{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5442670537010159,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 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
    }
  ],
  "logging_steps": 5,
  "max_steps": 110230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2706712035655680.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}