836,
      "step": 12000
    },
    {
      "entropy": 5.403762531280518,
      "epoch": 1.088987663280116,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004888251569483647,
      "loss": 5.0113,
      "mean_token_accuracy": 0.19987739324569703,
      "num_tokens": 24911845.0,
      "step": 12005
    },
    {
      "entropy": 5.361207818984985,
      "epoch": 1.0894412191582004,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888150843908288,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2027574062347412,
      "num_tokens": 24922211.0,
      "step": 12010
    },
    {
      "entropy": 5.489847183227539,
      "epoch": 1.0898947750362844,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888050074115426,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19196625500917436,
      "num_tokens": 24932129.0,
      "step": 12015
    },
    {
      "entropy": 5.494090127944946,
      "epoch": 1.0903483309143687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004887949260107143,
      "loss": 5.1334,
      "mean_token_accuracy": 0.1945919945836067,
      "num_tokens": 24942869.0,
      "step": 12020
    },
    {
      "entropy": 5.415091466903687,
      "epoch": 1.0908018867924527,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004887848401885524,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2002049967646599,
      "num_tokens": 24952621.0,
      "step": 12025
    },
    {
      "entropy": 5.534883880615235,
      "epoch": 1.091255442670537,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004887747499452656,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1875045657157898,
      "num_tokens": 24963587.0,
      "step": 12030
    },
    {
      "entropy": 5.480429410934448,
      "epoch": 1.0917089985486212,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887646552810624,
      "loss": 5.0306,
      "mean_token_accuracy": 0.19918842613697052,
      "num_tokens": 24974122.0,
      "step": 12035
    },
    {
      "entropy": 5.440608882904053,
      "epoch": 1.0921625544267053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887545561961517,
      "loss": 5.1152,
      "mean_token_accuracy": 0.19135283827781677,
      "num_tokens": 24985145.0,
      "step": 12040
    },
    {
      "entropy": 5.395934009552002,
      "epoch": 1.0926161103047896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887444526907424,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19766356348991393,
      "num_tokens": 24995511.0,
      "step": 12045
    },
    {
      "entropy": 5.450332164764404,
      "epoch": 1.0930696661828738,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004887343447650432,
      "loss": 5.061,
      "mean_token_accuracy": 0.1917600452899933,
      "num_tokens": 25006193.0,
      "step": 12050
    },
    {
      "entropy": 5.46754002571106,
      "epoch": 1.0935232220609579,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004887242324192634,
      "loss": 5.0732,
      "mean_token_accuracy": 0.1974289357662201,
      "num_tokens": 25016444.0,
      "step": 12055
    },
    {
      "entropy": 5.467090892791748,
      "epoch": 1.0939767779390421,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004887141156536119,
      "loss": 5.0745,
      "mean_token_accuracy": 0.1957561507821083,
      "num_tokens": 25027898.0,
      "step": 12060
    },
    {
      "entropy": 5.428858900070191,
      "epoch": 1.0944303338171262,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887039944682981,
      "loss": 4.9934,
      "mean_token_accuracy": 0.19940518587827682,
      "num_tokens": 25038201.0,
      "step": 12065
    },
    {
      "entropy": 5.53903865814209,
      "epoch": 1.0948838896952104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886938688635313,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18696350902318953,
      "num_tokens": 25048466.0,
      "step": 12070
    },
    {
      "entropy": 5.558661270141601,
      "epoch": 1.0953374455732947,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004886837388395207,
      "loss": 5.1272,
      "mean_token_accuracy": 0.197776260972023,
      "num_tokens": 25059064.0,
      "step": 12075
    },
    {
      "entropy": 5.527079391479492,
      "epoch": 1.0957910014513788,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488673604396476,
      "loss": 5.2062,
      "mean_token_accuracy": 0.18161393254995345,
      "num_tokens": 25068834.0,
      "step": 12080
    },
    {
      "entropy": 5.487286567687988,
      "epoch": 1.096244557329463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004886634655346067,
      "loss": 5.1407,
      "mean_token_accuracy": 0.19002335518598557,
      "num_tokens": 25080058.0,
      "step": 12085
    },
    {
      "entropy": 5.416148138046265,
      "epoch": 1.0966981132075473,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886533222541225,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20018682926893233,
      "num_tokens": 25089751.0,
      "step": 12090
    },
    {
      "entropy": 5.448303937911987,
      "epoch": 1.0971516690856313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886431745552331,
      "loss": 5.0962,
      "mean_token_accuracy": 0.19731398224830626,
      "num_tokens": 25099055.0,
      "step": 12095
    },
    {
      "entropy": 5.538101100921631,
      "epoch": 1.0976052249637156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004886330224381485,
      "loss": 5.1922,
      "mean_token_accuracy": 0.18713112622499467,
      "num_tokens": 25109793.0,
      "step": 12100
    },
    {
      "entropy": 5.500201988220215,
      "epoch": 1.0980587808417996,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886228659030785,
      "loss": 5.092,
      "mean_token_accuracy": 0.19966764748096466,
      "num_tokens": 25119814.0,
      "step": 12105
    },
    {
      "entropy": 5.498597908020019,
      "epoch": 1.098512336719884,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004886127049502331,
      "loss": 5.1462,
      "mean_token_accuracy": 0.19277023226022721,
      "num_tokens": 25130747.0,
      "step": 12110
    },
    {
      "entropy": 5.426317882537842,
      "epoch": 1.0989658925979682,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004886025395798228,
      "loss": 5.0515,
      "mean_token_accuracy": 0.1940399870276451,
      "num_tokens": 25142235.0,
      "step": 12115
    },
    {
      "entropy": 5.343563508987427,
      "epoch": 1.0994194484760522,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885923697920573,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2048935681581497,
      "num_tokens": 25152047.0,
      "step": 12120
    },
    {
      "entropy": 5.41880784034729,
      "epoch": 1.0998730043541365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885821955871472,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20211029052734375,
      "num_tokens": 25162678.0,
      "step": 12125
    },
    {
      "entropy": 5.490183734893799,
      "epoch": 1.1003265602322205,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004885720169653028,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19948583245277404,
      "num_tokens": 25172973.0,
      "step": 12130
    },
    {
      "entropy": 5.490416479110718,
      "epoch": 1.1007801161103048,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004885618339267348,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1911870390176773,
      "num_tokens": 25183458.0,
      "step": 12135
    },
    {
      "entropy": 5.395343351364136,
      "epoch": 1.101233671988389,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004885516464716537,
      "loss": 5.0467,
      "mean_token_accuracy": 0.19833064079284668,
      "num_tokens": 25193827.0,
      "step": 12140
    },
    {
      "entropy": 5.504849767684936,
      "epoch": 1.101687227866473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048854145460027,
      "loss": 5.0821,
      "mean_token_accuracy": 0.18926733285188674,
      "num_tokens": 25204448.0,
      "step": 12145
    },
    {
      "entropy": 5.47561469078064,
      "epoch": 1.1021407837445574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885312583127946,
      "loss": 5.0308,
      "mean_token_accuracy": 0.2008280038833618,
      "num_tokens": 25215041.0,
      "step": 12150
    },
    {
      "entropy": 5.443638896942138,
      "epoch": 1.1025943396226414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885210576094385,
      "loss": 5.0681,
      "mean_token_accuracy": 0.19479611963033677,
      "num_tokens": 25225937.0,
      "step": 12155
    },
    {
      "entropy": 5.407575702667236,
      "epoch": 1.1030478955007257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885108524904124,
      "loss": 4.986,
      "mean_token_accuracy": 0.20156417191028594,
      "num_tokens": 25235270.0,
      "step": 12160
    },
    {
      "entropy": 5.43190450668335,
      "epoch": 1.10350145137881,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885006429559275,
      "loss": 5.0864,
      "mean_token_accuracy": 0.18922053277492523,
      "num_tokens": 25247080.0,
      "step": 12165
    },
    {
      "entropy": 5.489084720611572,
      "epoch": 1.103955007256894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004884904290061949,
      "loss": 5.0778,
      "mean_token_accuracy": 0.18807390183210373,
      "num_tokens": 25256718.0,
      "step": 12170
    },
    {
      "entropy": 5.487191200256348,
      "epoch": 1.1044085631349783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004884802106414258,
      "loss": 5.1941,
      "mean_token_accuracy": 0.18723392188549043,
      "num_tokens": 25266632.0,
      "step": 12175
    },
    {
      "entropy": 5.464863109588623,
      "epoch": 1.1048621190130623,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884699878618316,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1978112742304802,
      "num_tokens": 25275598.0,
      "step": 12180
    },
    {
      "entropy": 5.517183065414429,
      "epoch": 1.1053156748911466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004884597606676236,
      "loss": 5.1461,
      "mean_token_accuracy": 0.1934637799859047,
      "num_tokens": 25286598.0,
      "step": 12185
    },
    {
      "entropy": 5.4533368110656735,
      "epoch": 1.1057692307692308,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004884495290590134,
      "loss": 5.0586,
      "mean_token_accuracy": 0.1986728146672249,
      "num_tokens": 25296730.0,
      "step": 12190
    },
    {
      "entropy": 5.412643527984619,
      "epoch": 1.1062227866473149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004884392930362124,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20431596785783768,
      "num_tokens": 25306326.0,
      "step": 12195
    },
    {
      "entropy": 5.470515298843384,
      "epoch": 1.1066763425253991,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004884290525994326,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19730253219604493,
      "num_tokens": 25316342.0,
      "step": 12200
    },
    {
      "entropy": 5.545531749725342,
      "epoch": 1.1071298984034832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884188077488855,
      "loss": 5.1291,
      "mean_token_accuracy": 0.18943889439105988,
      "num_tokens": 25326129.0,
      "step": 12205
    },
    {
      "entropy": 5.431648206710816,
      "epoch": 1.1075834542815675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884085584847831,
      "loss": 5.087,
      "mean_token_accuracy": 0.19083992540836334,
      "num_tokens": 25337156.0,
      "step": 12210
    },
    {
      "entropy": 5.507016372680664,
      "epoch": 1.1080370101596517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883983048073375,
      "loss": 5.2059,
      "mean_token_accuracy": 0.18340847343206407,
      "num_tokens": 25347632.0,
      "step": 12215
    },
    {
      "entropy": 5.433973884582519,
      "epoch": 1.1084905660377358,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004883880467167603,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20440714359283446,
      "num_tokens": 25359186.0,
      "step": 12220
    },
    {
      "entropy": 5.452834415435791,
      "epoch": 1.10894412191582,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000488377784213264,
      "loss": 5.0833,
      "mean_token_accuracy": 0.1990295484662056,
      "num_tokens": 25368993.0,
      "step": 12225
    },
    {
      "entropy": 5.404776620864868,
      "epoch": 1.1093976777939043,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004883675172970608,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19580052495002748,
      "num_tokens": 25379880.0,
      "step": 12230
    },
    {
      "entropy": 5.507901239395141,
      "epoch": 1.1098512336719883,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000488357245968363,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1886405259370804,
      "num_tokens": 25389942.0,
      "step": 12235
    },
    {
      "entropy": 5.460845708847046,
      "epoch": 1.1103047895500726,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004883469702273829,
      "loss": 5.0248,
      "mean_token_accuracy": 0.20065219849348068,
      "num_tokens": 25400836.0,
      "step": 12240
    },
    {
      "entropy": 5.439810752868652,
      "epoch": 1.1107583454281567,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004883366900743331,
      "loss": 5.0408,
      "mean_token_accuracy": 0.20459828972816468,
      "num_tokens": 25410502.0,
      "step": 12245
    },
    {
      "entropy": 5.412273931503296,
      "epoch": 1.111211901306241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004883264055094262,
      "loss": 4.9971,
      "mean_token_accuracy": 0.2025846540927887,
      "num_tokens": 25420301.0,
      "step": 12250
    },
    {
      "entropy": 5.423372793197632,
      "epoch": 1.1116654571843252,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883161165328749,
      "loss": 5.1022,
      "mean_token_accuracy": 0.19204796105623245,
      "num_tokens": 25431107.0,
      "step": 12255
    },
    {
      "entropy": 5.422839641571045,
      "epoch": 1.1121190130624092,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488305823144892,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19618319123983383,
      "num_tokens": 25441256.0,
      "step": 12260
    },
    {
      "entropy": 5.490327835083008,
      "epoch": 1.1125725689404935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004882955253456902,
      "loss": 5.0207,
      "mean_token_accuracy": 0.20296041369438172,
      "num_tokens": 25451627.0,
      "step": 12265
    },
    {
      "entropy": 5.496054744720459,
      "epoch": 1.1130261248185778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004882852231354826,
      "loss": 5.148,
      "mean_token_accuracy": 0.18812839686870575,
      "num_tokens": 25461273.0,
      "step": 12270
    },
    {
      "entropy": 5.3958149433135985,
      "epoch": 1.1134796806966618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004882749165144822,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20185253024101257,
      "num_tokens": 25471949.0,
      "step": 12275
    },
    {
      "entropy": 5.490109395980835,
      "epoch": 1.113933236574746,
      "grad_norm": 1.0,
      "learning_rate": 0.0004882646054829023,
      "loss": 5.056,
      "mean_token_accuracy": 0.19314002394676208,
      "num_tokens": 25482175.0,
      "step": 12280
    },
    {
      "entropy": 5.491643333435059,
      "epoch": 1.1143867924528301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048825429004095597,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19887231439352035,
      "num_tokens": 25491786.0,
      "step": 12285
    },
    {
      "entropy": 5.425946044921875,
      "epoch": 1.1148403483309144,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004882439701888566,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20476123839616775,
      "num_tokens": 25502103.0,
      "step": 12290
    },
    {
      "entropy": 5.43491415977478,
      "epoch": 1.1152939042089987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048823364592681754,
      "loss": 5.0132,
      "mean_token_accuracy": 0.1989473059773445,
      "num_tokens": 25512530.0,
      "step": 12295
    },
    {
      "entropy": 5.418636131286621,
      "epoch": 1.1157474600870827,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048822331725505234,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1979011669754982,
      "num_tokens": 25522905.0,
      "step": 12300
    },
    {
      "entropy": 5.410347890853882,
      "epoch": 1.116201015965167,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048821298417377456,
      "loss": 5.0001,
      "mean_token_accuracy": 0.20536549240350724,
      "num_tokens": 25532963.0,
      "step": 12305
    },
    {
      "entropy": 5.36260495185852,
      "epoch": 1.116654571843251,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488202646683198,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19717641025781632,
      "num_tokens": 25542583.0,
      "step": 12310
    },
    {
      "entropy": 5.430088186264038,
      "epoch": 1.1171081277213353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004881923047835364,
      "loss": 5.0414,
      "mean_token_accuracy": 0.19662002921104432,
      "num_tokens": 25553367.0,
      "step": 12315
    },
    {
      "entropy": 5.577167987823486,
      "epoch": 1.1175616835994195,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004881819584750036,
      "loss": 5.1457,
      "mean_token_accuracy": 0.1906937912106514,
      "num_tokens": 25564823.0,
      "step": 12320
    },
    {
      "entropy": 5.47358980178833,
      "epoch": 1.1180152394775036,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048817160775781356,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19616676568984986,
      "num_tokens": 25574946.0,
      "step": 12325
    },
    {
      "entropy": 5.454680919647217,
      "epoch": 1.1184687953555879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048816125263218023,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19523930549621582,
      "num_tokens": 25584779.0,
      "step": 12330
    },
    {
      "entropy": 5.5475927829742435,
      "epoch": 1.118922351233672,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048815089309831803,
      "loss": 5.1617,
      "mean_token_accuracy": 0.1924641713500023,
      "num_tokens": 25594725.0,
      "step": 12335
    },
    {
      "entropy": 5.4552511215209964,
      "epoch": 1.1193759071117562,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488140529156441,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19338975101709366,
      "num_tokens": 25605499.0,
      "step": 12340
    },
    {
      "entropy": 5.427136659622192,
      "epoch": 1.1198294629898404,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004881301608067635,
      "loss": 5.0939,
      "mean_token_accuracy": 0.1955274298787117,
      "num_tokens": 25616443.0,
      "step": 12345
    },
    {
      "entropy": 5.462292909622192,
      "epoch": 1.1202830188679245,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048811978804949997,
      "loss": 5.0569,
      "mean_token_accuracy": 0.19406415075063704,
      "num_tokens": 25627762.0,
      "step": 12350
    },
    {
      "entropy": 5.545721483230591,
      "epoch": 1.1207365747460087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488109410884865,
      "loss": 5.1339,
      "mean_token_accuracy": 0.19659575074911118,
      "num_tokens": 25637904.0,
      "step": 12355
    },
    {
      "entropy": 5.44553017616272,
      "epoch": 1.1211901306240928,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048809902931307294,
      "loss": 5.0098,
      "mean_token_accuracy": 0.19970433562994003,
      "num_tokens": 25649058.0,
      "step": 12360
    },
    {
      "entropy": 5.398366451263428,
      "epoch": 1.121643686502177,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004880886433343388,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20112952291965486,
      "num_tokens": 25659161.0,
      "step": 12365
    },
    {
      "entropy": 5.4166340827941895,
      "epoch": 1.1220972423802613,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004880782529488771,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20340972095727922,
      "num_tokens": 25669336.0,
      "step": 12370
    },
    {
      "entropy": 5.492840766906738,
      "epoch": 1.1225507982583454,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004880678581569029,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1954827070236206,
      "num_tokens": 25680443.0,
      "step": 12375
    },
    {
      "entropy": 5.4376270294189455,
      "epoch": 1.1230043541364296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048805745895863105,
      "loss": 5.0297,
      "mean_token_accuracy": 0.19975681155920028,
      "num_tokens": 25690202.0,
      "step": 12380
    },
    {
      "entropy": 5.4068975925445555,
      "epoch": 1.1234579100145137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004880470553542767,
      "loss": 5.0979,
      "mean_token_accuracy": 0.19868273437023162,
      "num_tokens": 25699255.0,
      "step": 12385
    },
    {
      "entropy": 5.454531192779541,
      "epoch": 1.123911465892598,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048803664734405495,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19199754446744918,
      "num_tokens": 25710714.0,
      "step": 12390
    },
    {
      "entropy": 5.463699913024902,
      "epoch": 1.1243650217706822,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000488026234928181,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1986502930521965,
      "num_tokens": 25721722.0,
      "step": 12395
    },
    {
      "entropy": 5.453849935531617,
      "epoch": 1.1248185776487662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004880158181068703,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2029099076986313,
      "num_tokens": 25732096.0,
      "step": 12400
    },
    {
      "entropy": 5.465212678909301,
      "epoch": 1.1252721335268505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048800539688033807,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19555932879447938,
      "num_tokens": 25741499.0,
      "step": 12405
    },
    {
      "entropy": 5.475707721710205,
      "epoch": 1.1257256894049348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004879949712488001,
      "loss": 5.1854,
      "mean_token_accuracy": 0.19263612776994704,
      "num_tokens": 25751913.0,
      "step": 12410
    },
    {
      "entropy": 5.4851563453674315,
      "epoch": 1.1261792452830188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048798454121247177,
      "loss": 5.1184,
      "mean_token_accuracy": 0.1930095836520195,
      "num_tokens": 25762602.0,
      "step": 12415
    },
    {
      "entropy": 5.414087772369385,
      "epoch": 1.126632801161103,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004879741067715688,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1986461728811264,
      "num_tokens": 25773872.0,
      "step": 12420
    },
    {
      "entropy": 5.476527690887451,
      "epoch": 1.1270863570391871,
      "grad_norm": 1.0,
      "learning_rate": 0.00048796366792630715,
      "loss": 5.072,
      "mean_token_accuracy": 0.19697064459323882,
      "num_tokens": 25784723.0,
      "step": 12425
    },
    {
      "entropy": 5.449933576583862,
      "epoch": 1.1275399129172714,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004879532246769025,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19810961186885834,
      "num_tokens": 25794897.0,
      "step": 12430
    },
    {
      "entropy": 5.416313743591308,
      "epoch": 1.1279934687953557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879427770235709,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20813877135515213,
      "num_tokens": 25804445.0,
      "step": 12435
    },
    {
      "entropy": 5.491560363769532,
      "epoch": 1.1284470246734397,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048793232496652846,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1873038589954376,
      "num_tokens": 25815495.0,
      "step": 12440
    },
    {
      "entropy": 5.480292654037475,
      "epoch": 1.128900580551524,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048792186850599124,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2019079625606537,
      "num_tokens": 25826120.0,
      "step": 12445
    },
    {
      "entropy": 5.44529709815979,
      "epoch": 1.1293541364296082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004879114076421755,
      "loss": 5.0644,
      "mean_token_accuracy": 0.20093466639518737,
      "num_tokens": 25835926.0,
      "step": 12450
    },
    {
      "entropy": 5.4184167861938475,
      "epoch": 1.1298076923076923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048790094237529763,
      "loss": 5.1001,
      "mean_token_accuracy": 0.1918276235461235,
      "num_tokens": 25846095.0,
      "step": 12455
    },
    {
      "entropy": 5.508404397964478,
      "epoch": 1.1302612481857766,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487890472705574,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19151043891906738,
      "num_tokens": 25858541.0,
      "step": 12460
    },
    {
      "entropy": 5.484144592285157,
      "epoch": 1.1307148040638606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048787999863322113,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2003431960940361,
      "num_tokens": 25867664.0,
      "step": 12465
    },
    {
      "entropy": 5.443308925628662,
      "epoch": 1.1311683599419449,
      "grad_norm": 1.0,
      "learning_rate": 0.0004878695201584557,
      "loss": 5.1018,
      "mean_token_accuracy": 0.1985948473215103,
      "num_tokens": 25877300.0,
      "step": 12470
    },
    {
      "entropy": 5.424301815032959,
      "epoch": 1.1316219158200291,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004878590372814943,
      "loss": 5.0575,
      "mean_token_accuracy": 0.20260118544101716,
      "num_tokens": 25888009.0,
      "step": 12475
    },
    {
      "entropy": 5.50917501449585,
      "epoch": 1.1320754716981132,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048784855000255374,
      "loss": 5.1167,
      "mean_token_accuracy": 0.19358260482549666,
      "num_tokens": 25897285.0,
      "step": 12480
    },
    {
      "entropy": 5.4255023956298825,
      "epoch": 1.1325290275761974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048783805832185097,
      "loss": 5.009,
      "mean_token_accuracy": 0.20263294130563736,
      "num_tokens": 25907355.0,
      "step": 12485
    },
    {
      "entropy": 5.472705078125,
      "epoch": 1.1329825834542815,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878275622396029,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20127519220113754,
      "num_tokens": 25918566.0,
      "step": 12490
    },
    {
      "entropy": 5.463294839859008,
      "epoch": 1.1334361393323658,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048781706175602665,
      "loss": 5.1393,
      "mean_token_accuracy": 0.19933076500892638,
      "num_tokens": 25929145.0,
      "step": 12495
    },
    {
      "entropy": 5.4120849609375,
      "epoch": 1.13388969521045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048780655687133927,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20007957518100739,
      "num_tokens": 25938163.0,
      "step": 12500
    },
    {
      "entropy": 5.403029680252075,
      "epoch": 1.134343251088534,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048779604758575806,
      "loss": 5.0394,
      "mean_token_accuracy": 0.1949640303850174,
      "num_tokens": 25948962.0,
      "step": 12505
    },
    {
      "entropy": 5.389069318771362,
      "epoch": 1.1347968069666183,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004877855338995004,
      "loss": 5.0095,
      "mean_token_accuracy": 0.19757837504148484,
      "num_tokens": 25959653.0,
      "step": 12510
    },
    {
      "entropy": 5.512422895431518,
      "epoch": 1.1352503628447024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048777501581278367,
      "loss": 5.1007,
      "mean_token_accuracy": 0.1961492881178856,
      "num_tokens": 25970268.0,
      "step": 12515
    },
    {
      "entropy": 5.425408792495728,
      "epoch": 1.1357039187227866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004877644933258254,
      "loss": 5.0506,
      "mean_token_accuracy": 0.19497006237506867,
      "num_tokens": 25980321.0,
      "step": 12520
    },
    {
      "entropy": 5.386345720291137,
      "epoch": 1.136157474600871,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004877539664388432,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19168413579463958,
      "num_tokens": 25992466.0,
      "step": 12525
    },
    {
      "entropy": 5.461630630493164,
      "epoch": 1.136611030478955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048774343515205464,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20293886661529542,
      "num_tokens": 26003553.0,
      "step": 12530
    },
    {
      "entropy": 5.46403136253357,
      "epoch": 1.1370645863570392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004877328994656778,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20149170607328415,
      "num_tokens": 26013299.0,
      "step": 12535
    },
    {
      "entropy": 5.416072607040405,
      "epoch": 1.1375181422351233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004877223593799303,
      "loss": 5.0718,
      "mean_token_accuracy": 0.19764264523983002,
      "num_tokens": 26022913.0,
      "step": 12540
    },
    {
      "entropy": 5.3480541706085205,
      "epoch": 1.1379716981132075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004877118148950302,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20063163489103317,
      "num_tokens": 26033191.0,
      "step": 12545
    },
    {
      "entropy": 5.589880514144897,
      "epoch": 1.1384252539912918,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048770126601119555,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1850689321756363,
      "num_tokens": 26043517.0,
      "step": 12550
    },
    {
      "entropy": 5.570900964736938,
      "epoch": 1.1388788098693758,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048769071272864454,
      "loss": 5.1594,
      "mean_token_accuracy": 0.19274065196514129,
      "num_tokens": 26054173.0,
      "step": 12555
    },
    {
      "entropy": 5.5047448635101315,
      "epoch": 1.13933236574746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048768015504759536,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1928042858839035,
      "num_tokens": 26064929.0,
      "step": 12560
    },
    {
      "entropy": 5.408150243759155,
      "epoch": 1.1397859216255442,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004876695929682665,
      "loss": 5.022,
      "mean_token_accuracy": 0.20337846577167512,
      "num_tokens": 26074872.0,
      "step": 12565
    },
    {
      "entropy": 5.414556503295898,
      "epoch": 1.1402394775036284,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048765902649087605,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20306820124387742,
      "num_tokens": 26086328.0,
      "step": 12570
    },
    {
      "entropy": 5.485406827926636,
      "epoch": 1.1406930333817127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048764845561564285,
      "loss": 5.0623,
      "mean_token_accuracy": 0.19497212320566176,
      "num_tokens": 26096477.0,
      "step": 12575
    },
    {
      "entropy": 5.446309804916382,
      "epoch": 1.1411465892597967,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048763788034278545,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20379463136196135,
      "num_tokens": 26106183.0,
      "step": 12580
    },
    {
      "entropy": 5.419105434417725,
      "epoch": 1.141600145137881,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004876273006725225,
      "loss": 5.0759,
      "mean_token_accuracy": 0.19942841231822966,
      "num_tokens": 26117932.0,
      "step": 12585
    },
    {
      "entropy": 5.397345113754272,
      "epoch": 1.1420537010159653,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048761671660507266,
      "loss": 5.0232,
      "mean_token_accuracy": 0.1995558112859726,
      "num_tokens": 26128616.0,
      "step": 12590
    },
    {
      "entropy": 5.464322662353515,
      "epoch": 1.1425072568940493,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487606128140655,
      "loss": 5.0477,
      "mean_token_accuracy": 0.19695389568805693,
      "num_tokens": 26138758.0,
      "step": 12595
    },
    {
      "entropy": 5.420771551132202,
      "epoch": 1.1429608127721336,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004875955352794885,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19705618023872376,
      "num_tokens": 26149199.0,
      "step": 12600
    },
    {
      "entropy": 5.4241548538208,
      "epoch": 1.1434143686502176,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004875849380217921,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20088202506303787,
      "num_tokens": 26160276.0,
      "step": 12605
    },
    {
      "entropy": 5.4528889656066895,
      "epoch": 1.1438679245283019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000487574336367785,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2016328141093254,
      "num_tokens": 26169564.0,
      "step": 12610
    },
    {
      "entropy": 5.366958856582642,
      "epoch": 1.1443214804063861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004875637303176865,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20262449383735656,
      "num_tokens": 26179135.0,
      "step": 12615
    },
    {
      "entropy": 5.459455728530884,
      "epoch": 1.1447750362844702,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048755311987171585,
      "loss": 5.1171,
      "mean_token_accuracy": 0.19743296205997468,
      "num_tokens": 26190116.0,
      "step": 12620
    },
    {
      "entropy": 5.479267597198486,
      "epoch": 1.1452285921625545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048754250503009244,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1911403179168701,
      "num_tokens": 26200847.0,
      "step": 12625
    },
    {
      "entropy": 5.545922040939331,
      "epoch": 1.1456821480406387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048753188579303596,
      "loss": 5.1417,
      "mean_token_accuracy": 0.1953682154417038,
      "num_tokens": 26211151.0,
      "step": 12630
    },
    {
      "entropy": 5.4858533382415775,
      "epoch": 1.1461357039187228,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004875212621607659,
      "loss": 5.0533,
      "mean_token_accuracy": 0.20282673686742783,
      "num_tokens": 26221273.0,
      "step": 12635
    },
    {
      "entropy": 5.466365957260132,
      "epoch": 1.146589259796807,
      "grad_norm": 1.0,
      "learning_rate": 0.00048751063413350203,
      "loss": 5.163,
      "mean_token_accuracy": 0.194965261220932,
      "num_tokens": 26231770.0,
      "step": 12640
    },
    {
      "entropy": 5.495484018325806,
      "epoch": 1.147042815674891,
      "grad_norm": 1.09375,
      "learning_rate": 0.000487500001711464,
      "loss": 5.1197,
      "mean_token_accuracy": 0.18718278259038926,
      "num_tokens": 26241386.0,
      "step": 12645
    },
    {
      "entropy": 5.574727201461792,
      "epoch": 1.1474963715529753,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004874893648948719,
      "loss": 5.1869,
      "mean_token_accuracy": 0.19222330152988434,
      "num_tokens": 26251991.0,
      "step": 12650
    },
    {
      "entropy": 5.574859189987182,
      "epoch": 1.1479499274310596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048747872368394546,
      "loss": 5.2441,
      "mean_token_accuracy": 0.19536986351013183,
      "num_tokens": 26262331.0,
      "step": 12655
    },
    {
      "entropy": 5.503897953033447,
      "epoch": 1.1484034833091437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048746807807890496,
      "loss": 5.1156,
      "mean_token_accuracy": 0.19601597040891647,
      "num_tokens": 26272939.0,
      "step": 12660
    },
    {
      "entropy": 5.452318811416626,
      "epoch": 1.148857039187228,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048745742807997046,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20496115684509278,
      "num_tokens": 26282849.0,
      "step": 12665
    },
    {
      "entropy": 5.424884271621704,
      "epoch": 1.149310595065312,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048744677368736216,
      "loss": 5.0783,
      "mean_token_accuracy": 0.20587577670812607,
      "num_tokens": 26292449.0,
      "step": 12670
    },
    {
      "entropy": 5.362032032012939,
      "epoch": 1.1497641509433962,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048743611490130044,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19552327394485475,
      "num_tokens": 26302506.0,
      "step": 12675
    },
    {
      "entropy": 5.492298984527588,
      "epoch": 1.1502177068214805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048742545172200576,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20193991661071778,
      "num_tokens": 26312346.0,
      "step": 12680
    },
    {
      "entropy": 5.507142162322998,
      "epoch": 1.1506712626995645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004874147841496986,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2021457850933075,
      "num_tokens": 26323173.0,
      "step": 12685
    },
    {
      "entropy": 5.518513441085815,
      "epoch": 1.1511248185776488,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004874041121845996,
      "loss": 5.1545,
      "mean_token_accuracy": 0.1896779328584671,
      "num_tokens": 26334371.0,
      "step": 12690
    },
    {
      "entropy": 5.42510199546814,
      "epoch": 1.1515783744557329,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873934358269294,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2023568034172058,
      "num_tokens": 26344506.0,
      "step": 12695
    },
    {
      "entropy": 5.5277252197265625,
      "epoch": 1.1520319303338171,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873827550769088,
      "loss": 5.1029,
      "mean_token_accuracy": 0.18706048280000687,
      "num_tokens": 26354534.0,
      "step": 12700
    },
    {
      "entropy": 5.45455904006958,
      "epoch": 1.1524854862119014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873720699347587,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19750375896692277,
      "num_tokens": 26365682.0,
      "step": 12705
    },
    {
      "entropy": 5.53352952003479,
      "epoch": 1.1529390420899854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004873613804007001,
      "loss": 5.1866,
      "mean_token_accuracy": 0.1976620763540268,
      "num_tokens": 26376702.0,
      "step": 12710
    },
    {
      "entropy": 5.440504360198974,
      "epoch": 1.1533925979680697,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004873506864749541,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20106828659772874,
      "num_tokens": 26387154.0,
      "step": 12715
    },
    {
      "entropy": 5.5413073062896725,
      "epoch": 1.1538461538461537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004873399881577417,
      "loss": 5.1618,
      "mean_token_accuracy": 0.19441997408866882,
      "num_tokens": 26397883.0,
      "step": 12720
    },
    {
      "entropy": 5.380645418167115,
      "epoch": 1.154299709724238,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004873292854492842,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20258790552616118,
      "num_tokens": 26407613.0,
      "step": 12725
    },
    {
      "entropy": 5.344910144805908,
      "epoch": 1.1547532656023223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873185783498031,
      "loss": 4.994,
      "mean_token_accuracy": 0.19510757476091384,
      "num_tokens": 26417221.0,
      "step": 12730
    },
    {
      "entropy": 5.446285247802734,
      "epoch": 1.1552068214804063,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004873078668595197,
      "loss": 5.0874,
      "mean_token_accuracy": 0.1940292626619339,
      "num_tokens": 26427976.0,
      "step": 12735
    },
    {
      "entropy": 5.473560428619384,
      "epoch": 1.1556603773584906,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048729715097865534,
      "loss": 5.1114,
      "mean_token_accuracy": 0.20027161091566087,
      "num_tokens": 26439213.0,
      "step": 12740
    },
    {
      "entropy": 5.435029172897339,
      "epoch": 1.1561139332365746,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004872864307074319,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20389191210269927,
      "num_tokens": 26449763.0,
      "step": 12745
    },
    {
      "entropy": 5.520690250396728,
      "epoch": 1.156567489114659,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048727570604607096,
      "loss": 5.1146,
      "mean_token_accuracy": 0.19712771773338317,
      "num_tokens": 26460133.0,
      "step": 12750
    },
    {
      "entropy": 5.5018998146057125,
      "epoch": 1.1570210449927432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048726497699479437,
      "loss": 5.0798,
      "mean_token_accuracy": 0.19398481249809266,
      "num_tokens": 26470164.0,
      "step": 12755
    },
    {
      "entropy": 5.442802476882934,
      "epoch": 1.1574746008708272,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000487254243553824,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19554944038391114,
      "num_tokens": 26480314.0,
      "step": 12760
    },
    {
      "entropy": 5.3819260597229,
      "epoch": 1.1579281567489115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872435057233817,
      "loss": 5.038,
      "mean_token_accuracy": 0.20284716337919234,
      "num_tokens": 26491184.0,
      "step": 12765
    },
    {
      "entropy": 5.406965637207032,
      "epoch": 1.1583817126269957,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048723276350368957,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20368421971797943,
      "num_tokens": 26501432.0,
      "step": 12770
    },
    {
      "entropy": 5.4483438491821286,
      "epoch": 1.1588352685050798,
      "grad_norm": 1.0,
      "learning_rate": 0.00048722201689496984,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19702692180871964,
      "num_tokens": 26510379.0,
      "step": 12775
    },
    {
      "entropy": 5.429786205291748,
      "epoch": 1.159288824383164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004872112658974447,
      "loss": 5.0561,
      "mean_token_accuracy": 0.20342138409614563,
      "num_tokens": 26519454.0,
      "step": 12780
    },
    {
      "entropy": 5.387827634811401,
      "epoch": 1.159742380261248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048720051051133657,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20228770673274993,
      "num_tokens": 26531325.0,
      "step": 12785
    },
    {
      "entropy": 5.440437602996826,
      "epoch": 1.1601959361393324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871897507368678,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20184554904699326,
      "num_tokens": 26541510.0,
      "step": 12790
    },
    {
      "entropy": 5.493928003311157,
      "epoch": 1.1606494920174166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048717898657426093,
      "loss": 5.2491,
      "mean_token_accuracy": 0.18356596529483796,
      "num_tokens": 26553400.0,
      "step": 12795
    },
    {
      "entropy": 5.481597852706909,
      "epoch": 1.1611030478955007,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048716821802373856,
      "loss": 5.0769,
      "mean_token_accuracy": 0.19700282961130142,
      "num_tokens": 26564340.0,
      "step": 12800
    },
    {
      "entropy": 5.448103237152099,
      "epoch": 1.161556603773585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048715744508552334,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19755611568689346,
      "num_tokens": 26573615.0,
      "step": 12805
    },
    {
      "entropy": 5.4913605690002445,
      "epoch": 1.1620101596516692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004871466677598381,
      "loss": 5.1896,
      "mean_token_accuracy": 0.18805204033851625,
      "num_tokens": 26584149.0,
      "step": 12810
    },
    {
      "entropy": 5.4462597370147705,
      "epoch": 1.1624637155297532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871358860469057,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1880527302622795,
      "num_tokens": 26595064.0,
      "step": 12815
    },
    {
      "entropy": 5.4592596054077145,
      "epoch": 1.1629172714078375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048712509994694903,
      "loss": 5.0957,
      "mean_token_accuracy": 0.19753992557525635,
      "num_tokens": 26605636.0,
      "step": 12820
    },
    {
      "entropy": 5.525445079803466,
      "epoch": 1.1633708272859216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004871143094601914,
      "loss": 5.1863,
      "mean_token_accuracy": 0.19185950756072997,
      "num_tokens": 26616065.0,
      "step": 12825
    },
    {
      "entropy": 5.518282413482666,
      "epoch": 1.1638243831640058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004871035145868557,
      "loss": 5.0995,
      "mean_token_accuracy": 0.19677670150995255,
      "num_tokens": 26627246.0,
      "step": 12830
    },
    {
      "entropy": 5.459026336669922,
      "epoch": 1.16427793904209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048709271532716524,
      "loss": 5.1024,
      "mean_token_accuracy": 0.19623770713806152,
      "num_tokens": 26637630.0,
      "step": 12835
    },
    {
      "entropy": 5.516048812866211,
      "epoch": 1.1647314949201741,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004870819116813434,
      "loss": 5.1509,
      "mean_token_accuracy": 0.19567939937114714,
      "num_tokens": 26648683.0,
      "step": 12840
    },
    {
      "entropy": 5.5269402980804445,
      "epoch": 1.1651850507982584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048707110364961363,
      "loss": 5.1139,
      "mean_token_accuracy": 0.1932143375277519,
      "num_tokens": 26659573.0,
      "step": 12845
    },
    {
      "entropy": 5.423376846313476,
      "epoch": 1.1656386066763424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048706029123219937,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2039925828576088,
      "num_tokens": 26669384.0,
      "step": 12850
    },
    {
      "entropy": 5.482315349578857,
      "epoch": 1.1660921625544267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048704947442932426,
      "loss": 5.1805,
      "mean_token_accuracy": 0.18662424385547638,
      "num_tokens": 26679721.0,
      "step": 12855
    },
    {
      "entropy": 5.531332302093506,
      "epoch": 1.166545718432511,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048703865324121196,
      "loss": 5.027,
      "mean_token_accuracy": 0.20461533665657045,
      "num_tokens": 26689923.0,
      "step": 12860
    },
    {
      "entropy": 5.401912021636963,
      "epoch": 1.166999274310595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048702782766808637,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19950828552246094,
      "num_tokens": 26699953.0,
      "step": 12865
    },
    {
      "entropy": 5.479466533660888,
      "epoch": 1.1674528301886793,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004870169977101711,
      "loss": 5.2542,
      "mean_token_accuracy": 0.185440394282341,
      "num_tokens": 26712099.0,
      "step": 12870
    },
    {
      "entropy": 5.518303060531617,
      "epoch": 1.1679063860667633,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048700616336769037,
      "loss": 5.087,
      "mean_token_accuracy": 0.1994853988289833,
      "num_tokens": 26722334.0,
      "step": 12875
    },
    {
      "entropy": 5.479982852935791,
      "epoch": 1.1683599419448476,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048699532464086817,
      "loss": 5.0563,
      "mean_token_accuracy": 0.199135123193264,
      "num_tokens": 26732470.0,
      "step": 12880
    },
    {
      "entropy": 5.441677236557007,
      "epoch": 1.1688134978229319,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048698448152992865,
      "loss": 5.0901,
      "mean_token_accuracy": 0.19643069058656693,
      "num_tokens": 26742015.0,
      "step": 12885
    },
    {
      "entropy": 5.417946290969849,
      "epoch": 1.169267053701016,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048697363403509605,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19372081607580185,
      "num_tokens": 26753231.0,
      "step": 12890
    },
    {
      "entropy": 5.489230966567993,
      "epoch": 1.1697206095791002,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869627821565946,
      "loss": 5.061,
      "mean_token_accuracy": 0.19762594401836395,
      "num_tokens": 26763895.0,
      "step": 12895
    },
    {
      "entropy": 5.52664737701416,
      "epoch": 1.1701741654571842,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869519258946489,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19777561277151107,
      "num_tokens": 26774208.0,
      "step": 12900
    },
    {
      "entropy": 5.434159612655639,
      "epoch": 1.1706277213352685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048694106524948324,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19534947276115416,
      "num_tokens": 26785554.0,
      "step": 12905
    },
    {
      "entropy": 5.466801881790161,
      "epoch": 1.1710812772133528,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048693020022132243,
      "loss": 5.0105,
      "mean_token_accuracy": 0.19920032918453218,
      "num_tokens": 26795488.0,
      "step": 12910
    },
    {
      "entropy": 5.522245454788208,
      "epoch": 1.1715348330914368,
      "grad_norm": 0.984375,
      "learning_rate": 0.000486919330810391,
      "loss": 5.1736,
      "mean_token_accuracy": 0.18677914440631865,
      "num_tokens": 26806801.0,
      "step": 12915
    },
    {
      "entropy": 5.49633207321167,
      "epoch": 1.171988388969521,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004869084570169139,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20189729630947112,
      "num_tokens": 26816869.0,
      "step": 12920
    },
    {
      "entropy": 5.430732583999633,
      "epoch": 1.172441944847605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048689757884111585,
      "loss": 5.032,
      "mean_token_accuracy": 0.19632718116044998,
      "num_tokens": 26827003.0,
      "step": 12925
    },
    {
      "entropy": 5.411961412429809,
      "epoch": 1.1728955007256894,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048688669628322185,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20101807564496993,
      "num_tokens": 26837065.0,
      "step": 12930
    },
    {
      "entropy": 5.389821910858155,
      "epoch": 1.1733490566037736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000486875809343457,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20275212675333024,
      "num_tokens": 26847487.0,
      "step": 12935
    },
    {
      "entropy": 5.447001123428345,
      "epoch": 1.1738026124818577,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048686491802204645,
      "loss": 5.041,
      "mean_token_accuracy": 0.2034755453467369,
      "num_tokens": 26858407.0,
      "step": 12940
    },
    {
      "entropy": 5.481881952285766,
      "epoch": 1.174256168359942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004868540223192154,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19423448741436006,
      "num_tokens": 26868469.0,
      "step": 12945
    },
    {
      "entropy": 5.493884038925171,
      "epoch": 1.1747097242380262,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004868431222351892,
      "loss": 5.1143,
      "mean_token_accuracy": 0.19713809788227082,
      "num_tokens": 26879580.0,
      "step": 12950
    },
    {
      "entropy": 5.488229417800904,
      "epoch": 1.1751632801161103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048683221777019313,
      "loss": 5.0896,
      "mean_token_accuracy": 0.20102750808000563,
      "num_tokens": 26889513.0,
      "step": 12955
    },
    {
      "entropy": 5.466822147369385,
      "epoch": 1.1756168359941945,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048682130892445293,
      "loss": 5.1661,
      "mean_token_accuracy": 0.1931549623608589,
      "num_tokens": 26900792.0,
      "step": 12960
    },
    {
      "entropy": 5.437531661987305,
      "epoch": 1.1760703918722786,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486810395698194,
      "loss": 5.0536,
      "mean_token_accuracy": 0.203791543841362,
      "num_tokens": 26911298.0,
      "step": 12965
    },
    {
      "entropy": 5.488604164123535,
      "epoch": 1.1765239477503628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048679947809164223,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19251493662595748,
      "num_tokens": 26921545.0,
      "step": 12970
    },
    {
      "entropy": 5.484994983673095,
      "epoch": 1.176977503628447,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004867885561050232,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20793627351522445,
      "num_tokens": 26932140.0,
      "step": 12975
    },
    {
      "entropy": 5.481871938705444,
      "epoch": 1.1774310595065312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867776297385629,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19122447669506074,
      "num_tokens": 26942634.0,
      "step": 12980
    },
    {
      "entropy": 5.495228290557861,
      "epoch": 1.1778846153846154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048676669899248723,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2001894935965538,
      "num_tokens": 26952758.0,
      "step": 12985
    },
    {
      "entropy": 5.39152045249939,
      "epoch": 1.1783381712626997,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004867557638670222,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20257681906223296,
      "num_tokens": 26963439.0,
      "step": 12990
    },
    {
      "entropy": 5.446252346038818,
      "epoch": 1.1787917271407837,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004867448243623941,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1936539277434349,
      "num_tokens": 26974331.0,
      "step": 12995
    },
    {
      "entropy": 5.441408300399781,
      "epoch": 1.179245283018868,
      "grad_norm": 1.078125,
      "learning_rate": 0.000486733880478829,
      "loss": 4.9579,
      "mean_token_accuracy": 0.20778458416461945,
      "num_tokens": 26983416.0,
      "step": 13000
    },
    {
      "entropy": 5.458328771591186,
      "epoch": 1.179698838896952,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048672293221655333,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20128298848867415,
      "num_tokens": 26994176.0,
      "step": 13005
    },
    {
      "entropy": 5.438782453536987,
      "epoch": 1.1801523947750363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867119795757935,
      "loss": 5.076,
      "mean_token_accuracy": 0.1955456778407097,
      "num_tokens": 27003295.0,
      "step": 13010
    },
    {
      "entropy": 5.462780618667603,
      "epoch": 1.1806059506531206,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000486701022556776,
      "loss": 5.0368,
      "mean_token_accuracy": 0.19585613012313843,
      "num_tokens": 27014018.0,
      "step": 13015
    },
    {
      "entropy": 5.444080924987793,
      "epoch": 1.1810595065312046,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004866900611597273,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20128905773162842,
      "num_tokens": 27024706.0,
      "step": 13020
    },
    {
      "entropy": 5.487481689453125,
      "epoch": 1.1815130624092889,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048667909538487427,
      "loss": 5.0813,
      "mean_token_accuracy": 0.20304662138223648,
      "num_tokens": 27035801.0,
      "step": 13025
    },
    {
      "entropy": 5.39069128036499,
      "epoch": 1.181966618287373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004866681252324436,
      "loss": 4.952,
      "mean_token_accuracy": 0.20496505051851271,
      "num_tokens": 27046596.0,
      "step": 13030
    },
    {
      "entropy": 5.428235149383545,
      "epoch": 1.1824201741654572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048665715070266214,
      "loss": 5.0642,
      "mean_token_accuracy": 0.19454382210969925,
      "num_tokens": 27057057.0,
      "step": 13035
    },
    {
      "entropy": 5.338984727859497,
      "epoch": 1.1828737300435415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048664617179575685,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20416733175516127,
      "num_tokens": 27067947.0,
      "step": 13040
    },
    {
      "entropy": 5.481903743743897,
      "epoch": 1.1833272859216255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004866351885119548,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19917620718479156,
      "num_tokens": 27077503.0,
      "step": 13045
    },
    {
      "entropy": 5.51445779800415,
      "epoch": 1.1837808417997098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004866242008514832,
      "loss": 5.1751,
      "mean_token_accuracy": 0.19417932480573655,
      "num_tokens": 27088298.0,
      "step": 13050
    },
    {
      "entropy": 5.456486558914184,
      "epoch": 1.1842343976777938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048661320881456904,
      "loss": 5.0368,
      "mean_token_accuracy": 0.19583414793014525,
      "num_tokens": 27097841.0,
      "step": 13055
    },
    {
      "entropy": 5.3624402523040775,
      "epoch": 1.184687953555878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004866022124014399,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20166653245687485,
      "num_tokens": 27108263.0,
      "step": 13060
    },
    {
      "entropy": 5.421962404251099,
      "epoch": 1.1851415094339623,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048659121161232313,
      "loss": 5.0483,
      "mean_token_accuracy": 0.20191064178943635,
      "num_tokens": 27119339.0,
      "step": 13065
    },
    {
      "entropy": 5.501571559906006,
      "epoch": 1.1855950653120464,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004865802064474461,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19581292271614076,
      "num_tokens": 27130393.0,
      "step": 13070
    },
    {
      "entropy": 5.4905561923980715,
      "epoch": 1.1860486211901307,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048656919690703653,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19397873133420945,
      "num_tokens": 27141156.0,
      "step": 13075
    },
    {
      "entropy": 5.503406047821045,
      "epoch": 1.1865021770682147,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486558182991322,
      "loss": 5.1199,
      "mean_token_accuracy": 0.19888761788606643,
      "num_tokens": 27151630.0,
      "step": 13080
    },
    {
      "entropy": 5.421214866638183,
      "epoch": 1.186955732946299,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004865471647005304,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20177824348211287,
      "num_tokens": 27160734.0,
      "step": 13085
    },
    {
      "entropy": 5.435471391677856,
      "epoch": 1.1874092888243832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048653614203488947,
      "loss": 5.0873,
      "mean_token_accuracy": 0.19972168803215026,
      "num_tokens": 27171584.0,
      "step": 13090
    },
    {
      "entropy": 5.436698722839355,
      "epoch": 1.1878628447024673,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004865251149946273,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2030612885951996,
      "num_tokens": 27180817.0,
      "step": 13095
    },
    {
      "entropy": 5.495074653625489,
      "epoch": 1.1883164005805515,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048651408357997173,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19619399309158325,
      "num_tokens": 27190363.0,
      "step": 13100
    },
    {
      "entropy": 5.454643440246582,
      "epoch": 1.1887699564586356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048650304779115114,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20756469517946244,
      "num_tokens": 27202181.0,
      "step": 13105
    },
    {
      "entropy": 5.412134408950806,
      "epoch": 1.1892235123367199,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004864920076283935,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19545914530754088,
      "num_tokens": 27212514.0,
      "step": 13110
    },
    {
      "entropy": 5.490122127532959,
      "epoch": 1.1896770682148041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864809630919273,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20648339837789537,
      "num_tokens": 27223352.0,
      "step": 13115
    },
    {
      "entropy": 5.463384580612183,
      "epoch": 1.1901306240928882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004864699141819809,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20400721579790115,
      "num_tokens": 27232643.0,
      "step": 13120
    },
    {
      "entropy": 5.384722566604614,
      "epoch": 1.1905841799709724,
      "grad_norm": 1.0,
      "learning_rate": 0.00048645886089878276,
      "loss": 5.0515,
      "mean_token_accuracy": 0.20001365542411803,
      "num_tokens": 27243793.0,
      "step": 13125
    },
    {
      "entropy": 5.4541233539581295,
      "epoch": 1.1910377358490567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048644780324256153,
      "loss": 5.0645,
      "mean_token_accuracy": 0.1945493534207344,
      "num_tokens": 27254554.0,
      "step": 13130
    },
    {
      "entropy": 5.475072622299194,
      "epoch": 1.1914912917271407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004864367412135458,
      "loss": 5.1231,
      "mean_token_accuracy": 0.19704449623823167,
      "num_tokens": 27265842.0,
      "step": 13135
    },
    {
      "entropy": 5.480845308303833,
      "epoch": 1.191944847605225,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048642567481196446,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1948697656393051,
      "num_tokens": 27277617.0,
      "step": 13140
    },
    {
      "entropy": 5.415107583999633,
      "epoch": 1.192398403483309,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004864146040380462,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1978158101439476,
      "num_tokens": 27287450.0,
      "step": 13145
    },
    {
      "entropy": 5.43112621307373,
      "epoch": 1.1928519593613933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004864035288920202,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19773907363414764,
      "num_tokens": 27297010.0,
      "step": 13150
    },
    {
      "entropy": 5.426964807510376,
      "epoch": 1.1933055152394776,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048639244937411516,
      "loss": 5.0932,
      "mean_token_accuracy": 0.19573701322078704,
      "num_tokens": 27306874.0,
      "step": 13155
    },
    {
      "entropy": 5.491222953796386,
      "epoch": 1.1937590711175616,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048638136548456045,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1987365409731865,
      "num_tokens": 27318762.0,
      "step": 13160
    },
    {
      "entropy": 5.4339982032775875,
      "epoch": 1.194212626995646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048637027722358536,
      "loss": 4.968,
      "mean_token_accuracy": 0.214048133790493,
      "num_tokens": 27327231.0,
      "step": 13165
    },
    {
      "entropy": 5.370849227905273,
      "epoch": 1.1946661828737302,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048635918459141893,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20344783663749694,
      "num_tokens": 27338517.0,
      "step": 13170
    },
    {
      "entropy": 5.4767357349395756,
      "epoch": 1.1951197387518142,
      "grad_norm": 1.0,
      "learning_rate": 0.00048634808758829073,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19354832619428636,
      "num_tokens": 27348861.0,
      "step": 13175
    },
    {
      "entropy": 5.46518726348877,
      "epoch": 1.1955732946298985,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048633698621443024,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20526906549930574,
      "num_tokens": 27358610.0,
      "step": 13180
    },
    {
      "entropy": 5.4275483131408695,
      "epoch": 1.1960268505079825,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000486325880470067,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19882804602384568,
      "num_tokens": 27369707.0,
      "step": 13185
    },
    {
      "entropy": 5.4541737079620365,
      "epoch": 1.1964804063860668,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004863147703554307,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19907967001199722,
      "num_tokens": 27381030.0,
      "step": 13190
    },
    {
      "entropy": 5.501768779754639,
      "epoch": 1.196933962264151,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048630365587075115,
      "loss": 5.0918,
      "mean_token_accuracy": 0.1956538215279579,
      "num_tokens": 27390866.0,
      "step": 13195
    },
    {
      "entropy": 5.478093004226684,
      "epoch": 1.197387518142235,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048629253701625816,
      "loss": 5.0222,
      "mean_token_accuracy": 0.205246639251709,
      "num_tokens": 27400724.0,
      "step": 13200
    },
    {
      "entropy": 5.450027704238892,
      "epoch": 1.1978410740203194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004862814137921816,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19982354640960692,
      "num_tokens": 27411212.0,
      "step": 13205
    },
    {
      "entropy": 5.4368507862091064,
      "epoch": 1.1982946298984034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048627028619875157,
      "loss": 5.074,
      "mean_token_accuracy": 0.19660384058952332,
      "num_tokens": 27421279.0,
      "step": 13210
    },
    {
      "entropy": 5.4540740013122555,
      "epoch": 1.1987481857764877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004862591542361982,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20009150803089143,
      "num_tokens": 27432520.0,
      "step": 13215
    },
    {
      "entropy": 5.392856073379517,
      "epoch": 1.199201741654572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862480179047516,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20842310637235642,
      "num_tokens": 27442246.0,
      "step": 13220
    },
    {
      "entropy": 5.449037837982178,
      "epoch": 1.199655297532656,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004862368772046423,
      "loss": 5.045,
      "mean_token_accuracy": 0.19735703915357589,
      "num_tokens": 27452134.0,
      "step": 13225
    },
    {
      "entropy": 5.461572074890137,
      "epoch": 1.2001088534107403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048622573213610057,
      "loss": 5.1387,
      "mean_token_accuracy": 0.1919248580932617,
      "num_tokens": 27462253.0,
      "step": 13230
    },
    {
      "entropy": 5.455710411071777,
      "epoch": 1.2005624092888243,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048621458269935676,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2015654996037483,
      "num_tokens": 27473131.0,
      "step": 13235
    },
    {
      "entropy": 5.46015772819519,
      "epoch": 1.2010159651669086,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048620342889464153,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20629900246858596,
      "num_tokens": 27483683.0,
      "step": 13240
    },
    {
      "entropy": 5.359331655502319,
      "epoch": 1.2014695210449928,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048619227072218565,
      "loss": 5.043,
      "mean_token_accuracy": 0.19450653046369554,
      "num_tokens": 27493679.0,
      "step": 13245
    },
    {
      "entropy": 5.400630521774292,
      "epoch": 1.2019230769230769,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004861811081822198,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20626620203256607,
      "num_tokens": 27503897.0,
      "step": 13250
    },
    {
      "entropy": 5.507361555099488,
      "epoch": 1.2023766328011611,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004861699412749748,
      "loss": 5.1281,
      "mean_token_accuracy": 0.192038594186306,
      "num_tokens": 27514560.0,
      "step": 13255
    },
    {
      "entropy": 5.431878185272216,
      "epoch": 1.2028301886792452,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004861587700006816,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19795208871364595,
      "num_tokens": 27525594.0,
      "step": 13260
    },
    {
      "entropy": 5.369587707519531,
      "epoch": 1.2032837445573294,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048614759435957114,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20556827783584594,
      "num_tokens": 27535676.0,
      "step": 13265
    },
    {
      "entropy": 5.386460971832276,
      "epoch": 1.2037373004354137,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004861364143518747,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19516339600086213,
      "num_tokens": 27545520.0,
      "step": 13270
    },
    {
      "entropy": 5.525118112564087,
      "epoch": 1.2041908563134978,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004861252299778234,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19029136896133422,
      "num_tokens": 27556866.0,
      "step": 13275
    },
    {
      "entropy": 5.426550960540771,
      "epoch": 1.204644412191582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048611404123764855,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20249955207109452,
      "num_tokens": 27566951.0,
      "step": 13280
    },
    {
      "entropy": 5.428686857223511,
      "epoch": 1.205097968069666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048610284813158154,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19744616746902466,
      "num_tokens": 27576796.0,
      "step": 13285
    },
    {
      "entropy": 5.387802028656006,
      "epoch": 1.2055515239477503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004860916506598538,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20070512294769288,
      "num_tokens": 27586902.0,
      "step": 13290
    },
    {
      "entropy": 5.557977724075317,
      "epoch": 1.2060050798258346,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048608044882269695,
      "loss": 5.1751,
      "mean_token_accuracy": 0.1885147735476494,
      "num_tokens": 27599232.0,
      "step": 13295
    },
    {
      "entropy": 5.423964977264404,
      "epoch": 1.2064586357039186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004860692426203427,
      "loss": 4.9695,
      "mean_token_accuracy": 0.19955171942710875,
      "num_tokens": 27609022.0,
      "step": 13300
    },
    {
      "entropy": 5.461498355865478,
      "epoch": 1.206912191582003,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004860580320530226,
      "loss": 5.1252,
      "mean_token_accuracy": 0.2006087601184845,
      "num_tokens": 27619829.0,
      "step": 13305
    },
    {
      "entropy": 5.458715867996216,
      "epoch": 1.2073657474600872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004860468171209687,
      "loss": 5.1272,
      "mean_token_accuracy": 0.205771704018116,
      "num_tokens": 27629337.0,
      "step": 13310
    },
    {
      "entropy": 5.517880821228028,
      "epoch": 1.2078193033381712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004860355978244128,
      "loss": 5.1399,
      "mean_token_accuracy": 0.19886357337236404,
      "num_tokens": 27639377.0,
      "step": 13315
    },
    {
      "entropy": 5.468346691131591,
      "epoch": 1.2082728592162555,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000486024374163587,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19874342381954194,
      "num_tokens": 27649989.0,
      "step": 13320
    },
    {
      "entropy": 5.4235265254974365,
      "epoch": 1.2087264150943395,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048601314613872334,
      "loss": 5.1085,
      "mean_token_accuracy": 0.18839994966983795,
      "num_tokens": 27659967.0,
      "step": 13325
    },
    {
      "entropy": 5.462990427017212,
      "epoch": 1.2091799709724238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486001913750054,
      "loss": 5.0514,
      "mean_token_accuracy": 0.2000204548239708,
      "num_tokens": 27670099.0,
      "step": 13330
    },
    {
      "entropy": 5.4266736030578615,
      "epoch": 1.209633526850508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859906769978113,
      "loss": 5.1068,
      "mean_token_accuracy": 0.190760999917984,
      "num_tokens": 27681497.0,
      "step": 13335
    },
    {
      "entropy": 5.539730072021484,
      "epoch": 1.2100870827285921,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048597943588222776,
      "loss": 5.1977,
      "mean_token_accuracy": 0.1931617110967636,
      "num_tokens": 27691828.0,
      "step": 13340
    },
    {
      "entropy": 5.5186830997467045,
      "epoch": 1.2105406386066764,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048596819040353564,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2051544338464737,
      "num_tokens": 27701426.0,
      "step": 13345
    },
    {
      "entropy": 5.463600921630859,
      "epoch": 1.2109941944847606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859569405619676,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20318714678287506,
      "num_tokens": 27711112.0,
      "step": 13350
    },
    {
      "entropy": 5.464934396743774,
      "epoch": 1.2114477503628447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004859456863577562,
      "loss": 5.126,
      "mean_token_accuracy": 0.19313183426856995,
      "num_tokens": 27723160.0,
      "step": 13355
    },
    {
      "entropy": 5.474137687683106,
      "epoch": 1.211901306240929,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048593442779113437,
      "loss": 5.0905,
      "mean_token_accuracy": 0.19417185932397843,
      "num_tokens": 27733143.0,
      "step": 13360
    },
    {
      "entropy": 5.44663667678833,
      "epoch": 1.212354862119013,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004859231648623348,
      "loss": 5.0624,
      "mean_token_accuracy": 0.19898829460144044,
      "num_tokens": 27743956.0,
      "step": 13365
    },
    {
      "entropy": 5.47234354019165,
      "epoch": 1.2128084179970973,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048591189757159035,
      "loss": 5.0764,
      "mean_token_accuracy": 0.2030792161822319,
      "num_tokens": 27754192.0,
      "step": 13370
    },
    {
      "entropy": 5.442453050613404,
      "epoch": 1.2132619738751815,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004859006259191342,
      "loss": 5.1129,
      "mean_token_accuracy": 0.18974043130874635,
      "num_tokens": 27763939.0,
      "step": 13375
    },
    {
      "entropy": 5.45769248008728,
      "epoch": 1.2137155297532656,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004858893499051993,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20038015097379686,
      "num_tokens": 27774743.0,
      "step": 13380
    },
    {
      "entropy": 5.4658379554748535,
      "epoch": 1.2141690856313498,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004858780695300189,
      "loss": 5.0665,
      "mean_token_accuracy": 0.19677289873361586,
      "num_tokens": 27785344.0,
      "step": 13385
    },
    {
      "entropy": 5.48243556022644,
      "epoch": 1.2146226415094339,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004858667847938263,
      "loss": 5.1281,
      "mean_token_accuracy": 0.1950848549604416,
      "num_tokens": 27797101.0,
      "step": 13390
    },
    {
      "entropy": 5.41499433517456,
      "epoch": 1.2150761973875182,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048585549569685486,
      "loss": 5.0171,
      "mean_token_accuracy": 0.2033864140510559,
      "num_tokens": 27807581.0,
      "step": 13395
    },
    {
      "entropy": 5.413809013366699,
      "epoch": 1.2155297532656024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048584420223933805,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20077995657920839,
      "num_tokens": 27818170.0,
      "step": 13400
    },
    {
      "entropy": 5.400589656829834,
      "epoch": 1.2159833091436865,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004858329044215094,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19965792447328568,
      "num_tokens": 27828153.0,
      "step": 13405
    },
    {
      "entropy": 5.457147598266602,
      "epoch": 1.2164368650217707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048582160224360253,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20848435014486313,
      "num_tokens": 27837681.0,
      "step": 13410
    },
    {
      "entropy": 5.440137815475464,
      "epoch": 1.2168904208998548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858102957058513,
      "loss": 5.0955,
      "mean_token_accuracy": 0.2026829332113266,
      "num_tokens": 27847447.0,
      "step": 13415
    },
    {
      "entropy": 5.371809101104736,
      "epoch": 1.217343976777939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048579898480848933,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2035052314400673,
      "num_tokens": 27858586.0,
      "step": 13420
    },
    {
      "entropy": 5.458023548126221,
      "epoch": 1.2177975326560233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004857876695517507,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19574685841798783,
      "num_tokens": 27868832.0,
      "step": 13425
    },
    {
      "entropy": 5.450125980377197,
      "epoch": 1.2182510885341074,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048577634993586923,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1965117871761322,
      "num_tokens": 27879339.0,
      "step": 13430
    },
    {
      "entropy": 5.49401478767395,
      "epoch": 1.2187046444121916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048576502596107924,
      "loss": 5.0731,
      "mean_token_accuracy": 0.19752632081508636,
      "num_tokens": 27889964.0,
      "step": 13435
    },
    {
      "entropy": 5.494561100006104,
      "epoch": 1.2191582002902757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857536976276148,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19315129667520523,
      "num_tokens": 27901004.0,
      "step": 13440
    },
    {
      "entropy": 5.422154235839844,
      "epoch": 1.21961175616836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857423649357102,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19650053530931472,
      "num_tokens": 27910788.0,
      "step": 13445
    },
    {
      "entropy": 5.49261589050293,
      "epoch": 1.2200653120464442,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048573102788559974,
      "loss": 5.082,
      "mean_token_accuracy": 0.20031410306692124,
      "num_tokens": 27921360.0,
      "step": 13450
    },
    {
      "entropy": 5.472883510589599,
      "epoch": 1.2205188679245282,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857196864775179,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20247834622859956,
      "num_tokens": 27931889.0,
      "step": 13455
    },
    {
      "entropy": 5.419800710678101,
      "epoch": 1.2209724238026125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048570834071169926,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20334334075450897,
      "num_tokens": 27942049.0,
      "step": 13460
    },
    {
      "entropy": 5.491580820083618,
      "epoch": 1.2214259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048569699058837856,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20272215604782104,
      "num_tokens": 27952416.0,
      "step": 13465
    },
    {
      "entropy": 5.4474114894866945,
      "epoch": 1.2218795355587808,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856856361077903,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20347577780485154,
      "num_tokens": 27962009.0,
      "step": 13470
    },
    {
      "entropy": 5.501205205917358,
      "epoch": 1.222333091436865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004856742772701694,
      "loss": 5.0844,
      "mean_token_accuracy": 0.19673027098178864,
      "num_tokens": 27972175.0,
      "step": 13475
    },
    {
      "entropy": 5.474965524673462,
      "epoch": 1.2227866473149491,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048566291407575073,
      "loss": 5.1027,
      "mean_token_accuracy": 0.2038879156112671,
      "num_tokens": 27982899.0,
      "step": 13480
    },
    {
      "entropy": 5.474923753738404,
      "epoch": 1.2232402031930334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048565154652476937,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20354434847831726,
      "num_tokens": 27993231.0,
      "step": 13485
    },
    {
      "entropy": 5.419674634933472,
      "epoch": 1.2236937590711174,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004856401746174603,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20534394532442093,
      "num_tokens": 28003623.0,
      "step": 13490
    },
    {
      "entropy": 5.452385377883911,
      "epoch": 1.2241473149492017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048562879835405874,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20308402925729752,
      "num_tokens": 28013878.0,
      "step": 13495
    },
    {
      "entropy": 5.40763578414917,
      "epoch": 1.224600870827286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004856174177348,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20048608928918837,
      "num_tokens": 28024224.0,
      "step": 13500
    },
    {
      "entropy": 5.432199096679687,
      "epoch": 1.22505442670537,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004856060327599194,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2000358670949936,
      "num_tokens": 28034819.0,
      "step": 13505
    },
    {
      "entropy": 5.412619400024414,
      "epoch": 1.2255079825834543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048559464342965233,
      "loss": 5.0721,
      "mean_token_accuracy": 0.19996053576469422,
      "num_tokens": 28044240.0,
      "step": 13510
    },
    {
      "entropy": 5.546529722213745,
      "epoch": 1.2259615384615385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004855832497442344,
      "loss": 5.2026,
      "mean_token_accuracy": 0.1869696170091629,
      "num_tokens": 28055046.0,
      "step": 13515
    },
    {
      "entropy": 5.427810525894165,
      "epoch": 1.2264150943396226,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004855718517039012,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20430445671081543,
      "num_tokens": 28065272.0,
      "step": 13520
    },
    {
      "entropy": 5.476324367523193,
      "epoch": 1.2268686502177069,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048556044930888844,
      "loss": 4.9943,
      "mean_token_accuracy": 0.20883602052927017,
      "num_tokens": 28076255.0,
      "step": 13525
    },
    {
      "entropy": 5.46715235710144,
      "epoch": 1.2273222060957911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004855490425594319,
      "loss": 5.1216,
      "mean_token_accuracy": 0.1979125589132309,
      "num_tokens": 28086258.0,
      "step": 13530
    },
    {
      "entropy": 5.460022735595703,
      "epoch": 1.2277757619738752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004855376314557675,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20612121671438216,
      "num_tokens": 28096497.0,
      "step": 13535
    },
    {
      "entropy": 5.485916566848755,
      "epoch": 1.2282293178519594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004855262159981313,
      "loss": 5.0505,
      "mean_token_accuracy": 0.20545612424612045,
      "num_tokens": 28106282.0,
      "step": 13540
    },
    {
      "entropy": 5.418528652191162,
      "epoch": 1.2286828737300435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048551479618675925,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19765927642583847,
      "num_tokens": 28115771.0,
      "step": 13545
    },
    {
      "entropy": 5.4977963924407955,
      "epoch": 1.2291364296081277,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048550337202188756,
      "loss": 5.02,
      "mean_token_accuracy": 0.20294783860445023,
      "num_tokens": 28125477.0,
      "step": 13550
    },
    {
      "entropy": 5.503011703491211,
      "epoch": 1.229589985486212,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004854919435037525,
      "loss": 5.064,
      "mean_token_accuracy": 0.19926833361387253,
      "num_tokens": 28135800.0,
      "step": 13555
    },
    {
      "entropy": 5.374794340133667,
      "epoch": 1.230043541364296,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004854805106325904,
      "loss": 4.9404,
      "mean_token_accuracy": 0.20187063962221147,
      "num_tokens": 28146779.0,
      "step": 13560
    },
    {
      "entropy": 5.445360422134399,
      "epoch": 1.2304970972423803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004854690734086378,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1955277666449547,
      "num_tokens": 28156819.0,
      "step": 13565
    },
    {
      "entropy": 5.462018251419067,
      "epoch": 1.2309506531204644,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004854576318321311,
      "loss": 5.0786,
      "mean_token_accuracy": 0.19740512669086457,
      "num_tokens": 28168553.0,
      "step": 13570
    },
    {
      "entropy": 5.506776571273804,
      "epoch": 1.2314042089985486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048544618590330685,
      "loss": 5.041,
      "mean_token_accuracy": 0.19851386845111846,
      "num_tokens": 28178064.0,
      "step": 13575
    },
    {
      "entropy": 5.536520624160767,
      "epoch": 1.231857764876633,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048543473562240194,
      "loss": 5.1733,
      "mean_token_accuracy": 0.19981323778629304,
      "num_tokens": 28188580.0,
      "step": 13580
    },
    {
      "entropy": 5.484105062484741,
      "epoch": 1.232311320754717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000485423280989653,
      "loss": 5.143,
      "mean_token_accuracy": 0.19147918820381166,
      "num_tokens": 28199955.0,
      "step": 13585
    },
    {
      "entropy": 5.497591495513916,
      "epoch": 1.2327648766328012,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485411822005297,
      "loss": 5.1,
      "mean_token_accuracy": 0.19627357721328736,
      "num_tokens": 28210506.0,
      "step": 13590
    },
    {
      "entropy": 5.454776954650879,
      "epoch": 1.2332184325108853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048540035866957097,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20769426971673965,
      "num_tokens": 28220486.0,
      "step": 13595
    },
    {
      "entropy": 5.52288236618042,
      "epoch": 1.2336719883889695,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004853888909827118,
      "loss": 5.1499,
      "mean_token_accuracy": 0.19164299666881562,
      "num_tokens": 28232668.0,
      "step": 13600
    },
    {
      "entropy": 5.452500867843628,
      "epoch": 1.2341255442670538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004853774189449568,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1957123190164566,
      "num_tokens": 28243733.0,
      "step": 13605
    },
    {
      "entropy": 5.523473739624023,
      "epoch": 1.2345791001451378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048536594255654315,
      "loss": 5.1259,
      "mean_token_accuracy": 0.19753974080085754,
      "num_tokens": 28254953.0,
      "step": 13610
    },
    {
      "entropy": 5.504914999008179,
      "epoch": 1.235032656023222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004853544618177083,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2029399037361145,
      "num_tokens": 28264959.0,
      "step": 13615
    },
    {
      "entropy": 5.5076741695404055,
      "epoch": 1.2354862119013061,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048534297672868945,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20373029857873917,
      "num_tokens": 28275509.0,
      "step": 13620
    },
    {
      "entropy": 5.466273307800293,
      "epoch": 1.2359397677793904,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048533148728972434,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20362340658903122,
      "num_tokens": 28285983.0,
      "step": 13625
    },
    {
      "entropy": 5.508877849578857,
      "epoch": 1.2363933236574747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048531999350105037,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19323109984397888,
      "num_tokens": 28296178.0,
      "step": 13630
    },
    {
      "entropy": 5.4554966449737545,
      "epoch": 1.2368468795355587,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048530849536290537,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19579949975013733,
      "num_tokens": 28305741.0,
      "step": 13635
    },
    {
      "entropy": 5.405771970748901,
      "epoch": 1.237300435413643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004852969928755271,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2004196122288704,
      "num_tokens": 28315607.0,
      "step": 13640
    },
    {
      "entropy": 5.41153917312622,
      "epoch": 1.237753991291727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048528548603915344,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20086603313684465,
      "num_tokens": 28327349.0,
      "step": 13645
    },
    {
      "entropy": 5.455974292755127,
      "epoch": 1.2382075471698113,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048527397485402236,
      "loss": 5.0228,
      "mean_token_accuracy": 0.19990852177143098,
      "num_tokens": 28339949.0,
      "step": 13650
    },
    {
      "entropy": 5.482010841369629,
      "epoch": 1.2386611030478956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048526245932037186,
      "loss": 5.068,
      "mean_token_accuracy": 0.202985480427742,
      "num_tokens": 28351340.0,
      "step": 13655
    },
    {
      "entropy": 5.355767774581909,
      "epoch": 1.2391146589259796,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048525093943844013,
      "loss": 4.919,
      "mean_token_accuracy": 0.209889517724514,
      "num_tokens": 28360929.0,
      "step": 13660
    },
    {
      "entropy": 5.409904336929321,
      "epoch": 1.2395682148040639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004852394152084654,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20112548023462296,
      "num_tokens": 28370477.0,
      "step": 13665
    },
    {
      "entropy": 5.431086206436158,
      "epoch": 1.240021770682148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000485227886630686,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21357503831386565,
      "num_tokens": 28380876.0,
      "step": 13670
    },
    {
      "entropy": 5.358803033828735,
      "epoch": 1.2404753265602322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004852163537053403,
      "loss": 4.9369,
      "mean_token_accuracy": 0.19906609058380126,
      "num_tokens": 28390685.0,
      "step": 13675
    },
    {
      "entropy": 5.3954808712005615,
      "epoch": 1.2409288824383164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004852048164326669,
      "loss": 5.0412,
      "mean_token_accuracy": 0.1940809279680252,
      "num_tokens": 28400236.0,
      "step": 13680
    },
    {
      "entropy": 5.483770275115967,
      "epoch": 1.2413824383164005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004851932748129043,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20474082380533218,
      "num_tokens": 28410353.0,
      "step": 13685
    },
    {
      "entropy": 5.49216742515564,
      "epoch": 1.2418359941944848,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004851817288462912,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20008640140295028,
      "num_tokens": 28420241.0,
      "step": 13690
    },
    {
      "entropy": 5.438311290740967,
      "epoch": 1.242289550072569,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004851701785330663,
      "loss": 5.0811,
      "mean_token_accuracy": 0.19697158932685851,
      "num_tokens": 28430610.0,
      "step": 13695
    },
    {
      "entropy": 5.428465366363525,
      "epoch": 1.242743105950653,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004851586238734687,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20403676927089692,
      "num_tokens": 28440026.0,
      "step": 13700
    },
    {
      "entropy": 5.454707479476928,
      "epoch": 1.2431966618287373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004851470648677371,
      "loss": 5.0842,
      "mean_token_accuracy": 0.18985789269208908,
      "num_tokens": 28450707.0,
      "step": 13705
    },
    {
      "entropy": 5.456290054321289,
      "epoch": 1.2436502177068216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851355015161107,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20259017944335939,
      "num_tokens": 28462378.0,
      "step": 13710
    },
    {
      "entropy": 5.535874938964843,
      "epoch": 1.2441037735849056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004851239338188286,
      "loss": 5.1117,
      "mean_token_accuracy": 0.1960500493645668,
      "num_tokens": 28473084.0,
      "step": 13715
    },
    {
      "entropy": 5.461009693145752,
      "epoch": 1.24455732946299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048511236177612994,
      "loss": 5.0429,
      "mean_token_accuracy": 0.19701926559209823,
      "num_tokens": 28482767.0,
      "step": 13720
    },
    {
      "entropy": 5.410448694229126,
      "epoch": 1.245010885341074,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048510078538825413,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20349868386983871,
      "num_tokens": 28493274.0,
      "step": 13725
    },
    {
      "entropy": 5.493898391723633,
      "epoch": 1.2454644412191582,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048508920465544056,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20370924472808838,
      "num_tokens": 28503823.0,
      "step": 13730
    },
    {
      "entropy": 5.419942474365234,
      "epoch": 1.2459179970972425,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850776195779286,
      "loss": 5.1138,
      "mean_token_accuracy": 0.1970915362238884,
      "num_tokens": 28514209.0,
      "step": 13735
    },
    {
      "entropy": 5.500412511825561,
      "epoch": 1.2463715529753265,
      "grad_norm": 1.0,
      "learning_rate": 0.000485066030155958,
      "loss": 5.0868,
      "mean_token_accuracy": 0.19385533630847931,
      "num_tokens": 28524665.0,
      "step": 13740
    },
    {
      "entropy": 5.434648895263672,
      "epoch": 1.2468251088534108,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004850544363897684,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2031462699174881,
      "num_tokens": 28533639.0,
      "step": 13745
    },
    {
      "entropy": 5.353348159790039,
      "epoch": 1.2472786647314948,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048504283827959947,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20723264366388322,
      "num_tokens": 28544469.0,
      "step": 13750
    },
    {
      "entropy": 5.4903758525848385,
      "epoch": 1.2477322206095791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850312358256912,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20680339336395265,
      "num_tokens": 28554637.0,
      "step": 13755
    },
    {
      "entropy": 5.423876667022705,
      "epoch": 1.2481857764876634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004850196290282834,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20637150704860688,
      "num_tokens": 28564752.0,
      "step": 13760
    },
    {
      "entropy": 5.469447374343872,
      "epoch": 1.2486393323657474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048500801788761613,
      "loss": 5.1582,
      "mean_token_accuracy": 0.19746088087558747,
      "num_tokens": 28575007.0,
      "step": 13765
    },
    {
      "entropy": 5.523878669738769,
      "epoch": 1.2490928882438317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004849964024039296,
      "loss": 5.0473,
      "mean_token_accuracy": 0.19712385386228562,
      "num_tokens": 28584907.0,
      "step": 13770
    },
    {
      "entropy": 5.538279438018799,
      "epoch": 1.2495464441219157,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048498478257746387,
      "loss": 5.1184,
      "mean_token_accuracy": 0.2048226147890091,
      "num_tokens": 28594244.0,
      "step": 13775
    },
    {
      "entropy": 5.463101148605347,
      "epoch": 1.25,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048497315840845934,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20071172267198562,
      "num_tokens": 28603816.0,
      "step": 13780
    },
    {
      "entropy": 5.521854114532471,
      "epoch": 1.2504535558780843,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004849615298971565,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19445410519838333,
      "num_tokens": 28613067.0,
      "step": 13785
    },
    {
      "entropy": 5.486951446533203,
      "epoch": 1.2509071117561683,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048494989704379556,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20105471462011337,
      "num_tokens": 28623752.0,
      "step": 13790
    },
    {
      "entropy": 5.413280344009399,
      "epoch": 1.2513606676342526,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004849382598486173,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20841633677482604,
      "num_tokens": 28633190.0,
      "step": 13795
    },
    {
      "entropy": 5.55495924949646,
      "epoch": 1.2518142235123366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048492661831186243,
      "loss": 5.1081,
      "mean_token_accuracy": 0.19206711351871492,
      "num_tokens": 28644567.0,
      "step": 13800
    },
    {
      "entropy": 5.522542428970337,
      "epoch": 1.2522677793904209,
      "grad_norm": 1.0,
      "learning_rate": 0.0004849149724337715,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1908365935087204,
      "num_tokens": 28654785.0,
      "step": 13805
    },
    {
      "entropy": 5.393986463546753,
      "epoch": 1.2527213352685052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048490332221458543,
      "loss": 4.9982,
      "mean_token_accuracy": 0.19900225549936296,
      "num_tokens": 28665286.0,
      "step": 13810
    },
    {
      "entropy": 5.48044376373291,
      "epoch": 1.2531748911465892,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848916676545452,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20103529691696168,
      "num_tokens": 28675830.0,
      "step": 13815
    },
    {
      "entropy": 5.53523325920105,
      "epoch": 1.2536284470246735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048488000875389174,
      "loss": 5.1604,
      "mean_token_accuracy": 0.1919257715344429,
      "num_tokens": 28686640.0,
      "step": 13820
    },
    {
      "entropy": 5.473912048339844,
      "epoch": 1.2540820029027575,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848683455128663,
      "loss": 5.0548,
      "mean_token_accuracy": 0.19861929714679719,
      "num_tokens": 28697677.0,
      "step": 13825
    },
    {
      "entropy": 5.460227012634277,
      "epoch": 1.2545355587808418,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048485667793170994,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19863663166761397,
      "num_tokens": 28708831.0,
      "step": 13830
    },
    {
      "entropy": 5.488374185562134,
      "epoch": 1.254989114658926,
      "grad_norm": 1.0,
      "learning_rate": 0.000484845006010664,
      "loss": 5.1024,
      "mean_token_accuracy": 0.19582780748605727,
      "num_tokens": 28718959.0,
      "step": 13835
    },
    {
      "entropy": 5.497612953186035,
      "epoch": 1.25544267053701,
      "grad_norm": 1.0,
      "learning_rate": 0.0004848333297499699,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2033858522772789,
      "num_tokens": 28728719.0,
      "step": 13840
    },
    {
      "entropy": 5.57854151725769,
      "epoch": 1.2558962264150944,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048482164914986903,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19002826511859894,
      "num_tokens": 28740032.0,
      "step": 13845
    },
    {
      "entropy": 5.421471977233887,
      "epoch": 1.2563497822931784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048480996421060307,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20632674247026445,
      "num_tokens": 28749374.0,
      "step": 13850
    },
    {
      "entropy": 5.39776201248169,
      "epoch": 1.2568033381712627,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048479827493241347,
      "loss": 5.0437,
      "mean_token_accuracy": 0.19689469933509826,
      "num_tokens": 28760038.0,
      "step": 13855
    },
    {
      "entropy": 5.452365922927856,
      "epoch": 1.257256894049347,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048478658131554216,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19886355251073837,
      "num_tokens": 28770247.0,
      "step": 13860
    },
    {
      "entropy": 5.440692329406739,
      "epoch": 1.2577104499274312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847748833602308,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20380997955799102,
      "num_tokens": 28780251.0,
      "step": 13865
    },
    {
      "entropy": 5.389808464050293,
      "epoch": 1.2581640058055152,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004847631810667215,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20636868774890899,
      "num_tokens": 28790829.0,
      "step": 13870
    },
    {
      "entropy": 5.435606670379639,
      "epoch": 1.2586175616835993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847514744352561,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20586081147193908,
      "num_tokens": 28801036.0,
      "step": 13875
    },
    {
      "entropy": 5.370066547393799,
      "epoch": 1.2590711175616836,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004847397634660768,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2021748438477516,
      "num_tokens": 28811013.0,
      "step": 13880
    },
    {
      "entropy": 5.3982508182525635,
      "epoch": 1.2595246734397678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048472804815942575,
      "loss": 5.002,
      "mean_token_accuracy": 0.19941286295652388,
      "num_tokens": 28821367.0,
      "step": 13885
    },
    {
      "entropy": 5.469687032699585,
      "epoch": 1.259978229317852,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048471632851554517,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20566950142383575,
      "num_tokens": 28831077.0,
      "step": 13890
    },
    {
      "entropy": 5.4493091106414795,
      "epoch": 1.2604317851959361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847046045346775,
      "loss": 5.0759,
      "mean_token_accuracy": 0.19188063442707062,
      "num_tokens": 28841490.0,
      "step": 13895
    },
    {
      "entropy": 5.484811115264892,
      "epoch": 1.2608853410740204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048469287621706516,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1937114790081978,
      "num_tokens": 28852510.0,
      "step": 13900
    },
    {
      "entropy": 5.460968494415283,
      "epoch": 1.2613388969521044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004846811435629507,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20472006350755692,
      "num_tokens": 28863388.0,
      "step": 13905
    },
    {
      "entropy": 5.3632509231567385,
      "epoch": 1.2617924528301887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004846694065725768,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2048925742506981,
      "num_tokens": 28873798.0,
      "step": 13910
    },
    {
      "entropy": 5.428742408752441,
      "epoch": 1.262246008708273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004846576652461861,
      "loss": 5.017,
      "mean_token_accuracy": 0.19878183901309968,
      "num_tokens": 28883482.0,
      "step": 13915
    },
    {
      "entropy": 5.449710321426392,
      "epoch": 1.262699564586357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048464591958402147,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20800818800926207,
      "num_tokens": 28893573.0,
      "step": 13920
    },
    {
      "entropy": 5.516650629043579,
      "epoch": 1.2631531204644413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048463416958632573,
      "loss": 5.1539,
      "mean_token_accuracy": 0.19524807929992677,
      "num_tokens": 28904064.0,
      "step": 13925
    },
    {
      "entropy": 5.480606603622436,
      "epoch": 1.2636066763425253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048462241525334206,
      "loss": 5.0882,
      "mean_token_accuracy": 0.19799073934555053,
      "num_tokens": 28915329.0,
      "step": 13930
    },
    {
      "entropy": 5.4654511451721195,
      "epoch": 1.2640602322206096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846106565853133,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19211496114730836,
      "num_tokens": 28925228.0,
      "step": 13935
    },
    {
      "entropy": 5.443673038482666,
      "epoch": 1.2645137880986939,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048459889358248284,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20181661546230317,
      "num_tokens": 28935891.0,
      "step": 13940
    },
    {
      "entropy": 5.448412036895752,
      "epoch": 1.264967343976778,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048458712624509383,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2014430433511734,
      "num_tokens": 28946372.0,
      "step": 13945
    },
    {
      "entropy": 5.474843120574951,
      "epoch": 1.2654208998548622,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048457535457338964,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19741925448179246,
      "num_tokens": 28957874.0,
      "step": 13950
    },
    {
      "entropy": 5.495023775100708,
      "epoch": 1.2658744557329462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004845635785676137,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1934681475162506,
      "num_tokens": 28968023.0,
      "step": 13955
    },
    {
      "entropy": 5.414947462081909,
      "epoch": 1.2663280116110305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048455179822800957,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20423877686262132,
      "num_tokens": 28977744.0,
      "step": 13960
    },
    {
      "entropy": 5.521876525878906,
      "epoch": 1.2667815674891147,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845400135548208,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19199081808328627,
      "num_tokens": 28988875.0,
      "step": 13965
    },
    {
      "entropy": 5.399349594116211,
      "epoch": 1.2672351233671988,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004845282245482912,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20364806205034255,
      "num_tokens": 28998562.0,
      "step": 13970
    },
    {
      "entropy": 5.461163949966431,
      "epoch": 1.267688679245283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004845164312086646,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2067327618598938,
      "num_tokens": 29008783.0,
      "step": 13975
    },
    {
      "entropy": 5.3779669284820555,
      "epoch": 1.268142235123367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845046335361847,
      "loss": 4.9629,
      "mean_token_accuracy": 0.19951594322919847,
      "num_tokens": 29019303.0,
      "step": 13980
    },
    {
      "entropy": 5.491613435745239,
      "epoch": 1.2685957910014514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048449283153109566,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1931134954094887,
      "num_tokens": 29029843.0,
      "step": 13985
    },
    {
      "entropy": 5.516561651229859,
      "epoch": 1.2690493468795356,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004844810251936413,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19759776145219804,
      "num_tokens": 29040603.0,
      "step": 13990
    },
    {
      "entropy": 5.44688777923584,
      "epoch": 1.2695029027576197,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004844692145240661,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2032545730471611,
      "num_tokens": 29050599.0,
      "step": 13995
    },
    {
      "entropy": 5.414919519424439,
      "epoch": 1.269956458635704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844573995226142,
      "loss": 5.0764,
      "mean_token_accuracy": 0.19551022350788116,
      "num_tokens": 29061292.0,
      "step": 14000
    },
    {
      "entropy": 5.406631278991699,
      "epoch": 1.270410014513788,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048444558018952984,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2096165671944618,
      "num_tokens": 29071779.0,
      "step": 14005
    },
    {
      "entropy": 5.416690397262573,
      "epoch": 1.2708635703918723,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048443375652505754,
      "loss": 4.9562,
      "mean_token_accuracy": 0.20473774522542953,
      "num_tokens": 29082073.0,
      "step": 14010
    },
    {
      "entropy": 5.43922872543335,
      "epoch": 1.2713171262699565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844219285294418,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20453754663467408,
      "num_tokens": 29092032.0,
      "step": 14015
    },
    {
      "entropy": 5.404605340957642,
      "epoch": 1.2717706821480406,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048441009620292714,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2056380331516266,
      "num_tokens": 29102936.0,
      "step": 14020
    },
    {
      "entropy": 5.442543697357178,
      "epoch": 1.2722242380261248,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004843982595457584,
      "loss": 5.0258,
      "mean_token_accuracy": 0.203692264854908,
      "num_tokens": 29113385.0,
      "step": 14025
    },
    {
      "entropy": 5.4288288116455075,
      "epoch": 1.2726777939042089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048438641855818026,
      "loss": 5.076,
      "mean_token_accuracy": 0.1968202218413353,
      "num_tokens": 29125075.0,
      "step": 14030
    },
    {
      "entropy": 5.4668896198272705,
      "epoch": 1.2731313497822931,
      "grad_norm": 1.0,
      "learning_rate": 0.0004843745732404376,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20056262612342834,
      "num_tokens": 29135098.0,
      "step": 14035
    },
    {
      "entropy": 5.551703929901123,
      "epoch": 1.2735849056603774,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843627235927755,
      "loss": 5.1132,
      "mean_token_accuracy": 0.19678700864315032,
      "num_tokens": 29145577.0,
      "step": 14040
    },
    {
      "entropy": 5.447263574600219,
      "epoch": 1.2740384615384617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004843508696154388,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2010553538799286,
      "num_tokens": 29155905.0,
      "step": 14045
    },
    {
      "entropy": 5.362573432922363,
      "epoch": 1.2744920174165457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048433901130867283,
      "loss": 4.9685,
      "mean_token_accuracy": 0.20433792024850844,
      "num_tokens": 29164926.0,
      "step": 14050
    },
    {
      "entropy": 5.383613729476929,
      "epoch": 1.2749455732946298,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004843271486727228,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20115193128585815,
      "num_tokens": 29175050.0,
      "step": 14055
    },
    {
      "entropy": 5.432337713241577,
      "epoch": 1.275399129172714,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048431528170783395,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19764190763235093,
      "num_tokens": 29186255.0,
      "step": 14060
    },
    {
      "entropy": 5.472193193435669,
      "epoch": 1.2758526850507983,
      "grad_norm": 1.0,
      "learning_rate": 0.0004843034104142517,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20842560082674028,
      "num_tokens": 29195639.0,
      "step": 14065
    },
    {
      "entropy": 5.45418119430542,
      "epoch": 1.2763062409288826,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004842915347922216,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19702211916446685,
      "num_tokens": 29206482.0,
      "step": 14070
    },
    {
      "entropy": 5.485725736618042,
      "epoch": 1.2767597968069666,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004842796548419893,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20077505707740784,
      "num_tokens": 29217433.0,
      "step": 14075
    },
    {
      "entropy": 5.506093978881836,
      "epoch": 1.2772133526850509,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048426777056380035,
      "loss": 5.1112,
      "mean_token_accuracy": 0.1898355156183243,
      "num_tokens": 29226193.0,
      "step": 14080
    },
    {
      "entropy": 5.413127422332764,
      "epoch": 1.277666908563135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004842558819579006,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2013655796647072,
      "num_tokens": 29237061.0,
      "step": 14085
    },
    {
      "entropy": 5.393286514282226,
      "epoch": 1.2781204644412192,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004842439890245358,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20250939279794694,
      "num_tokens": 29247689.0,
      "step": 14090
    },
    {
      "entropy": 5.413403701782227,
      "epoch": 1.2785740203193035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048423209176395204,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20171325504779816,
      "num_tokens": 29258235.0,
      "step": 14095
    },
    {
      "entropy": 5.405281209945679,
      "epoch": 1.2790275761973875,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004842201901763954,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2064704939723015,
      "num_tokens": 29268884.0,
      "step": 14100
    },
    {
      "entropy": 5.420878267288208,
      "epoch": 1.2794811320754718,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004842082842621118,
      "loss": 5.1251,
      "mean_token_accuracy": 0.19449977129697799,
      "num_tokens": 29280484.0,
      "step": 14105
    },
    {
      "entropy": 5.477148818969726,
      "epoch": 1.2799346879535558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841963740213476,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20096696317195892,
      "num_tokens": 29291199.0,
      "step": 14110
    },
    {
      "entropy": 5.512815713882446,
      "epoch": 1.28038824383164,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004841844594543491,
      "loss": 5.0608,
      "mean_token_accuracy": 0.1944354072213173,
      "num_tokens": 29302007.0,
      "step": 14115
    },
    {
      "entropy": 5.461197090148926,
      "epoch": 1.2808417997097243,
      "grad_norm": 1.0,
      "learning_rate": 0.0004841725405613626,
      "loss": 5.0319,
      "mean_token_accuracy": 0.2100828170776367,
      "num_tokens": 29312127.0,
      "step": 14120
    },
    {
      "entropy": 5.467263507843017,
      "epoch": 1.2812953555878084,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004841606173426347,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19667672961950303,
      "num_tokens": 29323691.0,
      "step": 14125
    },
    {
      "entropy": 5.434922504425049,
      "epoch": 1.2817489114658926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048414868979841197,
      "loss": 5.0647,
      "mean_token_accuracy": 0.19760261178016664,
      "num_tokens": 29335203.0,
      "step": 14130
    },
    {
      "entropy": 5.494853067398071,
      "epoch": 1.2822024673439767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048413675792894105,
      "loss": 5.0908,
      "mean_token_accuracy": 0.2013104259967804,
      "num_tokens": 29346422.0,
      "step": 14135
    },
    {
      "entropy": 5.44482274055481,
      "epoch": 1.282656023222061,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841248217344686,
      "loss": 5.0054,
      "mean_token_accuracy": 0.1991575077176094,
      "num_tokens": 29356084.0,
      "step": 14140
    },
    {
      "entropy": 5.40657606124878,
      "epoch": 1.2831095791001452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004841128812152415,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20386681109666824,
      "num_tokens": 29366093.0,
      "step": 14145
    },
    {
      "entropy": 5.477137041091919,
      "epoch": 1.2835631349782293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004841009363715068,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2001335069537163,
      "num_tokens": 29376621.0,
      "step": 14150
    },
    {
      "entropy": 5.5024326801300045,
      "epoch": 1.2840166908563135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004840889872035115,
      "loss": 5.1196,
      "mean_token_accuracy": 0.19854445606470109,
      "num_tokens": 29387119.0,
      "step": 14155
    },
    {
      "entropy": 5.455497598648071,
      "epoch": 1.2844702467343976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004840770337115026,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20350901484489442,
      "num_tokens": 29396981.0,
      "step": 14160
    },
    {
      "entropy": 5.488083553314209,
      "epoch": 1.2849238026124818,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048406507589572747,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20519718825817107,
      "num_tokens": 29406737.0,
      "step": 14165
    },
    {
      "entropy": 5.475179767608642,
      "epoch": 1.2853773584905661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004840531137564331,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20185788571834565,
      "num_tokens": 29417300.0,
      "step": 14170
    },
    {
      "entropy": 5.436309385299682,
      "epoch": 1.2858309143686502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004840411472938672,
      "loss": 4.966,
      "mean_token_accuracy": 0.20452154278755189,
      "num_tokens": 29426749.0,
      "step": 14175
    },
    {
      "entropy": 5.393154668807983,
      "epoch": 1.2862844702467344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004840291765082771,
      "loss": 4.9809,
      "mean_token_accuracy": 0.19870773255825042,
      "num_tokens": 29437264.0,
      "step": 14180
    },
    {
      "entropy": 5.422954511642456,
      "epoch": 1.2867380261248185,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048401720139991036,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1949951395392418,
      "num_tokens": 29448192.0,
      "step": 14185
    },
    {
      "entropy": 5.45281662940979,
      "epoch": 1.2871915820029027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004840052219690146,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2128344178199768,
      "num_tokens": 29457923.0,
      "step": 14190
    },
    {
      "entropy": 5.366837024688721,
      "epoch": 1.287645137880987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839932382158376,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21100290715694428,
      "num_tokens": 29466914.0,
      "step": 14195
    },
    {
      "entropy": 5.425601863861084,
      "epoch": 1.288098693759071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004839812501406272,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19851188957691193,
      "num_tokens": 29477648.0,
      "step": 14200
    },
    {
      "entropy": 5.4876306533813475,
      "epoch": 1.2885522496371553,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004839692577436313,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20532836019992828,
      "num_tokens": 29487947.0,
      "step": 14205
    },
    {
      "entropy": 5.553531694412231,
      "epoch": 1.2890058055152394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004839572610250978,
      "loss": 5.1424,
      "mean_token_accuracy": 0.1927253320813179,
      "num_tokens": 29497667.0,
      "step": 14210
    },
    {
      "entropy": 5.460983657836914,
      "epoch": 1.2894593613933236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048394525998527496,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19897037148475646,
      "num_tokens": 29507065.0,
      "step": 14215
    },
    {
      "entropy": 5.4825396060943605,
      "epoch": 1.2899129172714079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004839332546244109,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20061064064502715,
      "num_tokens": 29517752.0,
      "step": 14220
    },
    {
      "entropy": 5.454512977600098,
      "epoch": 1.2903664731494922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048392124494275385,
      "loss": 5.028,
      "mean_token_accuracy": 0.1958656743168831,
      "num_tokens": 29527894.0,
      "step": 14225
    },
    {
      "entropy": 5.423327112197876,
      "epoch": 1.2908200290275762,
      "grad_norm": 1.0,
      "learning_rate": 0.0004839092309405523,
      "loss": 5.0654,
      "mean_token_accuracy": 0.19844336658716202,
      "num_tokens": 29538136.0,
      "step": 14230
    },
    {
      "entropy": 5.452183866500855,
      "epoch": 1.2912735849056602,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048389721261805457,
      "loss": 5.047,
      "mean_token_accuracy": 0.19840533882379532,
      "num_tokens": 29549193.0,
      "step": 14235
    },
    {
      "entropy": 5.387442398071289,
      "epoch": 1.2917271407837445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048388518997550926,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20968773365020751,
      "num_tokens": 29558889.0,
      "step": 14240
    },
    {
      "entropy": 5.454190015792847,
      "epoch": 1.2921806966618288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048387316301316493,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19536134749650955,
      "num_tokens": 29569691.0,
      "step": 14245
    },
    {
      "entropy": 5.480500888824463,
      "epoch": 1.292634252539913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004838611317312705,
      "loss": 5.1387,
      "mean_token_accuracy": 0.19596640020608902,
      "num_tokens": 29580459.0,
      "step": 14250
    },
    {
      "entropy": 5.452585601806641,
      "epoch": 1.293087808417997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004838490961300745,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19633100181818008,
      "num_tokens": 29590652.0,
      "step": 14255
    },
    {
      "entropy": 5.474935817718506,
      "epoch": 1.2935413642960814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048383705620982606,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20783319175243378,
      "num_tokens": 29600353.0,
      "step": 14260
    },
    {
      "entropy": 5.4646180152893065,
      "epoch": 1.2939949201741654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048382501197077405,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20905023217201232,
      "num_tokens": 29610561.0,
      "step": 14265
    },
    {
      "entropy": 5.398465728759765,
      "epoch": 1.2944484760522497,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004838129634131675,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20370285362005233,
      "num_tokens": 29621741.0,
      "step": 14270
    },
    {
      "entropy": 5.408251619338989,
      "epoch": 1.294902031930334,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004838009105372558,
      "loss": 4.9551,
      "mean_token_accuracy": 0.20420809090137482,
      "num_tokens": 29633097.0,
      "step": 14275
    },
    {
      "entropy": 5.448822832107544,
      "epoch": 1.295355587808418,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048378885334328793,
      "loss": 5.0664,
      "mean_token_accuracy": 0.19318899959325792,
      "num_tokens": 29644290.0,
      "step": 14280
    },
    {
      "entropy": 5.518048477172852,
      "epoch": 1.2958091436865022,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004837767918315135,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19329611957073212,
      "num_tokens": 29654977.0,
      "step": 14285
    },
    {
      "entropy": 5.52895302772522,
      "epoch": 1.2962626995645863,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048376472600218174,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19857558757066726,
      "num_tokens": 29665020.0,
      "step": 14290
    },
    {
      "entropy": 5.4473686695098875,
      "epoch": 1.2967162554426706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837526558555422,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20379104614257812,
      "num_tokens": 29675098.0,
      "step": 14295
    },
    {
      "entropy": 5.395389080047607,
      "epoch": 1.2971698113207548,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004837405813918446,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2069588765501976,
      "num_tokens": 29685450.0,
      "step": 14300
    },
    {
      "entropy": 5.399443626403809,
      "epoch": 1.2976233671988389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048372850261133856,
      "loss": 4.9946,
      "mean_token_accuracy": 0.20608473420143128,
      "num_tokens": 29694818.0,
      "step": 14305
    },
    {
      "entropy": 5.476320552825928,
      "epoch": 1.2980769230769231,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048371641951427393,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19565918147563935,
      "num_tokens": 29705573.0,
      "step": 14310
    },
    {
      "entropy": 5.466921043395996,
      "epoch": 1.2985304789550072,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048370433210090055,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20001791417598724,
      "num_tokens": 29716366.0,
      "step": 14315
    },
    {
      "entropy": 5.4390482902526855,
      "epoch": 1.2989840348330914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004836922403714683,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20087531805038453,
      "num_tokens": 29727181.0,
      "step": 14320
    },
    {
      "entropy": 5.393351268768311,
      "epoch": 1.2994375907111757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048368014432622746,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20365506559610366,
      "num_tokens": 29736960.0,
      "step": 14325
    },
    {
      "entropy": 5.40035457611084,
      "epoch": 1.2998911465892597,
      "grad_norm": 1.03125,
      "learning_rate": 0.000483668043965428,
      "loss": 4.93,
      "mean_token_accuracy": 0.21387912034988404,
      "num_tokens": 29746354.0,
      "step": 14330
    },
    {
      "entropy": 5.445358324050903,
      "epoch": 1.300344702467344,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004836559392893203,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19463613033294677,
      "num_tokens": 29756627.0,
      "step": 14335
    },
    {
      "entropy": 5.4277238845825195,
      "epoch": 1.300798258345428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004836438302981545,
      "loss": 5.023,
      "mean_token_accuracy": 0.20784646421670913,
      "num_tokens": 29766684.0,
      "step": 14340
    },
    {
      "entropy": 5.427753973007202,
      "epoch": 1.3012518142235123,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004836317169921811,
      "loss": 5.055,
      "mean_token_accuracy": 0.20116524547338485,
      "num_tokens": 29776556.0,
      "step": 14345
    },
    {
      "entropy": 5.52408995628357,
      "epoch": 1.3017053701015966,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048361959937165075,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19422274082899094,
      "num_tokens": 29786896.0,
      "step": 14350
    },
    {
      "entropy": 5.439397239685059,
      "epoch": 1.3021589259796806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004836074774368138,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20644939094781875,
      "num_tokens": 29797332.0,
      "step": 14355
    },
    {
      "entropy": 5.44417028427124,
      "epoch": 1.302612481857765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048359535118792116,
      "loss": 5.0807,
      "mean_token_accuracy": 0.2013534665107727,
      "num_tokens": 29806428.0,
      "step": 14360
    },
    {
      "entropy": 5.437863159179687,
      "epoch": 1.303066037735849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004835832206252235,
      "loss": 5.055,
      "mean_token_accuracy": 0.2009737566113472,
      "num_tokens": 29816794.0,
      "step": 14365
    },
    {
      "entropy": 5.52238097190857,
      "epoch": 1.3035195936139332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004835710857489716,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20529444962739946,
      "num_tokens": 29827077.0,
      "step": 14370
    },
    {
      "entropy": 5.4323179721832275,
      "epoch": 1.3039731494920175,
      "grad_norm": 1.0,
      "learning_rate": 0.0004835589465594165,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19651621133089064,
      "num_tokens": 29837868.0,
      "step": 14375
    },
    {
      "entropy": 5.445604848861694,
      "epoch": 1.3044267053701015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048354680305680926,
      "loss": 5.0627,
      "mean_token_accuracy": 0.2027847409248352,
      "num_tokens": 29846929.0,
      "step": 14380
    },
    {
      "entropy": 5.41439905166626,
      "epoch": 1.3048802612481858,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048353465524140105,
      "loss": 5.0041,
      "mean_token_accuracy": 0.2011212959885597,
      "num_tokens": 29857905.0,
      "step": 14385
    },
    {
      "entropy": 5.437925243377686,
      "epoch": 1.3053338171262698,
      "grad_norm": 1.09375,
      "learning_rate": 0.000483522503113443,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19779178202152253,
      "num_tokens": 29868230.0,
      "step": 14390
    },
    {
      "entropy": 5.440568447113037,
      "epoch": 1.305787373004354,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048351034667318634,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2029603287577629,
      "num_tokens": 29879224.0,
      "step": 14395
    },
    {
      "entropy": 5.5207991123199465,
      "epoch": 1.3062409288824384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004834981859208827,
      "loss": 5.1658,
      "mean_token_accuracy": 0.19206029772758484,
      "num_tokens": 29890416.0,
      "step": 14400
    },
    {
      "entropy": 5.47591118812561,
      "epoch": 1.3066944847605226,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048348602085678346,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19681367427110671,
      "num_tokens": 29901064.0,
      "step": 14405
    },
    {
      "entropy": 5.4896162986755375,
      "epoch": 1.3071480406386067,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048347385148114004,
      "loss": 5.135,
      "mean_token_accuracy": 0.19611346125602722,
      "num_tokens": 29911084.0,
      "step": 14410
    },
    {
      "entropy": 5.397078657150269,
      "epoch": 1.3076015965166907,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048346167779420436,
      "loss": 4.986,
      "mean_token_accuracy": 0.21088134199380876,
      "num_tokens": 29921104.0,
      "step": 14415
    },
    {
      "entropy": 5.441667795181274,
      "epoch": 1.308055152394775,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048344949979622806,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20658447295427323,
      "num_tokens": 29932014.0,
      "step": 14420
    },
    {
      "entropy": 5.416403722763062,
      "epoch": 1.3085087082728593,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048343731748746294,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20720082372426987,
      "num_tokens": 29941707.0,
      "step": 14425
    },
    {
      "entropy": 5.453537797927856,
      "epoch": 1.3089622641509435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000483425130868161,
      "loss": 5.0319,
      "mean_token_accuracy": 0.19454951584339142,
      "num_tokens": 29952082.0,
      "step": 14430
    },
    {
      "entropy": 5.404423666000366,
      "epoch": 1.3094158200290276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004834129399385742,
      "loss": 4.8531,
      "mean_token_accuracy": 0.21807902902364731,
      "num_tokens": 29961744.0,
      "step": 14435
    },
    {
      "entropy": 5.438563823699951,
      "epoch": 1.3098693759071118,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834007446989548,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20898614972829818,
      "num_tokens": 29972611.0,
      "step": 14440
    },
    {
      "entropy": 5.407641267776489,
      "epoch": 1.3103229317851959,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048338854514955474,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2093118906021118,
      "num_tokens": 29982660.0,
      "step": 14445
    },
    {
      "entropy": 5.432706880569458,
      "epoch": 1.3107764876632801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048337634129062655,
      "loss": 5.0649,
      "mean_token_accuracy": 0.19014720171689986,
      "num_tokens": 29993349.0,
      "step": 14450
    },
    {
      "entropy": 5.524210405349732,
      "epoch": 1.3112300435413644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048336413312242255,
      "loss": 5.145,
      "mean_token_accuracy": 0.19443282335996628,
      "num_tokens": 30004474.0,
      "step": 14455
    },
    {
      "entropy": 5.499291038513183,
      "epoch": 1.3116835994194485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048335192064519514,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19702026546001433,
      "num_tokens": 30015189.0,
      "step": 14460
    },
    {
      "entropy": 5.472742652893066,
      "epoch": 1.3121371552975327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048333970385919695,
      "loss": 4.9927,
      "mean_token_accuracy": 0.19688308984041214,
      "num_tokens": 30025491.0,
      "step": 14465
    },
    {
      "entropy": 5.4534389019012455,
      "epoch": 1.3125907111756168,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048332748276468053,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2055317595601082,
      "num_tokens": 30034683.0,
      "step": 14470
    },
    {
      "entropy": 5.412027359008789,
      "epoch": 1.313044267053701,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004833152573618987,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20131270736455917,
      "num_tokens": 30044135.0,
      "step": 14475
    },
    {
      "entropy": 5.484115648269653,
      "epoch": 1.3134978229317853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048330302765110436,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20330243408679963,
      "num_tokens": 30054520.0,
      "step": 14480
    },
    {
      "entropy": 5.487312746047974,
      "epoch": 1.3139513788098693,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048329079363255024,
      "loss": 5.0749,
      "mean_token_accuracy": 0.19585316628217697,
      "num_tokens": 30065439.0,
      "step": 14485
    },
    {
      "entropy": 5.493306255340576,
      "epoch": 1.3144049346879536,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048327855530648934,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20039551556110383,
      "num_tokens": 30075362.0,
      "step": 14490
    },
    {
      "entropy": 5.432645845413208,
      "epoch": 1.3148584905660377,
      "grad_norm": 1.0,
      "learning_rate": 0.000483266312673175,
      "loss": 5.0265,
      "mean_token_accuracy": 0.19832777976989746,
      "num_tokens": 30084913.0,
      "step": 14495
    },
    {
      "entropy": 5.44658465385437,
      "epoch": 1.315312046444122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004832540657328602,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2005489856004715,
      "num_tokens": 30095704.0,
      "step": 14500
    },
    {
      "entropy": 5.435765361785888,
      "epoch": 1.3157656023222062,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004832418144857982,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20489873737096786,
      "num_tokens": 30105981.0,
      "step": 14505
    },
    {
      "entropy": 5.4482102394104,
      "epoch": 1.3162191582002902,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048322955893224245,
      "loss": 5.058,
      "mean_token_accuracy": 0.19121648818254472,
      "num_tokens": 30115772.0,
      "step": 14510
    },
    {
      "entropy": 5.466764068603515,
      "epoch": 1.3166727140783745,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048321729907244626,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20533003211021422,
      "num_tokens": 30127050.0,
      "step": 14515
    },
    {
      "entropy": 5.46243839263916,
      "epoch": 1.3171262699564585,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004832050349066634,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1961934819817543,
      "num_tokens": 30138221.0,
      "step": 14520
    },
    {
      "entropy": 5.4993661403656,
      "epoch": 1.3175798258345428,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004831927664351473,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1948306679725647,
      "num_tokens": 30150181.0,
      "step": 14525
    },
    {
      "entropy": 5.444983863830567,
      "epoch": 1.318033381712627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004831804936581518,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20480622947216034,
      "num_tokens": 30160521.0,
      "step": 14530
    },
    {
      "entropy": 5.518583297729492,
      "epoch": 1.3184869375907111,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004831682165759305,
      "loss": 5.0528,
      "mean_token_accuracy": 0.20147155076265336,
      "num_tokens": 30170611.0,
      "step": 14535
    },
    {
      "entropy": 5.483779048919677,
      "epoch": 1.3189404934687954,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048315593518873757,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2089202657341957,
      "num_tokens": 30179914.0,
      "step": 14540
    },
    {
      "entropy": 5.418671417236328,
      "epoch": 1.3193940493468794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004831436494968268,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20389744341373445,
      "num_tokens": 30191426.0,
      "step": 14545
    },
    {
      "entropy": 5.445282936096191,
      "epoch": 1.3198476052249637,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048313135950045234,
      "loss": 4.9902,
      "mean_token_accuracy": 0.20250587314367294,
      "num_tokens": 30202813.0,
      "step": 14550
    },
    {
      "entropy": 5.443851375579834,
      "epoch": 1.320301161103048,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048311906519986824,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20622701942920685,
      "num_tokens": 30213044.0,
      "step": 14555
    },
    {
      "entropy": 5.447887945175171,
      "epoch": 1.320754716981132,
      "grad_norm": 1.140625,
      "learning_rate": 0.000483106766595329,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20020903646945953,
      "num_tokens": 30223419.0,
      "step": 14560
    },
    {
      "entropy": 5.505366373062134,
      "epoch": 1.3212082728592163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004830944636870887,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19486852139234542,
      "num_tokens": 30233203.0,
      "step": 14565
    },
    {
      "entropy": 5.464370012283325,
      "epoch": 1.3216618287373003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048308215647540183,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2010621652007103,
      "num_tokens": 30244604.0,
      "step": 14570
    },
    {
      "entropy": 5.494146108627319,
      "epoch": 1.3221153846153846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048306984496052297,
      "loss": 5.1131,
      "mean_token_accuracy": 0.19592046290636062,
      "num_tokens": 30255355.0,
      "step": 14575
    },
    {
      "entropy": 5.54925045967102,
      "epoch": 1.3225689404934688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048305752914270665,
      "loss": 5.1373,
      "mean_token_accuracy": 0.19726582765579223,
      "num_tokens": 30265665.0,
      "step": 14580
    },
    {
      "entropy": 5.539267683029175,
      "epoch": 1.3230224963715531,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048304520902220767,
      "loss": 5.1185,
      "mean_token_accuracy": 0.20208002030849456,
      "num_tokens": 30274347.0,
      "step": 14585
    },
    {
      "entropy": 5.495355844497681,
      "epoch": 1.3234760522496372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004830328845992808,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20875369161367416,
      "num_tokens": 30284243.0,
      "step": 14590
    },
    {
      "entropy": 5.390436172485352,
      "epoch": 1.3239296081277212,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048302055587418074,
      "loss": 4.964,
      "mean_token_accuracy": 0.20331924855709077,
      "num_tokens": 30294327.0,
      "step": 14595
    },
    {
      "entropy": 5.410218811035156,
      "epoch": 1.3243831640058055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004830082228471626,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19879168272018433,
      "num_tokens": 30304188.0,
      "step": 14600
    },
    {
      "entropy": 5.45996503829956,
      "epoch": 1.3248367198838897,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048299588551848146,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2053225114941597,
      "num_tokens": 30314917.0,
      "step": 14605
    },
    {
      "entropy": 5.437911081314087,
      "epoch": 1.325290275761974,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048298354388839237,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20336007922887803,
      "num_tokens": 30325221.0,
      "step": 14610
    },
    {
      "entropy": 5.50159010887146,
      "epoch": 1.325743831640058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004829711979571506,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20005955100059508,
      "num_tokens": 30336196.0,
      "step": 14615
    },
    {
      "entropy": 5.423232793807983,
      "epoch": 1.3261973875181423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004829588477250115,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2151394471526146,
      "num_tokens": 30345546.0,
      "step": 14620
    },
    {
      "entropy": 5.445691871643066,
      "epoch": 1.3266509433962264,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004829464931922304,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20275977700948716,
      "num_tokens": 30355672.0,
      "step": 14625
    },
    {
      "entropy": 5.442801475524902,
      "epoch": 1.3271044992743106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048293413435906285,
      "loss": 5.0437,
      "mean_token_accuracy": 0.19868946373462676,
      "num_tokens": 30366233.0,
      "step": 14630
    },
    {
      "entropy": 5.471208190917968,
      "epoch": 1.327558055152395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829217712257644,
      "loss": 5.0438,
      "mean_token_accuracy": 0.20054723769426347,
      "num_tokens": 30377366.0,
      "step": 14635
    },
    {
      "entropy": 5.450821590423584,
      "epoch": 1.328011611030479,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048290940379259073,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20034591108560562,
      "num_tokens": 30388022.0,
      "step": 14640
    },
    {
      "entropy": 5.457063865661621,
      "epoch": 1.3284651669085632,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048289703205979756,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20205169171094894,
      "num_tokens": 30399235.0,
      "step": 14645
    },
    {
      "entropy": 5.57108793258667,
      "epoch": 1.3289187227866472,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048288465602764084,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20028597861528397,
      "num_tokens": 30409112.0,
      "step": 14650
    },
    {
      "entropy": 5.409694528579712,
      "epoch": 1.3293722786647315,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048287227569637645,
      "loss": 5.0199,
      "mean_token_accuracy": 0.20294926911592484,
      "num_tokens": 30420202.0,
      "step": 14655
    },
    {
      "entropy": 5.459199047088623,
      "epoch": 1.3298258345428158,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048285989106626036,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20262259989976883,
      "num_tokens": 30431210.0,
      "step": 14660
    },
    {
      "entropy": 5.454660701751709,
      "epoch": 1.3302793904208998,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004828475021375489,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20774459838867188,
      "num_tokens": 30441291.0,
      "step": 14665
    },
    {
      "entropy": 5.424936532974243,
      "epoch": 1.330732946298984,
      "grad_norm": 1.125,
      "learning_rate": 0.0004828351089104981,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21167429983615876,
      "num_tokens": 30450285.0,
      "step": 14670
    },
    {
      "entropy": 5.4106463432312015,
      "epoch": 1.3311865021770681,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048282271138536424,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20098748207092285,
      "num_tokens": 30460311.0,
      "step": 14675
    },
    {
      "entropy": 5.528666687011719,
      "epoch": 1.3316400580551524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004828103095624037,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2035180002450943,
      "num_tokens": 30470188.0,
      "step": 14680
    },
    {
      "entropy": 5.441724872589111,
      "epoch": 1.3320936139332367,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000482797903441873,
      "loss": 5.002,
      "mean_token_accuracy": 0.1982833057641983,
      "num_tokens": 30481406.0,
      "step": 14685
    },
    {
      "entropy": 5.374399185180664,
      "epoch": 1.3325471698113207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004827854930240288,
      "loss": 5.004,
      "mean_token_accuracy": 0.2057785615324974,
      "num_tokens": 30492314.0,
      "step": 14690
    },
    {
      "entropy": 5.532073640823365,
      "epoch": 1.333000725689405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004827730783091276,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19633910357952117,
      "num_tokens": 30503273.0,
      "step": 14695
    },
    {
      "entropy": 5.512577486038208,
      "epoch": 1.333454281567489,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048276065929742616,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19977344423532487,
      "num_tokens": 30513582.0,
      "step": 14700
    },
    {
      "entropy": 5.413820219039917,
      "epoch": 1.3339078374455733,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004827482359891814,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2092178627848625,
      "num_tokens": 30523261.0,
      "step": 14705
    },
    {
      "entropy": 5.468668031692505,
      "epoch": 1.3343613933236576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048273580838465016,
      "loss": 5.076,
      "mean_token_accuracy": 0.2011701285839081,
      "num_tokens": 30533774.0,
      "step": 14710
    },
    {
      "entropy": 5.464613723754883,
      "epoch": 1.3348149492017416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004827233764840894,
      "loss": 5.1199,
      "mean_token_accuracy": 0.19367510378360747,
      "num_tokens": 30544580.0,
      "step": 14715
    },
    {
      "entropy": 5.484494543075561,
      "epoch": 1.3352685050798259,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004827109402877564,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20472703725099564,
      "num_tokens": 30555105.0,
      "step": 14720
    },
    {
      "entropy": 5.46086688041687,
      "epoch": 1.33572206095791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004826984997959082,
      "loss": 5.1107,
      "mean_token_accuracy": 0.18970336467027665,
      "num_tokens": 30564912.0,
      "step": 14725
    },
    {
      "entropy": 5.424795866012573,
      "epoch": 1.3361756168359942,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000482686055008802,
      "loss": 5.029,
      "mean_token_accuracy": 0.19458736777305602,
      "num_tokens": 30575862.0,
      "step": 14730
    },
    {
      "entropy": 5.477669715881348,
      "epoch": 1.3366291727140784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048267360592669535,
      "loss": 5.15,
      "mean_token_accuracy": 0.19470789283514023,
      "num_tokens": 30586951.0,
      "step": 14735
    },
    {
      "entropy": 5.497680521011352,
      "epoch": 1.3370827285921625,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826611525498456,
      "loss": 5.0709,
      "mean_token_accuracy": 0.19571490734815597,
      "num_tokens": 30597604.0,
      "step": 14740
    },
    {
      "entropy": 5.504551362991333,
      "epoch": 1.3375362844702468,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826486948785103,
      "loss": 5.118,
      "mean_token_accuracy": 0.19376556128263472,
      "num_tokens": 30608157.0,
      "step": 14745
    },
    {
      "entropy": 5.462109375,
      "epoch": 1.3379898403483308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048263623291294706,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20001840591430664,
      "num_tokens": 30618527.0,
      "step": 14750
    },
    {
      "entropy": 5.5271852016448975,
      "epoch": 1.338443396226415,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048262376665341355,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19271586388349532,
      "num_tokens": 30629253.0,
      "step": 14755
    },
    {
      "entropy": 5.500084733963012,
      "epoch": 1.3388969521044993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004826112961001677,
      "loss": 5.107,
      "mean_token_accuracy": 0.18575113266706467,
      "num_tokens": 30639015.0,
      "step": 14760
    },
    {
      "entropy": 5.545000314712524,
      "epoch": 1.3393505079825834,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048259882125346735,
      "loss": 5.1807,
      "mean_token_accuracy": 0.18538891077041625,
      "num_tokens": 30650035.0,
      "step": 14765
    },
    {
      "entropy": 5.489197015762329,
      "epoch": 1.3398040638606676,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048258634211357045,
      "loss": 5.086,
      "mean_token_accuracy": 0.2026127129793167,
      "num_tokens": 30659813.0,
      "step": 14770
    },
    {
      "entropy": 5.461983060836792,
      "epoch": 1.3402576197387517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004825738586807351,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2014084503054619,
      "num_tokens": 30670243.0,
      "step": 14775
    },
    {
      "entropy": 5.427335739135742,
      "epoch": 1.340711175616836,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048256137095521945,
      "loss": 5.0189,
      "mean_token_accuracy": 0.20176693052053452,
      "num_tokens": 30680056.0,
      "step": 14780
    },
    {
      "entropy": 5.461502981185913,
      "epoch": 1.3411647314949202,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048254887893728175,
      "loss": 4.9573,
      "mean_token_accuracy": 0.20859941095113754,
      "num_tokens": 30690536.0,
      "step": 14785
    },
    {
      "entropy": 5.427439641952515,
      "epoch": 1.3416182873730045,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004825363826271804,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19469753801822662,
      "num_tokens": 30700200.0,
      "step": 14790
    },
    {
      "entropy": 5.4225091457366945,
      "epoch": 1.3420718432510885,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004825238820251737,
      "loss": 4.9502,
      "mean_token_accuracy": 0.20724240243434905,
      "num_tokens": 30710576.0,
      "step": 14795
    },
    {
      "entropy": 5.439230823516846,
      "epoch": 1.3425253991291728,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048251137713152027,
      "loss": 5.0538,
      "mean_token_accuracy": 0.19843484163284303,
      "num_tokens": 30721495.0,
      "step": 14800
    },
    {
      "entropy": 5.488902759552002,
      "epoch": 1.3429789550072568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004824988679464787,
      "loss": 5.0783,
      "mean_token_accuracy": 0.2005630537867546,
      "num_tokens": 30731700.0,
      "step": 14805
    },
    {
      "entropy": 5.503637075424194,
      "epoch": 1.343432510885341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048248635447030767,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19916742742061616,
      "num_tokens": 30743421.0,
      "step": 14810
    },
    {
      "entropy": 5.453477621078491,
      "epoch": 1.3438860667634254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004824738367032659,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20747420191764832,
      "num_tokens": 30754238.0,
      "step": 14815
    },
    {
      "entropy": 5.439966249465942,
      "epoch": 1.3443396226415094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004824613146456123,
      "loss": 5.0652,
      "mean_token_accuracy": 0.19901268929243088,
      "num_tokens": 30763938.0,
      "step": 14820
    },
    {
      "entropy": 5.524941539764404,
      "epoch": 1.3447931785195937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004824487882976059,
      "loss": 5.121,
      "mean_token_accuracy": 0.19707119315862656,
      "num_tokens": 30773654.0,
      "step": 14825
    },
    {
      "entropy": 5.533320617675781,
      "epoch": 1.3452467343976777,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048243625765950564,
      "loss": 5.103,
      "mean_token_accuracy": 0.19529523849487304,
      "num_tokens": 30784659.0,
      "step": 14830
    },
    {
      "entropy": 5.427193021774292,
      "epoch": 1.345700290275762,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004824237227315708,
      "loss": 4.9791,
      "mean_token_accuracy": 0.1971741423010826,
      "num_tokens": 30795194.0,
      "step": 14835
    },
    {
      "entropy": 5.3818628787994385,
      "epoch": 1.3461538461538463,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004824111835140604,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2113494724035263,
      "num_tokens": 30805623.0,
      "step": 14840
    },
    {
      "entropy": 5.438431215286255,
      "epoch": 1.3466074020319303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482398640007234,
      "loss": 4.9711,
      "mean_token_accuracy": 0.20331282466650008,
      "num_tokens": 30814899.0,
      "step": 14845
    },
    {
      "entropy": 5.443796539306641,
      "epoch": 1.3470609579100146,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048238609221135076,
      "loss": 5.0826,
      "mean_token_accuracy": 0.1992559924721718,
      "num_tokens": 30825546.0,
      "step": 14850
    },
    {
      "entropy": 5.39530725479126,
      "epoch": 1.3475145137880986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004823735401266704,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20329416990280152,
      "num_tokens": 30836406.0,
      "step": 14855
    },
    {
      "entropy": 5.427872180938721,
      "epoch": 1.3479680696661829,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048236098375345227,
      "loss": 5.073,
      "mean_token_accuracy": 0.20036971867084502,
      "num_tokens": 30847966.0,
      "step": 14860
    },
    {
      "entropy": 5.550434494018555,
      "epoch": 1.3484216255442671,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004823484230919563,
      "loss": 5.0939,
      "mean_token_accuracy": 0.19483736902475357,
      "num_tokens": 30856831.0,
      "step": 14865
    },
    {
      "entropy": 5.4240264892578125,
      "epoch": 1.3488751814223512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048233585814244195,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21467597633600236,
      "num_tokens": 30866076.0,
      "step": 14870
    },
    {
      "entropy": 5.4501073360443115,
      "epoch": 1.3493287373004355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004823232889051693,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20600891709327698,
      "num_tokens": 30876845.0,
      "step": 14875
    },
    {
      "entropy": 5.420187616348267,
      "epoch": 1.3497822931785195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048231071538039816,
      "loss": 5.041,
      "mean_token_accuracy": 0.19574736207723617,
      "num_tokens": 30887448.0,
      "step": 14880
    },
    {
      "entropy": 5.497925186157227,
      "epoch": 1.3502358490566038,
      "grad_norm": 1.0,
      "learning_rate": 0.00048229813756838863,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20030652433633805,
      "num_tokens": 30898414.0,
      "step": 14885
    },
    {
      "entropy": 5.445967340469361,
      "epoch": 1.350689404934688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004822855554694007,
      "loss": 5.0354,
      "mean_token_accuracy": 0.19963205456733704,
      "num_tokens": 30908559.0,
      "step": 14890
    },
    {
      "entropy": 5.457993984222412,
      "epoch": 1.351142960812772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048227296908369475,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20594154745340348,
      "num_tokens": 30918655.0,
      "step": 14895
    },
    {
      "entropy": 5.44824628829956,
      "epoch": 1.3515965166908563,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048226037841153104,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20258834958076477,
      "num_tokens": 30929188.0,
      "step": 14900
    },
    {
      "entropy": 5.501089763641358,
      "epoch": 1.3520500725689404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004822477834531698,
      "loss": 5.1109,
      "mean_token_accuracy": 0.19966886788606644,
      "num_tokens": 30939902.0,
      "step": 14905
    },
    {
      "entropy": 5.4724814891815186,
      "epoch": 1.3525036284470247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004822351842088716,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20224396884441376,
      "num_tokens": 30951289.0,
      "step": 14910
    },
    {
      "entropy": 5.541313600540161,
      "epoch": 1.352957184325109,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004822225806788969,
      "loss": 5.1729,
      "mean_token_accuracy": 0.18792562633752824,
      "num_tokens": 30961233.0,
      "step": 14915
    },
    {
      "entropy": 5.476714181900024,
      "epoch": 1.353410740203193,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048220997286350654,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20616774260997772,
      "num_tokens": 30971922.0,
      "step": 14920
    },
    {
      "entropy": 5.43091139793396,
      "epoch": 1.3538642960812772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048219736076296114,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20542879700660704,
      "num_tokens": 30981358.0,
      "step": 14925
    },
    {
      "entropy": 5.482666444778443,
      "epoch": 1.3543178519593613,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004821847443775214,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19539569616317748,
      "num_tokens": 30991984.0,
      "step": 14930
    },
    {
      "entropy": 5.469381523132324,
      "epoch": 1.3547714078374455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048217212370744853,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20533296167850495,
      "num_tokens": 31002790.0,
      "step": 14935
    },
    {
      "entropy": 5.367084264755249,
      "epoch": 1.3552249637155298,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048215949875300327,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2160383567214012,
      "num_tokens": 31012827.0,
      "step": 14940
    },
    {
      "entropy": 5.369181156158447,
      "epoch": 1.3556785195936139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821468695144467,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20431342273950576,
      "num_tokens": 31023075.0,
      "step": 14945
    },
    {
      "entropy": 5.48018445968628,
      "epoch": 1.3561320754716981,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048213423599204015,
      "loss": 5.055,
      "mean_token_accuracy": 0.20195306688547135,
      "num_tokens": 31032454.0,
      "step": 14950
    },
    {
      "entropy": 5.569077444076538,
      "epoch": 1.3565856313497822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048212159818604477,
      "loss": 5.1833,
      "mean_token_accuracy": 0.18612317144870758,
      "num_tokens": 31042498.0,
      "step": 14955
    },
    {
      "entropy": 5.434992074966431,
      "epoch": 1.3570391872278664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000482108956096722,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20031053572893143,
      "num_tokens": 31053188.0,
      "step": 14960
    },
    {
      "entropy": 5.452586126327515,
      "epoch": 1.3574927431059507,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048209630972433326,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20466638952493668,
      "num_tokens": 31062840.0,
      "step": 14965
    },
    {
      "entropy": 5.5303045272827145,
      "epoch": 1.357946298984035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048208365906914,
      "loss": 5.1531,
      "mean_token_accuracy": 0.2004319131374359,
      "num_tokens": 31072646.0,
      "step": 14970
    },
    {
      "entropy": 5.42915096282959,
      "epoch": 1.358399854862119,
      "grad_norm": 1.125,
      "learning_rate": 0.00048207100413140397,
      "loss": 4.9793,
      "mean_token_accuracy": 0.19790226519107817,
      "num_tokens": 31082700.0,
      "step": 14975
    },
    {
      "entropy": 5.508495330810547,
      "epoch": 1.3588534107402033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004820583449113868,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19902369678020476,
      "num_tokens": 31093743.0,
      "step": 14980
    },
    {
      "entropy": 5.455974435806274,
      "epoch": 1.3593069666182873,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048204568140935023,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2065230205655098,
      "num_tokens": 31104293.0,
      "step": 14985
    },
    {
      "entropy": 5.384525966644287,
      "epoch": 1.3597605224963716,
      "grad_norm": 1.0,
      "learning_rate": 0.0004820330136255563,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20574056953191758,
      "num_tokens": 31114244.0,
      "step": 14990
    },
    {
      "entropy": 5.420985412597656,
      "epoch": 1.3602140783744558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048202034156026683,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20137730985879898,
      "num_tokens": 31123596.0,
      "step": 14995
    },
    {
      "entropy": 5.50952730178833,
      "epoch": 1.36066763425254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048200766521374396,
      "loss": 5.0688,
      "mean_token_accuracy": 0.19272379130125045,
      "num_tokens": 31134258.0,
      "step": 15000
    },
    {
      "epoch": 1.36066763425254,
      "eval_entropy": 5.280845000733078,
      "eval_loss": 5.10365104675293,
      "eval_mean_token_accuracy": 0.20662164339011882,
      "eval_num_tokens": 31134258.0,
      "eval_runtime": 20.4823,
      "eval_samples_per_second": 1726.364,
      "eval_steps_per_second": 215.796,
      "step": 15000
    },
    {
      "entropy": 5.4724356174469,
      "epoch": 1.3611211901306242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004819949845862498,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20538970232009887,
      "num_tokens": 31144806.0,
      "step": 15005
    },
    {
      "entropy": 5.517314434051514,
      "epoch": 1.3615747460087082,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004819822996780466,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20090858042240142,
      "num_tokens": 31155642.0,
      "step": 15010
    },
    {
      "entropy": 5.508015060424805,
      "epoch": 1.3620283018867925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004819696104893967,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20463416576385499,
      "num_tokens": 31165034.0,
      "step": 15015
    },
    {
      "entropy": 5.478755807876587,
      "epoch": 1.3624818577648767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004819569170205625,
      "loss": 5.0517,
      "mean_token_accuracy": 0.19870482832193376,
      "num_tokens": 31176491.0,
      "step": 15020
    },
    {
      "entropy": 5.47834587097168,
      "epoch": 1.3629354136429608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048194421927180654,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19554078727960586,
      "num_tokens": 31186198.0,
      "step": 15025
    },
    {
      "entropy": 5.491092252731323,
      "epoch": 1.363388969521045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004819315172433914,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2019846871495247,
      "num_tokens": 31197094.0,
      "step": 15030
    },
    {
      "entropy": 5.501713037490845,
      "epoch": 1.363842525399129,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048191881093557977,
      "loss": 5.0588,
      "mean_token_accuracy": 0.19390265941619872,
      "num_tokens": 31207505.0,
      "step": 15035
    },
    {
      "entropy": 5.543484306335449,
      "epoch": 1.3642960812772134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048190610034863434,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1883800745010376,
      "num_tokens": 31218809.0,
      "step": 15040
    },
    {
      "entropy": 5.492035722732544,
      "epoch": 1.3647496371552976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004818933854828181,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19075336456298828,
      "num_tokens": 31230281.0,
      "step": 15045
    },
    {
      "entropy": 5.417478942871094,
      "epoch": 1.3652031930333817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004818806663383939,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20595351308584214,
      "num_tokens": 31239573.0,
      "step": 15050
    },
    {
      "entropy": 5.439194059371948,
      "epoch": 1.365656748911466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004818679429156248,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1995464190840721,
      "num_tokens": 31249671.0,
      "step": 15055
    },
    {
      "entropy": 5.4043097496032715,
      "epoch": 1.36611030478955,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048185521521477385,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19515294283628465,
      "num_tokens": 31259601.0,
      "step": 15060
    },
    {
      "entropy": 5.423627138137817,
      "epoch": 1.3665638606676342,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818424832361044,
      "loss": 5.0211,
      "mean_token_accuracy": 0.20258173048496247,
      "num_tokens": 31269085.0,
      "step": 15065
    },
    {
      "entropy": 5.443383264541626,
      "epoch": 1.3670174165457185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004818297469798797,
      "loss": 5.033,
      "mean_token_accuracy": 0.19837417751550673,
      "num_tokens": 31280177.0,
      "step": 15070
    },
    {
      "entropy": 5.475092983245849,
      "epoch": 1.3674709724238026,
      "grad_norm": 1.0,
      "learning_rate": 0.00048181700644636314,
      "loss": 4.959,
      "mean_token_accuracy": 0.20413506925106048,
      "num_tokens": 31290545.0,
      "step": 15075
    },
    {
      "entropy": 5.376138925552368,
      "epoch": 1.3679245283018868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048180426163581816,
      "loss": 4.8884,
      "mean_token_accuracy": 0.20839907675981523,
      "num_tokens": 31301143.0,
      "step": 15080
    },
    {
      "entropy": 5.407545137405395,
      "epoch": 1.3683780841799709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004817915125485083,
      "loss": 4.9994,
      "mean_token_accuracy": 0.19968184381723403,
      "num_tokens": 31311071.0,
      "step": 15085
    },
    {
      "entropy": 5.510735940933228,
      "epoch": 1.3688316400580551,
      "grad_norm": 1.0,
      "learning_rate": 0.00048177875918469735,
      "loss": 5.0838,
      "mean_token_accuracy": 0.2015668496489525,
      "num_tokens": 31322263.0,
      "step": 15090
    },
    {
      "entropy": 5.480931663513184,
      "epoch": 1.3692851959361394,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048176600154464897,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20168393552303315,
      "num_tokens": 31332975.0,
      "step": 15095
    },
    {
      "entropy": 5.530743408203125,
      "epoch": 1.3697387518142234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004817532396286269,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1979136660695076,
      "num_tokens": 31345252.0,
      "step": 15100
    },
    {
      "entropy": 5.473556995391846,
      "epoch": 1.3701923076923077,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004817404734368952,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20655723959207534,
      "num_tokens": 31355479.0,
      "step": 15105
    },
    {
      "entropy": 5.521392488479615,
      "epoch": 1.3706458635703918,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048172770296971785,
      "loss": 5.122,
      "mean_token_accuracy": 0.19433184564113617,
      "num_tokens": 31366020.0,
      "step": 15110
    },
    {
      "entropy": 5.481007862091064,
      "epoch": 1.371099419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004817149282273589,
      "loss": 5.045,
      "mean_token_accuracy": 0.19936117976903917,
      "num_tokens": 31376055.0,
      "step": 15115
    },
    {
      "entropy": 5.4966240406036375,
      "epoch": 1.3715529753265603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004817021492100825,
      "loss": 5.1654,
      "mean_token_accuracy": 0.19234600365161897,
      "num_tokens": 31386179.0,
      "step": 15120
    },
    {
      "entropy": 5.415863752365112,
      "epoch": 1.3720065312046443,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000481689365918153,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20340771079063416,
      "num_tokens": 31395850.0,
      "step": 15125
    },
    {
      "entropy": 5.457096195220947,
      "epoch": 1.3724600870827286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816765783518348,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21304902583360671,
      "num_tokens": 31405053.0,
      "step": 15130
    },
    {
      "entropy": 5.412125158309936,
      "epoch": 1.3729136429608126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004816637865113923,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21003830283880234,
      "num_tokens": 31415153.0,
      "step": 15135
    },
    {
      "entropy": 5.4197296619415285,
      "epoch": 1.373367198838897,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048165099039708994,
      "loss": 5.0389,
      "mean_token_accuracy": 0.199916335940361,
      "num_tokens": 31425451.0,
      "step": 15140
    },
    {
      "entropy": 5.41675443649292,
      "epoch": 1.3738207547169812,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004816381900091925,
      "loss": 5.0165,
      "mean_token_accuracy": 0.19917270839214324,
      "num_tokens": 31436298.0,
      "step": 15145
    },
    {
      "entropy": 5.454756546020508,
      "epoch": 1.3742743105950654,
      "grad_norm": 1.0,
      "learning_rate": 0.00048162538534796465,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20529780089855193,
      "num_tokens": 31446932.0,
      "step": 15150
    },
    {
      "entropy": 5.500065851211548,
      "epoch": 1.3747278664731495,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004816125764136711,
      "loss": 5.0792,
      "mean_token_accuracy": 0.19022306948900222,
      "num_tokens": 31457582.0,
      "step": 15155
    },
    {
      "entropy": 5.544494867324829,
      "epoch": 1.3751814223512338,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004815997632065769,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1995017260313034,
      "num_tokens": 31468168.0,
      "step": 15160
    },
    {
      "entropy": 5.470380640029907,
      "epoch": 1.3756349782293178,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048158694572694694,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2055066630244255,
      "num_tokens": 31478524.0,
      "step": 15165
    },
    {
      "entropy": 5.424317693710327,
      "epoch": 1.376088534107402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048157412397504623,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2064652994275093,
      "num_tokens": 31488608.0,
      "step": 15170
    },
    {
      "entropy": 5.419355297088623,
      "epoch": 1.3765420899854863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004815612979511401,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20743164122104646,
      "num_tokens": 31498984.0,
      "step": 15175
    },
    {
      "entropy": 5.518174743652343,
      "epoch": 1.3769956458635704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048154846765549353,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20261340588331223,
      "num_tokens": 31509110.0,
      "step": 15180
    },
    {
      "entropy": 5.468982219696045,
      "epoch": 1.3774492017416546,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048153563308837215,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20486426800489427,
      "num_tokens": 31519266.0,
      "step": 15185
    },
    {
      "entropy": 5.471644926071167,
      "epoch": 1.3779027576197387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048152279425004124,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20559466779232025,
      "num_tokens": 31529043.0,
      "step": 15190
    },
    {
      "entropy": 5.476575899124145,
      "epoch": 1.378356313497823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004815099511407663,
      "loss": 5.0181,
      "mean_token_accuracy": 0.20737977772951127,
      "num_tokens": 31538650.0,
      "step": 15195
    },
    {
      "entropy": 5.406746435165405,
      "epoch": 1.3788098693759072,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000481497103760813,
      "loss": 4.9612,
      "mean_token_accuracy": 0.20198217034339905,
      "num_tokens": 31549392.0,
      "step": 15200
    },
    {
      "entropy": 5.485914802551269,
      "epoch": 1.3792634252539913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048148425211044684,
      "loss": 5.1425,
      "mean_token_accuracy": 0.19214161485433578,
      "num_tokens": 31559999.0,
      "step": 15205
    },
    {
      "entropy": 5.496637058258057,
      "epoch": 1.3797169811320755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004814713961899338,
      "loss": 5.0311,
      "mean_token_accuracy": 0.19900211691856384,
      "num_tokens": 31570108.0,
      "step": 15210
    },
    {
      "entropy": 5.464858150482177,
      "epoch": 1.3801705370101596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048145853599953966,
      "loss": 5.0068,
      "mean_token_accuracy": 0.19859155416488647,
      "num_tokens": 31579698.0,
      "step": 15215
    },
    {
      "entropy": 5.431590795516968,
      "epoch": 1.3806240928882438,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048144567153953036,
      "loss": 5.0683,
      "mean_token_accuracy": 0.1957996219396591,
      "num_tokens": 31589292.0,
      "step": 15220
    },
    {
      "entropy": 5.504480648040771,
      "epoch": 1.381077648766328,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048143280281017197,
      "loss": 5.1295,
      "mean_token_accuracy": 0.19753700643777847,
      "num_tokens": 31599848.0,
      "step": 15225
    },
    {
      "entropy": 5.509224843978882,
      "epoch": 1.3815312046444121,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004814199298117307,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20320505499839783,
      "num_tokens": 31611109.0,
      "step": 15230
    },
    {
      "entropy": 5.423985385894776,
      "epoch": 1.3819847605224964,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048140705254447256,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20411137342453003,
      "num_tokens": 31621140.0,
      "step": 15235
    },
    {
      "entropy": 5.524167060852051,
      "epoch": 1.3824383164005805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048139417100866405,
      "loss": 5.0873,
      "mean_token_accuracy": 0.19728801548480987,
      "num_tokens": 31631268.0,
      "step": 15240
    },
    {
      "entropy": 5.503660774230957,
      "epoch": 1.3828918722786647,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048138128520457146,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2008575454354286,
      "num_tokens": 31640758.0,
      "step": 15245
    },
    {
      "entropy": 5.445882511138916,
      "epoch": 1.383345428156749,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004813683951324613,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20046461671590804,
      "num_tokens": 31651060.0,
      "step": 15250
    },
    {
      "entropy": 5.504006481170654,
      "epoch": 1.383798984034833,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048135550079260014,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19967885166406632,
      "num_tokens": 31662282.0,
      "step": 15255
    },
    {
      "entropy": 5.463409852981568,
      "epoch": 1.3842525399129173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004813426021852547,
      "loss": 5.0028,
      "mean_token_accuracy": 0.19269192218780518,
      "num_tokens": 31672364.0,
      "step": 15260
    },
    {
      "entropy": 5.475115585327148,
      "epoch": 1.3847060957910013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048132969931069156,
      "loss": 5.0618,
      "mean_token_accuracy": 0.19727637022733688,
      "num_tokens": 31683094.0,
      "step": 15265
    },
    {
      "entropy": 5.46751856803894,
      "epoch": 1.3851596516690856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004813167921691777,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20178418308496476,
      "num_tokens": 31693087.0,
      "step": 15270
    },
    {
      "entropy": 5.473338556289673,
      "epoch": 1.3856132075471699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048130388076098003,
      "loss": 5.02,
      "mean_token_accuracy": 0.19593427181243897,
      "num_tokens": 31703708.0,
      "step": 15275
    },
    {
      "entropy": 5.428715753555298,
      "epoch": 1.386066763425254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048129096508636547,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20823626965284348,
      "num_tokens": 31713640.0,
      "step": 15280
    },
    {
      "entropy": 5.467691898345947,
      "epoch": 1.3865203193033382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048127804514560126,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2034163922071457,
      "num_tokens": 31723526.0,
      "step": 15285
    },
    {
      "entropy": 5.414355897903443,
      "epoch": 1.3869738751814222,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048126512093895447,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20380796790122985,
      "num_tokens": 31733265.0,
      "step": 15290
    },
    {
      "entropy": 5.496120643615723,
      "epoch": 1.3874274310595065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004812521924666925,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2064056247472763,
      "num_tokens": 31743832.0,
      "step": 15295
    },
    {
      "entropy": 5.495451641082764,
      "epoch": 1.3878809869375908,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004812392597290825,
      "loss": 5.0604,
      "mean_token_accuracy": 0.209172622859478,
      "num_tokens": 31754443.0,
      "step": 15300
    },
    {
      "entropy": 5.372473287582397,
      "epoch": 1.3883345428156748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004812263227263921,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20174793899059296,
      "num_tokens": 31765304.0,
      "step": 15305
    },
    {
      "entropy": 5.432365036010742,
      "epoch": 1.388788098693759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812133814588888,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1972840517759323,
      "num_tokens": 31775809.0,
      "step": 15310
    },
    {
      "entropy": 5.494178009033203,
      "epoch": 1.3892416545718431,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812004359268402,
      "loss": 5.0189,
      "mean_token_accuracy": 0.20662343353033066,
      "num_tokens": 31786271.0,
      "step": 15315
    },
    {
      "entropy": 5.504286527633667,
      "epoch": 1.3896952104499274,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004811874861305141,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20229720920324326,
      "num_tokens": 31797358.0,
      "step": 15320
    },
    {
      "entropy": 5.4540262699127195,
      "epoch": 1.3901487663280117,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004811745320701782,
      "loss": 4.9942,
      "mean_token_accuracy": 0.21028562784194946,
      "num_tokens": 31808303.0,
      "step": 15325
    },
    {
      "entropy": 5.393069648742676,
      "epoch": 1.390602322206096,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004811615737461004,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20239959061145782,
      "num_tokens": 31818837.0,
      "step": 15330
    },
    {
      "entropy": 5.495754814147949,
      "epoch": 1.39105587808418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048114861115854877,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20156844109296798,
      "num_tokens": 31829729.0,
      "step": 15335
    },
    {
      "entropy": 5.44893627166748,
      "epoch": 1.3915094339622642,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048113564430779136,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20025082528591157,
      "num_tokens": 31839996.0,
      "step": 15340
    },
    {
      "entropy": 5.3745111465454105,
      "epoch": 1.3919629898403483,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004811226731940962,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2046112149953842,
      "num_tokens": 31850298.0,
      "step": 15345
    },
    {
      "entropy": 5.4414092063903805,
      "epoch": 1.3924165457184325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048110969781773173,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20993245840072633,
      "num_tokens": 31862170.0,
      "step": 15350
    },
    {
      "entropy": 5.513597440719605,
      "epoch": 1.3928701015965168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048109671817896614,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20092002004384996,
      "num_tokens": 31872081.0,
      "step": 15355
    },
    {
      "entropy": 5.440402698516846,
      "epoch": 1.3933236574746009,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048108373427806797,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20320018082857133,
      "num_tokens": 31882715.0,
      "step": 15360
    },
    {
      "entropy": 5.491623115539551,
      "epoch": 1.3937772133526851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048107074611530555,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20696883350610734,
      "num_tokens": 31891925.0,
      "step": 15365
    },
    {
      "entropy": 5.507652521133423,
      "epoch": 1.3942307692307692,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048105775369094756,
      "loss": 5.027,
      "mean_token_accuracy": 0.20350237488746642,
      "num_tokens": 31902320.0,
      "step": 15370
    },
    {
      "entropy": 5.429424095153808,
      "epoch": 1.3946843251088534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810447570052628,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2057958498597145,
      "num_tokens": 31912726.0,
      "step": 15375
    },
    {
      "entropy": 5.456639051437378,
      "epoch": 1.3951378809869377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048103175605851997,
      "loss": 5.0565,
      "mean_token_accuracy": 0.199539977312088,
      "num_tokens": 31924039.0,
      "step": 15380
    },
    {
      "entropy": 5.400138235092163,
      "epoch": 1.3955914368650217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004810187508509879,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2129800632596016,
      "num_tokens": 31933280.0,
      "step": 15385
    },
    {
      "entropy": 5.456555652618408,
      "epoch": 1.396044992743106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004810057413829355,
      "loss": 5.139,
      "mean_token_accuracy": 0.1994667336344719,
      "num_tokens": 31943182.0,
      "step": 15390
    },
    {
      "entropy": 5.525270462036133,
      "epoch": 1.39649854862119,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048099272765463193,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2036070257425308,
      "num_tokens": 31953256.0,
      "step": 15395
    },
    {
      "entropy": 5.555541372299194,
      "epoch": 1.3969521044992743,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004809797096663463,
      "loss": 5.1521,
      "mean_token_accuracy": 0.19529494643211365,
      "num_tokens": 31964231.0,
      "step": 15400
    },
    {
      "entropy": 5.475796127319336,
      "epoch": 1.3974056603773586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004809666874183477,
      "loss": 5.006,
      "mean_token_accuracy": 0.20255526006221772,
      "num_tokens": 31974322.0,
      "step": 15405
    },
    {
      "entropy": 5.4569597244262695,
      "epoch": 1.3978592162554426,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004809536609109056,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20720462054014205,
      "num_tokens": 31984239.0,
      "step": 15410
    },
    {
      "entropy": 5.445779752731323,
      "epoch": 1.398312772133527,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048094063014428925,
      "loss": 5.0569,
      "mean_token_accuracy": 0.19856468290090562,
      "num_tokens": 31994366.0,
      "step": 15415
    },
    {
      "entropy": 5.515515851974487,
      "epoch": 1.398766328011611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809275951187682,
      "loss": 5.1372,
      "mean_token_accuracy": 0.198408405482769,
      "num_tokens": 32004921.0,
      "step": 15420
    },
    {
      "entropy": 5.443058919906616,
      "epoch": 1.3992198838896952,
      "grad_norm": 1.0,
      "learning_rate": 0.00048091455583461204,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20184063017368317,
      "num_tokens": 32016385.0,
      "step": 15425
    },
    {
      "entropy": 5.46964020729065,
      "epoch": 1.3996734397677795,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048090151229209036,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19756338894367217,
      "num_tokens": 32027398.0,
      "step": 15430
    },
    {
      "entropy": 5.508984088897705,
      "epoch": 1.4001269956458635,
      "grad_norm": 1.0,
      "learning_rate": 0.0004808884644914729,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20654245913028718,
      "num_tokens": 32038326.0,
      "step": 15435
    },
    {
      "entropy": 5.473737478256226,
      "epoch": 1.4005805515239478,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004808754124330296,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19969897866249084,
      "num_tokens": 32048744.0,
      "step": 15440
    },
    {
      "entropy": 5.478224563598633,
      "epoch": 1.4010341074020318,
      "grad_norm": 1.125,
      "learning_rate": 0.0004808623561170302,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21029338836669922,
      "num_tokens": 32059133.0,
      "step": 15445
    },
    {
      "entropy": 5.460933017730713,
      "epoch": 1.401487663280116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808492955437449,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2009618729352951,
      "num_tokens": 32069277.0,
      "step": 15450
    },
    {
      "entropy": 5.456944704055786,
      "epoch": 1.4019412191582004,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048083623071344374,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2077607974410057,
      "num_tokens": 32079802.0,
      "step": 15455
    },
    {
      "entropy": 5.473855209350586,
      "epoch": 1.4023947750362844,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048082316162639685,
      "loss": 5.0919,
      "mean_token_accuracy": 0.19972393661737442,
      "num_tokens": 32090962.0,
      "step": 15460
    },
    {
      "entropy": 5.462399005889893,
      "epoch": 1.4028483309143687,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048081008828287447,
      "loss": 5.0661,
      "mean_token_accuracy": 0.19965442270040512,
      "num_tokens": 32101058.0,
      "step": 15465
    },
    {
      "entropy": 5.5409893035888675,
      "epoch": 1.4033018867924527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048079701068314706,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20538726300001145,
      "num_tokens": 32111763.0,
      "step": 15470
    },
    {
      "entropy": 5.487426519393921,
      "epoch": 1.403755442670537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048078392882748507,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20131370276212693,
      "num_tokens": 32123556.0,
      "step": 15475
    },
    {
      "entropy": 5.452586889266968,
      "epoch": 1.4042089985486212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048077084271615895,
      "loss": 4.997,
      "mean_token_accuracy": 0.20365253686904908,
      "num_tokens": 32134012.0,
      "step": 15480
    },
    {
      "entropy": 5.4020476818084715,
      "epoch": 1.4046625544267053,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004807577523494394,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21200866550207137,
      "num_tokens": 32144540.0,
      "step": 15485
    },
    {
      "entropy": 5.4838134765625,
      "epoch": 1.4051161103047896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048074465772759707,
      "loss": 5.0833,
      "mean_token_accuracy": 0.20107725113630295,
      "num_tokens": 32154655.0,
      "step": 15490
    },
    {
      "entropy": 5.434706163406372,
      "epoch": 1.4055696661828736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048073155885090276,
      "loss": 4.9749,
      "mean_token_accuracy": 0.205027037858963,
      "num_tokens": 32165191.0,
      "step": 15495
    },
    {
      "entropy": 5.492273426055908,
      "epoch": 1.4060232220609579,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048071845571962733,
      "loss": 5.0889,
      "mean_token_accuracy": 0.1927748829126358,
      "num_tokens": 32175961.0,
      "step": 15500
    },
    {
      "entropy": 5.532474040985107,
      "epoch": 1.4064767779390421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004807053483340419,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19928376972675324,
      "num_tokens": 32187073.0,
      "step": 15505
    },
    {
      "entropy": 5.467940092086792,
      "epoch": 1.4069303338171264,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004806922366944175,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20543879121541977,
      "num_tokens": 32197426.0,
      "step": 15510
    },
    {
      "entropy": 5.532773065567016,
      "epoch": 1.4073838896952104,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048067912080102513,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20775387585163116,
      "num_tokens": 32206847.0,
      "step": 15515
    },
    {
      "entropy": 5.436613607406616,
      "epoch": 1.4078374455732947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004806660006541362,
      "loss": 5.0496,
      "mean_token_accuracy": 0.19538942873477935,
      "num_tokens": 32217647.0,
      "step": 15520
    },
    {
      "entropy": 5.389906501770019,
      "epoch": 1.4082910014513788,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004806528762540219,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21112460196018218,
      "num_tokens": 32227133.0,
      "step": 15525
    },
    {
      "entropy": 5.488668537139892,
      "epoch": 1.408744557329463,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048063974760095374,
      "loss": 5.0988,
      "mean_token_accuracy": 0.1972224622964859,
      "num_tokens": 32237378.0,
      "step": 15530
    },
    {
      "entropy": 5.419837760925293,
      "epoch": 1.4091981132075473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048062661469520327,
      "loss": 4.987,
      "mean_token_accuracy": 0.20672549307346344,
      "num_tokens": 32247213.0,
      "step": 15535
    },
    {
      "entropy": 5.521755170822144,
      "epoch": 1.4096516690856313,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048061347753704193,
      "loss": 5.065,
      "mean_token_accuracy": 0.20156003534793854,
      "num_tokens": 32257100.0,
      "step": 15540
    },
    {
      "entropy": 5.488410186767578,
      "epoch": 1.4101052249637156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004806003361267415,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1978863701224327,
      "num_tokens": 32267456.0,
      "step": 15545
    },
    {
      "entropy": 5.429026746749878,
      "epoch": 1.4105587808417996,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004805871904645737,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20904991030693054,
      "num_tokens": 32278220.0,
      "step": 15550
    },
    {
      "entropy": 5.416950416564942,
      "epoch": 1.411012336719884,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004805740405508103,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2059457629919052,
      "num_tokens": 32289036.0,
      "step": 15555
    },
    {
      "entropy": 5.425758647918701,
      "epoch": 1.4114658925979682,
      "grad_norm": 1.125,
      "learning_rate": 0.0004805608863857236,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20169477611780168,
      "num_tokens": 32298909.0,
      "step": 15560
    },
    {
      "entropy": 5.475577926635742,
      "epoch": 1.4119194484760522,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048054772796958517,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2035790354013443,
      "num_tokens": 32309851.0,
      "step": 15565
    },
    {
      "entropy": 5.4505894660949705,
      "epoch": 1.4123730043541365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004805345653026675,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2048426628112793,
      "num_tokens": 32319336.0,
      "step": 15570
    },
    {
      "entropy": 5.487655544281006,
      "epoch": 1.4128265602322205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048052139838524263,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20487714558839798,
      "num_tokens": 32329898.0,
      "step": 15575
    },
    {
      "entropy": 5.4495281219482425,
      "epoch": 1.4132801161103048,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048050822721758274,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2070409268140793,
      "num_tokens": 32340100.0,
      "step": 15580
    },
    {
      "entropy": 5.434144020080566,
      "epoch": 1.413733671988389,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004804950517999604,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2006583109498024,
      "num_tokens": 32349663.0,
      "step": 15585
    },
    {
      "entropy": 5.507922792434693,
      "epoch": 1.414187227866473,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004804818721326481,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20222700536251068,
      "num_tokens": 32360461.0,
      "step": 15590
    },
    {
      "entropy": 5.479337215423584,
      "epoch": 1.4146407837445574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004804686882159183,
      "loss": 5.077,
      "mean_token_accuracy": 0.2038666307926178,
      "num_tokens": 32370910.0,
      "step": 15595
    },
    {
      "entropy": 5.5347740173339846,
      "epoch": 1.4150943396226414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048045550005004355,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20874321311712266,
      "num_tokens": 32381669.0,
      "step": 15600
    },
    {
      "entropy": 5.436280012130737,
      "epoch": 1.4155478955007257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004804423076352968,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20037585347890854,
      "num_tokens": 32392024.0,
      "step": 15605
    },
    {
      "entropy": 5.368772029876709,
      "epoch": 1.41600145137881,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048042911097195075,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20821220874786378,
      "num_tokens": 32401872.0,
      "step": 15610
    },
    {
      "entropy": 5.402716445922851,
      "epoch": 1.416455007256894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004804159100602784,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2087838977575302,
      "num_tokens": 32411749.0,
      "step": 15615
    },
    {
      "entropy": 5.559806823730469,
      "epoch": 1.4169085631349783,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048040270490055264,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19089394211769103,
      "num_tokens": 32421358.0,
      "step": 15620
    },
    {
      "entropy": 5.496081209182739,
      "epoch": 1.4173621190130623,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004803894954930465,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20114330798387528,
      "num_tokens": 32432322.0,
      "step": 15625
    },
    {
      "entropy": 5.460557270050049,
      "epoch": 1.4178156748911466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004803762818380334,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19781472980976106,
      "num_tokens": 32440930.0,
      "step": 15630
    },
    {
      "entropy": 5.4865857601165775,
      "epoch": 1.4182692307692308,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048036306393578644,
      "loss": 5.0588,
      "mean_token_accuracy": 0.19708771705627443,
      "num_tokens": 32451124.0,
      "step": 15635
    },
    {
      "entropy": 5.523217868804932,
      "epoch": 1.4187227866473149,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004803498417865789,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20069751739501954,
      "num_tokens": 32460636.0,
      "step": 15640
    },
    {
      "entropy": 5.494079637527466,
      "epoch": 1.4191763425253991,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803366153906844,
      "loss": 5.078,
      "mean_token_accuracy": 0.20009166449308396,
      "num_tokens": 32470501.0,
      "step": 15645
    },
    {
      "entropy": 5.4987688064575195,
      "epoch": 1.4196298984034832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004803233847483763,
      "loss": 5.1395,
      "mean_token_accuracy": 0.19809164553880693,
      "num_tokens": 32481559.0,
      "step": 15650
    },
    {
      "entropy": 5.607738590240478,
      "epoch": 1.4200834542815675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048031014985992825,
      "loss": 5.2143,
      "mean_token_accuracy": 0.1941840723156929,
      "num_tokens": 32493000.0,
      "step": 15655
    },
    {
      "entropy": 5.491443681716919,
      "epoch": 1.4205370101596517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000480296910725614,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21452815383672713,
      "num_tokens": 32502773.0,
      "step": 15660
    },
    {
      "entropy": 5.449616861343384,
      "epoch": 1.4209905660377358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004802836673457074,
      "loss": 5.1041,
      "mean_token_accuracy": 0.19487740099430084,
      "num_tokens": 32513430.0,
      "step": 15665
    },
    {
      "entropy": 5.401146984100341,
      "epoch": 1.42144412191582,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048027041972048225,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20312780737876893,
      "num_tokens": 32524409.0,
      "step": 15670
    },
    {
      "entropy": 5.512161064147949,
      "epoch": 1.421897677793904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004802571678502124,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1949110746383667,
      "num_tokens": 32534696.0,
      "step": 15675
    },
    {
      "entropy": 5.559273672103882,
      "epoch": 1.4223512336719883,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000480243911735172,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20404788851737976,
      "num_tokens": 32544782.0,
      "step": 15680
    },
    {
      "entropy": 5.447524356842041,
      "epoch": 1.4228047895500726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048023065137563534,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19569372087717057,
      "num_tokens": 32554818.0,
      "step": 15685
    },
    {
      "entropy": 5.436730623245239,
      "epoch": 1.4232583454281569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004802173867718764,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20380208194255828,
      "num_tokens": 32565007.0,
      "step": 15690
    },
    {
      "entropy": 5.4994275093078615,
      "epoch": 1.423711901306241,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004802041179241697,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20204194337129594,
      "num_tokens": 32575657.0,
      "step": 15695
    },
    {
      "entropy": 5.461858129501342,
      "epoch": 1.4241654571843252,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004801908448327895,
      "loss": 5.0609,
      "mean_token_accuracy": 0.19369821548461913,
      "num_tokens": 32586676.0,
      "step": 15700
    },
    {
      "entropy": 5.466211318969727,
      "epoch": 1.4246190130624092,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004801775674980103,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20252088010311126,
      "num_tokens": 32596191.0,
      "step": 15705
    },
    {
      "entropy": 5.533839464187622,
      "epoch": 1.4250725689404935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048016428592010676,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19708123356103896,
      "num_tokens": 32607178.0,
      "step": 15710
    },
    {
      "entropy": 5.522767639160156,
      "epoch": 1.4255261248185778,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048015100009935347,
      "loss": 5.0168,
      "mean_token_accuracy": 0.20005394518375397,
      "num_tokens": 32617301.0,
      "step": 15715
    },
    {
      "entropy": 5.391142988204956,
      "epoch": 1.4259796806966618,
      "grad_norm": 1.0,
      "learning_rate": 0.0004801377100360253,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2088123455643654,
      "num_tokens": 32627395.0,
      "step": 15720
    },
    {
      "entropy": 5.446273422241211,
      "epoch": 1.426433236574746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000480124415730397,
      "loss": 5.016,
      "mean_token_accuracy": 0.20472394526004792,
      "num_tokens": 32638460.0,
      "step": 15725
    },
    {
      "entropy": 5.551666641235352,
      "epoch": 1.4268867924528301,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801111171827434,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19856278747320175,
      "num_tokens": 32649319.0,
      "step": 15730
    },
    {
      "entropy": 5.487170839309693,
      "epoch": 1.4273403483309144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800978143933398,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2052985429763794,
      "num_tokens": 32659976.0,
      "step": 15735
    },
    {
      "entropy": 5.418977212905884,
      "epoch": 1.4277939042089987,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048008450736246094,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2020311713218689,
      "num_tokens": 32670772.0,
      "step": 15740
    },
    {
      "entropy": 5.513687801361084,
      "epoch": 1.4282474600870827,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048007119609038237,
      "loss": 5.0967,
      "mean_token_accuracy": 0.19420246332883834,
      "num_tokens": 32682368.0,
      "step": 15745
    },
    {
      "entropy": 5.397270441055298,
      "epoch": 1.428701015965167,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048005788057737907,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2074650853872299,
      "num_tokens": 32692649.0,
      "step": 15750
    },
    {
      "entropy": 5.482013940811157,
      "epoch": 1.429154571843251,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004800445608237266,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19943402111530303,
      "num_tokens": 32704075.0,
      "step": 15755
    },
    {
      "entropy": 5.5202741622924805,
      "epoch": 1.4296081277213353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048003123682970045,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19761826246976852,
      "num_tokens": 32714670.0,
      "step": 15760
    },
    {
      "entropy": 5.463421058654785,
      "epoch": 1.4300616835994195,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048001790859557596,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20732996463775635,
      "num_tokens": 32724601.0,
      "step": 15765
    },
    {
      "entropy": 5.4143373489379885,
      "epoch": 1.4305152394775036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004800045761216289,
      "loss": 4.9436,
      "mean_token_accuracy": 0.20634355694055556,
      "num_tokens": 32735015.0,
      "step": 15770
    },
    {
      "entropy": 5.4462672710418705,
      "epoch": 1.4309687953555879,
      "grad_norm": 1.03125,
      "learning_rate": 0.000479991239408135,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2087372660636902,
      "num_tokens": 32745899.0,
      "step": 15775
    },
    {
      "entropy": 5.463490533828735,
      "epoch": 1.431422351233672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047997789845537003,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2086031198501587,
      "num_tokens": 32756052.0,
      "step": 15780
    },
    {
      "entropy": 5.416792964935302,
      "epoch": 1.4318759071117562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004799645532636098,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20779118537902833,
      "num_tokens": 32767118.0,
      "step": 15785
    },
    {
      "entropy": 5.390763139724731,
      "epoch": 1.4323294629898404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047995120383313044,
      "loss": 4.8893,
      "mean_token_accuracy": 0.214095738530159,
      "num_tokens": 32777069.0,
      "step": 15790
    },
    {
      "entropy": 5.460057067871094,
      "epoch": 1.4327830188679245,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047993785016420796,
      "loss": 5.1078,
      "mean_token_accuracy": 0.19337689578533174,
      "num_tokens": 32786967.0,
      "step": 15795
    },
    {
      "entropy": 5.468468284606933,
      "epoch": 1.4332365747460087,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047992449225711854,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20572699904441832,
      "num_tokens": 32797563.0,
      "step": 15800
    },
    {
      "entropy": 5.476965618133545,
      "epoch": 1.4336901306240928,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047991113011213837,
      "loss": 4.9556,
      "mean_token_accuracy": 0.20637069642543793,
      "num_tokens": 32808730.0,
      "step": 15805
    },
    {
      "entropy": 5.433117055892945,
      "epoch": 1.434143686502177,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004798977637295438,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20631701201200486,
      "num_tokens": 32818050.0,
      "step": 15810
    },
    {
      "entropy": 5.427420234680175,
      "epoch": 1.4345972423802613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798843931096113,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20257741659879686,
      "num_tokens": 32828993.0,
      "step": 15815
    },
    {
      "entropy": 5.428350114822388,
      "epoch": 1.4350507982583454,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004798710182526173,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2052305147051811,
      "num_tokens": 32839331.0,
      "step": 15820
    },
    {
      "entropy": 5.46976866722107,
      "epoch": 1.4355043541364296,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047985763915883847,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20204367786645888,
      "num_tokens": 32849946.0,
      "step": 15825
    },
    {
      "entropy": 5.393970966339111,
      "epoch": 1.4359579100145137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798442558285514,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20497775375843047,
      "num_tokens": 32859960.0,
      "step": 15830
    },
    {
      "entropy": 5.424088191986084,
      "epoch": 1.436411465892598,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004798308682620329,
      "loss": 4.9742,
      "mean_token_accuracy": 0.19887613654136657,
      "num_tokens": 32870317.0,
      "step": 15835
    },
    {
      "entropy": 5.53906774520874,
      "epoch": 1.4368650217706822,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047981747645955996,
      "loss": 5.151,
      "mean_token_accuracy": 0.19574028700590135,
      "num_tokens": 32880906.0,
      "step": 15840
    },
    {
      "entropy": 5.56572732925415,
      "epoch": 1.4373185776487662,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004798040804214094,
      "loss": 5.1879,
      "mean_token_accuracy": 0.1897825300693512,
      "num_tokens": 32891038.0,
      "step": 15845
    },
    {
      "entropy": 5.569432735443115,
      "epoch": 1.4377721335268505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797906801478582,
      "loss": 5.1249,
      "mean_token_accuracy": 0.2030026987195015,
      "num_tokens": 32901013.0,
      "step": 15850
    },
    {
      "entropy": 5.516676425933838,
      "epoch": 1.4382256894049346,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004797772756391836,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20463033020496368,
      "num_tokens": 32911592.0,
      "step": 15855
    },
    {
      "entropy": 5.423782253265381,
      "epoch": 1.4386792452830188,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797638668956627,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20585375279188156,
      "num_tokens": 32922087.0,
      "step": 15860
    },
    {
      "entropy": 5.45404200553894,
      "epoch": 1.439132801161103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004797504539175729,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20368013083934783,
      "num_tokens": 32932039.0,
      "step": 15865
    },
    {
      "entropy": 5.473904275894165,
      "epoch": 1.4395863570391874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797370367051915,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20422232896089554,
      "num_tokens": 32941896.0,
      "step": 15870
    },
    {
      "entropy": 5.46373610496521,
      "epoch": 1.4400399129172714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000479723615258796,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1923507422208786,
      "num_tokens": 32951675.0,
      "step": 15875
    },
    {
      "entropy": 5.477837467193604,
      "epoch": 1.4404934687953557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047971018957866396,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20661236494779586,
      "num_tokens": 32961378.0,
      "step": 15880
    },
    {
      "entropy": 5.4550474166870115,
      "epoch": 1.4409470246734397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047969675966507314,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20333871394395828,
      "num_tokens": 32971770.0,
      "step": 15885
    },
    {
      "entropy": 5.438835620880127,
      "epoch": 1.441400580551524,
      "grad_norm": 1.0,
      "learning_rate": 0.000479683325518301,
      "loss": 5.002,
      "mean_token_accuracy": 0.2005586564540863,
      "num_tokens": 32982934.0,
      "step": 15890
    },
    {
      "entropy": 5.427374649047851,
      "epoch": 1.4418541364296082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004796698871386257,
      "loss": 5.0179,
      "mean_token_accuracy": 0.20118418335914612,
      "num_tokens": 32993252.0,
      "step": 15895
    },
    {
      "entropy": 5.490049886703491,
      "epoch": 1.4423076923076923,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004796564445263249,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20014355480670928,
      "num_tokens": 33003610.0,
      "step": 15900
    },
    {
      "entropy": 5.470461463928222,
      "epoch": 1.4427612481857766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047964299768167665,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20554691404104233,
      "num_tokens": 33014037.0,
      "step": 15905
    },
    {
      "entropy": 5.460197305679321,
      "epoch": 1.4432148040638606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004796295466049591,
      "loss": 5.0176,
      "mean_token_accuracy": 0.19628096073865892,
      "num_tokens": 33023299.0,
      "step": 15910
    },
    {
      "entropy": 5.513853693008423,
      "epoch": 1.4436683599419449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047961609129645036,
      "loss": 5.0764,
      "mean_token_accuracy": 0.19446345269680024,
      "num_tokens": 33034194.0,
      "step": 15915
    },
    {
      "entropy": 5.45906252861023,
      "epoch": 1.4441219158200291,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047960263175642867,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2084976240992546,
      "num_tokens": 33044715.0,
      "step": 15920
    },
    {
      "entropy": 5.386568164825439,
      "epoch": 1.4445754716981132,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795891679851724,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2160312220454216,
      "num_tokens": 33055214.0,
      "step": 15925
    },
    {
      "entropy": 5.455825710296631,
      "epoch": 1.4450290275761974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795756999829601,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20294891446828842,
      "num_tokens": 33065026.0,
      "step": 15930
    },
    {
      "entropy": 5.560147714614868,
      "epoch": 1.4454825834542815,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047956222775007016,
      "loss": 5.0568,
      "mean_token_accuracy": 0.1987060323357582,
      "num_tokens": 33075138.0,
      "step": 15935
    },
    {
      "entropy": 5.486181306838989,
      "epoch": 1.4459361393323658,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004795487512867812,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20502393692731857,
      "num_tokens": 33086625.0,
      "step": 15940
    },
    {
      "entropy": 5.515254735946655,
      "epoch": 1.44638969521045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047953527059337194,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19539907425642014,
      "num_tokens": 33097076.0,
      "step": 15945
    },
    {
      "entropy": 5.470223426818848,
      "epoch": 1.446843251088534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795217856701212,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20341595709323884,
      "num_tokens": 33107168.0,
      "step": 15950
    },
    {
      "entropy": 5.506333351135254,
      "epoch": 1.4472968069666183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047950829651730774,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20027395188808442,
      "num_tokens": 33117862.0,
      "step": 15955
    },
    {
      "entropy": 5.489769983291626,
      "epoch": 1.4477503628447024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004794948031352107,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20904716104269028,
      "num_tokens": 33127144.0,
      "step": 15960
    },
    {
      "entropy": 5.535410833358765,
      "epoch": 1.4482039187227866,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047948130552410897,
      "loss": 5.1125,
      "mean_token_accuracy": 0.19545748680830002,
      "num_tokens": 33139180.0,
      "step": 15965
    },
    {
      "entropy": 5.5159039974212645,
      "epoch": 1.448657474600871,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047946780368428175,
      "loss": 5.0065,
      "mean_token_accuracy": 0.19653345495462418,
      "num_tokens": 33149467.0,
      "step": 15970
    },
    {
      "entropy": 5.4328114032745365,
      "epoch": 1.449111030478955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004794542976160081,
      "loss": 4.8734,
      "mean_token_accuracy": 0.21548738479614257,
      "num_tokens": 33158752.0,
      "step": 15975
    },
    {
      "entropy": 5.412974882125854,
      "epoch": 1.4495645863570392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047944078731956765,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20715339183807374,
      "num_tokens": 33168242.0,
      "step": 15980
    },
    {
      "entropy": 5.409467077255249,
      "epoch": 1.4500181422351233,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047942727279523957,
      "loss": 4.9854,
      "mean_token_accuracy": 0.210733562707901,
      "num_tokens": 33179134.0,
      "step": 15985
    },
    {
      "entropy": 5.485414934158325,
      "epoch": 1.4504716981132075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794137540433034,
      "loss": 5.1127,
      "mean_token_accuracy": 0.1984705865383148,
      "num_tokens": 33188388.0,
      "step": 15990
    },
    {
      "entropy": 5.519913148880005,
      "epoch": 1.4509252539912918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004794002310640386,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2034293457865715,
      "num_tokens": 33198750.0,
      "step": 15995
    },
    {
      "entropy": 5.571501922607422,
      "epoch": 1.4513788098693758,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047938670385772503,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1909438416361809,
      "num_tokens": 33209532.0,
      "step": 16000
    },
    {
      "entropy": 5.500415992736817,
      "epoch": 1.45183236574746,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047937317242464244,
      "loss": 4.977,
      "mean_token_accuracy": 0.2058201715350151,
      "num_tokens": 33219675.0,
      "step": 16005
    },
    {
      "entropy": 5.509267568588257,
      "epoch": 1.4522859216255442,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793596367650704,
      "loss": 5.0847,
      "mean_token_accuracy": 0.19889651238918304,
      "num_tokens": 33230919.0,
      "step": 16010
    },
    {
      "entropy": 5.489873170852661,
      "epoch": 1.4527394775036284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004793460968792891,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20786553621292114,
      "num_tokens": 33241146.0,
      "step": 16015
    },
    {
      "entropy": 5.497219038009644,
      "epoch": 1.4531930333817127,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047933255276757836,
      "loss": 5.0404,
      "mean_token_accuracy": 0.19959067404270173,
      "num_tokens": 33251434.0,
      "step": 16020
    },
    {
      "entropy": 5.525360250473023,
      "epoch": 1.4536465892597967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004793190044302184,
      "loss": 5.1201,
      "mean_token_accuracy": 0.1891581580042839,
      "num_tokens": 33261868.0,
      "step": 16025
    },
    {
      "entropy": 5.469266128540039,
      "epoch": 1.454100145137881,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047930545186748936,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20658151358366011,
      "num_tokens": 33271798.0,
      "step": 16030
    },
    {
      "entropy": 5.464654397964478,
      "epoch": 1.454553701015965,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047929189507967153,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19962890148162843,
      "num_tokens": 33282614.0,
      "step": 16035
    },
    {
      "entropy": 5.420350074768066,
      "epoch": 1.4550072568940493,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047927833406704516,
      "loss": 4.975,
      "mean_token_accuracy": 0.20225861072540283,
      "num_tokens": 33292906.0,
      "step": 16040
    },
    {
      "entropy": 5.457213354110718,
      "epoch": 1.4554608127721336,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004792647688298909,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20239474773406982,
      "num_tokens": 33302720.0,
      "step": 16045
    },
    {
      "entropy": 5.417613506317139,
      "epoch": 1.4559143686502178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004792511993684891,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20726395696401595,
      "num_tokens": 33312959.0,
      "step": 16050
    },
    {
      "entropy": 5.456071567535401,
      "epoch": 1.4563679245283019,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047923762568312053,
      "loss": 4.923,
      "mean_token_accuracy": 0.20960447639226915,
      "num_tokens": 33322737.0,
      "step": 16055
    },
    {
      "entropy": 5.432339668273926,
      "epoch": 1.4568214804063861,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792240477740657,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2113322213292122,
      "num_tokens": 33333185.0,
      "step": 16060
    },
    {
      "entropy": 5.486125040054321,
      "epoch": 1.4572750362844702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004792104656416056,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19482650607824326,
      "num_tokens": 33343414.0,
      "step": 16065
    },
    {
      "entropy": 5.483291673660278,
      "epoch": 1.4577285921625545,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047919687928602104,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20450005680322647,
      "num_tokens": 33352527.0,
      "step": 16070
    },
    {
      "entropy": 5.509945487976074,
      "epoch": 1.4581821480406387,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047918328870759293,
      "loss": 5.0535,
      "mean_token_accuracy": 0.19920112788677216,
      "num_tokens": 33363526.0,
      "step": 16075
    },
    {
      "entropy": 5.483455085754395,
      "epoch": 1.4586357039187228,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004791696939066024,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19908895790576936,
      "num_tokens": 33374302.0,
      "step": 16080
    },
    {
      "entropy": 5.4868992328643795,
      "epoch": 1.459089259796807,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004791560948833306,
      "loss": 5.0481,
      "mean_token_accuracy": 0.20238014906644822,
      "num_tokens": 33384786.0,
      "step": 16085
    },
    {
      "entropy": 5.4957318782806395,
      "epoch": 1.459542815674891,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791424916380586,
      "loss": 5.0257,
      "mean_token_accuracy": 0.20975643545389175,
      "num_tokens": 33394464.0,
      "step": 16090
    },
    {
      "entropy": 5.490728235244751,
      "epoch": 1.4599963715529753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000479128884171068,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20181618332862855,
      "num_tokens": 33405476.0,
      "step": 16095
    },
    {
      "entropy": 5.509813213348389,
      "epoch": 1.4604499274310596,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047911527248264003,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2008212089538574,
      "num_tokens": 33415758.0,
      "step": 16100
    },
    {
      "entropy": 5.475230264663696,
      "epoch": 1.4609034833091437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047910165657305613,
      "loss": 5.0385,
      "mean_token_accuracy": 0.2024817645549774,
      "num_tokens": 33424845.0,
      "step": 16105
    },
    {
      "entropy": 5.4918581485748295,
      "epoch": 1.461357039187228,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047908803644259804,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20061122477054597,
      "num_tokens": 33435928.0,
      "step": 16110
    },
    {
      "entropy": 5.49398775100708,
      "epoch": 1.461810595065312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047907441209154726,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20125750750303267,
      "num_tokens": 33447307.0,
      "step": 16115
    },
    {
      "entropy": 5.486487150192261,
      "epoch": 1.4622641509433962,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047906078352018575,
      "loss": 5.0436,
      "mean_token_accuracy": 0.1945362851023674,
      "num_tokens": 33458116.0,
      "step": 16120
    },
    {
      "entropy": 5.35523943901062,
      "epoch": 1.4627177068214805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004790471507287952,
      "loss": 4.8962,
      "mean_token_accuracy": 0.20788645893335342,
      "num_tokens": 33469065.0,
      "step": 16125
    },
    {
      "entropy": 5.47483263015747,
      "epoch": 1.4631712626995645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047903351371765757,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2058606430888176,
      "num_tokens": 33479118.0,
      "step": 16130
    },
    {
      "entropy": 5.518635892868042,
      "epoch": 1.4636248185776488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047901987248705484,
      "loss": 5.0466,
      "mean_token_accuracy": 0.1991819068789482,
      "num_tokens": 33489642.0,
      "step": 16135
    },
    {
      "entropy": 5.476372289657593,
      "epoch": 1.4640783744557329,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004790062270372693,
      "loss": 5.0796,
      "mean_token_accuracy": 0.19231854677200316,
      "num_tokens": 33499771.0,
      "step": 16140
    },
    {
      "entropy": 5.460461521148682,
      "epoch": 1.4645319303338171,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004789925773685828,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20152659863233566,
      "num_tokens": 33509975.0,
      "step": 16145
    },
    {
      "entropy": 5.505796718597412,
      "epoch": 1.4649854862119014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789789234812779,
      "loss": 5.063,
      "mean_token_accuracy": 0.19863290190696717,
      "num_tokens": 33520017.0,
      "step": 16150
    },
    {
      "entropy": 5.512185573577881,
      "epoch": 1.4654390420899854,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047896526537563686,
      "loss": 5.1206,
      "mean_token_accuracy": 0.19162429571151735,
      "num_tokens": 33531977.0,
      "step": 16155
    },
    {
      "entropy": 5.497498464584351,
      "epoch": 1.4658925979680697,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047895160305194223,
      "loss": 5.0822,
      "mean_token_accuracy": 0.1961600065231323,
      "num_tokens": 33543660.0,
      "step": 16160
    },
    {
      "entropy": 5.485444498062134,
      "epoch": 1.4663461538461537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789379365104764,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2038754180073738,
      "num_tokens": 33553814.0,
      "step": 16165
    },
    {
      "entropy": 5.4929135799407955,
      "epoch": 1.466799709724238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004789242657515221,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20848118215799333,
      "num_tokens": 33564250.0,
      "step": 16170
    },
    {
      "entropy": 5.481130838394165,
      "epoch": 1.4672532656023223,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004789105907753621,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2035487875342369,
      "num_tokens": 33574755.0,
      "step": 16175
    },
    {
      "entropy": 5.451443719863891,
      "epoch": 1.4677068214804063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000478896911582279,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20966747105121614,
      "num_tokens": 33585268.0,
      "step": 16180
    },
    {
      "entropy": 5.468293523788452,
      "epoch": 1.4681603773584906,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788832281725558,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20301875174045564,
      "num_tokens": 33595753.0,
      "step": 16185
    },
    {
      "entropy": 5.443698072433472,
      "epoch": 1.4686139332365746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004788695405464755,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20774186849594117,
      "num_tokens": 33605270.0,
      "step": 16190
    },
    {
      "entropy": 5.50835337638855,
      "epoch": 1.469067489114659,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047885584870432127,
      "loss": 5.056,
      "mean_token_accuracy": 0.20199033319950105,
      "num_tokens": 33615096.0,
      "step": 16195
    },
    {
      "entropy": 5.489577198028565,
      "epoch": 1.4695210449927432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047884215264637606,
      "loss": 4.956,
      "mean_token_accuracy": 0.20772962123155594,
      "num_tokens": 33624770.0,
      "step": 16200
    },
    {
      "entropy": 5.520768451690674,
      "epoch": 1.4699746008708272,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004788284523729232,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1957830160856247,
      "num_tokens": 33634545.0,
      "step": 16205
    },
    {
      "entropy": 5.526898336410523,
      "epoch": 1.4704281567489115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047881474788424595,
      "loss": 5.0473,
      "mean_token_accuracy": 0.1975664019584656,
      "num_tokens": 33644950.0,
      "step": 16210
    },
    {
      "entropy": 5.510557317733765,
      "epoch": 1.4708817126269955,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047880103918062776,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20191543102264403,
      "num_tokens": 33654369.0,
      "step": 16215
    },
    {
      "entropy": 5.36762638092041,
      "epoch": 1.4713352685050798,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047878732626235225,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1993912026286125,
      "num_tokens": 33664687.0,
      "step": 16220
    },
    {
      "entropy": 5.4388457298278805,
      "epoch": 1.471788824383164,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004787736091297028,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20965267717838287,
      "num_tokens": 33675832.0,
      "step": 16225
    },
    {
      "entropy": 5.541381978988648,
      "epoch": 1.4722423802612483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004787598877829633,
      "loss": 5.0774,
      "mean_token_accuracy": 0.19990268498659133,
      "num_tokens": 33686788.0,
      "step": 16230
    },
    {
      "entropy": 5.459230518341064,
      "epoch": 1.4726959361393324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004787461622224173,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2072235971689224,
      "num_tokens": 33696384.0,
      "step": 16235
    },
    {
      "entropy": 5.405389881134033,
      "epoch": 1.4731494920174166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004787324324483488,
      "loss": 5.0039,
      "mean_token_accuracy": 0.19962379932403565,
      "num_tokens": 33707758.0,
      "step": 16240
    },
    {
      "entropy": 5.445700073242188,
      "epoch": 1.4736030478955007,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004787186984610416,
      "loss": 4.9762,
      "mean_token_accuracy": 0.20738447904586793,
      "num_tokens": 33717747.0,
      "step": 16245
    },
    {
      "entropy": 5.458881664276123,
      "epoch": 1.474056603773585,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047870496026077984,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20439188182353973,
      "num_tokens": 33727960.0,
      "step": 16250
    },
    {
      "entropy": 5.409978675842285,
      "epoch": 1.4745101596516692,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786912178478476,
      "loss": 4.9588,
      "mean_token_accuracy": 0.20595768839120865,
      "num_tokens": 33738874.0,
      "step": 16255
    },
    {
      "entropy": 5.426355361938477,
      "epoch": 1.4749637155297532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786774712225291,
      "loss": 4.961,
      "mean_token_accuracy": 0.2080053612589836,
      "num_tokens": 33749071.0,
      "step": 16260
    },
    {
      "entropy": 5.436343860626221,
      "epoch": 1.4754172714078375,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004786637203851086,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20291364938020706,
      "num_tokens": 33759509.0,
      "step": 16265
    },
    {
      "entropy": 5.503203725814819,
      "epoch": 1.4758708272859216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004786499653358705,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20802777409553527,
      "num_tokens": 33769974.0,
      "step": 16270
    },
    {
      "entropy": 5.541727781295776,
      "epoch": 1.4763243831640058,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004786362060750991,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20014467760920523,
      "num_tokens": 33780403.0,
      "step": 16275
    },
    {
      "entropy": 5.495248937606812,
      "epoch": 1.47677793904209,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047862244260307913,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20937622487545013,
      "num_tokens": 33791364.0,
      "step": 16280
    },
    {
      "entropy": 5.442668390274048,
      "epoch": 1.4772314949201741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786086749200951,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2074051097035408,
      "num_tokens": 33801964.0,
      "step": 16285
    },
    {
      "entropy": 5.502958011627197,
      "epoch": 1.4776850507982584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004785949030264319,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20473655313253403,
      "num_tokens": 33813132.0,
      "step": 16290
    },
    {
      "entropy": 5.482903337478637,
      "epoch": 1.4781386066763424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004785811269223741,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21504738479852675,
      "num_tokens": 33823069.0,
      "step": 16295
    },
    {
      "entropy": 5.526265478134155,
      "epoch": 1.4785921625544267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004785673466082068,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20451238751411438,
      "num_tokens": 33832626.0,
      "step": 16300
    },
    {
      "entropy": 5.423535013198853,
      "epoch": 1.479045718432511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047855356208421484,
      "loss": 4.9846,
      "mean_token_accuracy": 0.20531942695379257,
      "num_tokens": 33843384.0,
      "step": 16305
    },
    {
      "entropy": 5.421808290481567,
      "epoch": 1.479499274310595,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004785397733506834,
      "loss": 4.9899,
      "mean_token_accuracy": 0.20820764154195787,
      "num_tokens": 33854286.0,
      "step": 16310
    },
    {
      "entropy": 5.415575790405273,
      "epoch": 1.4799528301886793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004785259804078976,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20777007490396499,
      "num_tokens": 33863379.0,
      "step": 16315
    },
    {
      "entropy": 5.528962898254394,
      "epoch": 1.4804063860667633,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004785121832561426,
      "loss": 5.196,
      "mean_token_accuracy": 0.1923913463950157,
      "num_tokens": 33875641.0,
      "step": 16320
    },
    {
      "entropy": 5.500227355957032,
      "epoch": 1.4808599419448476,
      "grad_norm": 1.0,
      "learning_rate": 0.00047849838189570375,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2062392935156822,
      "num_tokens": 33885023.0,
      "step": 16325
    },
    {
      "entropy": 5.584210300445557,
      "epoch": 1.4813134978229319,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004784845763268666,
      "loss": 5.1143,
      "mean_token_accuracy": 0.19259014427661897,
      "num_tokens": 33895562.0,
      "step": 16330
    },
    {
      "entropy": 5.436470556259155,
      "epoch": 1.481767053701016,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004784707665499165,
      "loss": 5.0273,
      "mean_token_accuracy": 0.2102821946144104,
      "num_tokens": 33907095.0,
      "step": 16335
    },
    {
      "entropy": 5.387659215927124,
      "epoch": 1.4822206095791002,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004784569525651391,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2163789764046669,
      "num_tokens": 33917475.0,
      "step": 16340
    },
    {
      "entropy": 5.430476999282837,
      "epoch": 1.4826741654571842,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047844313437282005,
      "loss": 5.0141,
      "mean_token_accuracy": 0.2002805709838867,
      "num_tokens": 33927740.0,
      "step": 16345
    },
    {
      "entropy": 5.4267418384552,
      "epoch": 1.4831277213352685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047842931197324525,
      "loss": 4.952,
      "mean_token_accuracy": 0.20934444218873977,
      "num_tokens": 33937416.0,
      "step": 16350
    },
    {
      "entropy": 5.425552940368652,
      "epoch": 1.4835812772133528,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047841548536670033,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21422414779663085,
      "num_tokens": 33946968.0,
      "step": 16355
    },
    {
      "entropy": 5.5388099193573,
      "epoch": 1.4840348330914368,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047840165455347136,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19429281800985337,
      "num_tokens": 33957877.0,
      "step": 16360
    },
    {
      "entropy": 5.499372148513794,
      "epoch": 1.484488388969521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047838781953384436,
      "loss": 5.0504,
      "mean_token_accuracy": 0.19985658228397368,
      "num_tokens": 33967890.0,
      "step": 16365
    },
    {
      "entropy": 5.442298507690429,
      "epoch": 1.484941944847605,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783739803081054,
      "loss": 4.9927,
      "mean_token_accuracy": 0.20722893327474595,
      "num_tokens": 33977117.0,
      "step": 16370
    },
    {
      "entropy": 5.469358491897583,
      "epoch": 1.4853955007256894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047836013687654075,
      "loss": 5.0944,
      "mean_token_accuracy": 0.197372367978096,
      "num_tokens": 33986943.0,
      "step": 16375
    },
    {
      "entropy": 5.4732842445373535,
      "epoch": 1.4858490566037736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047834628923943654,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20302316546440125,
      "num_tokens": 33996499.0,
      "step": 16380
    },
    {
      "entropy": 5.5385230541229244,
      "epoch": 1.4863026124818577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004783324373970793,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20174136459827424,
      "num_tokens": 34006749.0,
      "step": 16385
    },
    {
      "entropy": 5.5679679870605465,
      "epoch": 1.486756168359942,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004783185813497555,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1920430764555931,
      "num_tokens": 34017194.0,
      "step": 16390
    },
    {
      "entropy": 5.541804218292237,
      "epoch": 1.487209724238026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004783047210977516,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2090506985783577,
      "num_tokens": 34027386.0,
      "step": 16395
    },
    {
      "entropy": 5.541867780685425,
      "epoch": 1.4876632801161103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047829085664135424,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19776039570569992,
      "num_tokens": 34037540.0,
      "step": 16400
    },
    {
      "entropy": 5.447138357162475,
      "epoch": 1.4881168359941945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047827698798085015,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20726717859506608,
      "num_tokens": 34048008.0,
      "step": 16405
    },
    {
      "entropy": 5.5058471202850345,
      "epoch": 1.4885703918722788,
      "grad_norm": 1.125,
      "learning_rate": 0.0004782631151165262,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20902367681264877,
      "num_tokens": 34058009.0,
      "step": 16410
    },
    {
      "entropy": 5.511841058731079,
      "epoch": 1.4890239477503628,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047824923804866914,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20172156542539596,
      "num_tokens": 34067625.0,
      "step": 16415
    },
    {
      "entropy": 5.414007949829101,
      "epoch": 1.4894775036284469,
      "grad_norm": 1.0,
      "learning_rate": 0.00047823535677756607,
      "loss": 5.0106,
      "mean_token_accuracy": 0.1978861317038536,
      "num_tokens": 34077349.0,
      "step": 16420
    },
    {
      "entropy": 5.532129621505737,
      "epoch": 1.4899310595065312,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004782214713035041,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19498357325792312,
      "num_tokens": 34088636.0,
      "step": 16425
    },
    {
      "entropy": 5.522089290618896,
      "epoch": 1.4903846153846154,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047820758162677026,
      "loss": 5.0411,
      "mean_token_accuracy": 0.20185174643993378,
      "num_tokens": 34098806.0,
      "step": 16430
    },
    {
      "entropy": 5.449494934082031,
      "epoch": 1.4908381712626997,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047819368774765187,
      "loss": 5.0152,
      "mean_token_accuracy": 0.20564057677984238,
      "num_tokens": 34109713.0,
      "step": 16435
    },
    {
      "entropy": 5.445164585113526,
      "epoch": 1.4912917271407837,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047817978966643626,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2078770637512207,
      "num_tokens": 34119585.0,
      "step": 16440
    },
    {
      "entropy": 5.544268798828125,
      "epoch": 1.491745283018868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004781658873834108,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19779797941446303,
      "num_tokens": 34131245.0,
      "step": 16445
    },
    {
      "entropy": 5.49971981048584,
      "epoch": 1.492198838896952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047815198089886296,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2045518785715103,
      "num_tokens": 34141229.0,
      "step": 16450
    },
    {
      "entropy": 5.47653374671936,
      "epoch": 1.4926523947750363,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004781380702130805,
      "loss": 5.0486,
      "mean_token_accuracy": 0.19726551175117493,
      "num_tokens": 34151781.0,
      "step": 16455
    },
    {
      "entropy": 5.417407560348511,
      "epoch": 1.4931059506531206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047812415532635087,
      "loss": 5.002,
      "mean_token_accuracy": 0.2048860549926758,
      "num_tokens": 34162305.0,
      "step": 16460
    },
    {
      "entropy": 5.498383903503418,
      "epoch": 1.4935595065312046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000478110236238962,
      "loss": 5.087,
      "mean_token_accuracy": 0.1978554055094719,
      "num_tokens": 34173613.0,
      "step": 16465
    },
    {
      "entropy": 5.472800397872925,
      "epoch": 1.4940130624092889,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047809631295120164,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20078982263803483,
      "num_tokens": 34183540.0,
      "step": 16470
    },
    {
      "entropy": 5.450933361053467,
      "epoch": 1.494466618287373,
      "grad_norm": 1.0,
      "learning_rate": 0.00047808238546335786,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20164911299943925,
      "num_tokens": 34193292.0,
      "step": 16475
    },
    {
      "entropy": 5.506211948394776,
      "epoch": 1.4949201741654572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047806845377571856,
      "loss": 5.07,
      "mean_token_accuracy": 0.20894212722778321,
      "num_tokens": 34202697.0,
      "step": 16480
    },
    {
      "entropy": 5.462220191955566,
      "epoch": 1.4953737300435415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047805451788857187,
      "loss": 4.965,
      "mean_token_accuracy": 0.20638453364372253,
      "num_tokens": 34213396.0,
      "step": 16485
    },
    {
      "entropy": 5.507080602645874,
      "epoch": 1.4958272859216255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000478040577802206,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2080722451210022,
      "num_tokens": 34223020.0,
      "step": 16490
    },
    {
      "entropy": 5.410426235198974,
      "epoch": 1.4962808417997098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004780266335169093,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20901964604854584,
      "num_tokens": 34233249.0,
      "step": 16495
    },
    {
      "entropy": 5.56065182685852,
      "epoch": 1.4967343976777938,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047801268503297,
      "loss": 5.0666,
      "mean_token_accuracy": 0.19952577948570252,
      "num_tokens": 34244091.0,
      "step": 16500
    },
    {
      "entropy": 5.526365327835083,
      "epoch": 1.497187953555878,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047799873235067667,
      "loss": 5.0661,
      "mean_token_accuracy": 0.20329954028129577,
      "num_tokens": 34255060.0,
      "step": 16505
    },
    {
      "entropy": 5.435563945770264,
      "epoch": 1.4976415094339623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779847754703178,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21159472316503525,
      "num_tokens": 34264361.0,
      "step": 16510
    },
    {
      "entropy": 5.440701484680176,
      "epoch": 1.4980950653120464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004779708143921821,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20453838855028153,
      "num_tokens": 34274480.0,
      "step": 16515
    },
    {
      "entropy": 5.460260534286499,
      "epoch": 1.4985486211901307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047795684911655824,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20553855001926422,
      "num_tokens": 34285951.0,
      "step": 16520
    },
    {
      "entropy": 5.507027626037598,
      "epoch": 1.4990021770682147,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477942879643735,
      "loss": 5.0732,
      "mean_token_accuracy": 0.1956965371966362,
      "num_tokens": 34296384.0,
      "step": 16525
    },
    {
      "entropy": 5.470954036712646,
      "epoch": 1.499455732946299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047792890597400125,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20878826379776,
      "num_tokens": 34306776.0,
      "step": 16530
    },
    {
      "entropy": 5.514407634735107,
      "epoch": 1.4999092888243832,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047791492810764604,
      "loss": 5.013,
      "mean_token_accuracy": 0.2054852917790413,
      "num_tokens": 34317047.0,
      "step": 16535
    },
    {
      "entropy": 5.449463796615601,
      "epoch": 1.5003628447024675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004779009460449584,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2083308905363083,
      "num_tokens": 34326392.0,
      "step": 16540
    },
    {
      "entropy": 5.503531122207642,
      "epoch": 1.5008164005805515,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047788695978622756,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21120216250419616,
      "num_tokens": 34336580.0,
      "step": 16545
    },
    {
      "entropy": 5.450557279586792,
      "epoch": 1.5012699564586356,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047787296933174265,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20305247008800506,
      "num_tokens": 34346792.0,
      "step": 16550
    },
    {
      "entropy": 5.5099790573120115,
      "epoch": 1.5017235123367199,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000477858974681793,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2033532068133354,
      "num_tokens": 34356675.0,
      "step": 16555
    },
    {
      "entropy": 5.521383571624756,
      "epoch": 1.5021770682148041,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778449758366681,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21447925865650178,
      "num_tokens": 34366357.0,
      "step": 16560
    },
    {
      "entropy": 5.44498348236084,
      "epoch": 1.5026306240928884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778309727966574,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20686540603637696,
      "num_tokens": 34377229.0,
      "step": 16565
    },
    {
      "entropy": 5.43284330368042,
      "epoch": 1.5030841799709724,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047781696556205045,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2040479376912117,
      "num_tokens": 34387873.0,
      "step": 16570
    },
    {
      "entropy": 5.49597134590149,
      "epoch": 1.5035377358490565,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047780295413313696,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20865806341171264,
      "num_tokens": 34398889.0,
      "step": 16575
    },
    {
      "entropy": 5.448739242553711,
      "epoch": 1.5039912917271407,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047778893851020676,
      "loss": 4.9186,
      "mean_token_accuracy": 0.20451200604438782,
      "num_tokens": 34408508.0,
      "step": 16580
    },
    {
      "entropy": 5.430432462692261,
      "epoch": 1.504444847605225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004777749186935496,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20057281851768494,
      "num_tokens": 34418928.0,
      "step": 16585
    },
    {
      "entropy": 5.500459241867065,
      "epoch": 1.5048984034833093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004777608946834555,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20214829593896866,
      "num_tokens": 34429789.0,
      "step": 16590
    },
    {
      "entropy": 5.56063404083252,
      "epoch": 1.5053519593613933,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004777468664802144,
      "loss": 5.1551,
      "mean_token_accuracy": 0.19416424483060837,
      "num_tokens": 34440795.0,
      "step": 16595
    },
    {
      "entropy": 5.541483449935913,
      "epoch": 1.5058055152394774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047773283408411636,
      "loss": 5.0441,
      "mean_token_accuracy": 0.19526468813419343,
      "num_tokens": 34450772.0,
      "step": 16600
    },
    {
      "entropy": 5.456420469284057,
      "epoch": 1.5062590711175616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047771879749545175,
      "loss": 5.003,
      "mean_token_accuracy": 0.20167505890130996,
      "num_tokens": 34461273.0,
      "step": 16605
    },
    {
      "entropy": 5.47718997001648,
      "epoch": 1.506712626995646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047770475671451073,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20287921130657197,
      "num_tokens": 34470942.0,
      "step": 16610
    },
    {
      "entropy": 5.515540313720703,
      "epoch": 1.5071661828737302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004776907117415837,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20255957096815108,
      "num_tokens": 34481117.0,
      "step": 16615
    },
    {
      "entropy": 5.4488598823547365,
      "epoch": 1.5076197387518142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047767666257696103,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20787885636091233,
      "num_tokens": 34490953.0,
      "step": 16620
    },
    {
      "entropy": 5.443657064437867,
      "epoch": 1.5080732946298983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047766260922093337,
      "loss": 4.916,
      "mean_token_accuracy": 0.205417500436306,
      "num_tokens": 34502765.0,
      "step": 16625
    },
    {
      "entropy": 5.56479024887085,
      "epoch": 1.5085268505079825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004776485516737913,
      "loss": 5.1003,
      "mean_token_accuracy": 0.19664451330900193,
      "num_tokens": 34514473.0,
      "step": 16630
    },
    {
      "entropy": 5.538726806640625,
      "epoch": 1.5089804063860668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004776344899358256,
      "loss": 5.1208,
      "mean_token_accuracy": 0.1967710256576538,
      "num_tokens": 34524218.0,
      "step": 16635
    },
    {
      "entropy": 5.490627384185791,
      "epoch": 1.509433962264151,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004776204240073269,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2014483705163002,
      "num_tokens": 34535697.0,
      "step": 16640
    },
    {
      "entropy": 5.509071969985962,
      "epoch": 1.509887518142235,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047760635388858624,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20891896039247512,
      "num_tokens": 34545130.0,
      "step": 16645
    },
    {
      "entropy": 5.454761981964111,
      "epoch": 1.5103410740203191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047759227957989465,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2079317569732666,
      "num_tokens": 34554974.0,
      "step": 16650
    },
    {
      "entropy": 5.485178279876709,
      "epoch": 1.5107946298984034,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477578201081543,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20100889652967452,
      "num_tokens": 34565660.0,
      "step": 16655
    },
    {
      "entropy": 5.537592601776123,
      "epoch": 1.5112481857764877,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004775641183938225,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20447785705327987,
      "num_tokens": 34576447.0,
      "step": 16660
    },
    {
      "entropy": 5.565694332122803,
      "epoch": 1.511701741654572,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047755003151702443,
      "loss": 5.173,
      "mean_token_accuracy": 0.19871526211500168,
      "num_tokens": 34588333.0,
      "step": 16665
    },
    {
      "entropy": 5.423282337188721,
      "epoch": 1.512155297532656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047753594045144015,
      "loss": 4.9193,
      "mean_token_accuracy": 0.20331991463899612,
      "num_tokens": 34598166.0,
      "step": 16670
    },
    {
      "entropy": 5.45153112411499,
      "epoch": 1.5126088534107403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004775218451973609,
      "loss": 4.9566,
      "mean_token_accuracy": 0.20150122046470642,
      "num_tokens": 34608615.0,
      "step": 16675
    },
    {
      "entropy": 5.474885368347168,
      "epoch": 1.5130624092888243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004775077457550784,
      "loss": 5.0232,
      "mean_token_accuracy": 0.2053156614303589,
      "num_tokens": 34618618.0,
      "step": 16680
    },
    {
      "entropy": 5.481243419647217,
      "epoch": 1.5135159651669086,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047749364212488403,
      "loss": 5.0388,
      "mean_token_accuracy": 0.19779873341321946,
      "num_tokens": 34629412.0,
      "step": 16685
    },
    {
      "entropy": 5.54996166229248,
      "epoch": 1.5139695210449928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004774795343070696,
      "loss": 5.1042,
      "mean_token_accuracy": 0.20105478912591934,
      "num_tokens": 34640487.0,
      "step": 16690
    },
    {
      "entropy": 5.509821653366089,
      "epoch": 1.5144230769230769,
      "grad_norm": 1.0,
      "learning_rate": 0.0004774654223019268,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20159787237644194,
      "num_tokens": 34650126.0,
      "step": 16695
    },
    {
      "entropy": 5.406566762924195,
      "epoch": 1.5148766328011611,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004774513061097474,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21117860376834868,
      "num_tokens": 34659721.0,
      "step": 16700
    },
    {
      "entropy": 5.463156604766846,
      "epoch": 1.5153301886792452,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774371857308235,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20736206471920013,
      "num_tokens": 34670611.0,
      "step": 16705
    },
    {
      "entropy": 5.461158514022827,
      "epoch": 1.5157837445573294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047742306116544693,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20645657032728196,
      "num_tokens": 34680173.0,
      "step": 16710
    },
    {
      "entropy": 5.4701244831085205,
      "epoch": 1.5162373004354137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047740893241390994,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20616694539785385,
      "num_tokens": 34690176.0,
      "step": 16715
    },
    {
      "entropy": 5.515377187728882,
      "epoch": 1.516690856313498,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004773947994765046,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2083370253443718,
      "num_tokens": 34700481.0,
      "step": 16720
    },
    {
      "entropy": 5.455897617340088,
      "epoch": 1.517144412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773806623535232,
      "loss": 5.1347,
      "mean_token_accuracy": 0.19965612143278122,
      "num_tokens": 34710643.0,
      "step": 16725
    },
    {
      "entropy": 5.507606935501099,
      "epoch": 1.517597968069666,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047736652104525816,
      "loss": 5.0098,
      "mean_token_accuracy": 0.1999215304851532,
      "num_tokens": 34720998.0,
      "step": 16730
    },
    {
      "entropy": 5.544073629379272,
      "epoch": 1.5180515239477503,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004773523755520019,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20022758543491365,
      "num_tokens": 34731071.0,
      "step": 16735
    },
    {
      "entropy": 5.468399381637573,
      "epoch": 1.5185050798258346,
      "grad_norm": 1.078125,
      "learning_rate": 0.000477338225874047,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20834554135799407,
      "num_tokens": 34741346.0,
      "step": 16740
    },
    {
      "entropy": 5.4984602451324465,
      "epoch": 1.5189586357039189,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047732407201168595,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2056797116994858,
      "num_tokens": 34751092.0,
      "step": 16745
    },
    {
      "entropy": 5.559841680526733,
      "epoch": 1.519412191582003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004773099139652116,
      "loss": 5.0843,
      "mean_token_accuracy": 0.2010665148496628,
      "num_tokens": 34761477.0,
      "step": 16750
    },
    {
      "entropy": 5.448795461654663,
      "epoch": 1.519865747460087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004772957517349167,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20591279715299607,
      "num_tokens": 34772491.0,
      "step": 16755
    },
    {
      "entropy": 5.4718138694763185,
      "epoch": 1.5203193033381712,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004772815853210939,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19933177679777145,
      "num_tokens": 34783192.0,
      "step": 16760
    },
    {
      "entropy": 5.4850438117980955,
      "epoch": 1.5207728592162555,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047726741472403655,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20672593712806703,
      "num_tokens": 34794164.0,
      "step": 16765
    },
    {
      "entropy": 5.508155679702758,
      "epoch": 1.5212264150943398,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047725323994403753,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19614226073026658,
      "num_tokens": 34804390.0,
      "step": 16770
    },
    {
      "entropy": 5.594349098205567,
      "epoch": 1.5216799709724238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047723906098138993,
      "loss": 5.155,
      "mean_token_accuracy": 0.1935633584856987,
      "num_tokens": 34814751.0,
      "step": 16775
    },
    {
      "entropy": 5.500635290145874,
      "epoch": 1.5221335268505078,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047722487783638697,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20784967839717866,
      "num_tokens": 34824977.0,
      "step": 16780
    },
    {
      "entropy": 5.4684912204742435,
      "epoch": 1.5225870827285921,
      "grad_norm": 1.140625,
      "learning_rate": 0.000477210690509322,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20255587100982667,
      "num_tokens": 34835540.0,
      "step": 16785
    },
    {
      "entropy": 5.525739097595215,
      "epoch": 1.5230406386066764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047719649900048843,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20400345474481582,
      "num_tokens": 34845233.0,
      "step": 16790
    },
    {
      "entropy": 5.545043039321899,
      "epoch": 1.5234941944847606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047718230331017975,
      "loss": 5.1223,
      "mean_token_accuracy": 0.19574718326330184,
      "num_tokens": 34855756.0,
      "step": 16795
    },
    {
      "entropy": 5.522501468658447,
      "epoch": 1.5239477503628447,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047716810343868956,
      "loss": 5.0383,
      "mean_token_accuracy": 0.191273033618927,
      "num_tokens": 34867869.0,
      "step": 16800
    },
    {
      "entropy": 5.5965899467468265,
      "epoch": 1.5244013062409287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047715389938631143,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1961804524064064,
      "num_tokens": 34878625.0,
      "step": 16805
    },
    {
      "entropy": 5.5731347560882565,
      "epoch": 1.524854862119013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047713969115333916,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1929438978433609,
      "num_tokens": 34889408.0,
      "step": 16810
    },
    {
      "entropy": 5.465712594985962,
      "epoch": 1.5253084179970973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004771254787400666,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19568723440170288,
      "num_tokens": 34899602.0,
      "step": 16815
    },
    {
      "entropy": 5.422050142288208,
      "epoch": 1.5257619738751815,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004771112621467876,
      "loss": 4.9375,
      "mean_token_accuracy": 0.20706008970737458,
      "num_tokens": 34910215.0,
      "step": 16820
    },
    {
      "entropy": 5.453781223297119,
      "epoch": 1.5262155297532656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004770970413737961,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20857041776180268,
      "num_tokens": 34919641.0,
      "step": 16825
    },
    {
      "entropy": 5.494102191925049,
      "epoch": 1.5266690856313496,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004770828164213865,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21438930034637452,
      "num_tokens": 34930243.0,
      "step": 16830
    },
    {
      "entropy": 5.524109840393066,
      "epoch": 1.5271226415094339,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004770685872898525,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20063476711511613,
      "num_tokens": 34941057.0,
      "step": 16835
    },
    {
      "entropy": 5.489181804656982,
      "epoch": 1.5275761973875182,
      "grad_norm": 0.875,
      "learning_rate": 0.00047705435397948876,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2134463220834732,
      "num_tokens": 34951150.0,
      "step": 16840
    },
    {
      "entropy": 5.508066415786743,
      "epoch": 1.5280297532656024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004770401164905895,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20280857831239701,
      "num_tokens": 34961648.0,
      "step": 16845
    },
    {
      "entropy": 5.499263668060303,
      "epoch": 1.5284833091436865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047702587482344915,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19851076304912568,
      "num_tokens": 34971573.0,
      "step": 16850
    },
    {
      "entropy": 5.529312515258789,
      "epoch": 1.5289368650217707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004770116289783622,
      "loss": 5.0283,
      "mean_token_accuracy": 0.19798604249954224,
      "num_tokens": 34981984.0,
      "step": 16855
    },
    {
      "entropy": 5.510770082473755,
      "epoch": 1.5293904208998548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004769973789556233,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20022655874490738,
      "num_tokens": 34990861.0,
      "step": 16860
    },
    {
      "entropy": 5.472228527069092,
      "epoch": 1.529843976777939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047698312475552714,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20059497505426407,
      "num_tokens": 35001504.0,
      "step": 16865
    },
    {
      "entropy": 5.514711380004883,
      "epoch": 1.5302975326560233,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769688663783684,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21026831716299058,
      "num_tokens": 35012425.0,
      "step": 16870
    },
    {
      "entropy": 5.482202053070068,
      "epoch": 1.5307510885341074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004769546038244421,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2071754142642021,
      "num_tokens": 35022854.0,
      "step": 16875
    },
    {
      "entropy": 5.526014947891236,
      "epoch": 1.5312046444121916,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047694033709404313,
      "loss": 5.0733,
      "mean_token_accuracy": 0.19844924956560134,
      "num_tokens": 35032694.0,
      "step": 16880
    },
    {
      "entropy": 5.488026809692383,
      "epoch": 1.5316582002902757,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047692606618746655,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2060248002409935,
      "num_tokens": 35042217.0,
      "step": 16885
    },
    {
      "entropy": 5.480239391326904,
      "epoch": 1.53211175616836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047691179110500735,
      "loss": 5.0745,
      "mean_token_accuracy": 0.19469793438911437,
      "num_tokens": 35052831.0,
      "step": 16890
    },
    {
      "entropy": 5.5425124168396,
      "epoch": 1.5325653120464442,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047689751184696097,
      "loss": 5.1061,
      "mean_token_accuracy": 0.19845710843801498,
      "num_tokens": 35061955.0,
      "step": 16895
    },
    {
      "entropy": 5.5257713317871096,
      "epoch": 1.5330188679245285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004768832284136226,
      "loss": 5.005,
      "mean_token_accuracy": 0.20163582265377045,
      "num_tokens": 35072526.0,
      "step": 16900
    },
    {
      "entropy": 5.4757163524627686,
      "epoch": 1.5334724238026125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768689408052876,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20861303061246872,
      "num_tokens": 35082415.0,
      "step": 16905
    },
    {
      "entropy": 5.436173772811889,
      "epoch": 1.5339259796806965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047685464902225144,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2085571765899658,
      "num_tokens": 35092805.0,
      "step": 16910
    },
    {
      "entropy": 5.4842204570770265,
      "epoch": 1.5343795355587808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004768403530648097,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20818058848381044,
      "num_tokens": 35102963.0,
      "step": 16915
    },
    {
      "entropy": 5.521869802474976,
      "epoch": 1.534833091436865,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047682605293325806,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20084257870912553,
      "num_tokens": 35114140.0,
      "step": 16920
    },
    {
      "entropy": 5.5732200145721436,
      "epoch": 1.5352866473149493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047681174862789215,
      "loss": 5.1061,
      "mean_token_accuracy": 0.2038816675543785,
      "num_tokens": 35125314.0,
      "step": 16925
    },
    {
      "entropy": 5.484570741653442,
      "epoch": 1.5357402031930334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047679744014900795,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20243679136037826,
      "num_tokens": 35134934.0,
      "step": 16930
    },
    {
      "entropy": 5.436294937133789,
      "epoch": 1.5361937590711174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767831274969011,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20469433665275574,
      "num_tokens": 35145341.0,
      "step": 16935
    },
    {
      "entropy": 5.412739849090576,
      "epoch": 1.5366473149492017,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047676881067186793,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20421936362981796,
      "num_tokens": 35156046.0,
      "step": 16940
    },
    {
      "entropy": 5.539416217803955,
      "epoch": 1.537100870827286,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047675448967420426,
      "loss": 5.0975,
      "mean_token_accuracy": 0.19606997519731523,
      "num_tokens": 35168232.0,
      "step": 16945
    },
    {
      "entropy": 5.580508279800415,
      "epoch": 1.5375544267053702,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004767401645042064,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2020314961671829,
      "num_tokens": 35178585.0,
      "step": 16950
    },
    {
      "entropy": 5.472460174560547,
      "epoch": 1.5380079825834543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047672583516217043,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20863931775093078,
      "num_tokens": 35189417.0,
      "step": 16955
    },
    {
      "entropy": 5.45528507232666,
      "epoch": 1.5384615384615383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767115016483928,
      "loss": 4.983,
      "mean_token_accuracy": 0.2048747405409813,
      "num_tokens": 35200167.0,
      "step": 16960
    },
    {
      "entropy": 5.474616384506225,
      "epoch": 1.5389150943396226,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004766971639631699,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20581676810979843,
      "num_tokens": 35210711.0,
      "step": 16965
    },
    {
      "entropy": 5.442872667312622,
      "epoch": 1.5393686502177069,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047668282210679825,
      "loss": 5.0311,
      "mean_token_accuracy": 0.19943596422672272,
      "num_tokens": 35220866.0,
      "step": 16970
    },
    {
      "entropy": 5.412336540222168,
      "epoch": 1.5398222060957911,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004766684760795745,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20932998210191728,
      "num_tokens": 35231478.0,
      "step": 16975
    },
    {
      "entropy": 5.415338563919067,
      "epoch": 1.5402757619738752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047665412588179524,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21175016909837724,
      "num_tokens": 35242233.0,
      "step": 16980
    },
    {
      "entropy": 5.461297941207886,
      "epoch": 1.5407293178519592,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004766397715137573,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1990339368581772,
      "num_tokens": 35253097.0,
      "step": 16985
    },
    {
      "entropy": 5.576567077636719,
      "epoch": 1.5411828737300435,
      "grad_norm": 1.0,
      "learning_rate": 0.00047662541297575743,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19975366592407226,
      "num_tokens": 35264258.0,
      "step": 16990
    },
    {
      "entropy": 5.605187034606933,
      "epoch": 1.5416364296081277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047661105026809274,
      "loss": 5.1254,
      "mean_token_accuracy": 0.19700509458780288,
      "num_tokens": 35273501.0,
      "step": 16995
    },
    {
      "entropy": 5.460659599304199,
      "epoch": 1.542089985486212,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004765966833910601,
      "loss": 5.0665,
      "mean_token_accuracy": 0.1999218925833702,
      "num_tokens": 35284125.0,
      "step": 17000
    },
    {
      "entropy": 5.564423942565918,
      "epoch": 1.542543541364296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004765823123449567,
      "loss": 5.0637,
      "mean_token_accuracy": 0.19853903353214264,
      "num_tokens": 35296039.0,
      "step": 17005
    },
    {
      "entropy": 5.506070804595947,
      "epoch": 1.54299709724238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047656793713007965,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20298447459936142,
      "num_tokens": 35306178.0,
      "step": 17010
    },
    {
      "entropy": 5.462584400177002,
      "epoch": 1.5434506531204644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047655355774672636,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20627882033586503,
      "num_tokens": 35316445.0,
      "step": 17015
    },
    {
      "entropy": 5.519912672042847,
      "epoch": 1.5439042089985486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047653917419519413,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1988663762807846,
      "num_tokens": 35327274.0,
      "step": 17020
    },
    {
      "entropy": 5.478893756866455,
      "epoch": 1.544357764876633,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004765247864757804,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20485960692167282,
      "num_tokens": 35339007.0,
      "step": 17025
    },
    {
      "entropy": 5.509318256378174,
      "epoch": 1.544811320754717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004765103945887827,
      "loss": 5.1123,
      "mean_token_accuracy": 0.19657397717237474,
      "num_tokens": 35349303.0,
      "step": 17030
    },
    {
      "entropy": 5.529847812652588,
      "epoch": 1.5452648766328012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004764959985344987,
      "loss": 4.9534,
      "mean_token_accuracy": 0.20820019841194154,
      "num_tokens": 35359556.0,
      "step": 17035
    },
    {
      "entropy": 5.45951886177063,
      "epoch": 1.5457184325108853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004764815983132261,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20989398658275604,
      "num_tokens": 35370043.0,
      "step": 17040
    },
    {
      "entropy": 5.461466217041016,
      "epoch": 1.5461719883889695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004764671939252628,
      "loss": 4.968,
      "mean_token_accuracy": 0.20857801139354706,
      "num_tokens": 35380031.0,
      "step": 17045
    },
    {
      "entropy": 5.488165378570557,
      "epoch": 1.5466255442670538,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004764527853709064,
      "loss": 5.0661,
      "mean_token_accuracy": 0.199637970328331,
      "num_tokens": 35390481.0,
      "step": 17050
    },
    {
      "entropy": 5.500989580154419,
      "epoch": 1.5470791001451378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004764383726504552,
      "loss": 5.0531,
      "mean_token_accuracy": 0.19923148900270463,
      "num_tokens": 35401756.0,
      "step": 17055
    },
    {
      "entropy": 5.526605939865112,
      "epoch": 1.547532656023222,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004764239557642071,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21369479596614838,
      "num_tokens": 35411045.0,
      "step": 17060
    },
    {
      "entropy": 5.458116579055786,
      "epoch": 1.5479862119013061,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004764095347124603,
      "loss": 5.023,
      "mean_token_accuracy": 0.1952745273709297,
      "num_tokens": 35421256.0,
      "step": 17065
    },
    {
      "entropy": 5.468409061431885,
      "epoch": 1.5484397677793904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004763951094955129,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20428329855203628,
      "num_tokens": 35430798.0,
      "step": 17070
    },
    {
      "entropy": 5.4707554340362545,
      "epoch": 1.5488933236574747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004763806801136634,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21311040967702866,
      "num_tokens": 35440459.0,
      "step": 17075
    },
    {
      "entropy": 5.475528955459595,
      "epoch": 1.549346879535559,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004763662465672101,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20399243980646134,
      "num_tokens": 35451586.0,
      "step": 17080
    },
    {
      "entropy": 5.487248945236206,
      "epoch": 1.549800435413643,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047635180885645146,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20108581632375716,
      "num_tokens": 35462197.0,
      "step": 17085
    },
    {
      "entropy": 5.514713144302368,
      "epoch": 1.550253991291727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004763373669816862,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20966029018163682,
      "num_tokens": 35471737.0,
      "step": 17090
    },
    {
      "entropy": 5.527065849304199,
      "epoch": 1.5507075471698113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004763229209432128,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1970732882618904,
      "num_tokens": 35482016.0,
      "step": 17095
    },
    {
      "entropy": 5.530784654617309,
      "epoch": 1.5511611030478956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763084707413301,
      "loss": 5.1574,
      "mean_token_accuracy": 0.19198521673679353,
      "num_tokens": 35492622.0,
      "step": 17100
    },
    {
      "entropy": 5.57908353805542,
      "epoch": 1.5516146589259798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047629401637633696,
      "loss": 5.1757,
      "mean_token_accuracy": 0.19465037435293198,
      "num_tokens": 35504468.0,
      "step": 17105
    },
    {
      "entropy": 5.486019611358643,
      "epoch": 1.5520682148040639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047627955784853226,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20591204166412352,
      "num_tokens": 35514272.0,
      "step": 17110
    },
    {
      "entropy": 5.4798156261444095,
      "epoch": 1.552521770682148,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476265095158215,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20031564831733703,
      "num_tokens": 35524110.0,
      "step": 17115
    },
    {
      "entropy": 5.497633266448974,
      "epoch": 1.5529753265602322,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004762506283056843,
      "loss": 4.914,
      "mean_token_accuracy": 0.21007648408412932,
      "num_tokens": 35535504.0,
      "step": 17120
    },
    {
      "entropy": 5.5904745101928714,
      "epoch": 1.5534288824383164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004762361572912393,
      "loss": 5.117,
      "mean_token_accuracy": 0.19613465964794158,
      "num_tokens": 35544885.0,
      "step": 17125
    },
    {
      "entropy": 5.519411087036133,
      "epoch": 1.5538824383164007,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047622168211517926,
      "loss": 5.1224,
      "mean_token_accuracy": 0.2027273118495941,
      "num_tokens": 35555485.0,
      "step": 17130
    },
    {
      "entropy": 5.4349846839904785,
      "epoch": 1.5543359941944848,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047620720277780353,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2088460147380829,
      "num_tokens": 35567197.0,
      "step": 17135
    },
    {
      "entropy": 5.523997163772583,
      "epoch": 1.5547895500725688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047619271927941165,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2054537519812584,
      "num_tokens": 35577576.0,
      "step": 17140
    },
    {
      "entropy": 5.511078262329102,
      "epoch": 1.555243105950653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000476178231620303,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2045543149113655,
      "num_tokens": 35588012.0,
      "step": 17145
    },
    {
      "entropy": 5.454594993591309,
      "epoch": 1.5556966618287373,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761637398007774,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2018841803073883,
      "num_tokens": 35598076.0,
      "step": 17150
    },
    {
      "entropy": 5.394225978851319,
      "epoch": 1.5561502177068216,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004761492438211342,
      "loss": 4.8495,
      "mean_token_accuracy": 0.21599810868501662,
      "num_tokens": 35607170.0,
      "step": 17155
    },
    {
      "entropy": 5.468464374542236,
      "epoch": 1.5566037735849056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047613474368167355,
      "loss": 4.9872,
      "mean_token_accuracy": 0.19680835902690888,
      "num_tokens": 35617865.0,
      "step": 17160
    },
    {
      "entropy": 5.472936391830444,
      "epoch": 1.5570573294629897,
      "grad_norm": 1.078125,
      "learning_rate": 0.000476120239382695,
      "loss": 5.0411,
      "mean_token_accuracy": 0.20120879709720613,
      "num_tokens": 35629329.0,
      "step": 17165
    },
    {
      "entropy": 5.4649515628814695,
      "epoch": 1.557510885341074,
      "grad_norm": 1.0,
      "learning_rate": 0.00047610573092449877,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20129372775554658,
      "num_tokens": 35640044.0,
      "step": 17170
    },
    {
      "entropy": 5.522956418991089,
      "epoch": 1.5579644412191582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004760912183073847,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20332991182804108,
      "num_tokens": 35650435.0,
      "step": 17175
    },
    {
      "entropy": 5.491788578033447,
      "epoch": 1.5584179970972425,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000476076701531653,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20425090044736863,
      "num_tokens": 35659613.0,
      "step": 17180
    },
    {
      "entropy": 5.427504205703736,
      "epoch": 1.5588715529753265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004760621805976039,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20563696473836898,
      "num_tokens": 35670987.0,
      "step": 17185
    },
    {
      "entropy": 5.478841972351074,
      "epoch": 1.5593251088534106,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047604765550553766,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2048560231924057,
      "num_tokens": 35682505.0,
      "step": 17190
    },
    {
      "entropy": 5.522268867492675,
      "epoch": 1.5597786647314948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047603312625575466,
      "loss": 4.9462,
      "mean_token_accuracy": 0.20475015491247178,
      "num_tokens": 35693056.0,
      "step": 17195
    },
    {
      "entropy": 5.4625657081604,
      "epoch": 1.5602322206095791,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047601859284855544,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21257547438144683,
      "num_tokens": 35703659.0,
      "step": 17200
    },
    {
      "entropy": 5.438641500473023,
      "epoch": 1.5606857764876634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760040552842403,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19803812056779863,
      "num_tokens": 35714367.0,
      "step": 17205
    },
    {
      "entropy": 5.425652980804443,
      "epoch": 1.5611393323657474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047598951356311035,
      "loss": 4.9711,
      "mean_token_accuracy": 0.1990872249007225,
      "num_tokens": 35724773.0,
      "step": 17210
    },
    {
      "entropy": 5.519612550735474,
      "epoch": 1.5615928882438317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004759749676854658,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20020296722650527,
      "num_tokens": 35735622.0,
      "step": 17215
    },
    {
      "entropy": 5.520531034469604,
      "epoch": 1.5620464441219157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004759604176516079,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2098139926791191,
      "num_tokens": 35744967.0,
      "step": 17220
    },
    {
      "entropy": 5.531719493865967,
      "epoch": 1.5625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759458634618372,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20392842590808868,
      "num_tokens": 35755647.0,
      "step": 17225
    },
    {
      "entropy": 5.503164672851563,
      "epoch": 1.5629535558780843,
      "grad_norm": 1.0,
      "learning_rate": 0.0004759313051164549,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2000377207994461,
      "num_tokens": 35765595.0,
      "step": 17230
    },
    {
      "entropy": 5.501796817779541,
      "epoch": 1.5634071117561683,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000475916742615762,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21368214190006257,
      "num_tokens": 35775804.0,
      "step": 17235
    },
    {
      "entropy": 5.46679916381836,
      "epoch": 1.5638606676342526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004759021759600597,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2093169242143631,
      "num_tokens": 35786227.0,
      "step": 17240
    },
    {
      "entropy": 5.532622194290161,
      "epoch": 1.5643142235123366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004758876051496492,
      "loss": 5.0919,
      "mean_token_accuracy": 0.20590397715568542,
      "num_tokens": 35797511.0,
      "step": 17245
    },
    {
      "entropy": 5.5053972721099855,
      "epoch": 1.5647677793904209,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047587303018483177,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20084312409162522,
      "num_tokens": 35807248.0,
      "step": 17250
    },
    {
      "entropy": 5.453809404373169,
      "epoch": 1.5652213352685052,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047585845106590897,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19742046445608138,
      "num_tokens": 35818164.0,
      "step": 17255
    },
    {
      "entropy": 5.467202234268188,
      "epoch": 1.5656748911465894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047584386779318214,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2104337528347969,
      "num_tokens": 35828746.0,
      "step": 17260
    },
    {
      "entropy": 5.519303274154663,
      "epoch": 1.5661284470246735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000475829280366953,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19842927008867264,
      "num_tokens": 35839325.0,
      "step": 17265
    },
    {
      "entropy": 5.502063179016114,
      "epoch": 1.5665820029027575,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047581468878752313,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20751725137233734,
      "num_tokens": 35849406.0,
      "step": 17270
    },
    {
      "entropy": 5.426310443878174,
      "epoch": 1.5670355587808418,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047580009305519427,
      "loss": 4.9205,
      "mean_token_accuracy": 0.20804960429668426,
      "num_tokens": 35858715.0,
      "step": 17275
    },
    {
      "entropy": 5.442573881149292,
      "epoch": 1.567489114658926,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004757854931702684,
      "loss": 5.0062,
      "mean_token_accuracy": 0.20239466428756714,
      "num_tokens": 35869063.0,
      "step": 17280
    },
    {
      "entropy": 5.444556140899659,
      "epoch": 1.5679426705370103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047577088913304733,
      "loss": 4.8804,
      "mean_token_accuracy": 0.20901139378547667,
      "num_tokens": 35878428.0,
      "step": 17285
    },
    {
      "entropy": 5.516884994506836,
      "epoch": 1.5683962264150944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047575628094383317,
      "loss": 5.0248,
      "mean_token_accuracy": 0.20472852289676666,
      "num_tokens": 35888801.0,
      "step": 17290
    },
    {
      "entropy": 5.444619464874267,
      "epoch": 1.5688497822931784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004757416686029279,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20317072570323944,
      "num_tokens": 35899381.0,
      "step": 17295
    },
    {
      "entropy": 5.528191423416137,
      "epoch": 1.5693033381712627,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004757270521106338,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20813838094472886,
      "num_tokens": 35909546.0,
      "step": 17300
    },
    {
      "entropy": 5.501978206634521,
      "epoch": 1.569756894049347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004757124314672531,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2017609715461731,
      "num_tokens": 35919301.0,
      "step": 17305
    },
    {
      "entropy": 5.591321849822998,
      "epoch": 1.5702104499274312,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047569780667308816,
      "loss": 5.0762,
      "mean_token_accuracy": 0.1994296759366989,
      "num_tokens": 35928603.0,
      "step": 17310
    },
    {
      "entropy": 5.540690517425537,
      "epoch": 1.5706640058055152,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047568317772844147,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20576718896627427,
      "num_tokens": 35939005.0,
      "step": 17315
    },
    {
      "entropy": 5.538856363296508,
      "epoch": 1.5711175616835993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047566854463361544,
      "loss": 5.015,
      "mean_token_accuracy": 0.21285732537508012,
      "num_tokens": 35948471.0,
      "step": 17320
    },
    {
      "entropy": 5.437993288040161,
      "epoch": 1.5715711175616836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004756539073889128,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2030189663171768,
      "num_tokens": 35958488.0,
      "step": 17325
    },
    {
      "entropy": 5.48625545501709,
      "epoch": 1.5720246734397678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047563926599463623,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19767525047063828,
      "num_tokens": 35968291.0,
      "step": 17330
    },
    {
      "entropy": 5.546093368530274,
      "epoch": 1.572478229317852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004756246204510886,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20795842111110688,
      "num_tokens": 35978565.0,
      "step": 17335
    },
    {
      "entropy": 5.52994179725647,
      "epoch": 1.5729317851959361,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047560997075857253,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20825816094875335,
      "num_tokens": 35987483.0,
      "step": 17340
    },
    {
      "entropy": 5.476396131515503,
      "epoch": 1.5733853410740202,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004755953169173913,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20181325376033782,
      "num_tokens": 35998355.0,
      "step": 17345
    },
    {
      "entropy": 5.467070770263672,
      "epoch": 1.5738388969521044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004755806589278477,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21285985708236693,
      "num_tokens": 36008365.0,
      "step": 17350
    },
    {
      "entropy": 5.51667366027832,
      "epoch": 1.5742924528301887,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047556599679024497,
      "loss": 5.0824,
      "mean_token_accuracy": 0.19374739676713942,
      "num_tokens": 36018880.0,
      "step": 17355
    },
    {
      "entropy": 5.522989702224732,
      "epoch": 1.574746008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047555133050488637,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19408987313508988,
      "num_tokens": 36029335.0,
      "step": 17360
    },
    {
      "entropy": 5.483440351486206,
      "epoch": 1.575199564586357,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047553666007207516,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2119162768125534,
      "num_tokens": 36039774.0,
      "step": 17365
    },
    {
      "entropy": 5.55950517654419,
      "epoch": 1.575653120464441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047552198549211463,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20450095236301422,
      "num_tokens": 36049864.0,
      "step": 17370
    },
    {
      "entropy": 5.49333701133728,
      "epoch": 1.5761066763425253,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004755073067653084,
      "loss": 5.0493,
      "mean_token_accuracy": 0.19705418944358827,
      "num_tokens": 36059224.0,
      "step": 17375
    },
    {
      "entropy": 5.421458721160889,
      "epoch": 1.5765602322206096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004754926238919599,
      "loss": 4.9525,
      "mean_token_accuracy": 0.211106775701046,
      "num_tokens": 36069243.0,
      "step": 17380
    },
    {
      "entropy": 5.504794073104859,
      "epoch": 1.5770137880986939,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004754779368723729,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2001550853252411,
      "num_tokens": 36079833.0,
      "step": 17385
    },
    {
      "entropy": 5.4966668605804445,
      "epoch": 1.577467343976778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004754632457068511,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2046745464205742,
      "num_tokens": 36090673.0,
      "step": 17390
    },
    {
      "entropy": 5.540458869934082,
      "epoch": 1.5779208998548622,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004754485503956982,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20388307571411132,
      "num_tokens": 36101719.0,
      "step": 17395
    },
    {
      "entropy": 5.503682947158813,
      "epoch": 1.5783744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754338509392183,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20253086239099502,
      "num_tokens": 36112352.0,
      "step": 17400
    },
    {
      "entropy": 5.374897909164429,
      "epoch": 1.5788280116110305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004754191473377153,
      "loss": 4.8921,
      "mean_token_accuracy": 0.20746100395917894,
      "num_tokens": 36121797.0,
      "step": 17405
    },
    {
      "entropy": 5.484429883956909,
      "epoch": 1.5792815674891147,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047540443959149316,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21143339276313783,
      "num_tokens": 36132094.0,
      "step": 17410
    },
    {
      "entropy": 5.499725246429444,
      "epoch": 1.5797351233671988,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004753897277008562,
      "loss": 4.9994,
      "mean_token_accuracy": 0.203934970498085,
      "num_tokens": 36142860.0,
      "step": 17415
    },
    {
      "entropy": 5.508707189559937,
      "epoch": 1.580188679245283,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004753750116661086,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20103585422039033,
      "num_tokens": 36154555.0,
      "step": 17420
    },
    {
      "entropy": 5.576639652252197,
      "epoch": 1.580642235123367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047536029148755477,
      "loss": 5.1314,
      "mean_token_accuracy": 0.19167136400938034,
      "num_tokens": 36165479.0,
      "step": 17425
    },
    {
      "entropy": 5.5136706829071045,
      "epoch": 1.5810957910014514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047534556716549893,
      "loss": 5.0417,
      "mean_token_accuracy": 0.2023715317249298,
      "num_tokens": 36175873.0,
      "step": 17430
    },
    {
      "entropy": 5.461419534683228,
      "epoch": 1.5815493468795356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047533083870024575,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2108200564980507,
      "num_tokens": 36186792.0,
      "step": 17435
    },
    {
      "entropy": 5.6181721687316895,
      "epoch": 1.58200290275762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004753161060920998,
      "loss": 5.1257,
      "mean_token_accuracy": 0.19894260317087173,
      "num_tokens": 36196691.0,
      "step": 17440
    },
    {
      "entropy": 5.569988870620728,
      "epoch": 1.582456458635704,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004753013693413657,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2052525535225868,
      "num_tokens": 36207173.0,
      "step": 17445
    },
    {
      "entropy": 5.460013198852539,
      "epoch": 1.582910014513788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047528662844834827,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20807620137929916,
      "num_tokens": 36217912.0,
      "step": 17450
    },
    {
      "entropy": 5.411827182769775,
      "epoch": 1.5833635703918723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047527188341335233,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2129110425710678,
      "num_tokens": 36227714.0,
      "step": 17455
    },
    {
      "entropy": 5.476075458526611,
      "epoch": 1.5838171262699565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004752571342366828,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20643447786569596,
      "num_tokens": 36239114.0,
      "step": 17460
    },
    {
      "entropy": 5.481729888916016,
      "epoch": 1.5842706821480408,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004752423809186446,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20117921978235245,
      "num_tokens": 36249436.0,
      "step": 17465
    },
    {
      "entropy": 5.498609352111816,
      "epoch": 1.5847242380261248,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047522762345954305,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2044916108250618,
      "num_tokens": 36259541.0,
      "step": 17470
    },
    {
      "entropy": 5.511799478530884,
      "epoch": 1.5851777939042089,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004752128618596832,
      "loss": 5.0275,
      "mean_token_accuracy": 0.19812653213739395,
      "num_tokens": 36270689.0,
      "step": 17475
    },
    {
      "entropy": 5.559856700897217,
      "epoch": 1.5856313497822931,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047519809611937036,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20168314427137374,
      "num_tokens": 36281191.0,
      "step": 17480
    },
    {
      "entropy": 5.437845468521118,
      "epoch": 1.5860849056603774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751833262389098,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21071845591068267,
      "num_tokens": 36291022.0,
      "step": 17485
    },
    {
      "entropy": 5.511733102798462,
      "epoch": 1.5865384615384617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047516855221860715,
      "loss": 5.042,
      "mean_token_accuracy": 0.20490602403879166,
      "num_tokens": 36301237.0,
      "step": 17490
    },
    {
      "entropy": 5.544945573806762,
      "epoch": 1.5869920174165457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047515377405876774,
      "loss": 5.0164,
      "mean_token_accuracy": 0.19902748465538025,
      "num_tokens": 36311792.0,
      "step": 17495
    },
    {
      "entropy": 5.4585517883300785,
      "epoch": 1.5874455732946298,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004751389917596973,
      "loss": 4.988,
      "mean_token_accuracy": 0.2170154094696045,
      "num_tokens": 36321847.0,
      "step": 17500
    },
    {
      "entropy": 5.476059007644653,
      "epoch": 1.587899129172714,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047512420532170154,
      "loss": 5.0971,
      "mean_token_accuracy": 0.19473932832479476,
      "num_tokens": 36332735.0,
      "step": 17505
    },
    {
      "entropy": 5.569549465179444,
      "epoch": 1.5883526850507983,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004751094147450862,
      "loss": 5.0814,
      "mean_token_accuracy": 0.20146389156579972,
      "num_tokens": 36343149.0,
      "step": 17510
    },
    {
      "entropy": 5.555501079559326,
      "epoch": 1.5888062409288826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004750946200301571,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2055591121315956,
      "num_tokens": 36353081.0,
      "step": 17515
    },
    {
      "entropy": 5.50593376159668,
      "epoch": 1.5892597968069666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047507982117722035,
      "loss": 5.0546,
      "mean_token_accuracy": 0.201623272895813,
      "num_tokens": 36363624.0,
      "step": 17520
    },
    {
      "entropy": 5.541685676574707,
      "epoch": 1.5897133526850507,
      "grad_norm": 1.03125,
      "learning_rate": 0.000475065018186582,
      "loss": 5.0761,
      "mean_token_accuracy": 0.19994409531354904,
      "num_tokens": 36374081.0,
      "step": 17525
    },
    {
      "entropy": 5.53523817062378,
      "epoch": 1.590166908563135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047505021105854794,
      "loss": 5.1406,
      "mean_token_accuracy": 0.1978500485420227,
      "num_tokens": 36384360.0,
      "step": 17530
    },
    {
      "entropy": 5.481768226623535,
      "epoch": 1.5906204644412192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047503539979342456,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2044894129037857,
      "num_tokens": 36395924.0,
      "step": 17535
    },
    {
      "entropy": 5.445667171478272,
      "epoch": 1.5910740203193035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750205843915182,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21521376967430114,
      "num_tokens": 36405246.0,
      "step": 17540
    },
    {
      "entropy": 5.48762264251709,
      "epoch": 1.5915275761973875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004750057648531351,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2125127375125885,
      "num_tokens": 36414964.0,
      "step": 17545
    },
    {
      "entropy": 5.468229579925537,
      "epoch": 1.5919811320754715,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004749909411785819,
      "loss": 5.052,
      "mean_token_accuracy": 0.19814764261245726,
      "num_tokens": 36425011.0,
      "step": 17550
    },
    {
      "entropy": 5.423689365386963,
      "epoch": 1.5924346879535558,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000474976113368165,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21135228127241135,
      "num_tokens": 36434734.0,
      "step": 17555
    },
    {
      "entropy": 5.501957988739013,
      "epoch": 1.59288824383164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004749612814221912,
      "loss": 5.038,
      "mean_token_accuracy": 0.20483655035495757,
      "num_tokens": 36444748.0,
      "step": 17560
    },
    {
      "entropy": 5.588495826721191,
      "epoch": 1.5933417997097243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047494644534096703,
      "loss": 5.0479,
      "mean_token_accuracy": 0.19796278774738313,
      "num_tokens": 36455076.0,
      "step": 17565
    },
    {
      "entropy": 5.512688446044922,
      "epoch": 1.5937953555878084,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004749316051247995,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2056950405240059,
      "num_tokens": 36465302.0,
      "step": 17570
    },
    {
      "entropy": 5.484334278106689,
      "epoch": 1.5942489114658926,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047491676077399543,
      "loss": 5.0783,
      "mean_token_accuracy": 0.20078891515731812,
      "num_tokens": 36476666.0,
      "step": 17575
    },
    {
      "entropy": 5.452722549438477,
      "epoch": 1.5947024673439767,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004749019122888618,
      "loss": 4.9335,
      "mean_token_accuracy": 0.208494570851326,
      "num_tokens": 36487379.0,
      "step": 17580
    },
    {
      "entropy": 5.5092161178588865,
      "epoch": 1.595156023222061,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047488705966970565,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21083510369062425,
      "num_tokens": 36497897.0,
      "step": 17585
    },
    {
      "entropy": 5.509284782409668,
      "epoch": 1.5956095791001452,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004748722029168342,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20730422884225846,
      "num_tokens": 36508528.0,
      "step": 17590
    },
    {
      "entropy": 5.514298009872436,
      "epoch": 1.5960631349782293,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748573420305546,
      "loss": 5.0609,
      "mean_token_accuracy": 0.19901353120803833,
      "num_tokens": 36519196.0,
      "step": 17595
    },
    {
      "entropy": 5.472837448120117,
      "epoch": 1.5965166908563135,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047484247701117427,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2019036278128624,
      "num_tokens": 36529871.0,
      "step": 17600
    },
    {
      "entropy": 5.600695705413818,
      "epoch": 1.5969702467343976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004748276078590006,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1906803548336029,
      "num_tokens": 36541188.0,
      "step": 17605
    },
    {
      "entropy": 5.5623010158538815,
      "epoch": 1.5974238026124818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004748127345743411,
      "loss": 5.0861,
      "mean_token_accuracy": 0.1991748109459877,
      "num_tokens": 36551968.0,
      "step": 17610
    },
    {
      "entropy": 5.518520832061768,
      "epoch": 1.5978773584905661,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004747978571575033,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20348055064678192,
      "num_tokens": 36562870.0,
      "step": 17615
    },
    {
      "entropy": 5.4489270687103275,
      "epoch": 1.5983309143686504,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004747829756087949,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21031002700328827,
      "num_tokens": 36573295.0,
      "step": 17620
    },
    {
      "entropy": 5.445514678955078,
      "epoch": 1.5987844702467344,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047476808992852364,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2091691941022873,
      "num_tokens": 36583830.0,
      "step": 17625
    },
    {
      "entropy": 5.48184494972229,
      "epoch": 1.5992380261248185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004747532001169974,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2106089934706688,
      "num_tokens": 36594052.0,
      "step": 17630
    },
    {
      "entropy": 5.509792280197144,
      "epoch": 1.5996915820029027,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047473830617452415,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20581653863191604,
      "num_tokens": 36604378.0,
      "step": 17635
    },
    {
      "entropy": 5.556970739364624,
      "epoch": 1.600145137880987,
      "grad_norm": 1.0,
      "learning_rate": 0.00047472340810141163,
      "loss": 5.147,
      "mean_token_accuracy": 0.19290078580379486,
      "num_tokens": 36615588.0,
      "step": 17640
    },
    {
      "entropy": 5.588248300552368,
      "epoch": 1.6005986937590713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047470850589796833,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20253516882658004,
      "num_tokens": 36626369.0,
      "step": 17645
    },
    {
      "entropy": 5.546917581558228,
      "epoch": 1.6010522496371553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047469359956450215,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20150643587112427,
      "num_tokens": 36636565.0,
      "step": 17650
    },
    {
      "entropy": 5.577430868148804,
      "epoch": 1.6015058055152394,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047467868910132144,
      "loss": 5.1484,
      "mean_token_accuracy": 0.1941498950123787,
      "num_tokens": 36647721.0,
      "step": 17655
    },
    {
      "entropy": 5.435346221923828,
      "epoch": 1.6019593613933236,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004746637745087346,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2122989535331726,
      "num_tokens": 36657887.0,
      "step": 17660
    },
    {
      "entropy": 5.555538558959961,
      "epoch": 1.6024129172714079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047464885578705,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20261109620332718,
      "num_tokens": 36667918.0,
      "step": 17665
    },
    {
      "entropy": 5.448113107681275,
      "epoch": 1.6028664731494922,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047463393293657616,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21008675992488862,
      "num_tokens": 36678491.0,
      "step": 17670
    },
    {
      "entropy": 5.493792295455933,
      "epoch": 1.6033200290275762,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004746190059576217,
      "loss": 5.052,
      "mean_token_accuracy": 0.20171189904212952,
      "num_tokens": 36688638.0,
      "step": 17675
    },
    {
      "entropy": 5.520508289337158,
      "epoch": 1.6037735849056602,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004746040748504954,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20498359203338623,
      "num_tokens": 36699344.0,
      "step": 17680
    },
    {
      "entropy": 5.533679056167602,
      "epoch": 1.6042271407837445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047458913961550593,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20418418049812317,
      "num_tokens": 36709606.0,
      "step": 17685
    },
    {
      "entropy": 5.444143676757813,
      "epoch": 1.6046806966618288,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745742002529622,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20646962225437165,
      "num_tokens": 36718740.0,
      "step": 17690
    },
    {
      "entropy": 5.486650180816651,
      "epoch": 1.605134252539913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004745592567631731,
      "loss": 5.0395,
      "mean_token_accuracy": 0.19889770448207855,
      "num_tokens": 36729960.0,
      "step": 17695
    },
    {
      "entropy": 5.449067878723144,
      "epoch": 1.605587808417997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047454430914644773,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21313123553991317,
      "num_tokens": 36740221.0,
      "step": 17700
    },
    {
      "entropy": 5.396981620788575,
      "epoch": 1.6060413642960811,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004745293574030952,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21185587346553802,
      "num_tokens": 36750841.0,
      "step": 17705
    },
    {
      "entropy": 5.435892915725708,
      "epoch": 1.6064949201741654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047451440153342477,
      "loss": 5.026,
      "mean_token_accuracy": 0.19739796966314316,
      "num_tokens": 36760944.0,
      "step": 17710
    },
    {
      "entropy": 5.519886016845703,
      "epoch": 1.6069484760522497,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004744994415377457,
      "loss": 5.0015,
      "mean_token_accuracy": 0.207135434448719,
      "num_tokens": 36770922.0,
      "step": 17715
    },
    {
      "entropy": 5.558530521392822,
      "epoch": 1.607402031930334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047448447741636724,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20146499425172806,
      "num_tokens": 36780702.0,
      "step": 17720
    },
    {
      "entropy": 5.543065357208252,
      "epoch": 1.607855587808418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047446950916959905,
      "loss": 5.08,
      "mean_token_accuracy": 0.19724457412958146,
      "num_tokens": 36791868.0,
      "step": 17725
    },
    {
      "entropy": 5.539865970611572,
      "epoch": 1.608309143686502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004744545367977505,
      "loss": 5.0413,
      "mean_token_accuracy": 0.20392582416534424,
      "num_tokens": 36802330.0,
      "step": 17730
    },
    {
      "entropy": 5.506350994110107,
      "epoch": 1.6087626995645863,
      "grad_norm": 1.125,
      "learning_rate": 0.00047443956030113134,
      "loss": 5.0658,
      "mean_token_accuracy": 0.20375836491584778,
      "num_tokens": 36812936.0,
      "step": 17735
    },
    {
      "entropy": 5.549493503570557,
      "epoch": 1.6092162554426706,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047442457968005123,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20139739960432052,
      "num_tokens": 36823123.0,
      "step": 17740
    },
    {
      "entropy": 5.447511148452759,
      "epoch": 1.6096698113207548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047440959493481996,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21495402604341507,
      "num_tokens": 36833050.0,
      "step": 17745
    },
    {
      "entropy": 5.500833415985108,
      "epoch": 1.6101233671988389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047439460606574745,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20559147894382476,
      "num_tokens": 36842520.0,
      "step": 17750
    },
    {
      "entropy": 5.4890522956848145,
      "epoch": 1.6105769230769231,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004743796130731437,
      "loss": 5.0053,
      "mean_token_accuracy": 0.19978885352611542,
      "num_tokens": 36853024.0,
      "step": 17755
    },
    {
      "entropy": 5.503660249710083,
      "epoch": 1.6110304789550072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004743646159573188,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21103161871433257,
      "num_tokens": 36862534.0,
      "step": 17760
    },
    {
      "entropy": 5.4798894882202145,
      "epoch": 1.6114840348330914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004743496147185828,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21533622294664384,
      "num_tokens": 36872356.0,
      "step": 17765
    },
    {
      "entropy": 5.476791620254517,
      "epoch": 1.6119375907111757,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047433460935724595,
      "loss": 5.009,
      "mean_token_accuracy": 0.2075714036822319,
      "num_tokens": 36882643.0,
      "step": 17770
    },
    {
      "entropy": 5.501908922195435,
      "epoch": 1.6123911465892597,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004743195998736186,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2041184648871422,
      "num_tokens": 36893349.0,
      "step": 17775
    },
    {
      "entropy": 5.556226587295532,
      "epoch": 1.612844702467344,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047430458626801107,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20778820514678956,
      "num_tokens": 36903713.0,
      "step": 17780
    },
    {
      "entropy": 5.4703563213348385,
      "epoch": 1.613298258345428,
      "grad_norm": 1.125,
      "learning_rate": 0.000474289568540734,
      "loss": 4.932,
      "mean_token_accuracy": 0.20787305682897567,
      "num_tokens": 36912999.0,
      "step": 17785
    },
    {
      "entropy": 5.516793489456177,
      "epoch": 1.6137518142235123,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047427454669209785,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20724982619285584,
      "num_tokens": 36923598.0,
      "step": 17790
    },
    {
      "entropy": 5.459460878372193,
      "epoch": 1.6142053701015966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047425952072241326,
      "loss": 4.972,
      "mean_token_accuracy": 0.20285503268241883,
      "num_tokens": 36933036.0,
      "step": 17795
    },
    {
      "entropy": 5.49089207649231,
      "epoch": 1.6146589259796809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000474244490631991,
      "loss": 5.0686,
      "mean_token_accuracy": 0.19884790927171708,
      "num_tokens": 36943137.0,
      "step": 17800
    },
    {
      "entropy": 5.613918256759644,
      "epoch": 1.615112481857765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004742294564211419,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19501156806945802,
      "num_tokens": 36955619.0,
      "step": 17805
    },
    {
      "entropy": 5.510789442062378,
      "epoch": 1.615566037735849,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004742144180901769,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21337293237447738,
      "num_tokens": 36966377.0,
      "step": 17810
    },
    {
      "entropy": 5.531083250045777,
      "epoch": 1.6160195936139332,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047419937563940697,
      "loss": 5.021,
      "mean_token_accuracy": 0.2078086480498314,
      "num_tokens": 36977100.0,
      "step": 17815
    },
    {
      "entropy": 5.509808015823364,
      "epoch": 1.6164731494920175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047418432906914314,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20011530071496964,
      "num_tokens": 36987120.0,
      "step": 17820
    },
    {
      "entropy": 5.542492389678955,
      "epoch": 1.6169267053701017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004741692783796967,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20618803203105926,
      "num_tokens": 36996885.0,
      "step": 17825
    },
    {
      "entropy": 5.513024425506591,
      "epoch": 1.6173802612481858,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047415422357137876,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21342350989580156,
      "num_tokens": 37006570.0,
      "step": 17830
    },
    {
      "entropy": 5.553627634048462,
      "epoch": 1.6178338171262698,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004741391646445008,
      "loss": 5.0692,
      "mean_token_accuracy": 0.19652530997991563,
      "num_tokens": 37017185.0,
      "step": 17835
    },
    {
      "entropy": 5.521615839004516,
      "epoch": 1.618287373004354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004741241015993741,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20495287030935289,
      "num_tokens": 37028386.0,
      "step": 17840
    },
    {
      "entropy": 5.459984064102173,
      "epoch": 1.6187409288824384,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047410903443631026,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20692460536956786,
      "num_tokens": 37038326.0,
      "step": 17845
    },
    {
      "entropy": 5.5151317596435545,
      "epoch": 1.6191944847605226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004740939631556209,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20376701205968856,
      "num_tokens": 37048570.0,
      "step": 17850
    },
    {
      "entropy": 5.456529808044434,
      "epoch": 1.6196480406386067,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004740788877576176,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2112232819199562,
      "num_tokens": 37058909.0,
      "step": 17855
    },
    {
      "entropy": 5.495297050476074,
      "epoch": 1.6201015965166907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004740638082426121,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21210072934627533,
      "num_tokens": 37068957.0,
      "step": 17860
    },
    {
      "entropy": 5.490260887145996,
      "epoch": 1.620555152394775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047404872461091643,
      "loss": 4.9561,
      "mean_token_accuracy": 0.20613258481025695,
      "num_tokens": 37080187.0,
      "step": 17865
    },
    {
      "entropy": 5.507491159439087,
      "epoch": 1.6210087082728593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047403363686284236,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20411915332078934,
      "num_tokens": 37090314.0,
      "step": 17870
    },
    {
      "entropy": 5.543750429153443,
      "epoch": 1.6214622641509435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047401854499870197,
      "loss": 5.074,
      "mean_token_accuracy": 0.20306196957826614,
      "num_tokens": 37099907.0,
      "step": 17875
    },
    {
      "entropy": 5.505069637298584,
      "epoch": 1.6219158200290276,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047400344901880735,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19933670312166213,
      "num_tokens": 37110495.0,
      "step": 17880
    },
    {
      "entropy": 5.524602603912354,
      "epoch": 1.6223693759071116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004739883489234707,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20316663831472398,
      "num_tokens": 37121957.0,
      "step": 17885
    },
    {
      "entropy": 5.5381800651550295,
      "epoch": 1.6228229317851959,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047397324471300424,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2094401389360428,
      "num_tokens": 37132743.0,
      "step": 17890
    },
    {
      "entropy": 5.520811605453491,
      "epoch": 1.6232764876632801,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047395813638772047,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20002538412809373,
      "num_tokens": 37143947.0,
      "step": 17895
    },
    {
      "entropy": 5.606742572784424,
      "epoch": 1.6237300435413644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047394302394793164,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1896388277411461,
      "num_tokens": 37154316.0,
      "step": 17900
    },
    {
      "entropy": 5.546631574630737,
      "epoch": 1.6241835994194485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739279073939504,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19914041161537172,
      "num_tokens": 37164999.0,
      "step": 17905
    },
    {
      "entropy": 5.53722620010376,
      "epoch": 1.6246371552975325,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739127867260894,
      "loss": 5.0524,
      "mean_token_accuracy": 0.2010467141866684,
      "num_tokens": 37174392.0,
      "step": 17910
    },
    {
      "entropy": 5.467570638656616,
      "epoch": 1.6250907111756168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738976619446613,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20052966475486755,
      "num_tokens": 37185341.0,
      "step": 17915
    },
    {
      "entropy": 5.5671929836273195,
      "epoch": 1.625544267053701,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047388253304997874,
      "loss": 5.0515,
      "mean_token_accuracy": 0.19656094610691072,
      "num_tokens": 37196725.0,
      "step": 17920
    },
    {
      "entropy": 5.5232336044311525,
      "epoch": 1.6259978229317853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004738674000423548,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2106102377176285,
      "num_tokens": 37206877.0,
      "step": 17925
    },
    {
      "entropy": 5.40243763923645,
      "epoch": 1.6264513788098693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004738522629221023,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2058248147368431,
      "num_tokens": 37217277.0,
      "step": 17930
    },
    {
      "entropy": 5.4153111457824705,
      "epoch": 1.6269049346879536,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004738371216895344,
      "loss": 4.925,
      "mean_token_accuracy": 0.21081706881523132,
      "num_tokens": 37227716.0,
      "step": 17935
    },
    {
      "entropy": 5.469329214096069,
      "epoch": 1.6273584905660377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738219763449641,
      "loss": 5.034,
      "mean_token_accuracy": 0.20435808897018432,
      "num_tokens": 37238178.0,
      "step": 17940
    },
    {
      "entropy": 5.489963245391846,
      "epoch": 1.627812046444122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738068268887047,
      "loss": 5.07,
      "mean_token_accuracy": 0.20421854108572007,
      "num_tokens": 37248408.0,
      "step": 17945
    },
    {
      "entropy": 5.5134241580963135,
      "epoch": 1.6282656023222062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004737916733210694,
      "loss": 5.1703,
      "mean_token_accuracy": 0.1944072127342224,
      "num_tokens": 37259176.0,
      "step": 17950
    },
    {
      "entropy": 5.532334995269776,
      "epoch": 1.6287191582002902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047377651564237165,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2068418338894844,
      "num_tokens": 37269205.0,
      "step": 17955
    },
    {
      "entropy": 5.529959297180175,
      "epoch": 1.6291727140783745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047376135385292494,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20907308608293534,
      "num_tokens": 37278841.0,
      "step": 17960
    },
    {
      "entropy": 5.490876150131226,
      "epoch": 1.6296262699564585,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004737461879530428,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2029963567852974,
      "num_tokens": 37288487.0,
      "step": 17965
    },
    {
      "entropy": 5.548338794708252,
      "epoch": 1.6300798258345428,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047373101794303884,
      "loss": 5.0538,
      "mean_token_accuracy": 0.20348800867795944,
      "num_tokens": 37300086.0,
      "step": 17970
    },
    {
      "entropy": 5.528116893768311,
      "epoch": 1.630533381712627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737158438232267,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19577748477458953,
      "num_tokens": 37310072.0,
      "step": 17975
    },
    {
      "entropy": 5.51202883720398,
      "epoch": 1.6309869375907113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047370066559392033,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21039992719888687,
      "num_tokens": 37320278.0,
      "step": 17980
    },
    {
      "entropy": 5.513598346710205,
      "epoch": 1.6314404934687954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047368548325543366,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2022240102291107,
      "num_tokens": 37330741.0,
      "step": 17985
    },
    {
      "entropy": 5.536466026306153,
      "epoch": 1.6318940493468794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004736702968080804,
      "loss": 5.0565,
      "mean_token_accuracy": 0.1967114433646202,
      "num_tokens": 37341213.0,
      "step": 17990
    },
    {
      "entropy": 5.467316484451294,
      "epoch": 1.6323476052249637,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047365510625217483,
      "loss": 4.923,
      "mean_token_accuracy": 0.20637372881174088,
      "num_tokens": 37350528.0,
      "step": 17995
    },
    {
      "entropy": 5.476673173904419,
      "epoch": 1.632801161103048,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047363991158803117,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20116400420665742,
      "num_tokens": 37359938.0,
      "step": 18000
    },
    {
      "epoch": 1.632801161103048,
      "eval_entropy": 5.256809419744155,
      "eval_loss": 5.060100555419922,
      "eval_mean_token_accuracy": 0.21065192328706991,
      "eval_num_tokens": 37359938.0,
      "eval_runtime": 20.499,
      "eval_samples_per_second": 1724.964,
      "eval_steps_per_second": 215.62,
      "step": 18000
    },
    {
      "entropy": 5.511171817779541,
      "epoch": 1.6332547169811322,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047362471281596334,
      "loss": 5.0259,
      "mean_token_accuracy": 0.20339687019586564,
      "num_tokens": 37369744.0,
      "step": 18005
    },
    {
      "entropy": 5.547972631454468,
      "epoch": 1.6337082728592163,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047360950993628585,
      "loss": 5.107,
      "mean_token_accuracy": 0.1997305393218994,
      "num_tokens": 37380578.0,
      "step": 18010
    },
    {
      "entropy": 5.515058755874634,
      "epoch": 1.6341618287373003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004735943029493132,
      "loss": 5.016,
      "mean_token_accuracy": 0.20206762105226517,
      "num_tokens": 37390237.0,
      "step": 18015
    },
    {
      "entropy": 5.5362725257873535,
      "epoch": 1.6346153846153846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047357909185535965,
      "loss": 5.061,
      "mean_token_accuracy": 0.20199340134859084,
      "num_tokens": 37400261.0,
      "step": 18020
    },
    {
      "entropy": 5.530751276016235,
      "epoch": 1.6350689404934688,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004735638766547399,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20671603828668594,
      "num_tokens": 37410012.0,
      "step": 18025
    },
    {
      "entropy": 5.550591659545899,
      "epoch": 1.6355224963715531,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047354865734776857,
      "loss": 5.1259,
      "mean_token_accuracy": 0.1921493247151375,
      "num_tokens": 37420530.0,
      "step": 18030
    },
    {
      "entropy": 5.527346992492676,
      "epoch": 1.6359760522496372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004735334339347605,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21734682321548462,
      "num_tokens": 37431146.0,
      "step": 18035
    },
    {
      "entropy": 5.5188727378845215,
      "epoch": 1.6364296081277212,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047351820641603035,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20577428489923477,
      "num_tokens": 37441253.0,
      "step": 18040
    },
    {
      "entropy": 5.41828088760376,
      "epoch": 1.6368831640058055,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004735029747918931,
      "loss": 5.0074,
      "mean_token_accuracy": 0.1972391352057457,
      "num_tokens": 37451326.0,
      "step": 18045
    },
    {
      "entropy": 5.448575639724732,
      "epoch": 1.6373367198838897,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047348773906266374,
      "loss": 4.9642,
      "mean_token_accuracy": 0.20745179802179337,
      "num_tokens": 37462462.0,
      "step": 18050
    },
    {
      "entropy": 5.457478761672974,
      "epoch": 1.637790275761974,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004734724992286574,
      "loss": 4.9649,
      "mean_token_accuracy": 0.20242210179567338,
      "num_tokens": 37473016.0,
      "step": 18055
    },
    {
      "entropy": 5.5375182151794435,
      "epoch": 1.638243831640058,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004734572552901892,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20301338881254197,
      "num_tokens": 37483737.0,
      "step": 18060
    },
    {
      "entropy": 5.480352878570557,
      "epoch": 1.638697387518142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004734420072475743,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20475003719329835,
      "num_tokens": 37493070.0,
      "step": 18065
    },
    {
      "entropy": 5.500450086593628,
      "epoch": 1.6391509433962264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004734267551011282,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20884710401296616,
      "num_tokens": 37503696.0,
      "step": 18070
    },
    {
      "entropy": 5.467238903045654,
      "epoch": 1.6396044992743106,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734114988511662,
      "loss": 4.9936,
      "mean_token_accuracy": 0.20570732206106185,
      "num_tokens": 37513468.0,
      "step": 18075
    },
    {
      "entropy": 5.403902006149292,
      "epoch": 1.640058055152395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004733962384980039,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21198950707912445,
      "num_tokens": 37523631.0,
      "step": 18080
    },
    {
      "entropy": 5.525754356384278,
      "epoch": 1.640511611030479,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004733809740419568,
      "loss": 5.0658,
      "mean_token_accuracy": 0.20697355270385742,
      "num_tokens": 37534602.0,
      "step": 18085
    },
    {
      "entropy": 5.4536527633667,
      "epoch": 1.640965166908563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047336570548334053,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2045159474015236,
      "num_tokens": 37544547.0,
      "step": 18090
    },
    {
      "entropy": 5.553745985031128,
      "epoch": 1.6414187227866472,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047335043282247105,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20612505972385406,
      "num_tokens": 37554903.0,
      "step": 18095
    },
    {
      "entropy": 5.479746913909912,
      "epoch": 1.6418722786647315,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047333515605966396,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2023545578122139,
      "num_tokens": 37565586.0,
      "step": 18100
    },
    {
      "entropy": 5.466207647323609,
      "epoch": 1.6423258345428158,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047331987519523533,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2078264221549034,
      "num_tokens": 37575996.0,
      "step": 18105
    },
    {
      "entropy": 5.462323617935181,
      "epoch": 1.6427793904208998,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004733045902295012,
      "loss": 4.9398,
      "mean_token_accuracy": 0.19835886359214783,
      "num_tokens": 37586441.0,
      "step": 18110
    },
    {
      "entropy": 5.536691761016845,
      "epoch": 1.643232946298984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047328930116277755,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20475217401981355,
      "num_tokens": 37598226.0,
      "step": 18115
    },
    {
      "entropy": 5.523672294616699,
      "epoch": 1.6436865021770681,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004732740079953806,
      "loss": 5.0482,
      "mean_token_accuracy": 0.20536264032125473,
      "num_tokens": 37609102.0,
      "step": 18120
    },
    {
      "entropy": 5.508391284942627,
      "epoch": 1.6441400580551524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004732587107276267,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20113572627305984,
      "num_tokens": 37619349.0,
      "step": 18125
    },
    {
      "entropy": 5.4941630363464355,
      "epoch": 1.6445936139332367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047324340935983225,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21285420358181,
      "num_tokens": 37629844.0,
      "step": 18130
    },
    {
      "entropy": 5.443892908096314,
      "epoch": 1.6450471698113207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047322810389231333,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2090269923210144,
      "num_tokens": 37639339.0,
      "step": 18135
    },
    {
      "entropy": 5.4682563781738285,
      "epoch": 1.645500725689405,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004732127943253868,
      "loss": 4.9682,
      "mean_token_accuracy": 0.20977348536252977,
      "num_tokens": 37648836.0,
      "step": 18140
    },
    {
      "entropy": 5.521398591995239,
      "epoch": 1.645954281567489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004731974806593693,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21081233620643616,
      "num_tokens": 37658424.0,
      "step": 18145
    },
    {
      "entropy": 5.499155282974243,
      "epoch": 1.6464078374455733,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047318216289457727,
      "loss": 4.9823,
      "mean_token_accuracy": 0.206360924243927,
      "num_tokens": 37669849.0,
      "step": 18150
    },
    {
      "entropy": 5.478052806854248,
      "epoch": 1.6468613933236576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047316684103132766,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20385076254606246,
      "num_tokens": 37678987.0,
      "step": 18155
    },
    {
      "entropy": 5.540914392471313,
      "epoch": 1.6473149492017418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004731515150699372,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20924765914678572,
      "num_tokens": 37689692.0,
      "step": 18160
    },
    {
      "entropy": 5.549307727813721,
      "epoch": 1.6477685050798259,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047313618501072303,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19636880308389665,
      "num_tokens": 37700092.0,
      "step": 18165
    },
    {
      "entropy": 5.50684289932251,
      "epoch": 1.64822206095791,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047312085085400194,
      "loss": 5.049,
      "mean_token_accuracy": 0.2032184660434723,
      "num_tokens": 37710305.0,
      "step": 18170
    },
    {
      "entropy": 5.629374217987061,
      "epoch": 1.6486756168359942,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047310551260009126,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19262295216321945,
      "num_tokens": 37720451.0,
      "step": 18175
    },
    {
      "entropy": 5.553662490844727,
      "epoch": 1.6491291727140784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000473090170249308,
      "loss": 5.024,
      "mean_token_accuracy": 0.2077767089009285,
      "num_tokens": 37731952.0,
      "step": 18180
    },
    {
      "entropy": 5.46588020324707,
      "epoch": 1.6495827285921627,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004730748238019696,
      "loss": 4.9644,
      "mean_token_accuracy": 0.1991961494088173,
      "num_tokens": 37742201.0,
      "step": 18185
    },
    {
      "entropy": 5.4644073963165285,
      "epoch": 1.6500362844702468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047305947325839337,
      "loss": 4.981,
      "mean_token_accuracy": 0.20674336403608323,
      "num_tokens": 37753317.0,
      "step": 18190
    },
    {
      "entropy": 5.529191160202027,
      "epoch": 1.6504898403483308,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004730441186188968,
      "loss": 5.0391,
      "mean_token_accuracy": 0.202471961081028,
      "num_tokens": 37763452.0,
      "step": 18195
    },
    {
      "entropy": 5.50238790512085,
      "epoch": 1.650943396226415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047302875988379734,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2040795624256134,
      "num_tokens": 37772804.0,
      "step": 18200
    },
    {
      "entropy": 5.5216504573822025,
      "epoch": 1.6513969521044993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047301339705341266,
      "loss": 5.0328,
      "mean_token_accuracy": 0.1995303675532341,
      "num_tokens": 37784441.0,
      "step": 18205
    },
    {
      "entropy": 5.498663234710693,
      "epoch": 1.6518505079825836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047299803012806056,
      "loss": 5.037,
      "mean_token_accuracy": 0.20703674852848053,
      "num_tokens": 37795103.0,
      "step": 18210
    },
    {
      "entropy": 5.489672660827637,
      "epoch": 1.6523040638606676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004729826591080586,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20495232492685317,
      "num_tokens": 37806654.0,
      "step": 18215
    },
    {
      "entropy": 5.523226261138916,
      "epoch": 1.6527576197387517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047296728399372483,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2009401485323906,
      "num_tokens": 37816673.0,
      "step": 18220
    },
    {
      "entropy": 5.539667224884033,
      "epoch": 1.653211175616836,
      "grad_norm": 1.0,
      "learning_rate": 0.00047295190478537725,
      "loss": 5.0383,
      "mean_token_accuracy": 0.19783694297075272,
      "num_tokens": 37827389.0,
      "step": 18225
    },
    {
      "entropy": 5.600225400924683,
      "epoch": 1.6536647314949202,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047293652148333376,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2023376613855362,
      "num_tokens": 37838087.0,
      "step": 18230
    },
    {
      "entropy": 5.41401104927063,
      "epoch": 1.6541182873730045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047292113408791266,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2029065489768982,
      "num_tokens": 37848630.0,
      "step": 18235
    },
    {
      "entropy": 5.410132217407226,
      "epoch": 1.6545718432510885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047290574259943195,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21658093780279158,
      "num_tokens": 37858689.0,
      "step": 18240
    },
    {
      "entropy": 5.547229766845703,
      "epoch": 1.6550253991291726,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047289034701821013,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2034848541021347,
      "num_tokens": 37869049.0,
      "step": 18245
    },
    {
      "entropy": 5.473025465011597,
      "epoch": 1.6554789550072568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047287494734456545,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20671870857477187,
      "num_tokens": 37879495.0,
      "step": 18250
    },
    {
      "entropy": 5.523059368133545,
      "epoch": 1.655932510885341,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047285954357881646,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20215840339660646,
      "num_tokens": 37889442.0,
      "step": 18255
    },
    {
      "entropy": 5.5659104347229,
      "epoch": 1.6563860667634254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004728441357212817,
      "loss": 5.17,
      "mean_token_accuracy": 0.18997660279273987,
      "num_tokens": 37900147.0,
      "step": 18260
    },
    {
      "entropy": 5.541626787185669,
      "epoch": 1.6568396226415094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004728287237722798,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20387801378965378,
      "num_tokens": 37910307.0,
      "step": 18265
    },
    {
      "entropy": 5.4466005802154545,
      "epoch": 1.6572931785195935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004728133077321295,
      "loss": 4.96,
      "mean_token_accuracy": 0.20442774891853333,
      "num_tokens": 37920838.0,
      "step": 18270
    },
    {
      "entropy": 5.528741884231567,
      "epoch": 1.6577467343976777,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047279788760114954,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20524655878543854,
      "num_tokens": 37930717.0,
      "step": 18275
    },
    {
      "entropy": 5.556101369857788,
      "epoch": 1.658200290275762,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004727824633796589,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20111787021160127,
      "num_tokens": 37941254.0,
      "step": 18280
    },
    {
      "entropy": 5.474472284317017,
      "epoch": 1.6586538461538463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004727670350679765,
      "loss": 5.0176,
      "mean_token_accuracy": 0.2114999607205391,
      "num_tokens": 37952132.0,
      "step": 18285
    },
    {
      "entropy": 5.490719175338745,
      "epoch": 1.6591074020319303,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004727516026664214,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2099560484290123,
      "num_tokens": 37962026.0,
      "step": 18290
    },
    {
      "entropy": 5.494945859909057,
      "epoch": 1.6595609579100146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004727361661753128,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2045206382870674,
      "num_tokens": 37972911.0,
      "step": 18295
    },
    {
      "entropy": 5.5300699234008786,
      "epoch": 1.6600145137880986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727207255949699,
      "loss": 4.9821,
      "mean_token_accuracy": 0.20639899522066116,
      "num_tokens": 37983577.0,
      "step": 18300
    },
    {
      "entropy": 5.53853759765625,
      "epoch": 1.6604680696661829,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047270528092571197,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20507221221923827,
      "num_tokens": 37994294.0,
      "step": 18305
    },
    {
      "entropy": 5.513009262084961,
      "epoch": 1.6609216255442671,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726898321678584,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20539045631885527,
      "num_tokens": 38005144.0,
      "step": 18310
    },
    {
      "entropy": 5.481329345703125,
      "epoch": 1.6613751814223512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004726743793217288,
      "loss": 5.0491,
      "mean_token_accuracy": 0.1994861990213394,
      "num_tokens": 38015277.0,
      "step": 18315
    },
    {
      "entropy": 5.52866039276123,
      "epoch": 1.6618287373004355,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047265892238764265,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20707998871803285,
      "num_tokens": 38027530.0,
      "step": 18320
    },
    {
      "entropy": 5.577838993072509,
      "epoch": 1.6622822931785195,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004726434613659196,
      "loss": 5.0722,
      "mean_token_accuracy": 0.20743615478277205,
      "num_tokens": 38036895.0,
      "step": 18325
    },
    {
      "entropy": 5.547301959991455,
      "epoch": 1.6627358490566038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726279962568794,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20268693417310715,
      "num_tokens": 38047001.0,
      "step": 18330
    },
    {
      "entropy": 5.52684211730957,
      "epoch": 1.663189404934688,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004726125270608419,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2136825889348984,
      "num_tokens": 38056815.0,
      "step": 18335
    },
    {
      "entropy": 5.51765546798706,
      "epoch": 1.6636429608127723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047259705377812703,
      "loss": 5.0644,
      "mean_token_accuracy": 0.20229309499263765,
      "num_tokens": 38067139.0,
      "step": 18340
    },
    {
      "entropy": 5.5113856315612795,
      "epoch": 1.6640965166908563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004725815764090546,
      "loss": 4.979,
      "mean_token_accuracy": 0.21099628508090973,
      "num_tokens": 38076824.0,
      "step": 18345
    },
    {
      "entropy": 5.525916671752929,
      "epoch": 1.6645500725689404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047256609495394494,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2043920114636421,
      "num_tokens": 38087517.0,
      "step": 18350
    },
    {
      "entropy": 5.540383386611938,
      "epoch": 1.6650036284470247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472550609413118,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20258957743644715,
      "num_tokens": 38098499.0,
      "step": 18355
    },
    {
      "entropy": 5.475964975357056,
      "epoch": 1.665457184325109,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004725351197868942,
      "loss": 5.002,
      "mean_token_accuracy": 0.21008841395378114,
      "num_tokens": 38108327.0,
      "step": 18360
    },
    {
      "entropy": 5.543871355056763,
      "epoch": 1.6659107402031932,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047251962607559373,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20743479281663896,
      "num_tokens": 38118667.0,
      "step": 18365
    },
    {
      "entropy": 5.459502553939819,
      "epoch": 1.6663642960812772,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047250412827953717,
      "loss": 4.9724,
      "mean_token_accuracy": 0.20886190384626388,
      "num_tokens": 38128571.0,
      "step": 18370
    },
    {
      "entropy": 5.55834002494812,
      "epoch": 1.6668178519593613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004724886263990448,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19585662037134172,
      "num_tokens": 38138921.0,
      "step": 18375
    },
    {
      "entropy": 5.57820463180542,
      "epoch": 1.6672714078374455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004724731204344372,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20742853879928588,
      "num_tokens": 38148858.0,
      "step": 18380
    },
    {
      "entropy": 5.438315725326538,
      "epoch": 1.6677249637155298,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047245761038603536,
      "loss": 4.95,
      "mean_token_accuracy": 0.21081405878067017,
      "num_tokens": 38159107.0,
      "step": 18385
    },
    {
      "entropy": 5.578572463989258,
      "epoch": 1.668178519593614,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047244209625415975,
      "loss": 5.0768,
      "mean_token_accuracy": 0.2027849793434143,
      "num_tokens": 38169228.0,
      "step": 18390
    },
    {
      "entropy": 5.588312578201294,
      "epoch": 1.6686320754716981,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004724265780391312,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20619215816259384,
      "num_tokens": 38179175.0,
      "step": 18395
    },
    {
      "entropy": 5.521461391448975,
      "epoch": 1.6690856313497822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004724110557412708,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2078677698969841,
      "num_tokens": 38188537.0,
      "step": 18400
    },
    {
      "entropy": 5.499007797241211,
      "epoch": 1.6695391872278664,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047239552936089943,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20057985186576843,
      "num_tokens": 38198972.0,
      "step": 18405
    },
    {
      "entropy": 5.501136779785156,
      "epoch": 1.6699927431059507,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047237999889833816,
      "loss": 4.9339,
      "mean_token_accuracy": 0.20863931775093078,
      "num_tokens": 38209472.0,
      "step": 18410
    },
    {
      "entropy": 5.422840404510498,
      "epoch": 1.670446298984035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047236446435390826,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2069828063249588,
      "num_tokens": 38219977.0,
      "step": 18415
    },
    {
      "entropy": 5.473114490509033,
      "epoch": 1.670899854862119,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047234892572793093,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20733280777931212,
      "num_tokens": 38230492.0,
      "step": 18420
    },
    {
      "entropy": 5.513241386413574,
      "epoch": 1.671353410740203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004723333830207275,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19537861049175262,
      "num_tokens": 38240799.0,
      "step": 18425
    },
    {
      "entropy": 5.512304449081421,
      "epoch": 1.6718069666182873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047231783623261954,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20044245421886445,
      "num_tokens": 38251439.0,
      "step": 18430
    },
    {
      "entropy": 5.5602222919464115,
      "epoch": 1.6722605224963716,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004723022853639283,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19738350361585616,
      "num_tokens": 38261795.0,
      "step": 18435
    },
    {
      "entropy": 5.462391996383667,
      "epoch": 1.6727140783744558,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004722867304149756,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20544118583202362,
      "num_tokens": 38272172.0,
      "step": 18440
    },
    {
      "entropy": 5.519338464736938,
      "epoch": 1.67316763425254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722711713860831,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19758394062519075,
      "num_tokens": 38282829.0,
      "step": 18445
    },
    {
      "entropy": 5.564418315887451,
      "epoch": 1.673621190130624,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004722556082775724,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20760981738567352,
      "num_tokens": 38293793.0,
      "step": 18450
    },
    {
      "entropy": 5.554120635986328,
      "epoch": 1.6740747460087082,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004722400410897655,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20341993421316146,
      "num_tokens": 38302577.0,
      "step": 18455
    },
    {
      "entropy": 5.5128473281860355,
      "epoch": 1.6745283018867925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004722244698229842,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20554074943065642,
      "num_tokens": 38312571.0,
      "step": 18460
    },
    {
      "entropy": 5.5175055980682375,
      "epoch": 1.6749818577648767,
      "grad_norm": 1.0,
      "learning_rate": 0.00047220889447755073,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21119394749403,
      "num_tokens": 38323544.0,
      "step": 18465
    },
    {
      "entropy": 5.557501602172851,
      "epoch": 1.6754354136429608,
      "grad_norm": 1.125,
      "learning_rate": 0.0004721933150537869,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20210747569799423,
      "num_tokens": 38334509.0,
      "step": 18470
    },
    {
      "entropy": 5.453472852706909,
      "epoch": 1.675888969521045,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047217773155201526,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20459824055433273,
      "num_tokens": 38344747.0,
      "step": 18475
    },
    {
      "entropy": 5.53552279472351,
      "epoch": 1.676342525399129,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047216214397255776,
      "loss": 5.0651,
      "mean_token_accuracy": 0.19655898809432984,
      "num_tokens": 38355671.0,
      "step": 18480
    },
    {
      "entropy": 5.517773103713989,
      "epoch": 1.6767960812772134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047214655231573693,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2041994571685791,
      "num_tokens": 38366368.0,
      "step": 18485
    },
    {
      "entropy": 5.513717985153198,
      "epoch": 1.6772496371552976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004721309565818752,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20049837082624436,
      "num_tokens": 38377322.0,
      "step": 18490
    },
    {
      "entropy": 5.584578275680542,
      "epoch": 1.6777031930333817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047211535677129496,
      "loss": 5.087,
      "mean_token_accuracy": 0.20294539034366607,
      "num_tokens": 38389204.0,
      "step": 18495
    },
    {
      "entropy": 5.510067462921143,
      "epoch": 1.678156748911466,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047209975288431893,
      "loss": 5.0408,
      "mean_token_accuracy": 0.20719282180070878,
      "num_tokens": 38399117.0,
      "step": 18500
    },
    {
      "entropy": 5.522906494140625,
      "epoch": 1.67861030478955,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047208414492126986,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19984093606472014,
      "num_tokens": 38410341.0,
      "step": 18505
    },
    {
      "entropy": 5.558302021026611,
      "epoch": 1.6790638606676342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047206853288247043,
      "loss": 5.0797,
      "mean_token_accuracy": 0.19582510143518447,
      "num_tokens": 38422742.0,
      "step": 18510
    },
    {
      "entropy": 5.5316057205200195,
      "epoch": 1.6795174165457185,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047205291676824345,
      "loss": 5.0714,
      "mean_token_accuracy": 0.197774276137352,
      "num_tokens": 38433022.0,
      "step": 18515
    },
    {
      "entropy": 5.5605991840362545,
      "epoch": 1.6799709724238028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472037296578912,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20111667215824128,
      "num_tokens": 38444325.0,
      "step": 18520
    },
    {
      "entropy": 5.5038533210754395,
      "epoch": 1.6804245283018868,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000472021672314799,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21152271330356598,
      "num_tokens": 38454576.0,
      "step": 18525
    },
    {
      "entropy": 5.488370323181153,
      "epoch": 1.6808780841799709,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047200604397622763,
      "loss": 4.9664,
      "mean_token_accuracy": 0.209333173930645,
      "num_tokens": 38465090.0,
      "step": 18530
    },
    {
      "entropy": 5.56174955368042,
      "epoch": 1.6813316400580551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719904115635211,
      "loss": 5.2117,
      "mean_token_accuracy": 0.1902170941233635,
      "num_tokens": 38477253.0,
      "step": 18535
    },
    {
      "entropy": 5.581843423843384,
      "epoch": 1.6817851959361394,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004719747750770026,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20014643520116807,
      "num_tokens": 38488040.0,
      "step": 18540
    },
    {
      "entropy": 5.487832021713257,
      "epoch": 1.6822387518142237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004719591345169956,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2080007091164589,
      "num_tokens": 38498016.0,
      "step": 18545
    },
    {
      "entropy": 5.582905292510986,
      "epoch": 1.6826923076923077,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047194348988382345,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20082142800092698,
      "num_tokens": 38508996.0,
      "step": 18550
    },
    {
      "entropy": 5.51684536933899,
      "epoch": 1.6831458635703918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004719278411778097,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20538436770439147,
      "num_tokens": 38519608.0,
      "step": 18555
    },
    {
      "entropy": 5.5028585433959964,
      "epoch": 1.683599419448476,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047191218839927817,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20028443336486818,
      "num_tokens": 38530737.0,
      "step": 18560
    },
    {
      "entropy": 5.497863245010376,
      "epoch": 1.6840529753265603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004718965315485523,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2058178350329399,
      "num_tokens": 38540925.0,
      "step": 18565
    },
    {
      "entropy": 5.513472843170166,
      "epoch": 1.6845065312046446,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004718808706259561,
      "loss": 5.0513,
      "mean_token_accuracy": 0.198410527408123,
      "num_tokens": 38553176.0,
      "step": 18570
    },
    {
      "entropy": 5.547786235809326,
      "epoch": 1.6849600870827286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047186520563181325,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1984600692987442,
      "num_tokens": 38563394.0,
      "step": 18575
    },
    {
      "entropy": 5.571151638031006,
      "epoch": 1.6854136429608126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004718495365664477,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20615155547857283,
      "num_tokens": 38573243.0,
      "step": 18580
    },
    {
      "entropy": 5.506847476959228,
      "epoch": 1.685867198838897,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047183386343018364,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20158745348453522,
      "num_tokens": 38584014.0,
      "step": 18585
    },
    {
      "entropy": 5.50274543762207,
      "epoch": 1.6863207547169812,
      "grad_norm": 1.0,
      "learning_rate": 0.0004718181862233451,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2084433391690254,
      "num_tokens": 38594467.0,
      "step": 18590
    },
    {
      "entropy": 5.584379100799561,
      "epoch": 1.6867743105950654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004718025049462563,
      "loss": 5.0571,
      "mean_token_accuracy": 0.20305939614772797,
      "num_tokens": 38604407.0,
      "step": 18595
    },
    {
      "entropy": 5.560861873626709,
      "epoch": 1.6872278664731495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004717868195992416,
      "loss": 5.0528,
      "mean_token_accuracy": 0.2088589459657669,
      "num_tokens": 38614508.0,
      "step": 18600
    },
    {
      "entropy": 5.575065040588379,
      "epoch": 1.6876814223512335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004717711301826253,
      "loss": 5.0651,
      "mean_token_accuracy": 0.20356738418340684,
      "num_tokens": 38626244.0,
      "step": 18605
    },
    {
      "entropy": 5.477683877944946,
      "epoch": 1.6881349782293178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047175543669673186,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20338715761899948,
      "num_tokens": 38636694.0,
      "step": 18610
    },
    {
      "entropy": 5.592471599578857,
      "epoch": 1.688588534107402,
      "grad_norm": 1.0,
      "learning_rate": 0.00047173973914188587,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19840389341115952,
      "num_tokens": 38647414.0,
      "step": 18615
    },
    {
      "entropy": 5.522599124908448,
      "epoch": 1.6890420899854863,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047172403751841196,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2026630625128746,
      "num_tokens": 38657425.0,
      "step": 18620
    },
    {
      "entropy": 5.5504255294799805,
      "epoch": 1.6894956458635704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047170833182663485,
      "loss": 5.0395,
      "mean_token_accuracy": 0.2048941120505333,
      "num_tokens": 38667751.0,
      "step": 18625
    },
    {
      "entropy": 5.4390911102294925,
      "epoch": 1.6899492017416544,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004716926220668793,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2103493943810463,
      "num_tokens": 38676918.0,
      "step": 18630
    },
    {
      "entropy": 5.449511766433716,
      "epoch": 1.6904027576197387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047167690823947015,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21317176818847655,
      "num_tokens": 38685887.0,
      "step": 18635
    },
    {
      "entropy": 5.46210298538208,
      "epoch": 1.690856313497823,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716611903447325,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20897857546806337,
      "num_tokens": 38696179.0,
      "step": 18640
    },
    {
      "entropy": 5.506293821334839,
      "epoch": 1.6913098693759072,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716454683829912,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20195146650075912,
      "num_tokens": 38705777.0,
      "step": 18645
    },
    {
      "entropy": 5.5022331237792965,
      "epoch": 1.6917634252539913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047162974235457163,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2052154392004013,
      "num_tokens": 38717709.0,
      "step": 18650
    },
    {
      "entropy": 5.513932704925537,
      "epoch": 1.6922169811320755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716140122597988,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20939967036247253,
      "num_tokens": 38729517.0,
      "step": 18655
    },
    {
      "entropy": 5.5043616771698,
      "epoch": 1.6926705370101596,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004715982780989981,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2061331108212471,
      "num_tokens": 38740878.0,
      "step": 18660
    },
    {
      "entropy": 5.506402587890625,
      "epoch": 1.6931240928882438,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047158253987249487,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20769711285829545,
      "num_tokens": 38751291.0,
      "step": 18665
    },
    {
      "entropy": 5.484677314758301,
      "epoch": 1.693577648766328,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047156679758061465,
      "loss": 4.948,
      "mean_token_accuracy": 0.21552773416042328,
      "num_tokens": 38760940.0,
      "step": 18670
    },
    {
      "entropy": 5.498112201690674,
      "epoch": 1.6940312046444121,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047155105122368297,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21133071333169937,
      "num_tokens": 38770240.0,
      "step": 18675
    },
    {
      "entropy": 5.448402261734008,
      "epoch": 1.6944847605224964,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004715353008020255,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20739692449569702,
      "num_tokens": 38780452.0,
      "step": 18680
    },
    {
      "entropy": 5.557797908782959,
      "epoch": 1.6949383164005805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047151954631596794,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2015103355050087,
      "num_tokens": 38790763.0,
      "step": 18685
    },
    {
      "entropy": 5.506512022018432,
      "epoch": 1.6953918722786647,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047150378776583603,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21901689022779464,
      "num_tokens": 38802164.0,
      "step": 18690
    },
    {
      "entropy": 5.507416725158691,
      "epoch": 1.695845428156749,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004714880251519557,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20090519338846208,
      "num_tokens": 38812901.0,
      "step": 18695
    },
    {
      "entropy": 5.484809684753418,
      "epoch": 1.6962989840348333,
      "grad_norm": 0.953125,
      "learning_rate": 0.000471472258474653,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20857700258493422,
      "num_tokens": 38823305.0,
      "step": 18700
    },
    {
      "entropy": 5.495634984970093,
      "epoch": 1.6967525399129173,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047145648773425384,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2041999951004982,
      "num_tokens": 38833602.0,
      "step": 18705
    },
    {
      "entropy": 5.514457893371582,
      "epoch": 1.6972060957910013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047144071293108456,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20269525349140166,
      "num_tokens": 38843950.0,
      "step": 18710
    },
    {
      "entropy": 5.519618082046509,
      "epoch": 1.6976596516690856,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004714249340654712,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2041005238890648,
      "num_tokens": 38853753.0,
      "step": 18715
    },
    {
      "entropy": 5.464105033874512,
      "epoch": 1.6981132075471699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004714091511377402,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2089301362633705,
      "num_tokens": 38863394.0,
      "step": 18720
    },
    {
      "entropy": 5.477627611160278,
      "epoch": 1.6985667634252541,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047139336414821784,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19901751428842546,
      "num_tokens": 38872972.0,
      "step": 18725
    },
    {
      "entropy": 5.461693048477173,
      "epoch": 1.6990203193033382,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047137757309723073,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21485331356525422,
      "num_tokens": 38883251.0,
      "step": 18730
    },
    {
      "entropy": 5.4876368045806885,
      "epoch": 1.6994738751814222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004713617779851053,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20316289067268373,
      "num_tokens": 38893274.0,
      "step": 18735
    },
    {
      "entropy": 5.559378385543823,
      "epoch": 1.6999274310595065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004713459788121683,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19765110164880753,
      "num_tokens": 38903870.0,
      "step": 18740
    },
    {
      "entropy": 5.5256421089172365,
      "epoch": 1.7003809869375908,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004713301755787464,
      "loss": 5.0435,
      "mean_token_accuracy": 0.2014593616127968,
      "num_tokens": 38914921.0,
      "step": 18745
    },
    {
      "entropy": 5.546377801895142,
      "epoch": 1.700834542815675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004713143682851665,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20599546134471894,
      "num_tokens": 38924630.0,
      "step": 18750
    },
    {
      "entropy": 5.454897737503051,
      "epoch": 1.701288098693759,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004712985569317553,
      "loss": 4.9742,
      "mean_token_accuracy": 0.204600790143013,
      "num_tokens": 38934926.0,
      "step": 18755
    },
    {
      "entropy": 5.56922607421875,
      "epoch": 1.7017416545718431,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004712827415188401,
      "loss": 5.1371,
      "mean_token_accuracy": 0.1945800855755806,
      "num_tokens": 38944513.0,
      "step": 18760
    },
    {
      "entropy": 5.561591768264771,
      "epoch": 1.7021952104499274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004712669220467476,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2074137046933174,
      "num_tokens": 38955918.0,
      "step": 18765
    },
    {
      "entropy": 5.548800992965698,
      "epoch": 1.7026487663280117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047125109851580533,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20577421486377717,
      "num_tokens": 38966221.0,
      "step": 18770
    },
    {
      "entropy": 5.49301290512085,
      "epoch": 1.703102322206096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047123527092634025,
      "loss": 4.9745,
      "mean_token_accuracy": 0.20071849822998047,
      "num_tokens": 38976463.0,
      "step": 18775
    },
    {
      "entropy": 5.4543568134307865,
      "epoch": 1.70355587808418,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004712194392786797,
      "loss": 4.9074,
      "mean_token_accuracy": 0.20934713929891585,
      "num_tokens": 38986375.0,
      "step": 18780
    },
    {
      "entropy": 5.595263910293579,
      "epoch": 1.704009433962264,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047120360357315127,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20466777235269545,
      "num_tokens": 38996116.0,
      "step": 18785
    },
    {
      "entropy": 5.468366289138794,
      "epoch": 1.7044629898403483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047118776381008216,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20442012697458267,
      "num_tokens": 39005846.0,
      "step": 18790
    },
    {
      "entropy": 5.541969347000122,
      "epoch": 1.7049165457184325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047117191998980026,
      "loss": 5.0147,
      "mean_token_accuracy": 0.19670799374580383,
      "num_tokens": 39016744.0,
      "step": 18795
    },
    {
      "entropy": 5.525983285903931,
      "epoch": 1.7053701015965168,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047115607211263295,
      "loss": 5.0568,
      "mean_token_accuracy": 0.19816389679908752,
      "num_tokens": 39026725.0,
      "step": 18800
    },
    {
      "entropy": 5.554859399795532,
      "epoch": 1.7058236574746009,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047114022017890814,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20423724949359895,
      "num_tokens": 39036851.0,
      "step": 18805
    },
    {
      "entropy": 5.502058696746826,
      "epoch": 1.706277213352685,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004711243641889536,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2052324578166008,
      "num_tokens": 39047179.0,
      "step": 18810
    },
    {
      "entropy": 5.449604272842407,
      "epoch": 1.7067307692307692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047110850414309714,
      "loss": 4.8444,
      "mean_token_accuracy": 0.21609296649694443,
      "num_tokens": 39057758.0,
      "step": 18815
    },
    {
      "entropy": 5.511652517318725,
      "epoch": 1.7071843251088534,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047109264004166696,
      "loss": 5.0617,
      "mean_token_accuracy": 0.19992961287498473,
      "num_tokens": 39068142.0,
      "step": 18820
    },
    {
      "entropy": 5.4653867244720455,
      "epoch": 1.7076378809869377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004710767718849909,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2027351438999176,
      "num_tokens": 39078157.0,
      "step": 18825
    },
    {
      "entropy": 5.488234376907348,
      "epoch": 1.7080914368650217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004710608996733973,
      "loss": 5.0097,
      "mean_token_accuracy": 0.211202010512352,
      "num_tokens": 39089090.0,
      "step": 18830
    },
    {
      "entropy": 5.56004695892334,
      "epoch": 1.708544992743106,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047104502340721433,
      "loss": 5.0614,
      "mean_token_accuracy": 0.19800453633069992,
      "num_tokens": 39099079.0,
      "step": 18835
    },
    {
      "entropy": 5.526583051681518,
      "epoch": 1.70899854862119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047102914308677033,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2017642602324486,
      "num_tokens": 39109927.0,
      "step": 18840
    },
    {
      "entropy": 5.538265323638916,
      "epoch": 1.7094521044992743,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047101325871239363,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2031232237815857,
      "num_tokens": 39119806.0,
      "step": 18845
    },
    {
      "entropy": 5.558316230773926,
      "epoch": 1.7099056603773586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004709973702844128,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20081631392240523,
      "num_tokens": 39130147.0,
      "step": 18850
    },
    {
      "entropy": 5.448058223724365,
      "epoch": 1.7103592162554426,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047098147780315645,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21496572643518447,
      "num_tokens": 39139937.0,
      "step": 18855
    },
    {
      "entropy": 5.492663812637329,
      "epoch": 1.710812772133527,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004709655812689532,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20803554952144623,
      "num_tokens": 39149977.0,
      "step": 18860
    },
    {
      "entropy": 5.475897932052613,
      "epoch": 1.711266328011611,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004709496806821318,
      "loss": 5.0302,
      "mean_token_accuracy": 0.20517557114362717,
      "num_tokens": 39161076.0,
      "step": 18865
    },
    {
      "entropy": 5.541501426696778,
      "epoch": 1.7117198838896952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047093377604302106,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19836925715208054,
      "num_tokens": 39170776.0,
      "step": 18870
    },
    {
      "entropy": 5.562673425674438,
      "epoch": 1.7121734397677795,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004709178673519499,
      "loss": 5.0397,
      "mean_token_accuracy": 0.19924242794513702,
      "num_tokens": 39181134.0,
      "step": 18875
    },
    {
      "entropy": 5.544519472122192,
      "epoch": 1.7126269956458637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004709019546092473,
      "loss": 5.1125,
      "mean_token_accuracy": 0.19395548850297928,
      "num_tokens": 39193331.0,
      "step": 18880
    },
    {
      "entropy": 5.582593297958374,
      "epoch": 1.7130805515239478,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047088603781524237,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19931646585464477,
      "num_tokens": 39204372.0,
      "step": 18885
    },
    {
      "entropy": 5.5269989490509035,
      "epoch": 1.7135341074020318,
      "grad_norm": 1.125,
      "learning_rate": 0.0004708701169702642,
      "loss": 4.988,
      "mean_token_accuracy": 0.2095421716570854,
      "num_tokens": 39214725.0,
      "step": 18890
    },
    {
      "entropy": 5.4547687530517575,
      "epoch": 1.713987663280116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004708541920746422,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2109682485461235,
      "num_tokens": 39225119.0,
      "step": 18895
    },
    {
      "entropy": 5.483250904083252,
      "epoch": 1.7144412191582004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004708382631287055,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20068428963422774,
      "num_tokens": 39235021.0,
      "step": 18900
    },
    {
      "entropy": 5.536379957199097,
      "epoch": 1.7148947750362846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004708223301327837,
      "loss": 5.1335,
      "mean_token_accuracy": 0.19601042419672013,
      "num_tokens": 39245830.0,
      "step": 18905
    },
    {
      "entropy": 5.585514068603516,
      "epoch": 1.7153483309143687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047080639308720617,
      "loss": 5.1009,
      "mean_token_accuracy": 0.20642118602991105,
      "num_tokens": 39255264.0,
      "step": 18910
    },
    {
      "entropy": 5.548873472213745,
      "epoch": 1.7158018867924527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047079045199230254,
      "loss": 4.9681,
      "mean_token_accuracy": 0.20718229115009307,
      "num_tokens": 39264966.0,
      "step": 18915
    },
    {
      "entropy": 5.505400085449219,
      "epoch": 1.716255442670537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047077450684840247,
      "loss": 5.0832,
      "mean_token_accuracy": 0.1983361214399338,
      "num_tokens": 39275082.0,
      "step": 18920
    },
    {
      "entropy": 5.57161955833435,
      "epoch": 1.7167089985486212,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004707585576558357,
      "loss": 5.1595,
      "mean_token_accuracy": 0.1953936904668808,
      "num_tokens": 39285426.0,
      "step": 18925
    },
    {
      "entropy": 5.6086994171142575,
      "epoch": 1.7171625544267055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004707426044149321,
      "loss": 5.0995,
      "mean_token_accuracy": 0.19788185954093934,
      "num_tokens": 39296633.0,
      "step": 18930
    },
    {
      "entropy": 5.510032510757446,
      "epoch": 1.7176161103047896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047072664712602153,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2131096675992012,
      "num_tokens": 39306775.0,
      "step": 18935
    },
    {
      "entropy": 5.508192873001098,
      "epoch": 1.7180696661828736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000470710685789434,
      "loss": 5.035,
      "mean_token_accuracy": 0.19883523732423783,
      "num_tokens": 39317917.0,
      "step": 18940
    },
    {
      "entropy": 5.535958242416382,
      "epoch": 1.7185232220609579,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047069472040549965,
      "loss": 5.036,
      "mean_token_accuracy": 0.2056172490119934,
      "num_tokens": 39328387.0,
      "step": 18945
    },
    {
      "entropy": 5.425024223327637,
      "epoch": 1.7189767779390421,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004706787509745486,
      "loss": 4.8894,
      "mean_token_accuracy": 0.2168424606323242,
      "num_tokens": 39337639.0,
      "step": 18950
    },
    {
      "entropy": 5.514598035812378,
      "epoch": 1.7194303338171264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047066277749691115,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20351004004478454,
      "num_tokens": 39347971.0,
      "step": 18955
    },
    {
      "entropy": 5.563658571243286,
      "epoch": 1.7198838896952104,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004706467999729175,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20269916355609893,
      "num_tokens": 39358038.0,
      "step": 18960
    },
    {
      "entropy": 5.555438899993897,
      "epoch": 1.7203374455732945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004706308184028983,
      "loss": 5.0489,
      "mean_token_accuracy": 0.202947998046875,
      "num_tokens": 39367834.0,
      "step": 18965
    },
    {
      "entropy": 5.551628732681275,
      "epoch": 1.7207910014513788,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004706148327871839,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20801008641719818,
      "num_tokens": 39378924.0,
      "step": 18970
    },
    {
      "entropy": 5.583839321136475,
      "epoch": 1.721244557329463,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047059884312610487,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19632723778486252,
      "num_tokens": 39390082.0,
      "step": 18975
    },
    {
      "entropy": 5.522844743728638,
      "epoch": 1.7216981132075473,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047058284941999185,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20150293707847594,
      "num_tokens": 39400485.0,
      "step": 18980
    },
    {
      "entropy": 5.5299999713897705,
      "epoch": 1.7221516690856313,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004705668516691758,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20307649224996566,
      "num_tokens": 39411601.0,
      "step": 18985
    },
    {
      "entropy": 5.523506689071655,
      "epoch": 1.7226052249637154,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047055084987398735,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2111507087945938,
      "num_tokens": 39421625.0,
      "step": 18990
    },
    {
      "entropy": 5.503273916244507,
      "epoch": 1.7230587808417996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004705348440347575,
      "loss": 5.001,
      "mean_token_accuracy": 0.20387445539236068,
      "num_tokens": 39432033.0,
      "step": 18995
    },
    {
      "entropy": 5.509365177154541,
      "epoch": 1.723512336719884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047051883415181724,
      "loss": 4.9769,
      "mean_token_accuracy": 0.20731743574142455,
      "num_tokens": 39442514.0,
      "step": 19000
    },
    {
      "entropy": 5.508768844604492,
      "epoch": 1.7239658925979682,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004705028202254977,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2113283485174179,
      "num_tokens": 39452497.0,
      "step": 19005
    },
    {
      "entropy": 5.4832415103912355,
      "epoch": 1.7244194484760522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047048680225612996,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20899880230426787,
      "num_tokens": 39462173.0,
      "step": 19010
    },
    {
      "entropy": 5.535536670684815,
      "epoch": 1.7248730043541363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704707802440454,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21252347975969316,
      "num_tokens": 39473123.0,
      "step": 19015
    },
    {
      "entropy": 5.562990760803222,
      "epoch": 1.7253265602322205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047045475418957524,
      "loss": 5.0343,
      "mean_token_accuracy": 0.19513361603021623,
      "num_tokens": 39484142.0,
      "step": 19020
    },
    {
      "entropy": 5.549053430557251,
      "epoch": 1.7257801161103048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004704387240930509,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20033205300569534,
      "num_tokens": 39494926.0,
      "step": 19025
    },
    {
      "entropy": 5.492986488342285,
      "epoch": 1.726233671988389,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704226899548041,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2030912771821022,
      "num_tokens": 39505153.0,
      "step": 19030
    },
    {
      "entropy": 5.480626821517944,
      "epoch": 1.726687227866473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047040665177516616,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2088142678141594,
      "num_tokens": 39515905.0,
      "step": 19035
    },
    {
      "entropy": 5.483314895629883,
      "epoch": 1.7271407837445574,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047039060955446886,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21209943741559983,
      "num_tokens": 39526762.0,
      "step": 19040
    },
    {
      "entropy": 5.484533214569092,
      "epoch": 1.7275943396226414,
      "grad_norm": 1.0625,
      "learning_rate": 0.000470374563293044,
      "loss": 4.9748,
      "mean_token_accuracy": 0.20307250767946244,
      "num_tokens": 39537340.0,
      "step": 19045
    },
    {
      "entropy": 5.537242794036866,
      "epoch": 1.7280478955007257,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004703585129912234,
      "loss": 4.991,
      "mean_token_accuracy": 0.20512688606977464,
      "num_tokens": 39547144.0,
      "step": 19050
    },
    {
      "entropy": 5.618268918991089,
      "epoch": 1.72850145137881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047034245864933894,
      "loss": 5.1177,
      "mean_token_accuracy": 0.1946992427110672,
      "num_tokens": 39557707.0,
      "step": 19055
    },
    {
      "entropy": 5.5895966529846195,
      "epoch": 1.728955007256894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047032640026772274,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19950711131095886,
      "num_tokens": 39567629.0,
      "step": 19060
    },
    {
      "entropy": 5.534259033203125,
      "epoch": 1.7294085631349783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004703103378467067,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20405916571617128,
      "num_tokens": 39577918.0,
      "step": 19065
    },
    {
      "entropy": 5.545049476623535,
      "epoch": 1.7298621190130623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047029427138662314,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20991769134998323,
      "num_tokens": 39587838.0,
      "step": 19070
    },
    {
      "entropy": 5.4847307205200195,
      "epoch": 1.7303156748911466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004702782008878042,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2125234305858612,
      "num_tokens": 39597353.0,
      "step": 19075
    },
    {
      "entropy": 5.498964595794678,
      "epoch": 1.7307692307692308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004702621263505824,
      "loss": 4.967,
      "mean_token_accuracy": 0.20899804085493087,
      "num_tokens": 39607349.0,
      "step": 19080
    },
    {
      "entropy": 5.524208688735962,
      "epoch": 1.731222786647315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004702460477752901,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20135973393917084,
      "num_tokens": 39617810.0,
      "step": 19085
    },
    {
      "entropy": 5.498603820800781,
      "epoch": 1.7316763425253991,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004702299651622597,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21558779627084732,
      "num_tokens": 39626805.0,
      "step": 19090
    },
    {
      "entropy": 5.4490362167358395,
      "epoch": 1.7321298984034832,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004702138785118239,
      "loss": 4.9573,
      "mean_token_accuracy": 0.20946362912654876,
      "num_tokens": 39637839.0,
      "step": 19095
    },
    {
      "entropy": 5.4664324760437015,
      "epoch": 1.7325834542815675,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004701977878243153,
      "loss": 4.9491,
      "mean_token_accuracy": 0.20754453390836716,
      "num_tokens": 39648697.0,
      "step": 19100
    },
    {
      "entropy": 5.464032983779907,
      "epoch": 1.7330370101596517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004701816931000668,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20650705695152283,
      "num_tokens": 39658154.0,
      "step": 19105
    },
    {
      "entropy": 5.502857494354248,
      "epoch": 1.733490566037736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047016559433941106,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20509715080261232,
      "num_tokens": 39667939.0,
      "step": 19110
    },
    {
      "entropy": 5.482723522186279,
      "epoch": 1.73394412191582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701494915426812,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2047874704003334,
      "num_tokens": 39678289.0,
      "step": 19115
    },
    {
      "entropy": 5.546759462356567,
      "epoch": 1.734397677793904,
      "grad_norm": 1.125,
      "learning_rate": 0.00047013338471021,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20536402910947799,
      "num_tokens": 39689901.0,
      "step": 19120
    },
    {
      "entropy": 5.515721130371094,
      "epoch": 1.7348512336719883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047011727384233074,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19690279215574263,
      "num_tokens": 39700831.0,
      "step": 19125
    },
    {
      "entropy": 5.5482320308685305,
      "epoch": 1.7353047895500726,
      "grad_norm": 1.0,
      "learning_rate": 0.0004701011589393765,
      "loss": 5.033,
      "mean_token_accuracy": 0.2001452073454857,
      "num_tokens": 39711794.0,
      "step": 19130
    },
    {
      "entropy": 5.52657151222229,
      "epoch": 1.7357583454281569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047008504000168054,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20627530217170714,
      "num_tokens": 39721889.0,
      "step": 19135
    },
    {
      "entropy": 5.483045530319214,
      "epoch": 1.736211901306241,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004700689170295763,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20222830325365065,
      "num_tokens": 39732562.0,
      "step": 19140
    },
    {
      "entropy": 5.572683954238892,
      "epoch": 1.736665457184325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047005279002339717,
      "loss": 5.018,
      "mean_token_accuracy": 0.20817613154649733,
      "num_tokens": 39742286.0,
      "step": 19145
    },
    {
      "entropy": 5.538823175430298,
      "epoch": 1.7371190130624092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047003665898347663,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2073296308517456,
      "num_tokens": 39754336.0,
      "step": 19150
    },
    {
      "entropy": 5.495729303359985,
      "epoch": 1.7375725689404935,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004700205239101483,
      "loss": 4.9188,
      "mean_token_accuracy": 0.21193399727344514,
      "num_tokens": 39763545.0,
      "step": 19155
    },
    {
      "entropy": 5.52259202003479,
      "epoch": 1.7380261248185778,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004700043848037458,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19503413289785385,
      "num_tokens": 39774536.0,
      "step": 19160
    },
    {
      "entropy": 5.509861278533935,
      "epoch": 1.7384796806966618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699882416646029,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21053527742624284,
      "num_tokens": 39785377.0,
      "step": 19165
    },
    {
      "entropy": 5.5219244956970215,
      "epoch": 1.7389332365747459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699720944930535,
      "loss": 5.0905,
      "mean_token_accuracy": 0.19307584017515184,
      "num_tokens": 39796692.0,
      "step": 19170
    },
    {
      "entropy": 5.461316347122192,
      "epoch": 1.7393867924528301,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046995594328943147,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20763515830039977,
      "num_tokens": 39807927.0,
      "step": 19175
    },
    {
      "entropy": 5.656646585464477,
      "epoch": 1.7398403483309144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004699397880540709,
      "loss": 5.1689,
      "mean_token_accuracy": 0.20154161900281906,
      "num_tokens": 39817617.0,
      "step": 19180
    },
    {
      "entropy": 5.548462867736816,
      "epoch": 1.7402939042089987,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004699236287873058,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21022000163793564,
      "num_tokens": 39827092.0,
      "step": 19185
    },
    {
      "entropy": 5.495359420776367,
      "epoch": 1.7407474600870827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699074654894705,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20519209057092666,
      "num_tokens": 39837378.0,
      "step": 19190
    },
    {
      "entropy": 5.51698932647705,
      "epoch": 1.7412010159651667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000469891298160899,
      "loss": 4.9807,
      "mean_token_accuracy": 0.1958746701478958,
      "num_tokens": 39848191.0,
      "step": 19195
    },
    {
      "entropy": 5.494634103775025,
      "epoch": 1.741654571843251,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046987512680192586,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2061051085591316,
      "num_tokens": 39857886.0,
      "step": 19200
    },
    {
      "entropy": 5.560719919204712,
      "epoch": 1.7421081277213353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046985895141288546,
      "loss": 5.0745,
      "mean_token_accuracy": 0.2028934597969055,
      "num_tokens": 39868348.0,
      "step": 19205
    },
    {
      "entropy": 5.568835687637329,
      "epoch": 1.7425616835994195,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046984277199411224,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2048014745116234,
      "num_tokens": 39879129.0,
      "step": 19210
    },
    {
      "entropy": 5.479235744476318,
      "epoch": 1.7430152394775036,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046982658854594083,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2069411188364029,
      "num_tokens": 39890239.0,
      "step": 19215
    },
    {
      "entropy": 5.446046924591064,
      "epoch": 1.7434687953555879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046981040106870595,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21154987961053848,
      "num_tokens": 39900907.0,
      "step": 19220
    },
    {
      "entropy": 5.530134630203247,
      "epoch": 1.743922351233672,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004697942095627423,
      "loss": 4.9989,
      "mean_token_accuracy": 0.209362630546093,
      "num_tokens": 39911564.0,
      "step": 19225
    },
    {
      "entropy": 5.51063084602356,
      "epoch": 1.7443759071117562,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697780140283848,
      "loss": 4.9491,
      "mean_token_accuracy": 0.20574220269918442,
      "num_tokens": 39922168.0,
      "step": 19230
    },
    {
      "entropy": 5.540814733505249,
      "epoch": 1.7448294629898404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004697618144659683,
      "loss": 4.9865,
      "mean_token_accuracy": 0.20843921601772308,
      "num_tokens": 39932690.0,
      "step": 19235
    },
    {
      "entropy": 5.622589206695556,
      "epoch": 1.7452830188679245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004697456108758278,
      "loss": 5.1385,
      "mean_token_accuracy": 0.2018663093447685,
      "num_tokens": 39942970.0,
      "step": 19240
    },
    {
      "entropy": 5.485025405883789,
      "epoch": 1.7457365747460087,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004697294032582986,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20647140443325043,
      "num_tokens": 39953255.0,
      "step": 19245
    },
    {
      "entropy": 5.441250514984131,
      "epoch": 1.7461901306240928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046971319161371564,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21407563537359237,
      "num_tokens": 39963521.0,
      "step": 19250
    },
    {
      "entropy": 5.515857315063476,
      "epoch": 1.746643686502177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046969697594241426,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20265513360500337,
      "num_tokens": 39973494.0,
      "step": 19255
    },
    {
      "entropy": 5.522301149368286,
      "epoch": 1.7470972423802613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004696807562447298,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20003973692655563,
      "num_tokens": 39984652.0,
      "step": 19260
    },
    {
      "entropy": 5.5341369152069095,
      "epoch": 1.7475507982583456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046966453252099763,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2052340716123581,
      "num_tokens": 39995490.0,
      "step": 19265
    },
    {
      "entropy": 5.476441144943237,
      "epoch": 1.7480043541364296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046964830477155335,
      "loss": 4.8282,
      "mean_token_accuracy": 0.22589887380599977,
      "num_tokens": 40006447.0,
      "step": 19270
    },
    {
      "entropy": 5.564656352996826,
      "epoch": 1.7484579100145137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046963207299673257,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20302298814058303,
      "num_tokens": 40015733.0,
      "step": 19275
    },
    {
      "entropy": 5.557398319244385,
      "epoch": 1.748911465892598,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046961583719687085,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20754121243953705,
      "num_tokens": 40025844.0,
      "step": 19280
    },
    {
      "entropy": 5.514825487136841,
      "epoch": 1.7493650217706822,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046959959737230414,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21481667459011078,
      "num_tokens": 40035942.0,
      "step": 19285
    },
    {
      "entropy": 5.42282395362854,
      "epoch": 1.7498185776487665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046958335352336806,
      "loss": 4.8831,
      "mean_token_accuracy": 0.21713285148143768,
      "num_tokens": 40045551.0,
      "step": 19290
    },
    {
      "entropy": 5.51376371383667,
      "epoch": 1.7502721335268505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004695671056503987,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20856801122426988,
      "num_tokens": 40055803.0,
      "step": 19295
    },
    {
      "entropy": 5.5142841815948485,
      "epoch": 1.7507256894049346,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469550853753732,
      "loss": 4.9963,
      "mean_token_accuracy": 0.20099260956048964,
      "num_tokens": 40066004.0,
      "step": 19300
    },
    {
      "entropy": 5.501036930084228,
      "epoch": 1.7511792452830188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469534597833704,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2118948593735695,
      "num_tokens": 40075522.0,
      "step": 19305
    },
    {
      "entropy": 5.496591711044312,
      "epoch": 1.751632801161103,
      "grad_norm": 1.078125,
      "learning_rate": 0.000469518337890651,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2103877067565918,
      "num_tokens": 40084839.0,
      "step": 19310
    },
    {
      "entropy": 5.559009075164795,
      "epoch": 1.7520863570391874,
      "grad_norm": 1.0,
      "learning_rate": 0.00046950207392490917,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20451847165822984,
      "num_tokens": 40094392.0,
      "step": 19315
    },
    {
      "entropy": 5.438671445846557,
      "epoch": 1.7525399129172714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046948580593681496,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20572173595428467,
      "num_tokens": 40105990.0,
      "step": 19320
    },
    {
      "entropy": 5.4502801418304445,
      "epoch": 1.7529934687953554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004694695339267047,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21390351206064223,
      "num_tokens": 40116143.0,
      "step": 19325
    },
    {
      "entropy": 5.614202833175659,
      "epoch": 1.7534470246734397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046945325789491483,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20463183969259263,
      "num_tokens": 40126849.0,
      "step": 19330
    },
    {
      "entropy": 5.510346460342407,
      "epoch": 1.753900580551524,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004694369778417822,
      "loss": 5.0399,
      "mean_token_accuracy": 0.2019796848297119,
      "num_tokens": 40137259.0,
      "step": 19335
    },
    {
      "entropy": 5.414547252655029,
      "epoch": 1.7543541364296082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004694206937676431,
      "loss": 4.8966,
      "mean_token_accuracy": 0.20899513810873033,
      "num_tokens": 40148578.0,
      "step": 19340
    },
    {
      "entropy": 5.565746641159057,
      "epoch": 1.7548076923076923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004694044056728347,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1964053377509117,
      "num_tokens": 40159754.0,
      "step": 19345
    },
    {
      "entropy": 5.538667440414429,
      "epoch": 1.7552612481857763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046938811355769363,
      "loss": 5.0042,
      "mean_token_accuracy": 0.2027725711464882,
      "num_tokens": 40169388.0,
      "step": 19350
    },
    {
      "entropy": 5.502602910995483,
      "epoch": 1.7557148040638606,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046937181742255673,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20214175283908845,
      "num_tokens": 40179271.0,
      "step": 19355
    },
    {
      "entropy": 5.557893991470337,
      "epoch": 1.7561683599419449,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004693555172677612,
      "loss": 5.0517,
      "mean_token_accuracy": 0.20480775088071823,
      "num_tokens": 40189292.0,
      "step": 19360
    },
    {
      "entropy": 5.472114896774292,
      "epoch": 1.7566219158200291,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000469339213093644,
      "loss": 5.009,
      "mean_token_accuracy": 0.20178278386592866,
      "num_tokens": 40199271.0,
      "step": 19365
    },
    {
      "entropy": 5.507282257080078,
      "epoch": 1.7570754716981132,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046932290490054235,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20328470319509506,
      "num_tokens": 40210386.0,
      "step": 19370
    },
    {
      "entropy": 5.564381408691406,
      "epoch": 1.7575290275761972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004693065926887936,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20100197196006775,
      "num_tokens": 40220892.0,
      "step": 19375
    },
    {
      "entropy": 5.565685033798218,
      "epoch": 1.7579825834542815,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046929027645873484,
      "loss": 4.9877,
      "mean_token_accuracy": 0.20461044311523438,
      "num_tokens": 40231737.0,
      "step": 19380
    },
    {
      "entropy": 5.487754487991333,
      "epoch": 1.7584361393323658,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004692739562107038,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20050707459449768,
      "num_tokens": 40242218.0,
      "step": 19385
    },
    {
      "entropy": 5.499060773849488,
      "epoch": 1.75888969521045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046925763194503775,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20986446738243103,
      "num_tokens": 40253659.0,
      "step": 19390
    },
    {
      "entropy": 5.450934648513794,
      "epoch": 1.759343251088534,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004692413036620744,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2149718463420868,
      "num_tokens": 40264328.0,
      "step": 19395
    },
    {
      "entropy": 5.464368915557861,
      "epoch": 1.7597968069666183,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004692249713621514,
      "loss": 4.946,
      "mean_token_accuracy": 0.2071563258767128,
      "num_tokens": 40275188.0,
      "step": 19400
    },
    {
      "entropy": 5.575522470474243,
      "epoch": 1.7602503628447024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046920863504560647,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19950825870037078,
      "num_tokens": 40287094.0,
      "step": 19405
    },
    {
      "entropy": 5.535629320144653,
      "epoch": 1.7607039187227866,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004691922947127775,
      "loss": 4.9732,
      "mean_token_accuracy": 0.20288341492414474,
      "num_tokens": 40296551.0,
      "step": 19410
    },
    {
      "entropy": 5.493284082412719,
      "epoch": 1.761157474600871,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004691759503640024,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20166113674640657,
      "num_tokens": 40306920.0,
      "step": 19415
    },
    {
      "entropy": 5.486135101318359,
      "epoch": 1.761611030478955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004691596019996191,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2151423826813698,
      "num_tokens": 40316797.0,
      "step": 19420
    },
    {
      "entropy": 5.435636186599732,
      "epoch": 1.7620645863570392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004691432496199658,
      "loss": 4.945,
      "mean_token_accuracy": 0.203757406771183,
      "num_tokens": 40327002.0,
      "step": 19425
    },
    {
      "entropy": 5.617444562911987,
      "epoch": 1.7625181422351233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046912689322538064,
      "loss": 5.084,
      "mean_token_accuracy": 0.20587912946939468,
      "num_tokens": 40337247.0,
      "step": 19430
    },
    {
      "entropy": 5.523028898239136,
      "epoch": 1.7629716981132075,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046911053281620187,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20531492978334426,
      "num_tokens": 40346960.0,
      "step": 19435
    },
    {
      "entropy": 5.554391193389892,
      "epoch": 1.7634252539912918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046909416839276775,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2060078427195549,
      "num_tokens": 40357169.0,
      "step": 19440
    },
    {
      "entropy": 5.494356679916382,
      "epoch": 1.763878809869376,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004690777999554168,
      "loss": 4.871,
      "mean_token_accuracy": 0.21701277941465377,
      "num_tokens": 40367660.0,
      "step": 19445
    },
    {
      "entropy": 5.48631534576416,
      "epoch": 1.76433236574746,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004690614275044875,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2141926497220993,
      "num_tokens": 40378217.0,
      "step": 19450
    },
    {
      "entropy": 5.511997699737549,
      "epoch": 1.7647859216255442,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004690450510403184,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20661132037639618,
      "num_tokens": 40388249.0,
      "step": 19455
    },
    {
      "entropy": 5.5403870105743405,
      "epoch": 1.7652394775036284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690286705632483,
      "loss": 5.0429,
      "mean_token_accuracy": 0.2022801861166954,
      "num_tokens": 40398239.0,
      "step": 19460
    },
    {
      "entropy": 5.527664661407471,
      "epoch": 1.7656930333817127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004690122860736157,
      "loss": 5.0476,
      "mean_token_accuracy": 0.20684996545314788,
      "num_tokens": 40408473.0,
      "step": 19465
    },
    {
      "entropy": 5.639688348770141,
      "epoch": 1.766146589259797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046899589757175964,
      "loss": 5.128,
      "mean_token_accuracy": 0.19599819481372832,
      "num_tokens": 40418674.0,
      "step": 19470
    },
    {
      "entropy": 5.489195823669434,
      "epoch": 1.766600145137881,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000468979505058019,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2068178355693817,
      "num_tokens": 40429815.0,
      "step": 19475
    },
    {
      "entropy": 5.607937955856324,
      "epoch": 1.767053701015965,
      "grad_norm": 1.125,
      "learning_rate": 0.00046896310853273274,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20456432402133942,
      "num_tokens": 40439809.0,
      "step": 19480
    },
    {
      "entropy": 5.617741441726684,
      "epoch": 1.7675072568940493,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046894670799624,
      "loss": 5.0626,
      "mean_token_accuracy": 0.19883812963962555,
      "num_tokens": 40449503.0,
      "step": 19485
    },
    {
      "entropy": 5.5056475639343265,
      "epoch": 1.7679608127721336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046893030344887995,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19412409514188766,
      "num_tokens": 40459794.0,
      "step": 19490
    },
    {
      "entropy": 5.484907579421997,
      "epoch": 1.7684143686502178,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046891389489099166,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21248606443405152,
      "num_tokens": 40470044.0,
      "step": 19495
    },
    {
      "entropy": 5.530512571334839,
      "epoch": 1.7688679245283019,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046889748232291476,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2080856442451477,
      "num_tokens": 40480709.0,
      "step": 19500
    },
    {
      "entropy": 5.438140058517456,
      "epoch": 1.769321480406386,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046888106574498843,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2188742369413376,
      "num_tokens": 40490906.0,
      "step": 19505
    },
    {
      "entropy": 5.464437818527221,
      "epoch": 1.7697750362844702,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004688646451575522,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2143838033080101,
      "num_tokens": 40501971.0,
      "step": 19510
    },
    {
      "entropy": 5.481470727920533,
      "epoch": 1.7702285921625545,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046884822056094587,
      "loss": 5.0354,
      "mean_token_accuracy": 0.19920152574777603,
      "num_tokens": 40512685.0,
      "step": 19515
    },
    {
      "entropy": 5.446528577804566,
      "epoch": 1.7706821480406387,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046883179195550883,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20782757103443145,
      "num_tokens": 40523460.0,
      "step": 19520
    },
    {
      "entropy": 5.5545446395874025,
      "epoch": 1.7711357039187228,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046881535934158094,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2085803419351578,
      "num_tokens": 40533594.0,
      "step": 19525
    },
    {
      "entropy": 5.516283655166626,
      "epoch": 1.7715892597968068,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004687989227195021,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20703327655792236,
      "num_tokens": 40543504.0,
      "step": 19530
    },
    {
      "entropy": 5.50254077911377,
      "epoch": 1.772042815674891,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046878248208961204,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20424250811338424,
      "num_tokens": 40553711.0,
      "step": 19535
    },
    {
      "entropy": 5.4620030403137205,
      "epoch": 1.7724963715529753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000468766037452251,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2051798552274704,
      "num_tokens": 40563731.0,
      "step": 19540
    },
    {
      "entropy": 5.559663248062134,
      "epoch": 1.7729499274310596,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046874958880775886,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20399201661348343,
      "num_tokens": 40574235.0,
      "step": 19545
    },
    {
      "entropy": 5.561195611953735,
      "epoch": 1.7734034833091437,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004687331361564758,
      "loss": 4.9539,
      "mean_token_accuracy": 0.20946568101644517,
      "num_tokens": 40584459.0,
      "step": 19550
    },
    {
      "entropy": 5.494787502288818,
      "epoch": 1.7738570391872277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004687166794987422,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20427904427051544,
      "num_tokens": 40594268.0,
      "step": 19555
    },
    {
      "entropy": 5.467495536804199,
      "epoch": 1.774310595065312,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004687002188348983,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2154368594288826,
      "num_tokens": 40604673.0,
      "step": 19560
    },
    {
      "entropy": 5.448693799972534,
      "epoch": 1.7747641509433962,
      "grad_norm": 1.125,
      "learning_rate": 0.00046868375416528443,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21056706607341766,
      "num_tokens": 40614998.0,
      "step": 19565
    },
    {
      "entropy": 5.502387046813965,
      "epoch": 1.7752177068214805,
      "grad_norm": 1.125,
      "learning_rate": 0.00046866728549024123,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20484382063150405,
      "num_tokens": 40624729.0,
      "step": 19570
    },
    {
      "entropy": 5.545193481445312,
      "epoch": 1.7756712626995645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004686508128101092,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20403294563293456,
      "num_tokens": 40635673.0,
      "step": 19575
    },
    {
      "entropy": 5.559050893783569,
      "epoch": 1.7761248185776488,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468634336125229,
      "loss": 5.058,
      "mean_token_accuracy": 0.20020204931497573,
      "num_tokens": 40645990.0,
      "step": 19580
    },
    {
      "entropy": 5.572917890548706,
      "epoch": 1.7765783744557329,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046861785543594143,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21186477690935135,
      "num_tokens": 40656339.0,
      "step": 19585
    },
    {
      "entropy": 5.50078330039978,
      "epoch": 1.7770319303338171,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004686013707425873,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21024974882602693,
      "num_tokens": 40666371.0,
      "step": 19590
    },
    {
      "entropy": 5.512684726715088,
      "epoch": 1.7774854862119014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046858488204550735,
      "loss": 4.957,
      "mean_token_accuracy": 0.20563161373138428,
      "num_tokens": 40676423.0,
      "step": 19595
    },
    {
      "entropy": 5.495868921279907,
      "epoch": 1.7779390420899854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004685683893450428,
      "loss": 4.916,
      "mean_token_accuracy": 0.2111999660730362,
      "num_tokens": 40686331.0,
      "step": 19600
    },
    {
      "entropy": 5.439804172515869,
      "epoch": 1.7783925979680697,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004685518926415346,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2131078854203224,
      "num_tokens": 40697408.0,
      "step": 19605
    },
    {
      "entropy": 5.423649454116822,
      "epoch": 1.7788461538461537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000468535391935324,
      "loss": 4.856,
      "mean_token_accuracy": 0.22234155088663102,
      "num_tokens": 40707078.0,
      "step": 19610
    },
    {
      "entropy": 5.535253810882568,
      "epoch": 1.779299709724238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685188872267521,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20636095255613326,
      "num_tokens": 40718116.0,
      "step": 19615
    },
    {
      "entropy": 5.523155736923218,
      "epoch": 1.7797532656023223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685023785161603,
      "loss": 5.0187,
      "mean_token_accuracy": 0.20456632524728774,
      "num_tokens": 40727737.0,
      "step": 19620
    },
    {
      "entropy": 5.5154128074646,
      "epoch": 1.7802068214804065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004684858658038901,
      "loss": 4.9649,
      "mean_token_accuracy": 0.202775239944458,
      "num_tokens": 40737455.0,
      "step": 19625
    },
    {
      "entropy": 5.590185356140137,
      "epoch": 1.7806603773584906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004684693490902828,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21528024077415467,
      "num_tokens": 40746715.0,
      "step": 19630
    },
    {
      "entropy": 5.465564012527466,
      "epoch": 1.7811139332365746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046845282837568,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21505769342184067,
      "num_tokens": 40756432.0,
      "step": 19635
    },
    {
      "entropy": 5.411591958999634,
      "epoch": 1.781567489114659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046843630366042355,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2048589125275612,
      "num_tokens": 40766629.0,
      "step": 19640
    },
    {
      "entropy": 5.582585382461548,
      "epoch": 1.7820210449927432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046841977494485506,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1974947214126587,
      "num_tokens": 40776650.0,
      "step": 19645
    },
    {
      "entropy": 5.601768112182617,
      "epoch": 1.7824746008708274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004684032422293162,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20934145748615265,
      "num_tokens": 40786722.0,
      "step": 19650
    },
    {
      "entropy": 5.398625421524048,
      "epoch": 1.7829281567489115,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468386705514149,
      "loss": 4.8343,
      "mean_token_accuracy": 0.2224738508462906,
      "num_tokens": 40797058.0,
      "step": 19655
    },
    {
      "entropy": 5.437365961074829,
      "epoch": 1.7833817126269955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004683701647996956,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20820920020341874,
      "num_tokens": 40808207.0,
      "step": 19660
    },
    {
      "entropy": 5.570583391189575,
      "epoch": 1.7838352685050798,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683536200862978,
      "loss": 5.089,
      "mean_token_accuracy": 0.20439439564943312,
      "num_tokens": 40819655.0,
      "step": 19665
    },
    {
      "entropy": 5.5647047996521,
      "epoch": 1.784288824383164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004683370713742979,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20476250648498534,
      "num_tokens": 40831015.0,
      "step": 19670
    },
    {
      "entropy": 5.48437008857727,
      "epoch": 1.7847423802612483,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046832051866403814,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21035304218530654,
      "num_tokens": 40841763.0,
      "step": 19675
    },
    {
      "entropy": 5.4942773342132565,
      "epoch": 1.7851959361393324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004683039619558607,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20394841432571412,
      "num_tokens": 40852479.0,
      "step": 19680
    },
    {
      "entropy": 5.536971139907837,
      "epoch": 1.7856494920174164,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046828740125010813,
      "loss": 5.0006,
      "mean_token_accuracy": 0.20805347412824632,
      "num_tokens": 40862486.0,
      "step": 19685
    },
    {
      "entropy": 5.5686804294586185,
      "epoch": 1.7861030478955007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004682708365471228,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20404157787561417,
      "num_tokens": 40872182.0,
      "step": 19690
    },
    {
      "entropy": 5.5247588634490965,
      "epoch": 1.786556603773585,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046825426784724734,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20590101182460785,
      "num_tokens": 40883036.0,
      "step": 19695
    },
    {
      "entropy": 5.5683763980865475,
      "epoch": 1.7870101596516692,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004682376951508244,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19138999432325363,
      "num_tokens": 40893684.0,
      "step": 19700
    },
    {
      "entropy": 5.57241907119751,
      "epoch": 1.7874637155297532,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046822111845819665,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20308079421520234,
      "num_tokens": 40904540.0,
      "step": 19705
    },
    {
      "entropy": 5.557100105285644,
      "epoch": 1.7879172714078373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046820453776970697,
      "loss": 5.054,
      "mean_token_accuracy": 0.205801622569561,
      "num_tokens": 40915269.0,
      "step": 19710
    },
    {
      "entropy": 5.52552809715271,
      "epoch": 1.7883708272859216,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046818795308569815,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20193398147821426,
      "num_tokens": 40925987.0,
      "step": 19715
    },
    {
      "entropy": 5.469293165206909,
      "epoch": 1.7888243831640058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681713644065133,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2131742060184479,
      "num_tokens": 40936130.0,
      "step": 19720
    },
    {
      "entropy": 5.439738702774048,
      "epoch": 1.78927793904209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004681547717324954,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2143712431192398,
      "num_tokens": 40945476.0,
      "step": 19725
    },
    {
      "entropy": 5.4216382026672365,
      "epoch": 1.7897314949201741,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046813817506398755,
      "loss": 4.8598,
      "mean_token_accuracy": 0.21695656776428224,
      "num_tokens": 40956492.0,
      "step": 19730
    },
    {
      "entropy": 5.554205465316772,
      "epoch": 1.7901850507982582,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046812157440133303,
      "loss": 5.0703,
      "mean_token_accuracy": 0.204265259206295,
      "num_tokens": 40966990.0,
      "step": 19735
    },
    {
      "entropy": 5.551835060119629,
      "epoch": 1.7906386066763424,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681049697448752,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20988238602876663,
      "num_tokens": 40977065.0,
      "step": 19740
    },
    {
      "entropy": 5.4898449897766115,
      "epoch": 1.7910921625544267,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046808836109495734,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20649953335523605,
      "num_tokens": 40987207.0,
      "step": 19745
    },
    {
      "entropy": 5.484203433990478,
      "epoch": 1.791545718432511,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004680717484519229,
      "loss": 4.9492,
      "mean_token_accuracy": 0.20586683303117753,
      "num_tokens": 40997541.0,
      "step": 19750
    },
    {
      "entropy": 5.488790988922119,
      "epoch": 1.791999274310595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004680551318161156,
      "loss": 4.975,
      "mean_token_accuracy": 0.21097332686185838,
      "num_tokens": 41009044.0,
      "step": 19755
    },
    {
      "entropy": 5.525434255599976,
      "epoch": 1.7924528301886793,
      "grad_norm": 1.0,
      "learning_rate": 0.00046803851118787897,
      "loss": 4.9968,
      "mean_token_accuracy": 0.20461712777614594,
      "num_tokens": 41018165.0,
      "step": 19760
    },
    {
      "entropy": 5.474305772781372,
      "epoch": 1.7929063860667633,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004680218865675567,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2071532815694809,
      "num_tokens": 41028972.0,
      "step": 19765
    },
    {
      "entropy": 5.486135721206665,
      "epoch": 1.7933599419448476,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004680052579554926,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20814012885093688,
      "num_tokens": 41039552.0,
      "step": 19770
    },
    {
      "entropy": 5.596327066421509,
      "epoch": 1.7938134978229319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004679886253520307,
      "loss": 5.058,
      "mean_token_accuracy": 0.19694051742553711,
      "num_tokens": 41049808.0,
      "step": 19775
    },
    {
      "entropy": 5.517556190490723,
      "epoch": 1.794267053701016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046797198875751474,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2051583707332611,
      "num_tokens": 41060574.0,
      "step": 19780
    },
    {
      "entropy": 5.589335012435913,
      "epoch": 1.7947206095791002,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004679553481722889,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1991354376077652,
      "num_tokens": 41071239.0,
      "step": 19785
    },
    {
      "entropy": 5.57397494316101,
      "epoch": 1.7951741654571842,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046793870359669733,
      "loss": 5.1171,
      "mean_token_accuracy": 0.2064829632639885,
      "num_tokens": 41080819.0,
      "step": 19790
    },
    {
      "entropy": 5.55178394317627,
      "epoch": 1.7956277213352685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004679220550310842,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2165278434753418,
      "num_tokens": 41090457.0,
      "step": 19795
    },
    {
      "entropy": 5.514302015304565,
      "epoch": 1.7960812772133528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679054024757937,
      "loss": 4.9891,
      "mean_token_accuracy": 0.20996480882167817,
      "num_tokens": 41101210.0,
      "step": 19800
    },
    {
      "entropy": 5.510220861434936,
      "epoch": 1.796534833091437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004678887459311704,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20979314297437668,
      "num_tokens": 41110566.0,
      "step": 19805
    },
    {
      "entropy": 5.441089677810669,
      "epoch": 1.796988388969521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004678720853975587,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2132904589176178,
      "num_tokens": 41120000.0,
      "step": 19810
    },
    {
      "entropy": 5.455682373046875,
      "epoch": 1.797441944847605,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046785542087530307,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2143726408481598,
      "num_tokens": 41129944.0,
      "step": 19815
    },
    {
      "entropy": 5.472288274765015,
      "epoch": 1.7978955007256894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004678387523647483,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2070723608136177,
      "num_tokens": 41138803.0,
      "step": 19820
    },
    {
      "entropy": 5.520013809204102,
      "epoch": 1.7983490566037736,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046782207986623885,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19517021626234055,
      "num_tokens": 41150365.0,
      "step": 19825
    },
    {
      "entropy": 5.610871505737305,
      "epoch": 1.798802612481858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046780540338011974,
      "loss": 5.02,
      "mean_token_accuracy": 0.2008575588464737,
      "num_tokens": 41160310.0,
      "step": 19830
    },
    {
      "entropy": 5.503529834747314,
      "epoch": 1.799256168359942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004677887229067358,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2091690182685852,
      "num_tokens": 41170506.0,
      "step": 19835
    },
    {
      "entropy": 5.534892129898071,
      "epoch": 1.799709724238026,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046777203844643184,
      "loss": 5.0202,
      "mean_token_accuracy": 0.19795064479112626,
      "num_tokens": 41181293.0,
      "step": 19840
    },
    {
      "entropy": 5.4575417041778564,
      "epoch": 1.8001632801161103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046775534999955303,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2056989386677742,
      "num_tokens": 41191637.0,
      "step": 19845
    },
    {
      "entropy": 5.48751482963562,
      "epoch": 1.8006168359941945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004677386575664445,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21417139172554017,
      "num_tokens": 41201833.0,
      "step": 19850
    },
    {
      "entropy": 5.5553830623626705,
      "epoch": 1.8010703918722788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046772196114745133,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21154498308897018,
      "num_tokens": 41211243.0,
      "step": 19855
    },
    {
      "entropy": 5.537686347961426,
      "epoch": 1.8015239477503628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004677052607429189,
      "loss": 4.972,
      "mean_token_accuracy": 0.21050416827201843,
      "num_tokens": 41221111.0,
      "step": 19860
    },
    {
      "entropy": 5.473995018005371,
      "epoch": 1.8019775036284469,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004676885563531927,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20612288415431976,
      "num_tokens": 41231578.0,
      "step": 19865
    },
    {
      "entropy": 5.500777292251587,
      "epoch": 1.8024310595065312,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046767184797861795,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20313949137926102,
      "num_tokens": 41242130.0,
      "step": 19870
    },
    {
      "entropy": 5.556730842590332,
      "epoch": 1.8028846153846154,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046765513561954026,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2023453250527382,
      "num_tokens": 41252885.0,
      "step": 19875
    },
    {
      "entropy": 5.555192470550537,
      "epoch": 1.8033381712626997,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046763841927630534,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20067857652902604,
      "num_tokens": 41263532.0,
      "step": 19880
    },
    {
      "entropy": 5.549077892303467,
      "epoch": 1.8037917271407837,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046762169894925876,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2067990019917488,
      "num_tokens": 41273469.0,
      "step": 19885
    },
    {
      "entropy": 5.524668884277344,
      "epoch": 1.8042452830188678,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004676049746387464,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2050783470273018,
      "num_tokens": 41283663.0,
      "step": 19890
    },
    {
      "entropy": 5.477745485305786,
      "epoch": 1.804698838896952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675882463451141,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2139314129948616,
      "num_tokens": 41294657.0,
      "step": 19895
    },
    {
      "entropy": 5.536194944381714,
      "epoch": 1.8051523947750363,
      "grad_norm": 1.125,
      "learning_rate": 0.0004675715140687078,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21073366701602936,
      "num_tokens": 41305124.0,
      "step": 19900
    },
    {
      "entropy": 5.465384435653687,
      "epoch": 1.8056059506531206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046755477780987344,
      "loss": 4.9387,
      "mean_token_accuracy": 0.21205850392580033,
      "num_tokens": 41315839.0,
      "step": 19905
    },
    {
      "entropy": 5.5487367630004885,
      "epoch": 1.8060595065312046,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046753803756895726,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20375055372714995,
      "num_tokens": 41326388.0,
      "step": 19910
    },
    {
      "entropy": 5.518201732635498,
      "epoch": 1.8065130624092887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004675212933463054,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19393086135387422,
      "num_tokens": 41337333.0,
      "step": 19915
    },
    {
      "entropy": 5.582690715789795,
      "epoch": 1.806966618287373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675045451422641,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19874513745307923,
      "num_tokens": 41348353.0,
      "step": 19920
    },
    {
      "entropy": 5.561363458633423,
      "epoch": 1.8074201741654572,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046748779295717973,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21072236597537994,
      "num_tokens": 41358797.0,
      "step": 19925
    },
    {
      "entropy": 5.538416910171509,
      "epoch": 1.8078737300435415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004674710367913989,
      "loss": 5.015,
      "mean_token_accuracy": 0.20656365007162095,
      "num_tokens": 41368818.0,
      "step": 19930
    },
    {
      "entropy": 5.585899019241333,
      "epoch": 1.8083272859216255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004674542766452678,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20926343500614167,
      "num_tokens": 41378949.0,
      "step": 19935
    },
    {
      "entropy": 5.547982454299927,
      "epoch": 1.8087808417997098,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046743751251913327,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2045910596847534,
      "num_tokens": 41388892.0,
      "step": 19940
    },
    {
      "entropy": 5.550196695327759,
      "epoch": 1.8092343976777938,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046742074441334197,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2078236907720566,
      "num_tokens": 41399197.0,
      "step": 19945
    },
    {
      "entropy": 5.521719932556152,
      "epoch": 1.809687953555878,
      "grad_norm": 1.125,
      "learning_rate": 0.0004674039723282406,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20002949237823486,
      "num_tokens": 41409551.0,
      "step": 19950
    },
    {
      "entropy": 5.481037569046021,
      "epoch": 1.8101415094339623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046738719626417607,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2104635313153267,
      "num_tokens": 41418900.0,
      "step": 19955
    },
    {
      "entropy": 5.483044624328613,
      "epoch": 1.8105950653120464,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046737041622149536,
      "loss": 4.9378,
      "mean_token_accuracy": 0.20926041752099991,
      "num_tokens": 41428204.0,
      "step": 19960
    },
    {
      "entropy": 5.52245979309082,
      "epoch": 1.8110486211901307,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004673536322005454,
      "loss": 5.017,
      "mean_token_accuracy": 0.20543889254331588,
      "num_tokens": 41437922.0,
      "step": 19965
    },
    {
      "entropy": 5.540520906448364,
      "epoch": 1.8115021770682147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004673368442016732,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20341960489749908,
      "num_tokens": 41447411.0,
      "step": 19970
    },
    {
      "entropy": 5.578132677078247,
      "epoch": 1.811955732946299,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004673200522252261,
      "loss": 5.0826,
      "mean_token_accuracy": 0.20498971045017242,
      "num_tokens": 41458756.0,
      "step": 19975
    },
    {
      "entropy": 5.543388175964355,
      "epoch": 1.8124092888243832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004673032562715514,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20887200832366942,
      "num_tokens": 41468610.0,
      "step": 19980
    },
    {
      "entropy": 5.527509593963623,
      "epoch": 1.8128628447024675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046728645634099627,
      "loss": 5.0901,
      "mean_token_accuracy": 0.1987825319170952,
      "num_tokens": 41479849.0,
      "step": 19985
    },
    {
      "entropy": 5.515531539916992,
      "epoch": 1.8133164005805515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046726965243390825,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20240487307310104,
      "num_tokens": 41490945.0,
      "step": 19990
    },
    {
      "entropy": 5.470834827423095,
      "epoch": 1.8137699564586356,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004672528445506348,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20127549171447753,
      "num_tokens": 41501479.0,
      "step": 19995
    },
    {
      "entropy": 5.51609525680542,
      "epoch": 1.8142235123367199,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004672360326915236,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21051958948373795,
      "num_tokens": 41512447.0,
      "step": 20000
    },
    {
      "entropy": 5.462694454193115,
      "epoch": 1.8146770682148041,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004672192168569222,
      "loss": 4.9075,
      "mean_token_accuracy": 0.21520113497972487,
      "num_tokens": 41521905.0,
      "step": 20005
    },
    {
      "entropy": 5.471878480911255,
      "epoch": 1.8151306240928884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046720239704717845,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21198217272758485,
      "num_tokens": 41533143.0,
      "step": 20010
    },
    {
      "entropy": 5.482673501968383,
      "epoch": 1.8155841799709724,
      "grad_norm": 1.125,
      "learning_rate": 0.00046718557326264,
      "loss": 4.9443,
      "mean_token_accuracy": 0.20825288891792298,
      "num_tokens": 41542234.0,
      "step": 20015
    },
    {
      "entropy": 5.545076704025268,
      "epoch": 1.8160377358490565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004671687455036551,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20356175005435945,
      "num_tokens": 41552632.0,
      "step": 20020
    },
    {
      "entropy": 5.518356990814209,
      "epoch": 1.8164912917271407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004671519137705715,
      "loss": 5.0037,
      "mean_token_accuracy": 0.20862066149711608,
      "num_tokens": 41563723.0,
      "step": 20025
    },
    {
      "entropy": 5.527810955047608,
      "epoch": 1.816944847605225,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046713507806373736,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20548459142446518,
      "num_tokens": 41573680.0,
      "step": 20030
    },
    {
      "entropy": 5.520146417617798,
      "epoch": 1.8173984034833093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046711823838350086,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20444160848855972,
      "num_tokens": 41584722.0,
      "step": 20035
    },
    {
      "entropy": 5.511092567443848,
      "epoch": 1.8178519593613933,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046710139473021025,
      "loss": 5.0741,
      "mean_token_accuracy": 0.19642874598503113,
      "num_tokens": 41595785.0,
      "step": 20040
    },
    {
      "entropy": 5.545671176910401,
      "epoch": 1.8183055152394774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046708454710421376,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20113495439291001,
      "num_tokens": 41606023.0,
      "step": 20045
    },
    {
      "entropy": 5.514336347579956,
      "epoch": 1.8187590711175616,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046706769550586,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20624094307422638,
      "num_tokens": 41616010.0,
      "step": 20050
    },
    {
      "entropy": 5.530388689041137,
      "epoch": 1.819212626995646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046705083993549723,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21033221781253814,
      "num_tokens": 41625105.0,
      "step": 20055
    },
    {
      "entropy": 5.495745611190796,
      "epoch": 1.8196661828737302,
      "grad_norm": 1.125,
      "learning_rate": 0.0004670339803934743,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2080562338232994,
      "num_tokens": 41634989.0,
      "step": 20060
    },
    {
      "entropy": 5.465115642547607,
      "epoch": 1.8201197387518142,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004670171168801396,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2078110843896866,
      "num_tokens": 41645736.0,
      "step": 20065
    },
    {
      "entropy": 5.54468560218811,
      "epoch": 1.8205732946298983,
      "grad_norm": 1.078125,
      "learning_rate": 0.000467000249395842,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20836457312107087,
      "num_tokens": 41655505.0,
      "step": 20070
    },
    {
      "entropy": 5.486360883712768,
      "epoch": 1.8210268505079825,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669833779409303,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2087858200073242,
      "num_tokens": 41665732.0,
      "step": 20075
    },
    {
      "entropy": 5.493438863754273,
      "epoch": 1.8214804063860668,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004669665025157535,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2020236074924469,
      "num_tokens": 41676050.0,
      "step": 20080
    },
    {
      "entropy": 5.591602611541748,
      "epoch": 1.821933962264151,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046694962312066046,
      "loss": 5.1363,
      "mean_token_accuracy": 0.2001710593700409,
      "num_tokens": 41686787.0,
      "step": 20085
    },
    {
      "entropy": 5.499366092681885,
      "epoch": 1.822387518142235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046693273975600036,
      "loss": 4.9289,
      "mean_token_accuracy": 0.20800776779651642,
      "num_tokens": 41697168.0,
      "step": 20090
    },
    {
      "entropy": 5.476192235946655,
      "epoch": 1.8228410740203191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004669158524221222,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21236103475093843,
      "num_tokens": 41707524.0,
      "step": 20095
    },
    {
      "entropy": 5.556261444091797,
      "epoch": 1.8232946298984034,
      "grad_norm": 1.0,
      "learning_rate": 0.00046689896111937537,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20973631888628005,
      "num_tokens": 41718109.0,
      "step": 20100
    },
    {
      "entropy": 5.610829162597656,
      "epoch": 1.8237481857764877,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004668820658481091,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20574668049812317,
      "num_tokens": 41728862.0,
      "step": 20105
    },
    {
      "entropy": 5.622160243988037,
      "epoch": 1.824201741654572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046686516660867283,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1992063581943512,
      "num_tokens": 41739024.0,
      "step": 20110
    },
    {
      "entropy": 5.53470139503479,
      "epoch": 1.824655297532656,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004668482634014159,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20258474349975586,
      "num_tokens": 41748979.0,
      "step": 20115
    },
    {
      "entropy": 5.479855966567993,
      "epoch": 1.8251088534107403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004668313562266882,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20993564575910567,
      "num_tokens": 41759972.0,
      "step": 20120
    },
    {
      "entropy": 5.501933145523071,
      "epoch": 1.8255624092888243,
      "grad_norm": 1.0625,
      "learning_rate": 0.000466814445084839,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21612322181463242,
      "num_tokens": 41770150.0,
      "step": 20125
    },
    {
      "entropy": 5.525883293151855,
      "epoch": 1.8260159651669086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004667975299762183,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2107188180088997,
      "num_tokens": 41780405.0,
      "step": 20130
    },
    {
      "entropy": 5.475947761535645,
      "epoch": 1.8264695210449928,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046678061090117576,
      "loss": 4.903,
      "mean_token_accuracy": 0.21209864616394042,
      "num_tokens": 41791269.0,
      "step": 20135
    },
    {
      "entropy": 5.428743648529053,
      "epoch": 1.8269230769230769,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004667636878600613,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21087191104888917,
      "num_tokens": 41802513.0,
      "step": 20140
    },
    {
      "entropy": 5.56286849975586,
      "epoch": 1.8273766328011611,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004667467608532249,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19700977504253386,
      "num_tokens": 41814149.0,
      "step": 20145
    },
    {
      "entropy": 5.561482286453247,
      "epoch": 1.8278301886792452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004667298298810167,
      "loss": 5.024,
      "mean_token_accuracy": 0.20162852555513383,
      "num_tokens": 41825763.0,
      "step": 20150
    },
    {
      "entropy": 5.500921678543091,
      "epoch": 1.8282837445573294,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004667128949437867,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20902491807937623,
      "num_tokens": 41834543.0,
      "step": 20155
    },
    {
      "entropy": 5.523082876205445,
      "epoch": 1.8287373004354137,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004666959560418852,
      "loss": 4.9123,
      "mean_token_accuracy": 0.219485867023468,
      "num_tokens": 41844940.0,
      "step": 20160
    },
    {
      "entropy": 5.4727842807769775,
      "epoch": 1.829190856313498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046667901317566245,
      "loss": 4.9188,
      "mean_token_accuracy": 0.21303580403327943,
      "num_tokens": 41855532.0,
      "step": 20165
    },
    {
      "entropy": 5.5545471668243405,
      "epoch": 1.829644412191582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004666620663454689,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20356351137161255,
      "num_tokens": 41865661.0,
      "step": 20170
    },
    {
      "entropy": 5.5023308277130125,
      "epoch": 1.830097968069666,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046664511555165496,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2042405366897583,
      "num_tokens": 41875662.0,
      "step": 20175
    },
    {
      "entropy": 5.495385313034058,
      "epoch": 1.8305515239477503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046662816079457114,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21413073390722276,
      "num_tokens": 41886211.0,
      "step": 20180
    },
    {
      "entropy": 5.605622100830078,
      "epoch": 1.8310050798258346,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004666112020745682,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19857825338840485,
      "num_tokens": 41896737.0,
      "step": 20185
    },
    {
      "entropy": 5.522352790832519,
      "epoch": 1.8314586357039189,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004665942393919967,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2082372471690178,
      "num_tokens": 41907768.0,
      "step": 20190
    },
    {
      "entropy": 5.516893434524536,
      "epoch": 1.831912191582003,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046657727274720755,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20389573127031327,
      "num_tokens": 41917587.0,
      "step": 20195
    },
    {
      "entropy": 5.451837539672852,
      "epoch": 1.832365747460087,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004665603021405516,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21552098989486695,
      "num_tokens": 41927842.0,
      "step": 20200
    },
    {
      "entropy": 5.450876522064209,
      "epoch": 1.8328193033381712,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004665433275723798,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21344021558761597,
      "num_tokens": 41939899.0,
      "step": 20205
    },
    {
      "entropy": 5.500430679321289,
      "epoch": 1.8332728592162555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004665263490430431,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20242294818162918,
      "num_tokens": 41950932.0,
      "step": 20210
    },
    {
      "entropy": 5.546673393249511,
      "epoch": 1.8337264150943398,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004665093665528928,
      "loss": 5.0306,
      "mean_token_accuracy": 0.20848318785429001,
      "num_tokens": 41963512.0,
      "step": 20215
    },
    {
      "entropy": 5.593548536300659,
      "epoch": 1.8341799709724238,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046649238010227983,
      "loss": 5.027,
      "mean_token_accuracy": 0.20841220915317535,
      "num_tokens": 41974677.0,
      "step": 20220
    },
    {
      "entropy": 5.452464294433594,
      "epoch": 1.8346335268505078,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004664753896915558,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21078637689352037,
      "num_tokens": 41984582.0,
      "step": 20225
    },
    {
      "entropy": 5.492018222808838,
      "epoch": 1.8350870827285921,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664583953210719,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2051735356450081,
      "num_tokens": 41994624.0,
      "step": 20230
    },
    {
      "entropy": 5.518248462677002,
      "epoch": 1.8355406386066764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664413969911795,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21068610548973082,
      "num_tokens": 42004835.0,
      "step": 20235
    },
    {
      "entropy": 5.446583843231201,
      "epoch": 1.8359941944847606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004664243947022303,
      "loss": 4.95,
      "mean_token_accuracy": 0.2101311355829239,
      "num_tokens": 42014727.0,
      "step": 20240
    },
    {
      "entropy": 5.512728643417359,
      "epoch": 1.8364477503628447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046640738845457586,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20640296787023543,
      "num_tokens": 42023628.0,
      "step": 20245
    },
    {
      "entropy": 5.55571870803833,
      "epoch": 1.8369013062409287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004663903782485678,
      "loss": 5.0451,
      "mean_token_accuracy": 0.2015094056725502,
      "num_tokens": 42033587.0,
      "step": 20250
    },
    {
      "entropy": 5.564340829849243,
      "epoch": 1.837354862119013,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046637336408455793,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2039337232708931,
      "num_tokens": 42043765.0,
      "step": 20255
    },
    {
      "entropy": 5.581139183044433,
      "epoch": 1.8378084179970973,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046635634596289816,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2069259002804756,
      "num_tokens": 42054205.0,
      "step": 20260
    },
    {
      "entropy": 5.458800506591797,
      "epoch": 1.8382619738751815,
      "grad_norm": 1.0,
      "learning_rate": 0.0004663393238839404,
      "loss": 4.8914,
      "mean_token_accuracy": 0.21075702756643294,
      "num_tokens": 42064512.0,
      "step": 20265
    },
    {
      "entropy": 5.478949785232544,
      "epoch": 1.8387155297532656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004663222978480366,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20276922285556792,
      "num_tokens": 42074954.0,
      "step": 20270
    },
    {
      "entropy": 5.5394398212432865,
      "epoch": 1.8391690856313496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466305267855539,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2032952278852463,
      "num_tokens": 42084914.0,
      "step": 20275
    },
    {
      "entropy": 5.5595615863800045,
      "epoch": 1.8396226415094339,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004662882339067996,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20879424512386321,
      "num_tokens": 42095454.0,
      "step": 20280
    },
    {
      "entropy": 5.524030351638794,
      "epoch": 1.8400761973875182,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004662711960021709,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1993553966283798,
      "num_tokens": 42106960.0,
      "step": 20285
    },
    {
      "entropy": 5.489474582672119,
      "epoch": 1.8405297532656024,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046625415414200513,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2108674466609955,
      "num_tokens": 42117009.0,
      "step": 20290
    },
    {
      "entropy": 5.548690843582153,
      "epoch": 1.8409833091436865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004662371083266546,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20959905236959459,
      "num_tokens": 42126371.0,
      "step": 20295
    },
    {
      "entropy": 5.560553932189942,
      "epoch": 1.8414368650217707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000466220058556472,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20469729453325272,
      "num_tokens": 42135890.0,
      "step": 20300
    },
    {
      "entropy": 5.5765221118927,
      "epoch": 1.8418904208998548,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046620300483180986,
      "loss": 5.1176,
      "mean_token_accuracy": 0.19918676763772963,
      "num_tokens": 42145696.0,
      "step": 20305
    },
    {
      "entropy": 5.524145030975342,
      "epoch": 1.842343976777939,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004661859471530209,
      "loss": 5.0355,
      "mean_token_accuracy": 0.20542675703763963,
      "num_tokens": 42156236.0,
      "step": 20310
    },
    {
      "entropy": 5.587287187576294,
      "epoch": 1.8427975326560233,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004661688855204578,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2009128749370575,
      "num_tokens": 42166179.0,
      "step": 20315
    },
    {
      "entropy": 5.493165826797485,
      "epoch": 1.8432510885341074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046615181993447344,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21054396480321885,
      "num_tokens": 42176643.0,
      "step": 20320
    },
    {
      "entropy": 5.557625150680542,
      "epoch": 1.8437046444121916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004661347503954207,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19340878576040268,
      "num_tokens": 42188328.0,
      "step": 20325
    },
    {
      "entropy": 5.5738866329193115,
      "epoch": 1.8441582002902757,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046611767690365265,
      "loss": 5.0129,
      "mean_token_accuracy": 0.2106790453195572,
      "num_tokens": 42198110.0,
      "step": 20330
    },
    {
      "entropy": 5.481729221343994,
      "epoch": 1.84461175616836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004661005994595223,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20133370757102967,
      "num_tokens": 42208438.0,
      "step": 20335
    },
    {
      "entropy": 5.525497341156006,
      "epoch": 1.8450653120464442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004660835180633829,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20635638386011124,
      "num_tokens": 42220052.0,
      "step": 20340
    },
    {
      "entropy": 5.51782603263855,
      "epoch": 1.8455188679245285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004660664327155876,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20845620036125184,
      "num_tokens": 42229497.0,
      "step": 20345
    },
    {
      "entropy": 5.49897723197937,
      "epoch": 1.8459724238026125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004660493434164898,
      "loss": 5.0291,
      "mean_token_accuracy": 0.1970988929271698,
      "num_tokens": 42240725.0,
      "step": 20350
    },
    {
      "entropy": 5.582056856155395,
      "epoch": 1.8464259796806965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004660322501664429,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20600544661283493,
      "num_tokens": 42251114.0,
      "step": 20355
    },
    {
      "entropy": 5.61389651298523,
      "epoch": 1.8468795355587808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004660151529658004,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1986292600631714,
      "num_tokens": 42262825.0,
      "step": 20360
    },
    {
      "entropy": 5.571655797958374,
      "epoch": 1.847333091436865,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659980518149158,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2176257774233818,
      "num_tokens": 42272176.0,
      "step": 20365
    },
    {
      "entropy": 5.56455545425415,
      "epoch": 1.8477866473149493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046598094671414285,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2081517606973648,
      "num_tokens": 42281934.0,
      "step": 20370
    },
    {
      "entropy": 5.537998533248901,
      "epoch": 1.8482402031930334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046596383766383533,
      "loss": 4.9471,
      "mean_token_accuracy": 0.20919624269008635,
      "num_tokens": 42291927.0,
      "step": 20375
    },
    {
      "entropy": 5.476874446868896,
      "epoch": 1.8486937590711174,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004659467246643469,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20838529020547866,
      "num_tokens": 42301407.0,
      "step": 20380
    },
    {
      "entropy": 5.437132501602173,
      "epoch": 1.8491473149492017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046592960771603154,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21864065825939177,
      "num_tokens": 42311199.0,
      "step": 20385
    },
    {
      "entropy": 5.498507404327393,
      "epoch": 1.849600870827286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004659124868192433,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21190747767686843,
      "num_tokens": 42320725.0,
      "step": 20390
    },
    {
      "entropy": 5.598555183410644,
      "epoch": 1.8500544267053702,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004658953619743361,
      "loss": 5.0838,
      "mean_token_accuracy": 0.20678532719612122,
      "num_tokens": 42330857.0,
      "step": 20395
    },
    {
      "entropy": 5.48690185546875,
      "epoch": 1.8505079825834543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046587823318166427,
      "loss": 4.9036,
      "mean_token_accuracy": 0.20804093033075333,
      "num_tokens": 42340241.0,
      "step": 20400
    },
    {
      "entropy": 5.549601602554321,
      "epoch": 1.8509615384615383,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046586110044158184,
      "loss": 5.1084,
      "mean_token_accuracy": 0.1954648569226265,
      "num_tokens": 42351756.0,
      "step": 20405
    },
    {
      "entropy": 5.530306148529053,
      "epoch": 1.8514150943396226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046584396375444335,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21770662665367127,
      "num_tokens": 42361649.0,
      "step": 20410
    },
    {
      "entropy": 5.52125186920166,
      "epoch": 1.8518686502177069,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046582682312060295,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21299203336238862,
      "num_tokens": 42371273.0,
      "step": 20415
    },
    {
      "entropy": 5.460595750808716,
      "epoch": 1.8523222060957911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004658096785404152,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2060428649187088,
      "num_tokens": 42382275.0,
      "step": 20420
    },
    {
      "entropy": 5.58466420173645,
      "epoch": 1.8527757619738752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046579253001423473,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20317630022764205,
      "num_tokens": 42392360.0,
      "step": 20425
    },
    {
      "entropy": 5.54338641166687,
      "epoch": 1.8532293178519592,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046577537754241604,
      "loss": 5.0026,
      "mean_token_accuracy": 0.20023627877235411,
      "num_tokens": 42403695.0,
      "step": 20430
    },
    {
      "entropy": 5.518084383010864,
      "epoch": 1.8536828737300435,
      "grad_norm": 1.0,
      "learning_rate": 0.000465758221125314,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21301815062761306,
      "num_tokens": 42414199.0,
      "step": 20435
    },
    {
      "entropy": 5.533279275894165,
      "epoch": 1.8541364296081277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046574106076328333,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20730967223644256,
      "num_tokens": 42423655.0,
      "step": 20440
    },
    {
      "entropy": 5.498628425598144,
      "epoch": 1.854589985486212,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004657238964566789,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2135024517774582,
      "num_tokens": 42434235.0,
      "step": 20445
    },
    {
      "entropy": 5.472362375259399,
      "epoch": 1.855043541364296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657067282058557,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20742813795804976,
      "num_tokens": 42444701.0,
      "step": 20450
    },
    {
      "entropy": 5.465218162536621,
      "epoch": 1.85549709724238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656895560111687,
      "loss": 4.981,
      "mean_token_accuracy": 0.2064000517129898,
      "num_tokens": 42454499.0,
      "step": 20455
    },
    {
      "entropy": 5.561305713653565,
      "epoch": 1.8559506531204644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046567237987297323,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19911492466926575,
      "num_tokens": 42465391.0,
      "step": 20460
    },
    {
      "entropy": 5.579448318481445,
      "epoch": 1.8564042089985486,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004656551997916243,
      "loss": 4.9686,
      "mean_token_accuracy": 0.19929346442222595,
      "num_tokens": 42476126.0,
      "step": 20465
    },
    {
      "entropy": 5.5623602867126465,
      "epoch": 1.856857764876633,
      "grad_norm": 1.0,
      "learning_rate": 0.00046563801576747716,
      "loss": 5.044,
      "mean_token_accuracy": 0.20417959094047547,
      "num_tokens": 42486833.0,
      "step": 20470
    },
    {
      "entropy": 5.613506937026978,
      "epoch": 1.857311320754717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046562082780088733,
      "loss": 5.071,
      "mean_token_accuracy": 0.20036802142858506,
      "num_tokens": 42496986.0,
      "step": 20475
    },
    {
      "entropy": 5.545318651199341,
      "epoch": 1.8577648766328012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656036358922103,
      "loss": 4.9858,
      "mean_token_accuracy": 0.206824791431427,
      "num_tokens": 42507352.0,
      "step": 20480
    },
    {
      "entropy": 5.529227924346924,
      "epoch": 1.8582184325108853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004655864400418014,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2064192160964012,
      "num_tokens": 42517655.0,
      "step": 20485
    },
    {
      "entropy": 5.485933399200439,
      "epoch": 1.8586719883889695,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004655692402500163,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21793061941862107,
      "num_tokens": 42527502.0,
      "step": 20490
    },
    {
      "entropy": 5.499330377578735,
      "epoch": 1.8591255442670538,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046555203651721085,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20745202898979187,
      "num_tokens": 42538893.0,
      "step": 20495
    },
    {
      "entropy": 5.418507289886475,
      "epoch": 1.8595791001451378,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004655348288437407,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22026781141757965,
      "num_tokens": 42550372.0,
      "step": 20500
    },
    {
      "entropy": 5.5770313262939455,
      "epoch": 1.860032656023222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004655176172299618,
      "loss": 5.0655,
      "mean_token_accuracy": 0.20463417172431947,
      "num_tokens": 42560814.0,
      "step": 20505
    },
    {
      "entropy": 5.595072937011719,
      "epoch": 1.8604862119013061,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046550040167622986,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1982729837298393,
      "num_tokens": 42570993.0,
      "step": 20510
    },
    {
      "entropy": 5.5297078609466555,
      "epoch": 1.8609397677793904,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046548318218290124,
      "loss": 4.9382,
      "mean_token_accuracy": 0.20696137696504593,
      "num_tokens": 42581357.0,
      "step": 20515
    },
    {
      "entropy": 5.518084716796875,
      "epoch": 1.8613933236574747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004654659587503318,
      "loss": 5.068,
      "mean_token_accuracy": 0.19770045429468155,
      "num_tokens": 42591653.0,
      "step": 20520
    },
    {
      "entropy": 5.593356609344482,
      "epoch": 1.861846879535559,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004654487313788778,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1982532650232315,
      "num_tokens": 42602232.0,
      "step": 20525
    },
    {
      "entropy": 5.536996698379516,
      "epoch": 1.862300435413643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046543150006889546,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21202207803726197,
      "num_tokens": 42613283.0,
      "step": 20530
    },
    {
      "entropy": 5.507025051116943,
      "epoch": 1.862753991291727,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004654142648207412,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2139369398355484,
      "num_tokens": 42623594.0,
      "step": 20535
    },
    {
      "entropy": 5.530036115646363,
      "epoch": 1.8632075471698113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653970256347715,
      "loss": 5.041,
      "mean_token_accuracy": 0.20130124390125276,
      "num_tokens": 42633587.0,
      "step": 20540
    },
    {
      "entropy": 5.548380708694458,
      "epoch": 1.8636611030478956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046537978251134265,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20500110536813737,
      "num_tokens": 42643426.0,
      "step": 20545
    },
    {
      "entropy": 5.490745925903321,
      "epoch": 1.8641146589259798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653625354508115,
      "loss": 4.9734,
      "mean_token_accuracy": 0.20890772193670273,
      "num_tokens": 42653521.0,
      "step": 20550
    },
    {
      "entropy": 5.5804942607879635,
      "epoch": 1.8645682148040639,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004653452844535346,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20665431767702103,
      "num_tokens": 42664818.0,
      "step": 20555
    },
    {
      "entropy": 5.613869380950928,
      "epoch": 1.865021770682148,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046532802951986863,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20891079157590867,
      "num_tokens": 42674738.0,
      "step": 20560
    },
    {
      "entropy": 5.470247459411621,
      "epoch": 1.8654753265602322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004653107706501705,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2177759200334549,
      "num_tokens": 42685003.0,
      "step": 20565
    },
    {
      "entropy": 5.454487228393555,
      "epoch": 1.8659288824383164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004652935078447973,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21224944144487382,
      "num_tokens": 42695428.0,
      "step": 20570
    },
    {
      "entropy": 5.55644907951355,
      "epoch": 1.8663824383164007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046527624110410574,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21364315897226333,
      "num_tokens": 42704901.0,
      "step": 20575
    },
    {
      "entropy": 5.598369312286377,
      "epoch": 1.8668359941944848,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000465258970428453,
      "loss": 5.0522,
      "mean_token_accuracy": 0.2018121674656868,
      "num_tokens": 42714379.0,
      "step": 20580
    },
    {
      "entropy": 5.59432315826416,
      "epoch": 1.8672895500725688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046524169581819633,
      "loss": 5.153,
      "mean_token_accuracy": 0.19972582757472992,
      "num_tokens": 42724813.0,
      "step": 20585
    },
    {
      "entropy": 5.480282020568848,
      "epoch": 1.867743105950653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004652244172736929,
      "loss": 4.994,
      "mean_token_accuracy": 0.20874302089214325,
      "num_tokens": 42736590.0,
      "step": 20590
    },
    {
      "entropy": 5.586142301559448,
      "epoch": 1.8681966618287373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004652071347953,
      "loss": 5.0821,
      "mean_token_accuracy": 0.19929955899715424,
      "num_tokens": 42746718.0,
      "step": 20595
    },
    {
      "entropy": 5.540536260604858,
      "epoch": 1.8686502177068216,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046518984838337513,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20269346237182617,
      "num_tokens": 42756484.0,
      "step": 20600
    },
    {
      "entropy": 5.574197816848755,
      "epoch": 1.8691037735849056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004651725580382757,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2045188307762146,
      "num_tokens": 42768512.0,
      "step": 20605
    },
    {
      "entropy": 5.516998624801635,
      "epoch": 1.8695573294629897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046515526376035934,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2083277702331543,
      "num_tokens": 42779812.0,
      "step": 20610
    },
    {
      "entropy": 5.427085828781128,
      "epoch": 1.870010885341074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004651379655499837,
      "loss": 4.9156,
      "mean_token_accuracy": 0.21540848016738892,
      "num_tokens": 42791163.0,
      "step": 20615
    },
    {
      "entropy": 5.494724988937378,
      "epoch": 1.8704644412191582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046512066340750636,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20653896182775497,
      "num_tokens": 42801810.0,
      "step": 20620
    },
    {
      "entropy": 5.5778765201568605,
      "epoch": 1.8709179970972425,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004651033573332853,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2088393285870552,
      "num_tokens": 42813358.0,
      "step": 20625
    },
    {
      "entropy": 5.4334807872772215,
      "epoch": 1.8713715529753265,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004650860473276784,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2156376376748085,
      "num_tokens": 42823519.0,
      "step": 20630
    },
    {
      "entropy": 5.545387315750122,
      "epoch": 1.8718251088534106,
      "grad_norm": 1.125,
      "learning_rate": 0.0004650687333910436,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20554910302162172,
      "num_tokens": 42833257.0,
      "step": 20635
    },
    {
      "entropy": 5.571887588500976,
      "epoch": 1.8722786647314948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046505141552373897,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21233807355165482,
      "num_tokens": 42843536.0,
      "step": 20640
    },
    {
      "entropy": 5.523766040802002,
      "epoch": 1.8727322206095791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004650340937261226,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20724039971828462,
      "num_tokens": 42853911.0,
      "step": 20645
    },
    {
      "entropy": 5.6062156677246096,
      "epoch": 1.8731857764876634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004650167679985528,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19527529627084733,
      "num_tokens": 42864173.0,
      "step": 20650
    },
    {
      "entropy": 5.5203619480133055,
      "epoch": 1.8736393323657474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004649994383413877,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21267371773719787,
      "num_tokens": 42873470.0,
      "step": 20655
    },
    {
      "entropy": 5.506966209411621,
      "epoch": 1.8740928882438317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046498210475498595,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21166199296712876,
      "num_tokens": 42882496.0,
      "step": 20660
    },
    {
      "entropy": 5.506556510925293,
      "epoch": 1.8745464441219157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046496476723970574,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20904543846845627,
      "num_tokens": 42893163.0,
      "step": 20665
    },
    {
      "entropy": 5.50634994506836,
      "epoch": 1.875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649474257959058,
      "loss": 4.965,
      "mean_token_accuracy": 0.20617899894714356,
      "num_tokens": 42903406.0,
      "step": 20670
    },
    {
      "entropy": 5.544222402572632,
      "epoch": 1.8754535558780843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649300804239447,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20702796578407287,
      "num_tokens": 42913702.0,
      "step": 20675
    },
    {
      "entropy": 5.484886598587036,
      "epoch": 1.8759071117561683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046491273112418105,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21404954195022582,
      "num_tokens": 42923877.0,
      "step": 20680
    },
    {
      "entropy": 5.499910020828247,
      "epoch": 1.8763606676342526,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046489537789697386,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20722677558660507,
      "num_tokens": 42934393.0,
      "step": 20685
    },
    {
      "entropy": 5.5295860290527346,
      "epoch": 1.8768142235123366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046487802074268177,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21113871335983275,
      "num_tokens": 42945803.0,
      "step": 20690
    },
    {
      "entropy": 5.5067907810211185,
      "epoch": 1.8772677793904209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046486065966166377,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20922113656997682,
      "num_tokens": 42955829.0,
      "step": 20695
    },
    {
      "entropy": 5.534265613555908,
      "epoch": 1.8777213352685052,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004648432946542791,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20132580101490022,
      "num_tokens": 42967080.0,
      "step": 20700
    },
    {
      "entropy": 5.487653923034668,
      "epoch": 1.8781748911465894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046482592572088654,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20034733563661575,
      "num_tokens": 42978082.0,
      "step": 20705
    },
    {
      "entropy": 5.573316764831543,
      "epoch": 1.8786284470246735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046480855286184565,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20434023290872574,
      "num_tokens": 42988982.0,
      "step": 20710
    },
    {
      "entropy": 5.554499721527099,
      "epoch": 1.8790820029027575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046479117607751535,
      "loss": 4.9838,
      "mean_token_accuracy": 0.19854393154382705,
      "num_tokens": 42998982.0,
      "step": 20715
    },
    {
      "entropy": 5.526091575622559,
      "epoch": 1.8795355587808418,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046477379536825534,
      "loss": 4.9236,
      "mean_token_accuracy": 0.20879895836114884,
      "num_tokens": 43009574.0,
      "step": 20720
    },
    {
      "entropy": 5.548022603988647,
      "epoch": 1.879989114658926,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004647564107344247,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20474085807800294,
      "num_tokens": 43019335.0,
      "step": 20725
    },
    {
      "entropy": 5.468513154983521,
      "epoch": 1.8804426705370103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046473902217638327,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21021214723587037,
      "num_tokens": 43029264.0,
      "step": 20730
    },
    {
      "entropy": 5.4808114528656,
      "epoch": 1.8808962264150944,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046472162969449037,
      "loss": 4.9756,
      "mean_token_accuracy": 0.2108668640255928,
      "num_tokens": 43040519.0,
      "step": 20735
    },
    {
      "entropy": 5.57932825088501,
      "epoch": 1.8813497822931784,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046470423328910594,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21168433129787445,
      "num_tokens": 43051146.0,
      "step": 20740
    },
    {
      "entropy": 5.596247911453247,
      "epoch": 1.8818033381712627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046468683296058957,
      "loss": 5.0823,
      "mean_token_accuracy": 0.20713951736688613,
      "num_tokens": 43061339.0,
      "step": 20745
    },
    {
      "entropy": 5.497734260559082,
      "epoch": 1.882256894049347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046466942870930126,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2102876588702202,
      "num_tokens": 43072624.0,
      "step": 20750
    },
    {
      "entropy": 5.545460033416748,
      "epoch": 1.8827104499274312,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004646520205356007,
      "loss": 4.9513,
      "mean_token_accuracy": 0.20594087541103362,
      "num_tokens": 43082120.0,
      "step": 20755
    },
    {
      "entropy": 5.609503507614136,
      "epoch": 1.8831640058055152,
      "grad_norm": 1.25,
      "learning_rate": 0.00046463460843984805,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20164793282747268,
      "num_tokens": 43093202.0,
      "step": 20760
    },
    {
      "entropy": 5.674992609024048,
      "epoch": 1.8836175616835993,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004646171924224034,
      "loss": 5.1496,
      "mean_token_accuracy": 0.20784732401371003,
      "num_tokens": 43104801.0,
      "step": 20765
    },
    {
      "entropy": 5.5489527702331545,
      "epoch": 1.8840711175616836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046459977248362687,
      "loss": 4.9622,
      "mean_token_accuracy": 0.20929950177669526,
      "num_tokens": 43114855.0,
      "step": 20770
    },
    {
      "entropy": 5.533009958267212,
      "epoch": 1.8845246734397678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004645823486238788,
      "loss": 4.974,
      "mean_token_accuracy": 0.20246503204107286,
      "num_tokens": 43124759.0,
      "step": 20775
    },
    {
      "entropy": 5.508164787292481,
      "epoch": 1.884978229317852,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004645649208435193,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21345612406730652,
      "num_tokens": 43135526.0,
      "step": 20780
    },
    {
      "entropy": 5.540684652328491,
      "epoch": 1.8854317851959361,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046454748914290906,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20299570262432098,
      "num_tokens": 43145605.0,
      "step": 20785
    },
    {
      "entropy": 5.623768997192383,
      "epoch": 1.8858853410740202,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046453005352240835,
      "loss": 5.1192,
      "mean_token_accuracy": 0.19875621050596237,
      "num_tokens": 43156578.0,
      "step": 20790
    },
    {
      "entropy": 5.578191328048706,
      "epoch": 1.8863388969521044,
      "grad_norm": 1.078125,
      "learning_rate": 0.000464512613982378,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20559273660182953,
      "num_tokens": 43166870.0,
      "step": 20795
    },
    {
      "entropy": 5.5336119651794435,
      "epoch": 1.8867924528301887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004644951705231783,
      "loss": 4.947,
      "mean_token_accuracy": 0.21584009975194932,
      "num_tokens": 43176406.0,
      "step": 20800
    },
    {
      "entropy": 5.625538349151611,
      "epoch": 1.887246008708273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046447772314517024,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20117899626493455,
      "num_tokens": 43187502.0,
      "step": 20805
    },
    {
      "entropy": 5.471494913101196,
      "epoch": 1.887699564586357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046446027184871464,
      "loss": 4.9044,
      "mean_token_accuracy": 0.21427054107189178,
      "num_tokens": 43197928.0,
      "step": 20810
    },
    {
      "entropy": 5.504427337646485,
      "epoch": 1.888153120464441,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046444281663417236,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21687044948339462,
      "num_tokens": 43209083.0,
      "step": 20815
    },
    {
      "entropy": 5.551455545425415,
      "epoch": 1.8886066763425253,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004644253575019043,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20451149493455886,
      "num_tokens": 43219089.0,
      "step": 20820
    },
    {
      "entropy": 5.584152936935425,
      "epoch": 1.8890602322206096,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004644078944522716,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20532690286636351,
      "num_tokens": 43229440.0,
      "step": 20825
    },
    {
      "entropy": 5.519352865219116,
      "epoch": 1.8895137880986939,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004643904274856353,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20806962847709656,
      "num_tokens": 43239995.0,
      "step": 20830
    },
    {
      "entropy": 5.4865211009979244,
      "epoch": 1.889967343976778,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046437295660235674,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21096756607294082,
      "num_tokens": 43249722.0,
      "step": 20835
    },
    {
      "entropy": 5.475346326828003,
      "epoch": 1.8904208998548622,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004643554818027972,
      "loss": 5.011,
      "mean_token_accuracy": 0.2061210662126541,
      "num_tokens": 43260029.0,
      "step": 20840
    },
    {
      "entropy": 5.555378055572509,
      "epoch": 1.8908744557329462,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004643380030873181,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2109837368130684,
      "num_tokens": 43270571.0,
      "step": 20845
    },
    {
      "entropy": 5.538992643356323,
      "epoch": 1.8913280116110305,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004643205204562808,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2108249172568321,
      "num_tokens": 43281310.0,
      "step": 20850
    },
    {
      "entropy": 5.512079238891602,
      "epoch": 1.8917815674891147,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046430303391004687,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2100333169102669,
      "num_tokens": 43292904.0,
      "step": 20855
    },
    {
      "entropy": 5.523993253707886,
      "epoch": 1.8922351233671988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004642855434489779,
      "loss": 4.9749,
      "mean_token_accuracy": 0.20685950964689254,
      "num_tokens": 43303745.0,
      "step": 20860
    },
    {
      "entropy": 5.570832300186157,
      "epoch": 1.892688679245283,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046426804907343574,
      "loss": 5.0057,
      "mean_token_accuracy": 0.20969969779253006,
      "num_tokens": 43314652.0,
      "step": 20865
    },
    {
      "entropy": 5.496357393264771,
      "epoch": 1.893142235123367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046425055078378206,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2160039648413658,
      "num_tokens": 43324242.0,
      "step": 20870
    },
    {
      "entropy": 5.48613805770874,
      "epoch": 1.8935957910014514,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046423304858037885,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21549898535013198,
      "num_tokens": 43334054.0,
      "step": 20875
    },
    {
      "entropy": 5.474690866470337,
      "epoch": 1.8940493468795356,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004642155424635878,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22431363314390182,
      "num_tokens": 43343389.0,
      "step": 20880
    },
    {
      "entropy": 5.483912801742553,
      "epoch": 1.89450290275762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046419803243377125,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21698027998209,
      "num_tokens": 43353892.0,
      "step": 20885
    },
    {
      "entropy": 5.466065549850464,
      "epoch": 1.894956458635704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046418051849129115,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21455236077308654,
      "num_tokens": 43363912.0,
      "step": 20890
    },
    {
      "entropy": 5.513872575759888,
      "epoch": 1.895410014513788,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004641630006365097,
      "loss": 5.0315,
      "mean_token_accuracy": 0.202163402736187,
      "num_tokens": 43374719.0,
      "step": 20895
    },
    {
      "entropy": 5.570595026016235,
      "epoch": 1.8958635703918723,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046414547886978926,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20297456681728362,
      "num_tokens": 43385549.0,
      "step": 20900
    },
    {
      "entropy": 5.568581295013428,
      "epoch": 1.8963171262699565,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046412795319149204,
      "loss": 5.0238,
      "mean_token_accuracy": 0.216174978017807,
      "num_tokens": 43395503.0,
      "step": 20905
    },
    {
      "entropy": 5.515105676651001,
      "epoch": 1.8967706821480408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046411042360198053,
      "loss": 5.026,
      "mean_token_accuracy": 0.20338638573884965,
      "num_tokens": 43406913.0,
      "step": 20910
    },
    {
      "entropy": 5.531669187545776,
      "epoch": 1.8972242380261248,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640928901016173,
      "loss": 4.9338,
      "mean_token_accuracy": 0.20720363557338714,
      "num_tokens": 43416459.0,
      "step": 20915
    },
    {
      "entropy": 5.509416341781616,
      "epoch": 1.8976777939042089,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046407535269076493,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20892471969127654,
      "num_tokens": 43426280.0,
      "step": 20920
    },
    {
      "entropy": 5.583643436431885,
      "epoch": 1.8981313497822931,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046405781136978606,
      "loss": 5.0225,
      "mean_token_accuracy": 0.20613077729940416,
      "num_tokens": 43436676.0,
      "step": 20925
    },
    {
      "entropy": 5.63284592628479,
      "epoch": 1.8985849056603774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046404026613904357,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20198507159948348,
      "num_tokens": 43447019.0,
      "step": 20930
    },
    {
      "entropy": 5.572532033920288,
      "epoch": 1.8990384615384617,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004640227169989001,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20607016682624818,
      "num_tokens": 43456609.0,
      "step": 20935
    },
    {
      "entropy": 5.550652933120728,
      "epoch": 1.8994920174165457,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004640051639497187,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20605168044567107,
      "num_tokens": 43467616.0,
      "step": 20940
    },
    {
      "entropy": 5.454766130447387,
      "epoch": 1.8999455732946298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004639876069918624,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20363384038209914,
      "num_tokens": 43478209.0,
      "step": 20945
    },
    {
      "entropy": 5.518501901626587,
      "epoch": 1.900399129172714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046397004612569416,
      "loss": 4.8878,
      "mean_token_accuracy": 0.21243149936199188,
      "num_tokens": 43487278.0,
      "step": 20950
    },
    {
      "entropy": 5.4941329002380375,
      "epoch": 1.9008526850507983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046395248135157715,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2183908686041832,
      "num_tokens": 43497457.0,
      "step": 20955
    },
    {
      "entropy": 5.543423986434936,
      "epoch": 1.9013062409288826,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046393491266987484,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20089227706193924,
      "num_tokens": 43507761.0,
      "step": 20960
    },
    {
      "entropy": 5.468902492523194,
      "epoch": 1.9017597968069666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004639173400809502,
      "loss": 4.9417,
      "mean_token_accuracy": 0.20726902931928634,
      "num_tokens": 43517198.0,
      "step": 20965
    },
    {
      "entropy": 5.527848100662231,
      "epoch": 1.9022133526850507,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004638997635851669,
      "loss": 5.0474,
      "mean_token_accuracy": 0.20085853487253189,
      "num_tokens": 43527429.0,
      "step": 20970
    },
    {
      "entropy": 5.603863906860352,
      "epoch": 1.902666908563135,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046388218318288844,
      "loss": 5.0512,
      "mean_token_accuracy": 0.2058670088648796,
      "num_tokens": 43537867.0,
      "step": 20975
    },
    {
      "entropy": 5.544486427307129,
      "epoch": 1.9031204644412192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004638645988744782,
      "loss": 4.977,
      "mean_token_accuracy": 0.2110976591706276,
      "num_tokens": 43547284.0,
      "step": 20980
    },
    {
      "entropy": 5.5339678764343265,
      "epoch": 1.9035740203193035,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638470106603,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20343973636627197,
      "num_tokens": 43557693.0,
      "step": 20985
    },
    {
      "entropy": 5.572199535369873,
      "epoch": 1.9040275761973875,
      "grad_norm": 1.125,
      "learning_rate": 0.0004638294185407174,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2162001460790634,
      "num_tokens": 43566909.0,
      "step": 20990
    },
    {
      "entropy": 5.541551828384399,
      "epoch": 1.9044811320754715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638118225160944,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2148605614900589,
      "num_tokens": 43576963.0,
      "step": 20995
    },
    {
      "entropy": 5.45386323928833,
      "epoch": 1.9049346879535558,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046379422258679477,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20698957294225692,
      "num_tokens": 43586649.0,
      "step": 21000
    },
    {
      "epoch": 1.9049346879535558,
      "eval_entropy": 5.257148189674136,
      "eval_loss": 5.037424087524414,
      "eval_mean_token_accuracy": 0.21348288062437357,
      "eval_num_tokens": 43586649.0,
      "eval_runtime": 20.2464,
      "eval_samples_per_second": 1746.479,
      "eval_steps_per_second": 218.31,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9472777408696320.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}