{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.3472803347280333,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742597675323486,
      "epoch": 0.0041841004184100415,
      "grad_norm": 5.46875,
      "learning_rate": 2e-06,
      "loss": 10.7608,
      "mean_token_accuracy": 9.661835501901805e-05,
      "num_tokens": 9848.0,
      "step": 5
    },
    {
      "entropy": 10.74262809753418,
      "epoch": 0.008368200836820083,
      "grad_norm": 5.5,
      "learning_rate": 4.5e-06,
      "loss": 10.723,
      "mean_token_accuracy": 0.0,
      "num_tokens": 18196.0,
      "step": 10
    },
    {
      "entropy": 10.742603778839111,
      "epoch": 0.012552301255230125,
      "grad_norm": 5.34375,
      "learning_rate": 7e-06,
      "loss": 10.703,
      "mean_token_accuracy": 0.0,
      "num_tokens": 26927.0,
      "step": 15
    },
    {
      "entropy": 10.742668437957764,
      "epoch": 0.016736401673640166,
      "grad_norm": 5.65625,
      "learning_rate": 9.5e-06,
      "loss": 10.6357,
      "mean_token_accuracy": 0.0005675206426531077,
      "num_tokens": 34987.0,
      "step": 20
    },
    {
      "entropy": 10.742566776275634,
      "epoch": 0.02092050209205021,
      "grad_norm": 5.03125,
      "learning_rate": 1.2e-05,
      "loss": 10.5588,
      "mean_token_accuracy": 0.010127833811566234,
      "num_tokens": 44936.0,
      "step": 25
    },
    {
      "entropy": 10.742234230041504,
      "epoch": 0.02510460251046025,
      "grad_norm": 4.3125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.427,
      "mean_token_accuracy": 0.0348518967628479,
      "num_tokens": 54562.0,
      "step": 30
    },
    {
      "entropy": 10.741086196899413,
      "epoch": 0.029288702928870293,
      "grad_norm": 3.359375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.2814,
      "mean_token_accuracy": 0.03771395143121481,
      "num_tokens": 63844.0,
      "step": 35
    },
    {
      "entropy": 10.738179874420165,
      "epoch": 0.03347280334728033,
      "grad_norm": 2.734375,
      "learning_rate": 1.95e-05,
      "loss": 10.1587,
      "mean_token_accuracy": 0.04109239801764488,
      "num_tokens": 74331.0,
      "step": 40
    },
    {
      "entropy": 10.734289169311523,
      "epoch": 0.03765690376569038,
      "grad_norm": 2.484375,
      "learning_rate": 2.2e-05,
      "loss": 9.995,
      "mean_token_accuracy": 0.04218972772359848,
      "num_tokens": 83028.0,
      "step": 45
    },
    {
      "entropy": 10.730850219726562,
      "epoch": 0.04184100418410042,
      "grad_norm": 2.15625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9353,
      "mean_token_accuracy": 0.040206428803503516,
      "num_tokens": 91395.0,
      "step": 50
    },
    {
      "entropy": 10.729630088806152,
      "epoch": 0.04602510460251046,
      "grad_norm": 2.28125,
      "learning_rate": 2.7e-05,
      "loss": 9.8674,
      "mean_token_accuracy": 0.048422401398420335,
      "num_tokens": 101868.0,
      "step": 55
    },
    {
      "entropy": 10.727583599090575,
      "epoch": 0.0502092050209205,
      "grad_norm": 2.015625,
      "learning_rate": 2.95e-05,
      "loss": 9.7933,
      "mean_token_accuracy": 0.04613714665174484,
      "num_tokens": 110403.0,
      "step": 60
    },
    {
      "entropy": 10.723635578155518,
      "epoch": 0.05439330543933055,
      "grad_norm": 1.921875,
      "learning_rate": 3.2e-05,
      "loss": 9.8021,
      "mean_token_accuracy": 0.041691455990076065,
      "num_tokens": 119557.0,
      "step": 65
    },
    {
      "entropy": 10.719999694824219,
      "epoch": 0.058577405857740586,
      "grad_norm": 2.0,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.676,
      "mean_token_accuracy": 0.04646212197840214,
      "num_tokens": 128903.0,
      "step": 70
    },
    {
      "entropy": 10.71647481918335,
      "epoch": 0.06276150627615062,
      "grad_norm": 1.8359375,
      "learning_rate": 3.7e-05,
      "loss": 9.6628,
      "mean_token_accuracy": 0.041718531772494315,
      "num_tokens": 137630.0,
      "step": 75
    },
    {
      "entropy": 10.713119316101075,
      "epoch": 0.06694560669456066,
      "grad_norm": 2.0,
      "learning_rate": 3.95e-05,
      "loss": 9.5313,
      "mean_token_accuracy": 0.0421627514064312,
      "num_tokens": 146335.0,
      "step": 80
    },
    {
      "entropy": 10.708419799804688,
      "epoch": 0.07112970711297072,
      "grad_norm": 2.0,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.4686,
      "mean_token_accuracy": 0.04617839008569717,
      "num_tokens": 155766.0,
      "step": 85
    },
    {
      "entropy": 10.702109909057617,
      "epoch": 0.07531380753138076,
      "grad_norm": 1.859375,
      "learning_rate": 4.45e-05,
      "loss": 9.412,
      "mean_token_accuracy": 0.04476847723126411,
      "num_tokens": 165844.0,
      "step": 90
    },
    {
      "entropy": 10.693535137176514,
      "epoch": 0.0794979079497908,
      "grad_norm": 1.90625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.3394,
      "mean_token_accuracy": 0.05125464424490929,
      "num_tokens": 174858.0,
      "step": 95
    },
    {
      "entropy": 10.678814315795899,
      "epoch": 0.08368200836820083,
      "grad_norm": 1.9140625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.2166,
      "mean_token_accuracy": 0.06135832369327545,
      "num_tokens": 184020.0,
      "step": 100
    },
    {
      "entropy": 10.653692531585694,
      "epoch": 0.08786610878661087,
      "grad_norm": 1.8828125,
      "learning_rate": 5.2e-05,
      "loss": 9.115,
      "mean_token_accuracy": 0.05314192697405815,
      "num_tokens": 193714.0,
      "step": 105
    },
    {
      "entropy": 10.62697982788086,
      "epoch": 0.09205020920502092,
      "grad_norm": 1.90625,
      "learning_rate": 5.45e-05,
      "loss": 9.0408,
      "mean_token_accuracy": 0.05419124960899353,
      "num_tokens": 203139.0,
      "step": 110
    },
    {
      "entropy": 10.581832027435302,
      "epoch": 0.09623430962343096,
      "grad_norm": 1.875,
      "learning_rate": 5.7e-05,
      "loss": 8.8865,
      "mean_token_accuracy": 0.05798161998391151,
      "num_tokens": 212446.0,
      "step": 115
    },
    {
      "entropy": 10.510433387756347,
      "epoch": 0.100418410041841,
      "grad_norm": 1.8046875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.7392,
      "mean_token_accuracy": 0.06025223545730114,
      "num_tokens": 220910.0,
      "step": 120
    },
    {
      "entropy": 10.451266765594482,
      "epoch": 0.10460251046025104,
      "grad_norm": 1.6953125,
      "learning_rate": 6.2e-05,
      "loss": 8.6729,
      "mean_token_accuracy": 0.055778010934591296,
      "num_tokens": 230522.0,
      "step": 125
    },
    {
      "entropy": 10.406130599975587,
      "epoch": 0.1087866108786611,
      "grad_norm": 1.7578125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5065,
      "mean_token_accuracy": 0.05544214770197868,
      "num_tokens": 240092.0,
      "step": 130
    },
    {
      "entropy": 10.331532001495361,
      "epoch": 0.11297071129707113,
      "grad_norm": 1.5625,
      "learning_rate": 6.7e-05,
      "loss": 8.4083,
      "mean_token_accuracy": 0.06165078543126583,
      "num_tokens": 249478.0,
      "step": 135
    },
    {
      "entropy": 10.237152481079102,
      "epoch": 0.11715481171548117,
      "grad_norm": 1.71875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.2,
      "mean_token_accuracy": 0.06240398585796356,
      "num_tokens": 257714.0,
      "step": 140
    },
    {
      "entropy": 10.116984844207764,
      "epoch": 0.12133891213389121,
      "grad_norm": 1.4296875,
      "learning_rate": 7.2e-05,
      "loss": 8.1928,
      "mean_token_accuracy": 0.05669947192072868,
      "num_tokens": 266569.0,
      "step": 145
    },
    {
      "entropy": 10.02002773284912,
      "epoch": 0.12552301255230125,
      "grad_norm": 1.375,
      "learning_rate": 7.45e-05,
      "loss": 8.0619,
      "mean_token_accuracy": 0.06153124496340752,
      "num_tokens": 276688.0,
      "step": 150
    },
    {
      "entropy": 9.869840335845947,
      "epoch": 0.1297071129707113,
      "grad_norm": 1.3359375,
      "learning_rate": 7.7e-05,
      "loss": 7.8676,
      "mean_token_accuracy": 0.06409889869391919,
      "num_tokens": 285430.0,
      "step": 155
    },
    {
      "entropy": 9.67565851211548,
      "epoch": 0.13389121338912133,
      "grad_norm": 1.4375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.789,
      "mean_token_accuracy": 0.06454115808010101,
      "num_tokens": 295727.0,
      "step": 160
    },
    {
      "entropy": 9.464604949951172,
      "epoch": 0.13807531380753138,
      "grad_norm": 1.1171875,
      "learning_rate": 8.2e-05,
      "loss": 7.6957,
      "mean_token_accuracy": 0.06615287810564041,
      "num_tokens": 304854.0,
      "step": 165
    },
    {
      "entropy": 9.24670057296753,
      "epoch": 0.14225941422594143,
      "grad_norm": 1.3046875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.6093,
      "mean_token_accuracy": 0.07150843031704426,
      "num_tokens": 313579.0,
      "step": 170
    },
    {
      "entropy": 9.048341655731202,
      "epoch": 0.14644351464435146,
      "grad_norm": 1.0234375,
      "learning_rate": 8.7e-05,
      "loss": 7.6128,
      "mean_token_accuracy": 0.07077031470835209,
      "num_tokens": 323310.0,
      "step": 175
    },
    {
      "entropy": 8.823628616333007,
      "epoch": 0.1506276150627615,
      "grad_norm": 1.03125,
      "learning_rate": 8.95e-05,
      "loss": 7.4519,
      "mean_token_accuracy": 0.06803311333060265,
      "num_tokens": 331720.0,
      "step": 180
    },
    {
      "entropy": 8.658255577087402,
      "epoch": 0.15481171548117154,
      "grad_norm": 1.15625,
      "learning_rate": 9.2e-05,
      "loss": 7.3797,
      "mean_token_accuracy": 0.07455825209617614,
      "num_tokens": 340331.0,
      "step": 185
    },
    {
      "entropy": 8.487533950805664,
      "epoch": 0.1589958158995816,
      "grad_norm": 1.2421875,
      "learning_rate": 9.45e-05,
      "loss": 7.3713,
      "mean_token_accuracy": 0.07890852093696595,
      "num_tokens": 349677.0,
      "step": 190
    },
    {
      "entropy": 8.328230381011963,
      "epoch": 0.16317991631799164,
      "grad_norm": 1.3046875,
      "learning_rate": 9.7e-05,
      "loss": 7.3358,
      "mean_token_accuracy": 0.06953225210309029,
      "num_tokens": 358674.0,
      "step": 195
    },
    {
      "entropy": 8.279506969451905,
      "epoch": 0.16736401673640167,
      "grad_norm": 0.9609375,
      "learning_rate": 9.95e-05,
      "loss": 7.33,
      "mean_token_accuracy": 0.07827548906207085,
      "num_tokens": 368258.0,
      "step": 200
    },
    {
      "entropy": 8.202318572998047,
      "epoch": 0.17154811715481172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000102,
      "loss": 7.2291,
      "mean_token_accuracy": 0.08247470743954181,
      "num_tokens": 377083.0,
      "step": 205
    },
    {
      "entropy": 8.081873416900635,
      "epoch": 0.17573221757322174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2799,
      "mean_token_accuracy": 0.07478422187268734,
      "num_tokens": 385456.0,
      "step": 210
    },
    {
      "entropy": 8.072468662261963,
      "epoch": 0.1799163179916318,
      "grad_norm": 1.46875,
      "learning_rate": 0.000107,
      "loss": 7.1697,
      "mean_token_accuracy": 0.0813635230064392,
      "num_tokens": 395295.0,
      "step": 215
    },
    {
      "entropy": 8.06587061882019,
      "epoch": 0.18410041841004185,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001095,
      "loss": 7.2016,
      "mean_token_accuracy": 0.07997569367289543,
      "num_tokens": 405023.0,
      "step": 220
    },
    {
      "entropy": 7.917844343185425,
      "epoch": 0.18828451882845187,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000112,
      "loss": 7.2075,
      "mean_token_accuracy": 0.07806592285633088,
      "num_tokens": 414224.0,
      "step": 225
    },
    {
      "entropy": 7.906620788574219,
      "epoch": 0.19246861924686193,
      "grad_norm": 0.9375,
      "learning_rate": 0.0001145,
      "loss": 7.0979,
      "mean_token_accuracy": 0.07944381982088089,
      "num_tokens": 423951.0,
      "step": 230
    },
    {
      "entropy": 7.995381021499634,
      "epoch": 0.19665271966527198,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1259,
      "mean_token_accuracy": 0.08254904597997666,
      "num_tokens": 432427.0,
      "step": 235
    },
    {
      "entropy": 7.85433292388916,
      "epoch": 0.200836820083682,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.0351,
      "mean_token_accuracy": 0.08393027260899544,
      "num_tokens": 441560.0,
      "step": 240
    },
    {
      "entropy": 7.8878974437713625,
      "epoch": 0.20502092050209206,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000122,
      "loss": 7.1383,
      "mean_token_accuracy": 0.0793025080114603,
      "num_tokens": 450364.0,
      "step": 245
    },
    {
      "entropy": 7.774558973312378,
      "epoch": 0.20920502092050208,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0001245,
      "loss": 7.0707,
      "mean_token_accuracy": 0.08014305979013443,
      "num_tokens": 460076.0,
      "step": 250
    },
    {
      "entropy": 7.792851877212525,
      "epoch": 0.21338912133891214,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000127,
      "loss": 7.023,
      "mean_token_accuracy": 0.08774993643164634,
      "num_tokens": 468718.0,
      "step": 255
    },
    {
      "entropy": 7.835711526870727,
      "epoch": 0.2175732217573222,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001295,
      "loss": 7.0841,
      "mean_token_accuracy": 0.08281839713454246,
      "num_tokens": 478436.0,
      "step": 260
    },
    {
      "entropy": 7.807744693756104,
      "epoch": 0.2217573221757322,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000132,
      "loss": 7.0932,
      "mean_token_accuracy": 0.08015878945589065,
      "num_tokens": 486833.0,
      "step": 265
    },
    {
      "entropy": 7.68460259437561,
      "epoch": 0.22594142259414227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.9589,
      "mean_token_accuracy": 0.09470265135169029,
      "num_tokens": 495119.0,
      "step": 270
    },
    {
      "entropy": 7.6987611770629885,
      "epoch": 0.2301255230125523,
      "grad_norm": 1.515625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1042,
      "mean_token_accuracy": 0.09113300144672394,
      "num_tokens": 505396.0,
      "step": 275
    },
    {
      "entropy": 7.621922779083252,
      "epoch": 0.23430962343096234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001395,
      "loss": 6.9717,
      "mean_token_accuracy": 0.08859988525509835,
      "num_tokens": 515259.0,
      "step": 280
    },
    {
      "entropy": 7.760473251342773,
      "epoch": 0.2384937238493724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.9769,
      "mean_token_accuracy": 0.09006874561309815,
      "num_tokens": 523608.0,
      "step": 285
    },
    {
      "entropy": 7.605395030975342,
      "epoch": 0.24267782426778242,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001445,
      "loss": 6.9736,
      "mean_token_accuracy": 0.08525220081210136,
      "num_tokens": 534613.0,
      "step": 290
    },
    {
      "entropy": 7.54643816947937,
      "epoch": 0.24686192468619247,
      "grad_norm": 1.015625,
      "learning_rate": 0.000147,
      "loss": 6.9884,
      "mean_token_accuracy": 0.09405484721064568,
      "num_tokens": 543347.0,
      "step": 295
    },
    {
      "entropy": 7.628460121154785,
      "epoch": 0.2510460251046025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0001495,
      "loss": 6.956,
      "mean_token_accuracy": 0.08765949010848999,
      "num_tokens": 552371.0,
      "step": 300
    },
    {
      "entropy": 7.5912059307098385,
      "epoch": 0.25523012552301255,
      "grad_norm": 1.0625,
      "learning_rate": 0.000152,
      "loss": 6.8822,
      "mean_token_accuracy": 0.08999307677149773,
      "num_tokens": 560742.0,
      "step": 305
    },
    {
      "entropy": 7.553208541870117,
      "epoch": 0.2594142259414226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9822,
      "mean_token_accuracy": 0.09437551945447922,
      "num_tokens": 569960.0,
      "step": 310
    },
    {
      "entropy": 7.580251216888428,
      "epoch": 0.26359832635983266,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000157,
      "loss": 6.9347,
      "mean_token_accuracy": 0.08774123415350914,
      "num_tokens": 578859.0,
      "step": 315
    },
    {
      "entropy": 7.558048486709595,
      "epoch": 0.26778242677824265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001595,
      "loss": 6.9406,
      "mean_token_accuracy": 0.09432831630110741,
      "num_tokens": 588146.0,
      "step": 320
    },
    {
      "entropy": 7.445785140991211,
      "epoch": 0.2719665271966527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000162,
      "loss": 6.8822,
      "mean_token_accuracy": 0.0896220214664936,
      "num_tokens": 597268.0,
      "step": 325
    },
    {
      "entropy": 7.3907698631286625,
      "epoch": 0.27615062761506276,
      "grad_norm": 1.0,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.8154,
      "mean_token_accuracy": 0.0945595309138298,
      "num_tokens": 606199.0,
      "step": 330
    },
    {
      "entropy": 7.577031898498535,
      "epoch": 0.2803347280334728,
      "grad_norm": 1.15625,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9307,
      "mean_token_accuracy": 0.08274017758667469,
      "num_tokens": 615966.0,
      "step": 335
    },
    {
      "entropy": 7.524800825119018,
      "epoch": 0.28451882845188287,
      "grad_norm": 1.0625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9039,
      "mean_token_accuracy": 0.08914825171232224,
      "num_tokens": 625689.0,
      "step": 340
    },
    {
      "entropy": 7.424820232391357,
      "epoch": 0.28870292887029286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.9367,
      "mean_token_accuracy": 0.08863812088966369,
      "num_tokens": 634877.0,
      "step": 345
    },
    {
      "entropy": 7.479735708236694,
      "epoch": 0.2928870292887029,
      "grad_norm": 1.125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.8576,
      "mean_token_accuracy": 0.09228913635015487,
      "num_tokens": 644042.0,
      "step": 350
    },
    {
      "entropy": 7.3995614528656,
      "epoch": 0.29707112970711297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000177,
      "loss": 6.8397,
      "mean_token_accuracy": 0.09121191054582596,
      "num_tokens": 653072.0,
      "step": 355
    },
    {
      "entropy": 7.390188837051392,
      "epoch": 0.301255230125523,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0001795,
      "loss": 6.8429,
      "mean_token_accuracy": 0.0936873622238636,
      "num_tokens": 663686.0,
      "step": 360
    },
    {
      "entropy": 7.433347082138061,
      "epoch": 0.3054393305439331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000182,
      "loss": 6.7824,
      "mean_token_accuracy": 0.09806219711899758,
      "num_tokens": 672293.0,
      "step": 365
    },
    {
      "entropy": 7.224409723281861,
      "epoch": 0.30962343096234307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001845,
      "loss": 6.7495,
      "mean_token_accuracy": 0.09524724259972572,
      "num_tokens": 681758.0,
      "step": 370
    },
    {
      "entropy": 7.373961687088013,
      "epoch": 0.3138075313807531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000187,
      "loss": 6.7705,
      "mean_token_accuracy": 0.0979244202375412,
      "num_tokens": 690727.0,
      "step": 375
    },
    {
      "entropy": 7.338280773162841,
      "epoch": 0.3179916317991632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001895,
      "loss": 6.7246,
      "mean_token_accuracy": 0.09411645382642746,
      "num_tokens": 701210.0,
      "step": 380
    },
    {
      "entropy": 7.166968154907226,
      "epoch": 0.32217573221757323,
      "grad_norm": 1.375,
      "learning_rate": 0.000192,
      "loss": 6.7198,
      "mean_token_accuracy": 0.0960633248090744,
      "num_tokens": 710869.0,
      "step": 385
    },
    {
      "entropy": 7.326888370513916,
      "epoch": 0.3263598326359833,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001945,
      "loss": 6.8285,
      "mean_token_accuracy": 0.09366477653384209,
      "num_tokens": 721455.0,
      "step": 390
    },
    {
      "entropy": 7.2980750560760494,
      "epoch": 0.3305439330543933,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7142,
      "mean_token_accuracy": 0.10093756765127182,
      "num_tokens": 730917.0,
      "step": 395
    },
    {
      "entropy": 7.274618291854859,
      "epoch": 0.33472803347280333,
      "grad_norm": 1.09375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7402,
      "mean_token_accuracy": 0.09866241216659546,
      "num_tokens": 739302.0,
      "step": 400
    },
    {
      "entropy": 7.29223780632019,
      "epoch": 0.3389121338912134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000202,
      "loss": 6.7097,
      "mean_token_accuracy": 0.09633656591176987,
      "num_tokens": 748785.0,
      "step": 405
    },
    {
      "entropy": 7.214037179946899,
      "epoch": 0.34309623430962344,
      "grad_norm": 1.125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.6545,
      "mean_token_accuracy": 0.10300204828381539,
      "num_tokens": 757215.0,
      "step": 410
    },
    {
      "entropy": 7.234053421020508,
      "epoch": 0.3472803347280335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000207,
      "loss": 6.5793,
      "mean_token_accuracy": 0.10455078259110451,
      "num_tokens": 765682.0,
      "step": 415
    },
    {
      "entropy": 7.289216184616089,
      "epoch": 0.3514644351464435,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002095,
      "loss": 6.7796,
      "mean_token_accuracy": 0.0947409562766552,
      "num_tokens": 774814.0,
      "step": 420
    },
    {
      "entropy": 7.198604440689087,
      "epoch": 0.35564853556485354,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000212,
      "loss": 6.6991,
      "mean_token_accuracy": 0.1013828955590725,
      "num_tokens": 784019.0,
      "step": 425
    },
    {
      "entropy": 7.1404036521911625,
      "epoch": 0.3598326359832636,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002145,
      "loss": 6.5758,
      "mean_token_accuracy": 0.10198531225323677,
      "num_tokens": 793237.0,
      "step": 430
    },
    {
      "entropy": 7.1974271774292,
      "epoch": 0.36401673640167365,
      "grad_norm": 1.109375,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6162,
      "mean_token_accuracy": 0.10484573319554329,
      "num_tokens": 801548.0,
      "step": 435
    },
    {
      "entropy": 7.14229302406311,
      "epoch": 0.3682008368200837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002195,
      "loss": 6.6647,
      "mean_token_accuracy": 0.10372263044118882,
      "num_tokens": 810974.0,
      "step": 440
    },
    {
      "entropy": 7.127859115600586,
      "epoch": 0.3723849372384937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000222,
      "loss": 6.5601,
      "mean_token_accuracy": 0.1046792708337307,
      "num_tokens": 820023.0,
      "step": 445
    },
    {
      "entropy": 7.167815637588501,
      "epoch": 0.37656903765690375,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002245,
      "loss": 6.644,
      "mean_token_accuracy": 0.10097603350877762,
      "num_tokens": 828297.0,
      "step": 450
    },
    {
      "entropy": 7.183742904663086,
      "epoch": 0.3807531380753138,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.693,
      "mean_token_accuracy": 0.0922306977212429,
      "num_tokens": 838268.0,
      "step": 455
    },
    {
      "entropy": 7.183096885681152,
      "epoch": 0.38493723849372385,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6117,
      "mean_token_accuracy": 0.10094281509518624,
      "num_tokens": 846491.0,
      "step": 460
    },
    {
      "entropy": 7.094890165328979,
      "epoch": 0.3891213389121339,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.5546,
      "mean_token_accuracy": 0.10523919835686683,
      "num_tokens": 856501.0,
      "step": 465
    },
    {
      "entropy": 7.092893457412719,
      "epoch": 0.39330543933054396,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.5958,
      "mean_token_accuracy": 0.10569338127970695,
      "num_tokens": 867054.0,
      "step": 470
    },
    {
      "entropy": 7.0160187721252445,
      "epoch": 0.39748953974895396,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000237,
      "loss": 6.6239,
      "mean_token_accuracy": 0.10482471361756325,
      "num_tokens": 875956.0,
      "step": 475
    },
    {
      "entropy": 7.013614892959595,
      "epoch": 0.401673640167364,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002395,
      "loss": 6.5334,
      "mean_token_accuracy": 0.1062661923468113,
      "num_tokens": 884427.0,
      "step": 480
    },
    {
      "entropy": 7.170659828186035,
      "epoch": 0.40585774058577406,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000242,
      "loss": 6.6249,
      "mean_token_accuracy": 0.10283279418945312,
      "num_tokens": 894181.0,
      "step": 485
    },
    {
      "entropy": 6.948815584182739,
      "epoch": 0.4100418410041841,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002445,
      "loss": 6.4707,
      "mean_token_accuracy": 0.10783380344510078,
      "num_tokens": 904076.0,
      "step": 490
    },
    {
      "entropy": 7.047258281707764,
      "epoch": 0.41422594142259417,
      "grad_norm": 0.953125,
      "learning_rate": 0.000247,
      "loss": 6.6131,
      "mean_token_accuracy": 0.10789759010076523,
      "num_tokens": 913915.0,
      "step": 495
    },
    {
      "entropy": 7.123381805419922,
      "epoch": 0.41841004184100417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002495,
      "loss": 6.6688,
      "mean_token_accuracy": 0.1050674706697464,
      "num_tokens": 922876.0,
      "step": 500
    },
    {
      "epoch": 0.41841004184100417,
      "eval_entropy": 7.012504460621465,
      "eval_loss": 6.6101908683776855,
      "eval_mean_token_accuracy": 0.10924081287226795,
      "eval_num_tokens": 922876.0,
      "eval_runtime": 2.0948,
      "eval_samples_per_second": 1865.089,
      "eval_steps_per_second": 233.434,
      "step": 500
    },
    {
      "entropy": 7.033219575881958,
      "epoch": 0.4225941422594142,
      "grad_norm": 1.125,
      "learning_rate": 0.000252,
      "loss": 6.4668,
      "mean_token_accuracy": 0.11332256868481635,
      "num_tokens": 931998.0,
      "step": 505
    },
    {
      "entropy": 7.079867506027222,
      "epoch": 0.42677824267782427,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002545,
      "loss": 6.5197,
      "mean_token_accuracy": 0.11432325989007949,
      "num_tokens": 941387.0,
      "step": 510
    },
    {
      "entropy": 6.93487229347229,
      "epoch": 0.4309623430962343,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000257,
      "loss": 6.5834,
      "mean_token_accuracy": 0.10249554514884948,
      "num_tokens": 951866.0,
      "step": 515
    },
    {
      "entropy": 6.953582286834717,
      "epoch": 0.4351464435146444,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002595,
      "loss": 6.5018,
      "mean_token_accuracy": 0.10587486550211907,
      "num_tokens": 961252.0,
      "step": 520
    },
    {
      "entropy": 7.056652307510376,
      "epoch": 0.4393305439330544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000262,
      "loss": 6.5132,
      "mean_token_accuracy": 0.1105636864900589,
      "num_tokens": 969857.0,
      "step": 525
    },
    {
      "entropy": 6.941860485076904,
      "epoch": 0.4435146443514644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5023,
      "mean_token_accuracy": 0.10602044016122818,
      "num_tokens": 978754.0,
      "step": 530
    },
    {
      "entropy": 7.038395500183105,
      "epoch": 0.4476987447698745,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.4919,
      "mean_token_accuracy": 0.110679891705513,
      "num_tokens": 987935.0,
      "step": 535
    },
    {
      "entropy": 6.966084718704224,
      "epoch": 0.45188284518828453,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.4664,
      "mean_token_accuracy": 0.10865552499890327,
      "num_tokens": 996873.0,
      "step": 540
    },
    {
      "entropy": 6.99370813369751,
      "epoch": 0.4560669456066946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5515,
      "mean_token_accuracy": 0.1058430440723896,
      "num_tokens": 1006487.0,
      "step": 545
    },
    {
      "entropy": 6.852114534378051,
      "epoch": 0.4602510460251046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002745,
      "loss": 6.3853,
      "mean_token_accuracy": 0.11713317707180977,
      "num_tokens": 1015715.0,
      "step": 550
    },
    {
      "entropy": 6.969930839538574,
      "epoch": 0.46443514644351463,
      "grad_norm": 1.328125,
      "learning_rate": 0.000277,
      "loss": 6.5013,
      "mean_token_accuracy": 0.11022178530693054,
      "num_tokens": 1025091.0,
      "step": 555
    },
    {
      "entropy": 6.90451455116272,
      "epoch": 0.4686192468619247,
      "grad_norm": 1.125,
      "learning_rate": 0.0002795,
      "loss": 6.4309,
      "mean_token_accuracy": 0.10915803164243698,
      "num_tokens": 1034174.0,
      "step": 560
    },
    {
      "entropy": 7.026072406768799,
      "epoch": 0.47280334728033474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6341,
      "mean_token_accuracy": 0.1005947694182396,
      "num_tokens": 1043777.0,
      "step": 565
    },
    {
      "entropy": 6.970922946929932,
      "epoch": 0.4769874476987448,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002845,
      "loss": 6.472,
      "mean_token_accuracy": 0.11195325776934624,
      "num_tokens": 1052749.0,
      "step": 570
    },
    {
      "entropy": 6.93495135307312,
      "epoch": 0.4811715481171548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000287,
      "loss": 6.4883,
      "mean_token_accuracy": 0.1101924866437912,
      "num_tokens": 1061564.0,
      "step": 575
    },
    {
      "entropy": 6.911145734786987,
      "epoch": 0.48535564853556484,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002895,
      "loss": 6.4042,
      "mean_token_accuracy": 0.1145945630967617,
      "num_tokens": 1070006.0,
      "step": 580
    },
    {
      "entropy": 6.821494293212891,
      "epoch": 0.4895397489539749,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000292,
      "loss": 6.4755,
      "mean_token_accuracy": 0.10867053493857384,
      "num_tokens": 1079799.0,
      "step": 585
    },
    {
      "entropy": 6.9094232559204105,
      "epoch": 0.49372384937238495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002945,
      "loss": 6.3056,
      "mean_token_accuracy": 0.12089367508888245,
      "num_tokens": 1088842.0,
      "step": 590
    },
    {
      "entropy": 6.798662376403809,
      "epoch": 0.497907949790795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000297,
      "loss": 6.3798,
      "mean_token_accuracy": 0.12022144198417664,
      "num_tokens": 1097960.0,
      "step": 595
    },
    {
      "entropy": 6.814728689193726,
      "epoch": 0.502092050209205,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002995,
      "loss": 6.3724,
      "mean_token_accuracy": 0.11288540661334992,
      "num_tokens": 1106797.0,
      "step": 600
    },
    {
      "entropy": 6.8617775440216064,
      "epoch": 0.5062761506276151,
      "grad_norm": 1.140625,
      "learning_rate": 0.000302,
      "loss": 6.5201,
      "mean_token_accuracy": 0.1136062651872635,
      "num_tokens": 1116384.0,
      "step": 605
    },
    {
      "entropy": 6.71177659034729,
      "epoch": 0.5104602510460251,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003045,
      "loss": 6.28,
      "mean_token_accuracy": 0.11779255121946335,
      "num_tokens": 1125206.0,
      "step": 610
    },
    {
      "entropy": 6.948096656799317,
      "epoch": 0.5146443514644351,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000307,
      "loss": 6.5012,
      "mean_token_accuracy": 0.11721058636903763,
      "num_tokens": 1134110.0,
      "step": 615
    },
    {
      "entropy": 6.975963926315307,
      "epoch": 0.5188284518828452,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003095,
      "loss": 6.498,
      "mean_token_accuracy": 0.11025303602218628,
      "num_tokens": 1142833.0,
      "step": 620
    },
    {
      "entropy": 6.840864133834839,
      "epoch": 0.5230125523012552,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000312,
      "loss": 6.411,
      "mean_token_accuracy": 0.11678725853562355,
      "num_tokens": 1152006.0,
      "step": 625
    },
    {
      "entropy": 6.82530255317688,
      "epoch": 0.5271966527196653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003145,
      "loss": 6.3572,
      "mean_token_accuracy": 0.11277682185173035,
      "num_tokens": 1161061.0,
      "step": 630
    },
    {
      "entropy": 6.856273174285889,
      "epoch": 0.5313807531380753,
      "grad_norm": 1.0,
      "learning_rate": 0.000317,
      "loss": 6.3817,
      "mean_token_accuracy": 0.11423582211136818,
      "num_tokens": 1170754.0,
      "step": 635
    },
    {
      "entropy": 6.753575038909912,
      "epoch": 0.5355648535564853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003195,
      "loss": 6.3207,
      "mean_token_accuracy": 0.11631909161806106,
      "num_tokens": 1180854.0,
      "step": 640
    },
    {
      "entropy": 6.778096914291382,
      "epoch": 0.5397489539748954,
      "grad_norm": 1.140625,
      "learning_rate": 0.000322,
      "loss": 6.399,
      "mean_token_accuracy": 0.11601272895932198,
      "num_tokens": 1190567.0,
      "step": 645
    },
    {
      "entropy": 6.753851079940796,
      "epoch": 0.5439330543933054,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.2537,
      "mean_token_accuracy": 0.12245947867631912,
      "num_tokens": 1200247.0,
      "step": 650
    },
    {
      "entropy": 6.746957540512085,
      "epoch": 0.5481171548117155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.3527,
      "mean_token_accuracy": 0.1171742670238018,
      "num_tokens": 1209358.0,
      "step": 655
    },
    {
      "entropy": 6.7102185726165775,
      "epoch": 0.5523012552301255,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.3628,
      "mean_token_accuracy": 0.11740919798612595,
      "num_tokens": 1217746.0,
      "step": 660
    },
    {
      "entropy": 6.741930294036865,
      "epoch": 0.5564853556485355,
      "grad_norm": 1.375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.3404,
      "mean_token_accuracy": 0.12082380875945091,
      "num_tokens": 1226829.0,
      "step": 665
    },
    {
      "entropy": 6.716346597671508,
      "epoch": 0.5606694560669456,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.2708,
      "mean_token_accuracy": 0.1217438243329525,
      "num_tokens": 1236642.0,
      "step": 670
    },
    {
      "entropy": 6.851208734512329,
      "epoch": 0.5648535564853556,
      "grad_norm": 1.265625,
      "learning_rate": 0.000337,
      "loss": 6.3657,
      "mean_token_accuracy": 0.12304765060544014,
      "num_tokens": 1244934.0,
      "step": 675
    },
    {
      "entropy": 6.569645929336548,
      "epoch": 0.5690376569037657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003395,
      "loss": 6.2686,
      "mean_token_accuracy": 0.11759066805243493,
      "num_tokens": 1254786.0,
      "step": 680
    },
    {
      "entropy": 6.745736312866211,
      "epoch": 0.5732217573221757,
      "grad_norm": 1.046875,
      "learning_rate": 0.000342,
      "loss": 6.2219,
      "mean_token_accuracy": 0.12392263934016227,
      "num_tokens": 1263745.0,
      "step": 685
    },
    {
      "entropy": 6.629066610336304,
      "epoch": 0.5774058577405857,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.3768,
      "mean_token_accuracy": 0.12062641084194184,
      "num_tokens": 1273101.0,
      "step": 690
    },
    {
      "entropy": 6.80110445022583,
      "epoch": 0.5815899581589958,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000347,
      "loss": 6.3366,
      "mean_token_accuracy": 0.1182647056877613,
      "num_tokens": 1282547.0,
      "step": 695
    },
    {
      "entropy": 6.721027660369873,
      "epoch": 0.5857740585774058,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003495,
      "loss": 6.361,
      "mean_token_accuracy": 0.11650118827819825,
      "num_tokens": 1291594.0,
      "step": 700
    },
    {
      "entropy": 6.686480140686035,
      "epoch": 0.5899581589958159,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000352,
      "loss": 6.2755,
      "mean_token_accuracy": 0.12708859369158745,
      "num_tokens": 1300282.0,
      "step": 705
    },
    {
      "entropy": 6.703722047805786,
      "epoch": 0.5941422594142259,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003545,
      "loss": 6.3668,
      "mean_token_accuracy": 0.12298200204968453,
      "num_tokens": 1309304.0,
      "step": 710
    },
    {
      "entropy": 6.666914319992065,
      "epoch": 0.5983263598326359,
      "grad_norm": 1.171875,
      "learning_rate": 0.000357,
      "loss": 6.2053,
      "mean_token_accuracy": 0.12288029938936233,
      "num_tokens": 1317638.0,
      "step": 715
    },
    {
      "entropy": 6.816176033020019,
      "epoch": 0.602510460251046,
      "grad_norm": 1.125,
      "learning_rate": 0.0003595,
      "loss": 6.4085,
      "mean_token_accuracy": 0.1202589988708496,
      "num_tokens": 1327242.0,
      "step": 720
    },
    {
      "entropy": 6.689037799835205,
      "epoch": 0.606694560669456,
      "grad_norm": 1.046875,
      "learning_rate": 0.000362,
      "loss": 6.2151,
      "mean_token_accuracy": 0.12098749801516533,
      "num_tokens": 1335880.0,
      "step": 725
    },
    {
      "entropy": 6.657412004470825,
      "epoch": 0.6108786610878661,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003645,
      "loss": 6.298,
      "mean_token_accuracy": 0.12196487411856652,
      "num_tokens": 1344909.0,
      "step": 730
    },
    {
      "entropy": 6.623217487335205,
      "epoch": 0.6150627615062761,
      "grad_norm": 1.109375,
      "learning_rate": 0.000367,
      "loss": 6.1665,
      "mean_token_accuracy": 0.12766727730631827,
      "num_tokens": 1353997.0,
      "step": 735
    },
    {
      "entropy": 6.612822961807251,
      "epoch": 0.6192468619246861,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003695,
      "loss": 6.3168,
      "mean_token_accuracy": 0.12191536054015159,
      "num_tokens": 1363293.0,
      "step": 740
    },
    {
      "entropy": 6.85352053642273,
      "epoch": 0.6234309623430963,
      "grad_norm": 1.171875,
      "learning_rate": 0.000372,
      "loss": 6.335,
      "mean_token_accuracy": 0.11951717808842659,
      "num_tokens": 1372174.0,
      "step": 745
    },
    {
      "entropy": 6.609316825866699,
      "epoch": 0.6276150627615062,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003745,
      "loss": 6.2327,
      "mean_token_accuracy": 0.12466636523604394,
      "num_tokens": 1381067.0,
      "step": 750
    },
    {
      "entropy": 6.585425615310669,
      "epoch": 0.6317991631799164,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377,
      "loss": 6.2179,
      "mean_token_accuracy": 0.12963995933532715,
      "num_tokens": 1390072.0,
      "step": 755
    },
    {
      "entropy": 6.66300778388977,
      "epoch": 0.6359832635983264,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003795,
      "loss": 6.2147,
      "mean_token_accuracy": 0.13023070693016053,
      "num_tokens": 1399391.0,
      "step": 760
    },
    {
      "entropy": 6.558363771438598,
      "epoch": 0.6401673640167364,
      "grad_norm": 1.203125,
      "learning_rate": 0.000382,
      "loss": 6.2236,
      "mean_token_accuracy": 0.12857722416520118,
      "num_tokens": 1408054.0,
      "step": 765
    },
    {
      "entropy": 6.632496404647827,
      "epoch": 0.6443514644351465,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003845,
      "loss": 6.2303,
      "mean_token_accuracy": 0.11898199543356895,
      "num_tokens": 1417406.0,
      "step": 770
    },
    {
      "entropy": 6.595701122283936,
      "epoch": 0.6485355648535565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.2822,
      "mean_token_accuracy": 0.11752806529402733,
      "num_tokens": 1426907.0,
      "step": 775
    },
    {
      "entropy": 6.572923707962036,
      "epoch": 0.6527196652719666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3368,
      "mean_token_accuracy": 0.12002972811460495,
      "num_tokens": 1436085.0,
      "step": 780
    },
    {
      "entropy": 6.790693521499634,
      "epoch": 0.6569037656903766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.233,
      "mean_token_accuracy": 0.1243402674794197,
      "num_tokens": 1444823.0,
      "step": 785
    },
    {
      "entropy": 6.5481942653656,
      "epoch": 0.6610878661087866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.2919,
      "mean_token_accuracy": 0.12223824709653855,
      "num_tokens": 1454621.0,
      "step": 790
    },
    {
      "entropy": 6.545780181884766,
      "epoch": 0.6652719665271967,
      "grad_norm": 1.125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.1532,
      "mean_token_accuracy": 0.1284099817276001,
      "num_tokens": 1463784.0,
      "step": 795
    },
    {
      "entropy": 6.611534309387207,
      "epoch": 0.6694560669456067,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003995,
      "loss": 6.2818,
      "mean_token_accuracy": 0.12604539170861245,
      "num_tokens": 1473257.0,
      "step": 800
    },
    {
      "entropy": 6.62378249168396,
      "epoch": 0.6736401673640168,
      "grad_norm": 1.140625,
      "learning_rate": 0.000402,
      "loss": 6.23,
      "mean_token_accuracy": 0.12763578295707703,
      "num_tokens": 1482195.0,
      "step": 805
    },
    {
      "entropy": 6.484483909606934,
      "epoch": 0.6778242677824268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004045,
      "loss": 6.2622,
      "mean_token_accuracy": 0.12974331825971602,
      "num_tokens": 1491590.0,
      "step": 810
    },
    {
      "entropy": 6.641168451309204,
      "epoch": 0.6820083682008368,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.1307,
      "mean_token_accuracy": 0.13457018807530402,
      "num_tokens": 1500208.0,
      "step": 815
    },
    {
      "entropy": 6.524597692489624,
      "epoch": 0.6861924686192469,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004095,
      "loss": 6.2619,
      "mean_token_accuracy": 0.12744015082716942,
      "num_tokens": 1509412.0,
      "step": 820
    },
    {
      "entropy": 6.619950962066651,
      "epoch": 0.6903765690376569,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000412,
      "loss": 6.2651,
      "mean_token_accuracy": 0.12576703503727912,
      "num_tokens": 1518426.0,
      "step": 825
    },
    {
      "entropy": 6.689215564727784,
      "epoch": 0.694560669456067,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004145,
      "loss": 6.2173,
      "mean_token_accuracy": 0.12354342117905617,
      "num_tokens": 1528605.0,
      "step": 830
    },
    {
      "entropy": 6.413666105270385,
      "epoch": 0.698744769874477,
      "grad_norm": 1.1875,
      "learning_rate": 0.000417,
      "loss": 6.1746,
      "mean_token_accuracy": 0.13070485070347787,
      "num_tokens": 1537449.0,
      "step": 835
    },
    {
      "entropy": 6.594534254074096,
      "epoch": 0.702928870292887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004195,
      "loss": 6.1963,
      "mean_token_accuracy": 0.13175581246614457,
      "num_tokens": 1546743.0,
      "step": 840
    },
    {
      "entropy": 6.484565734863281,
      "epoch": 0.7071129707112971,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000422,
      "loss": 6.1817,
      "mean_token_accuracy": 0.12574852034449577,
      "num_tokens": 1555691.0,
      "step": 845
    },
    {
      "entropy": 6.489085245132446,
      "epoch": 0.7112970711297071,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004245,
      "loss": 6.1334,
      "mean_token_accuracy": 0.1334412842988968,
      "num_tokens": 1564957.0,
      "step": 850
    },
    {
      "entropy": 6.588404178619385,
      "epoch": 0.7154811715481172,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000427,
      "loss": 6.1974,
      "mean_token_accuracy": 0.13589535504579545,
      "num_tokens": 1573901.0,
      "step": 855
    },
    {
      "entropy": 6.4834239959716795,
      "epoch": 0.7196652719665272,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295,
      "loss": 6.1916,
      "mean_token_accuracy": 0.131468053907156,
      "num_tokens": 1582644.0,
      "step": 860
    },
    {
      "entropy": 6.61418342590332,
      "epoch": 0.7238493723849372,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432,
      "loss": 6.3198,
      "mean_token_accuracy": 0.12452186271548271,
      "num_tokens": 1592363.0,
      "step": 865
    },
    {
      "entropy": 6.546977949142456,
      "epoch": 0.7280334728033473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004345,
      "loss": 6.2498,
      "mean_token_accuracy": 0.12673747837543486,
      "num_tokens": 1602012.0,
      "step": 870
    },
    {
      "entropy": 6.463743782043457,
      "epoch": 0.7322175732217573,
      "grad_norm": 1.171875,
      "learning_rate": 0.000437,
      "loss": 6.141,
      "mean_token_accuracy": 0.13073898032307624,
      "num_tokens": 1611625.0,
      "step": 875
    },
    {
      "entropy": 6.395287895202637,
      "epoch": 0.7364016736401674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004395,
      "loss": 6.1141,
      "mean_token_accuracy": 0.12709902375936508,
      "num_tokens": 1620941.0,
      "step": 880
    },
    {
      "entropy": 6.560024118423462,
      "epoch": 0.7405857740585774,
      "grad_norm": 1.265625,
      "learning_rate": 0.000442,
      "loss": 6.2078,
      "mean_token_accuracy": 0.1278709888458252,
      "num_tokens": 1629225.0,
      "step": 885
    },
    {
      "entropy": 6.4868265151977536,
      "epoch": 0.7447698744769874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004445,
      "loss": 6.1722,
      "mean_token_accuracy": 0.12686478942632676,
      "num_tokens": 1639551.0,
      "step": 890
    },
    {
      "entropy": 6.418560361862182,
      "epoch": 0.7489539748953975,
      "grad_norm": 1.078125,
      "learning_rate": 0.000447,
      "loss": 6.1591,
      "mean_token_accuracy": 0.11997205838561058,
      "num_tokens": 1648963.0,
      "step": 895
    },
    {
      "entropy": 6.467530012130737,
      "epoch": 0.7531380753138075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.0847,
      "mean_token_accuracy": 0.1315011478960514,
      "num_tokens": 1657735.0,
      "step": 900
    },
    {
      "entropy": 6.564916038513184,
      "epoch": 0.7573221757322176,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1211,
      "mean_token_accuracy": 0.13231319710612297,
      "num_tokens": 1666349.0,
      "step": 905
    },
    {
      "entropy": 6.3584685802459715,
      "epoch": 0.7615062761506276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.0491,
      "mean_token_accuracy": 0.1330125942826271,
      "num_tokens": 1675920.0,
      "step": 910
    },
    {
      "entropy": 6.345976114273071,
      "epoch": 0.7656903765690377,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.0834,
      "mean_token_accuracy": 0.13028152361512185,
      "num_tokens": 1685626.0,
      "step": 915
    },
    {
      "entropy": 6.5751933574676515,
      "epoch": 0.7698744769874477,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2104,
      "mean_token_accuracy": 0.13214068561792375,
      "num_tokens": 1694972.0,
      "step": 920
    },
    {
      "entropy": 6.388721895217896,
      "epoch": 0.7740585774058577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000462,
      "loss": 6.1102,
      "mean_token_accuracy": 0.14101462960243225,
      "num_tokens": 1703806.0,
      "step": 925
    },
    {
      "entropy": 6.467752361297608,
      "epoch": 0.7782426778242678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004645,
      "loss": 6.1868,
      "mean_token_accuracy": 0.13027679920196533,
      "num_tokens": 1712949.0,
      "step": 930
    },
    {
      "entropy": 6.603727626800537,
      "epoch": 0.7824267782426778,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000467,
      "loss": 6.3193,
      "mean_token_accuracy": 0.12327147945761681,
      "num_tokens": 1722328.0,
      "step": 935
    },
    {
      "entropy": 6.54327802658081,
      "epoch": 0.7866108786610879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004695,
      "loss": 6.2759,
      "mean_token_accuracy": 0.1266297660768032,
      "num_tokens": 1731880.0,
      "step": 940
    },
    {
      "entropy": 6.298931503295899,
      "epoch": 0.7907949790794979,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000472,
      "loss": 6.0979,
      "mean_token_accuracy": 0.1360059678554535,
      "num_tokens": 1740719.0,
      "step": 945
    },
    {
      "entropy": 6.476912546157837,
      "epoch": 0.7949790794979079,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004745,
      "loss": 6.1403,
      "mean_token_accuracy": 0.13098946809768677,
      "num_tokens": 1749753.0,
      "step": 950
    },
    {
      "entropy": 6.520664882659912,
      "epoch": 0.799163179916318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000477,
      "loss": 6.1454,
      "mean_token_accuracy": 0.12673419713974,
      "num_tokens": 1758913.0,
      "step": 955
    },
    {
      "entropy": 6.467524242401123,
      "epoch": 0.803347280334728,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004795,
      "loss": 6.1628,
      "mean_token_accuracy": 0.12942118048667908,
      "num_tokens": 1768536.0,
      "step": 960
    },
    {
      "entropy": 6.351734304428101,
      "epoch": 0.8075313807531381,
      "grad_norm": 1.1875,
      "learning_rate": 0.000482,
      "loss": 6.1028,
      "mean_token_accuracy": 0.129199555516243,
      "num_tokens": 1777291.0,
      "step": 965
    },
    {
      "entropy": 6.46312575340271,
      "epoch": 0.8117154811715481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004845,
      "loss": 6.2004,
      "mean_token_accuracy": 0.13277156054973602,
      "num_tokens": 1787186.0,
      "step": 970
    },
    {
      "entropy": 6.409244060516357,
      "epoch": 0.8158995815899581,
      "grad_norm": 1.171875,
      "learning_rate": 0.000487,
      "loss": 5.9866,
      "mean_token_accuracy": 0.13967311680316924,
      "num_tokens": 1795581.0,
      "step": 975
    },
    {
      "entropy": 6.252563381195069,
      "epoch": 0.8200836820083682,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004895,
      "loss": 5.9885,
      "mean_token_accuracy": 0.141060072183609,
      "num_tokens": 1804627.0,
      "step": 980
    },
    {
      "entropy": 6.4267253398895265,
      "epoch": 0.8242677824267782,
      "grad_norm": 1.125,
      "learning_rate": 0.000492,
      "loss": 6.1131,
      "mean_token_accuracy": 0.13135791718959808,
      "num_tokens": 1813808.0,
      "step": 985
    },
    {
      "entropy": 6.388511610031128,
      "epoch": 0.8284518828451883,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004945,
      "loss": 6.1245,
      "mean_token_accuracy": 0.1352509319782257,
      "num_tokens": 1823422.0,
      "step": 990
    },
    {
      "entropy": 6.442827367782593,
      "epoch": 0.8326359832635983,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000497,
      "loss": 6.1643,
      "mean_token_accuracy": 0.13328976929187775,
      "num_tokens": 1832437.0,
      "step": 995
    },
    {
      "entropy": 6.325851535797119,
      "epoch": 0.8368200836820083,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995,
      "loss": 6.0927,
      "mean_token_accuracy": 0.13116609752178193,
      "num_tokens": 1842063.0,
      "step": 1000
    },
    {
      "epoch": 0.8368200836820083,
      "eval_entropy": 6.231119002788832,
      "eval_loss": 6.111203670501709,
      "eval_mean_token_accuracy": 0.1427788064493236,
      "eval_num_tokens": 1842063.0,
      "eval_runtime": 2.2784,
      "eval_samples_per_second": 1714.779,
      "eval_steps_per_second": 214.622,
      "step": 1000
    },
    {
      "entropy": 6.437012720108032,
      "epoch": 0.8410041841004184,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999998518355666,
      "loss": 6.2293,
      "mean_token_accuracy": 0.13318051919341087,
      "num_tokens": 1851043.0,
      "step": 1005
    },
    {
      "entropy": 6.348101234436035,
      "epoch": 0.8451882845188284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999992499178905,
      "loss": 6.0751,
      "mean_token_accuracy": 0.13775536939501762,
      "num_tokens": 1860268.0,
      "step": 1010
    },
    {
      "entropy": 6.429720544815064,
      "epoch": 0.8493723849372385,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999981849879322,
      "loss": 6.1844,
      "mean_token_accuracy": 0.13298494592309,
      "num_tokens": 1870096.0,
      "step": 1015
    },
    {
      "entropy": 6.338539743423462,
      "epoch": 0.8535564853556485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999966570478833,
      "loss": 6.0722,
      "mean_token_accuracy": 0.13555127531290054,
      "num_tokens": 1879164.0,
      "step": 1020
    },
    {
      "entropy": 6.346892070770264,
      "epoch": 0.8577405857740585,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999946661008878,
      "loss": 6.0087,
      "mean_token_accuracy": 0.1377266153693199,
      "num_tokens": 1888135.0,
      "step": 1025
    },
    {
      "entropy": 6.362658977508545,
      "epoch": 0.8619246861924686,
      "grad_norm": 1.265625,
      "learning_rate": 0.000499992212151043,
      "loss": 6.0797,
      "mean_token_accuracy": 0.1345998913049698,
      "num_tokens": 1898302.0,
      "step": 1030
    },
    {
      "entropy": 6.250446605682373,
      "epoch": 0.8661087866108786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999892952033987,
      "loss": 6.0507,
      "mean_token_accuracy": 0.1350796677172184,
      "num_tokens": 1907753.0,
      "step": 1035
    },
    {
      "entropy": 6.385110569000244,
      "epoch": 0.8702928870292888,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999859152639573,
      "loss": 6.1031,
      "mean_token_accuracy": 0.13561567589640616,
      "num_tokens": 1917543.0,
      "step": 1040
    },
    {
      "entropy": 6.414516115188599,
      "epoch": 0.8744769874476988,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999820723396745,
      "loss": 6.1416,
      "mean_token_accuracy": 0.13240931779146195,
      "num_tokens": 1927335.0,
      "step": 1045
    },
    {
      "entropy": 6.312007570266724,
      "epoch": 0.8786610878661087,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999777664384581,
      "loss": 6.0641,
      "mean_token_accuracy": 0.13389642611145974,
      "num_tokens": 1936381.0,
      "step": 1050
    },
    {
      "entropy": 6.345766687393189,
      "epoch": 0.8828451882845189,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999729975691692,
      "loss": 6.0975,
      "mean_token_accuracy": 0.13339410722255707,
      "num_tokens": 1945559.0,
      "step": 1055
    },
    {
      "entropy": 6.25928635597229,
      "epoch": 0.8870292887029289,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999677657416212,
      "loss": 5.9715,
      "mean_token_accuracy": 0.13333128094673158,
      "num_tokens": 1954709.0,
      "step": 1060
    },
    {
      "entropy": 6.3643087387084964,
      "epoch": 0.891213389121339,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999620709665804,
      "loss": 6.0994,
      "mean_token_accuracy": 0.13211975023150443,
      "num_tokens": 1963664.0,
      "step": 1065
    },
    {
      "entropy": 6.317386150360107,
      "epoch": 0.895397489539749,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999559132557659,
      "loss": 6.1559,
      "mean_token_accuracy": 0.1293097920715809,
      "num_tokens": 1974103.0,
      "step": 1070
    },
    {
      "entropy": 6.498347997665405,
      "epoch": 0.899581589958159,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999492926218489,
      "loss": 6.025,
      "mean_token_accuracy": 0.13773725554347038,
      "num_tokens": 1982924.0,
      "step": 1075
    },
    {
      "entropy": 6.243421602249145,
      "epoch": 0.9037656903765691,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499942209078454,
      "loss": 5.9663,
      "mean_token_accuracy": 0.1428417220711708,
      "num_tokens": 1992033.0,
      "step": 1080
    },
    {
      "entropy": 6.1911012649536135,
      "epoch": 0.9079497907949791,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999346626401578,
      "loss": 5.9586,
      "mean_token_accuracy": 0.14146680235862732,
      "num_tokens": 2001218.0,
      "step": 1085
    },
    {
      "entropy": 6.288367033004761,
      "epoch": 0.9121338912133892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999266533224897,
      "loss": 6.111,
      "mean_token_accuracy": 0.1368482619524002,
      "num_tokens": 2010463.0,
      "step": 1090
    },
    {
      "entropy": 6.326322317123413,
      "epoch": 0.9163179916317992,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999181811419314,
      "loss": 5.9808,
      "mean_token_accuracy": 0.13955054804682732,
      "num_tokens": 2019246.0,
      "step": 1095
    },
    {
      "entropy": 6.388986015319825,
      "epoch": 0.9205020920502092,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999092461159177,
      "loss": 6.0683,
      "mean_token_accuracy": 0.14024771228432656,
      "num_tokens": 2028083.0,
      "step": 1100
    },
    {
      "entropy": 6.223995351791382,
      "epoch": 0.9246861924686193,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998998482628351,
      "loss": 5.8752,
      "mean_token_accuracy": 0.1448050245642662,
      "num_tokens": 2036631.0,
      "step": 1105
    },
    {
      "entropy": 6.252813816070557,
      "epoch": 0.9288702928870293,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004998899876020229,
      "loss": 6.0088,
      "mean_token_accuracy": 0.13815638795495033,
      "num_tokens": 2045910.0,
      "step": 1110
    },
    {
      "entropy": 6.219067764282227,
      "epoch": 0.9330543933054394,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499879664153773,
      "loss": 6.0618,
      "mean_token_accuracy": 0.14408401846885682,
      "num_tokens": 2055959.0,
      "step": 1115
    },
    {
      "entropy": 6.39411563873291,
      "epoch": 0.9372384937238494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998688779393291,
      "loss": 6.1148,
      "mean_token_accuracy": 0.1365758113563061,
      "num_tokens": 2065475.0,
      "step": 1120
    },
    {
      "entropy": 6.1112316131591795,
      "epoch": 0.9414225941422594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998576289808877,
      "loss": 5.885,
      "mean_token_accuracy": 0.14670267850160598,
      "num_tokens": 2074469.0,
      "step": 1125
    },
    {
      "entropy": 6.235973262786866,
      "epoch": 0.9456066945606695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998459173015972,
      "loss": 6.0622,
      "mean_token_accuracy": 0.14163898378610612,
      "num_tokens": 2084604.0,
      "step": 1130
    },
    {
      "entropy": 6.387408065795898,
      "epoch": 0.9497907949790795,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998337429255584,
      "loss": 6.0332,
      "mean_token_accuracy": 0.141568361222744,
      "num_tokens": 2093377.0,
      "step": 1135
    },
    {
      "entropy": 6.254647159576416,
      "epoch": 0.9539748953974896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998211058778243,
      "loss": 6.0613,
      "mean_token_accuracy": 0.1345399834215641,
      "num_tokens": 2102798.0,
      "step": 1140
    },
    {
      "entropy": 6.338140487670898,
      "epoch": 0.9581589958158996,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998080061843999,
      "loss": 6.034,
      "mean_token_accuracy": 0.1373809665441513,
      "num_tokens": 2112516.0,
      "step": 1145
    },
    {
      "entropy": 6.290405893325806,
      "epoch": 0.9623430962343096,
      "grad_norm": 1.125,
      "learning_rate": 0.000499794443872242,
      "loss": 6.1161,
      "mean_token_accuracy": 0.13683309629559517,
      "num_tokens": 2121677.0,
      "step": 1150
    },
    {
      "entropy": 6.194741535186767,
      "epoch": 0.9665271966527197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004997804189692601,
      "loss": 5.9541,
      "mean_token_accuracy": 0.14311629980802537,
      "num_tokens": 2131220.0,
      "step": 1155
    },
    {
      "entropy": 6.297001314163208,
      "epoch": 0.9707112970711297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997659315043149,
      "loss": 6.1041,
      "mean_token_accuracy": 0.134016689658165,
      "num_tokens": 2142104.0,
      "step": 1160
    },
    {
      "entropy": 6.216556453704834,
      "epoch": 0.9748953974895398,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997509815072191,
      "loss": 6.0076,
      "mean_token_accuracy": 0.1369858868420124,
      "num_tokens": 2151970.0,
      "step": 1165
    },
    {
      "entropy": 6.243878126144409,
      "epoch": 0.9790794979079498,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997355690087379,
      "loss": 5.9732,
      "mean_token_accuracy": 0.13861552774906158,
      "num_tokens": 2161371.0,
      "step": 1170
    },
    {
      "entropy": 6.119789552688599,
      "epoch": 0.9832635983263598,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997196940405874,
      "loss": 5.9401,
      "mean_token_accuracy": 0.13875976651906968,
      "num_tokens": 2170650.0,
      "step": 1175
    },
    {
      "entropy": 6.23285436630249,
      "epoch": 0.9874476987447699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997033566354358,
      "loss": 5.9983,
      "mean_token_accuracy": 0.14316384941339494,
      "num_tokens": 2180934.0,
      "step": 1180
    },
    {
      "entropy": 6.242339372634888,
      "epoch": 0.9916317991631799,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996865568269028,
      "loss": 5.9674,
      "mean_token_accuracy": 0.1459835335612297,
      "num_tokens": 2189894.0,
      "step": 1185
    },
    {
      "entropy": 6.221574783325195,
      "epoch": 0.99581589958159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996692946495597,
      "loss": 5.9109,
      "mean_token_accuracy": 0.15525714308023453,
      "num_tokens": 2200208.0,
      "step": 1190
    },
    {
      "entropy": 6.222945690155029,
      "epoch": 1.0,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004996515701389295,
      "loss": 5.9637,
      "mean_token_accuracy": 0.1387474164366722,
      "num_tokens": 2209322.0,
      "step": 1195
    },
    {
      "entropy": 6.12794132232666,
      "epoch": 1.00418410041841,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004996333833314862,
      "loss": 5.7841,
      "mean_token_accuracy": 0.1540491595864296,
      "num_tokens": 2217663.0,
      "step": 1200
    },
    {
      "entropy": 6.149629306793213,
      "epoch": 1.00836820083682,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996147342646554,
      "loss": 5.8162,
      "mean_token_accuracy": 0.14273207560181617,
      "num_tokens": 2226559.0,
      "step": 1205
    },
    {
      "entropy": 6.246478652954101,
      "epoch": 1.0125523012552302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995956229768137,
      "loss": 5.8536,
      "mean_token_accuracy": 0.15705856084823608,
      "num_tokens": 2235468.0,
      "step": 1210
    },
    {
      "entropy": 6.064305543899536,
      "epoch": 1.0167364016736402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995760495072893,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1609800159931183,
      "num_tokens": 2245176.0,
      "step": 1215
    },
    {
      "entropy": 6.180075407028198,
      "epoch": 1.0209205020920502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995560138963611,
      "loss": 5.8354,
      "mean_token_accuracy": 0.15163842141628264,
      "num_tokens": 2255123.0,
      "step": 1220
    },
    {
      "entropy": 6.143266677856445,
      "epoch": 1.0251046025104602,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995355161852593,
      "loss": 5.7615,
      "mean_token_accuracy": 0.14811853021383287,
      "num_tokens": 2263596.0,
      "step": 1225
    },
    {
      "entropy": 6.131743717193603,
      "epoch": 1.0292887029288702,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004995145564161648,
      "loss": 5.7905,
      "mean_token_accuracy": 0.14918514639139174,
      "num_tokens": 2272228.0,
      "step": 1230
    },
    {
      "entropy": 6.156940174102783,
      "epoch": 1.0334728033472804,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994931346322094,
      "loss": 5.8357,
      "mean_token_accuracy": 0.1530950203537941,
      "num_tokens": 2281248.0,
      "step": 1235
    },
    {
      "entropy": 6.178971242904663,
      "epoch": 1.0376569037656904,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994712508774757,
      "loss": 5.813,
      "mean_token_accuracy": 0.14753878712654114,
      "num_tokens": 2290958.0,
      "step": 1240
    },
    {
      "entropy": 6.062367248535156,
      "epoch": 1.0418410041841004,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499448905196997,
      "loss": 5.7476,
      "mean_token_accuracy": 0.1606424257159233,
      "num_tokens": 2299763.0,
      "step": 1245
    },
    {
      "entropy": 6.122745561599731,
      "epoch": 1.0460251046025104,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994260976367573,
      "loss": 5.7576,
      "mean_token_accuracy": 0.1608661964535713,
      "num_tokens": 2309847.0,
      "step": 1250
    },
    {
      "entropy": 6.17245512008667,
      "epoch": 1.0502092050209204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004994028282436905,
      "loss": 5.8114,
      "mean_token_accuracy": 0.14902583807706832,
      "num_tokens": 2317556.0,
      "step": 1255
    },
    {
      "entropy": 6.0724709033966064,
      "epoch": 1.0543933054393306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993790970656814,
      "loss": 5.7246,
      "mean_token_accuracy": 0.15202396363019943,
      "num_tokens": 2327151.0,
      "step": 1260
    },
    {
      "entropy": 6.026160478591919,
      "epoch": 1.0585774058577406,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993549041515652,
      "loss": 5.7251,
      "mean_token_accuracy": 0.15616882890462874,
      "num_tokens": 2336119.0,
      "step": 1265
    },
    {
      "entropy": 6.0959837436676025,
      "epoch": 1.0627615062761506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993302495511268,
      "loss": 5.7422,
      "mean_token_accuracy": 0.15904839485883712,
      "num_tokens": 2345262.0,
      "step": 1270
    },
    {
      "entropy": 6.198559761047363,
      "epoch": 1.0669456066945606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993051333151014,
      "loss": 5.8823,
      "mean_token_accuracy": 0.14653945490717887,
      "num_tokens": 2355210.0,
      "step": 1275
    },
    {
      "entropy": 6.012920284271241,
      "epoch": 1.0711297071129706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992795554951742,
      "loss": 5.6244,
      "mean_token_accuracy": 0.15309498757123946,
      "num_tokens": 2364599.0,
      "step": 1280
    },
    {
      "entropy": 5.942229127883911,
      "epoch": 1.0753138075313808,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004992535161439803,
      "loss": 5.8123,
      "mean_token_accuracy": 0.14621696844697,
      "num_tokens": 2374143.0,
      "step": 1285
    },
    {
      "entropy": 6.222422313690186,
      "epoch": 1.0794979079497908,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992270153151045,
      "loss": 5.8022,
      "mean_token_accuracy": 0.1560079947113991,
      "num_tokens": 2383155.0,
      "step": 1290
    },
    {
      "entropy": 6.197408390045166,
      "epoch": 1.0836820083682008,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004992000530630812,
      "loss": 5.8354,
      "mean_token_accuracy": 0.1492285281419754,
      "num_tokens": 2392542.0,
      "step": 1295
    },
    {
      "entropy": 6.031621360778809,
      "epoch": 1.0878661087866108,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991726294433943,
      "loss": 5.8021,
      "mean_token_accuracy": 0.14778647273778917,
      "num_tokens": 2401851.0,
      "step": 1300
    },
    {
      "entropy": 6.112837648391723,
      "epoch": 1.0920502092050208,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991447445124774,
      "loss": 5.7584,
      "mean_token_accuracy": 0.15441246777772905,
      "num_tokens": 2410726.0,
      "step": 1305
    },
    {
      "entropy": 6.012867641448975,
      "epoch": 1.096234309623431,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499116398327713,
      "loss": 5.8041,
      "mean_token_accuracy": 0.14813676327466965,
      "num_tokens": 2420702.0,
      "step": 1310
    },
    {
      "entropy": 6.022567844390869,
      "epoch": 1.100418410041841,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990875909474331,
      "loss": 5.7489,
      "mean_token_accuracy": 0.15360454171895982,
      "num_tokens": 2431335.0,
      "step": 1315
    },
    {
      "entropy": 6.095766305923462,
      "epoch": 1.104602510460251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990583224309187,
      "loss": 5.7173,
      "mean_token_accuracy": 0.15707379579544067,
      "num_tokens": 2440722.0,
      "step": 1320
    },
    {
      "entropy": 6.091929769515991,
      "epoch": 1.108786610878661,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990285928383997,
      "loss": 5.7968,
      "mean_token_accuracy": 0.14727665334939957,
      "num_tokens": 2451230.0,
      "step": 1325
    },
    {
      "entropy": 6.078146171569824,
      "epoch": 1.112970711297071,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989984022310545,
      "loss": 5.8262,
      "mean_token_accuracy": 0.14682078659534453,
      "num_tokens": 2461442.0,
      "step": 1330
    },
    {
      "entropy": 6.035668563842774,
      "epoch": 1.1171548117154813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989677506710109,
      "loss": 5.6587,
      "mean_token_accuracy": 0.15804804116487503,
      "num_tokens": 2471381.0,
      "step": 1335
    },
    {
      "entropy": 6.023805093765259,
      "epoch": 1.1213389121338913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989366382213449,
      "loss": 5.7038,
      "mean_token_accuracy": 0.1592436045408249,
      "num_tokens": 2479937.0,
      "step": 1340
    },
    {
      "entropy": 6.033054256439209,
      "epoch": 1.1255230125523012,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989050649460806,
      "loss": 5.6689,
      "mean_token_accuracy": 0.16124916672706605,
      "num_tokens": 2488974.0,
      "step": 1345
    },
    {
      "entropy": 6.041873598098755,
      "epoch": 1.1297071129707112,
      "grad_norm": 1.0,
      "learning_rate": 0.000498873030910191,
      "loss": 5.8188,
      "mean_token_accuracy": 0.1526327461004257,
      "num_tokens": 2498800.0,
      "step": 1350
    },
    {
      "entropy": 5.99496603012085,
      "epoch": 1.1338912133891212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004988405361795969,
      "loss": 5.7115,
      "mean_token_accuracy": 0.15621206909418106,
      "num_tokens": 2506580.0,
      "step": 1355
    },
    {
      "entropy": 6.047221612930298,
      "epoch": 1.1380753138075315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988075808211672,
      "loss": 5.756,
      "mean_token_accuracy": 0.1572654902935028,
      "num_tokens": 2516342.0,
      "step": 1360
    },
    {
      "entropy": 5.983310604095459,
      "epoch": 1.1422594142259415,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987741649027189,
      "loss": 5.8001,
      "mean_token_accuracy": 0.1560211196541786,
      "num_tokens": 2526054.0,
      "step": 1365
    },
    {
      "entropy": 6.1061523914337155,
      "epoch": 1.1464435146443515,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987402884930163,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1552029624581337,
      "num_tokens": 2535556.0,
      "step": 1370
    },
    {
      "entropy": 6.062144947052002,
      "epoch": 1.1506276150627615,
      "grad_norm": 1.125,
      "learning_rate": 0.0004987059516617717,
      "loss": 5.7273,
      "mean_token_accuracy": 0.15809612423181535,
      "num_tokens": 2544930.0,
      "step": 1375
    },
    {
      "entropy": 5.931625747680664,
      "epoch": 1.1548117154811715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986711544796449,
      "loss": 5.6373,
      "mean_token_accuracy": 0.1627542957663536,
      "num_tokens": 2554362.0,
      "step": 1380
    },
    {
      "entropy": 6.023519992828369,
      "epoch": 1.1589958158995817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498635897018243,
      "loss": 5.7605,
      "mean_token_accuracy": 0.1554619088768959,
      "num_tokens": 2563499.0,
      "step": 1385
    },
    {
      "entropy": 6.004778623580933,
      "epoch": 1.1631799163179917,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986001793501198,
      "loss": 5.7618,
      "mean_token_accuracy": 0.15196048617362976,
      "num_tokens": 2573346.0,
      "step": 1390
    },
    {
      "entropy": 6.028854703903198,
      "epoch": 1.1673640167364017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004985640015487768,
      "loss": 5.6471,
      "mean_token_accuracy": 0.16079026758670806,
      "num_tokens": 2582541.0,
      "step": 1395
    },
    {
      "entropy": 6.001078462600708,
      "epoch": 1.1715481171548117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985273636886621,
      "loss": 5.7365,
      "mean_token_accuracy": 0.15553900450468064,
      "num_tokens": 2592015.0,
      "step": 1400
    },
    {
      "entropy": 6.000535202026367,
      "epoch": 1.1757322175732217,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004984902658451704,
      "loss": 5.7402,
      "mean_token_accuracy": 0.1547416031360626,
      "num_tokens": 2601998.0,
      "step": 1405
    },
    {
      "entropy": 6.00449857711792,
      "epoch": 1.1799163179916319,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984527080946432,
      "loss": 5.6785,
      "mean_token_accuracy": 0.16148286759853364,
      "num_tokens": 2611435.0,
      "step": 1410
    },
    {
      "entropy": 6.018415307998657,
      "epoch": 1.1841004184100419,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984146905143683,
      "loss": 5.7759,
      "mean_token_accuracy": 0.15475187003612517,
      "num_tokens": 2621639.0,
      "step": 1415
    },
    {
      "entropy": 6.0044273853302,
      "epoch": 1.1882845188284519,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004983762131825797,
      "loss": 5.7389,
      "mean_token_accuracy": 0.1585055634379387,
      "num_tokens": 2631390.0,
      "step": 1420
    },
    {
      "entropy": 5.9931258201599125,
      "epoch": 1.1924686192468619,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983372761784576,
      "loss": 5.7802,
      "mean_token_accuracy": 0.15404947847127914,
      "num_tokens": 2640769.0,
      "step": 1425
    },
    {
      "entropy": 6.096167278289795,
      "epoch": 1.196652719665272,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982978795821281,
      "loss": 5.779,
      "mean_token_accuracy": 0.15546282827854158,
      "num_tokens": 2651137.0,
      "step": 1430
    },
    {
      "entropy": 5.995174264907837,
      "epoch": 1.200836820083682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982580234746632,
      "loss": 5.7072,
      "mean_token_accuracy": 0.15209258496761321,
      "num_tokens": 2660530.0,
      "step": 1435
    },
    {
      "entropy": 5.99799337387085,
      "epoch": 1.205020920502092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049821770793808,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1549163356423378,
      "num_tokens": 2669541.0,
      "step": 1440
    },
    {
      "entropy": 6.023744249343872,
      "epoch": 1.209205020920502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004981769330553418,
      "loss": 5.7502,
      "mean_token_accuracy": 0.15705211609601974,
      "num_tokens": 2678542.0,
      "step": 1445
    },
    {
      "entropy": 6.029311323165894,
      "epoch": 1.213389121338912,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004981356989103566,
      "loss": 5.8299,
      "mean_token_accuracy": 0.15689381957054138,
      "num_tokens": 2687720.0,
      "step": 1450
    },
    {
      "entropy": 6.031181144714355,
      "epoch": 1.217573221757322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980940055879774,
      "loss": 5.6934,
      "mean_token_accuracy": 0.15815790593624116,
      "num_tokens": 2697641.0,
      "step": 1455
    },
    {
      "entropy": 5.9612993717193605,
      "epoch": 1.2217573221757323,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980518531740027,
      "loss": 5.7072,
      "mean_token_accuracy": 0.1583000048995018,
      "num_tokens": 2706600.0,
      "step": 1460
    },
    {
      "entropy": 6.063562631607056,
      "epoch": 1.2259414225941423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980092417551752,
      "loss": 5.8514,
      "mean_token_accuracy": 0.14898592680692674,
      "num_tokens": 2716633.0,
      "step": 1465
    },
    {
      "entropy": 6.047478342056275,
      "epoch": 1.2301255230125523,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004979661714191825,
      "loss": 5.628,
      "mean_token_accuracy": 0.16600804030895233,
      "num_tokens": 2724829.0,
      "step": 1470
    },
    {
      "entropy": 5.925992202758789,
      "epoch": 1.2343096234309623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979226422546563,
      "loss": 5.7374,
      "mean_token_accuracy": 0.1585901975631714,
      "num_tokens": 2734276.0,
      "step": 1475
    },
    {
      "entropy": 6.035736894607544,
      "epoch": 1.2384937238493725,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004978786543511726,
      "loss": 5.8273,
      "mean_token_accuracy": 0.157758666574955,
      "num_tokens": 2744413.0,
      "step": 1480
    },
    {
      "entropy": 6.014146232604981,
      "epoch": 1.2426778242677825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978342077992515,
      "loss": 5.7065,
      "mean_token_accuracy": 0.15562080591917038,
      "num_tokens": 2753242.0,
      "step": 1485
    },
    {
      "entropy": 5.832571172714234,
      "epoch": 1.2468619246861925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004977893026903568,
      "loss": 5.6248,
      "mean_token_accuracy": 0.16905127316713334,
      "num_tokens": 2761811.0,
      "step": 1490
    },
    {
      "entropy": 6.082716751098633,
      "epoch": 1.2510460251046025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977439391168963,
      "loss": 5.7302,
      "mean_token_accuracy": 0.15641461163759232,
      "num_tokens": 2770560.0,
      "step": 1495
    },
    {
      "entropy": 6.041548204421997,
      "epoch": 1.2552301255230125,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976981171722205,
      "loss": 5.7379,
      "mean_token_accuracy": 0.15544732064008712,
      "num_tokens": 2780131.0,
      "step": 1500
    },
    {
      "epoch": 1.2552301255230125,
      "eval_entropy": 5.6769954109972,
      "eval_loss": 5.8483991622924805,
      "eval_mean_token_accuracy": 0.15941667107290042,
      "eval_num_tokens": 2780131.0,
      "eval_runtime": 2.0884,
      "eval_samples_per_second": 1870.827,
      "eval_steps_per_second": 234.153,
      "step": 1500
    },
    {
      "entropy": 6.092925977706909,
      "epoch": 1.2594142259414225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004976518369506237,
      "loss": 5.8435,
      "mean_token_accuracy": 0.14601506665349007,
      "num_tokens": 2789319.0,
      "step": 1505
    },
    {
      "entropy": 5.92581090927124,
      "epoch": 1.2635983263598327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976050985473433,
      "loss": 5.7037,
      "mean_token_accuracy": 0.1522383913397789,
      "num_tokens": 2799253.0,
      "step": 1510
    },
    {
      "entropy": 6.032972526550293,
      "epoch": 1.2677824267782427,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004975579020585593,
      "loss": 5.8124,
      "mean_token_accuracy": 0.15160835534334183,
      "num_tokens": 2808968.0,
      "step": 1515
    },
    {
      "entropy": 6.074542236328125,
      "epoch": 1.2719665271966527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004975102475813947,
      "loss": 5.6999,
      "mean_token_accuracy": 0.1575356364250183,
      "num_tokens": 2817282.0,
      "step": 1520
    },
    {
      "entropy": 6.00976939201355,
      "epoch": 1.2761506276150627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974621352139144,
      "loss": 5.7538,
      "mean_token_accuracy": 0.15369689762592315,
      "num_tokens": 2826395.0,
      "step": 1525
    },
    {
      "entropy": 5.989823675155639,
      "epoch": 1.280334728033473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974135650551262,
      "loss": 5.8344,
      "mean_token_accuracy": 0.15231696516275406,
      "num_tokens": 2836195.0,
      "step": 1530
    },
    {
      "entropy": 6.010791635513305,
      "epoch": 1.284518828451883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004973645372049796,
      "loss": 5.6702,
      "mean_token_accuracy": 0.16355077773332596,
      "num_tokens": 2846020.0,
      "step": 1535
    },
    {
      "entropy": 5.891250371932983,
      "epoch": 1.288702928870293,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973150517643659,
      "loss": 5.6186,
      "mean_token_accuracy": 0.16435685753822327,
      "num_tokens": 2854895.0,
      "step": 1540
    },
    {
      "entropy": 5.977266883850097,
      "epoch": 1.292887029288703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972651088351185,
      "loss": 5.6632,
      "mean_token_accuracy": 0.16027058362960817,
      "num_tokens": 2864491.0,
      "step": 1545
    },
    {
      "entropy": 5.980721950531006,
      "epoch": 1.297071129707113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972147085200116,
      "loss": 5.6816,
      "mean_token_accuracy": 0.15872606933116912,
      "num_tokens": 2873962.0,
      "step": 1550
    },
    {
      "entropy": 6.017466497421265,
      "epoch": 1.301255230125523,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004971638509227611,
      "loss": 5.7151,
      "mean_token_accuracy": 0.16112489104270936,
      "num_tokens": 2883530.0,
      "step": 1555
    },
    {
      "entropy": 5.925843286514282,
      "epoch": 1.3054393305439331,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004971125361480236,
      "loss": 5.6906,
      "mean_token_accuracy": 0.15688816756010054,
      "num_tokens": 2892867.0,
      "step": 1560
    },
    {
      "entropy": 5.94919638633728,
      "epoch": 1.3096234309623431,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970607643013969,
      "loss": 5.6645,
      "mean_token_accuracy": 0.15854619294404984,
      "num_tokens": 2902320.0,
      "step": 1565
    },
    {
      "entropy": 5.903445148468018,
      "epoch": 1.3138075313807531,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497008535489419,
      "loss": 5.5946,
      "mean_token_accuracy": 0.16695220172405242,
      "num_tokens": 2911390.0,
      "step": 1570
    },
    {
      "entropy": 5.889070701599121,
      "epoch": 1.3179916317991631,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004969558498195684,
      "loss": 5.6532,
      "mean_token_accuracy": 0.16906433552503586,
      "num_tokens": 2921295.0,
      "step": 1575
    },
    {
      "entropy": 5.958902788162232,
      "epoch": 1.3221757322175733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969027074002638,
      "loss": 5.6603,
      "mean_token_accuracy": 0.15776332914829255,
      "num_tokens": 2931328.0,
      "step": 1580
    },
    {
      "entropy": 5.8630468368530275,
      "epoch": 1.3263598326359833,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004968491083408636,
      "loss": 5.6133,
      "mean_token_accuracy": 0.1640845373272896,
      "num_tokens": 2940193.0,
      "step": 1585
    },
    {
      "entropy": 5.861121034622192,
      "epoch": 1.3305439330543933,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004967950527516663,
      "loss": 5.6416,
      "mean_token_accuracy": 0.16769741475582123,
      "num_tokens": 2949759.0,
      "step": 1590
    },
    {
      "entropy": 5.887998342514038,
      "epoch": 1.3347280334728033,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967405407439092,
      "loss": 5.577,
      "mean_token_accuracy": 0.15625791251659393,
      "num_tokens": 2959185.0,
      "step": 1595
    },
    {
      "entropy": 5.965415334701538,
      "epoch": 1.3389121338912133,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004966855724297698,
      "loss": 5.7328,
      "mean_token_accuracy": 0.15423317551612853,
      "num_tokens": 2968945.0,
      "step": 1600
    },
    {
      "entropy": 5.944570827484131,
      "epoch": 1.3430962343096233,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004966301479223637,
      "loss": 5.6716,
      "mean_token_accuracy": 0.159438119828701,
      "num_tokens": 2978395.0,
      "step": 1605
    },
    {
      "entropy": 5.858027172088623,
      "epoch": 1.3472803347280335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004965742673357456,
      "loss": 5.5804,
      "mean_token_accuracy": 0.1700463503599167,
      "num_tokens": 2987030.0,
      "step": 1610
    },
    {
      "entropy": 5.9461335182189945,
      "epoch": 1.3514644351464435,
      "grad_norm": 1.125,
      "learning_rate": 0.0004965179307849088,
      "loss": 5.6511,
      "mean_token_accuracy": 0.16362962871789932,
      "num_tokens": 2995736.0,
      "step": 1615
    },
    {
      "entropy": 5.962363815307617,
      "epoch": 1.3556485355648535,
      "grad_norm": 1.09375,
      "learning_rate": 0.000496461138385785,
      "loss": 5.7402,
      "mean_token_accuracy": 0.15967344343662263,
      "num_tokens": 3005105.0,
      "step": 1620
    },
    {
      "entropy": 5.873168802261352,
      "epoch": 1.3598326359832635,
      "grad_norm": 1.125,
      "learning_rate": 0.0004964038902552438,
      "loss": 5.6571,
      "mean_token_accuracy": 0.16216485053300858,
      "num_tokens": 3013857.0,
      "step": 1625
    },
    {
      "entropy": 5.950975131988526,
      "epoch": 1.3640167364016738,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963461865110926,
      "loss": 5.689,
      "mean_token_accuracy": 0.1605420783162117,
      "num_tokens": 3023972.0,
      "step": 1630
    },
    {
      "entropy": 5.836272764205932,
      "epoch": 1.3682008368200838,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004962880272720764,
      "loss": 5.6589,
      "mean_token_accuracy": 0.16736868917942047,
      "num_tokens": 3032705.0,
      "step": 1635
    },
    {
      "entropy": 5.930795669555664,
      "epoch": 1.3723849372384938,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004962294126578775,
      "loss": 5.5945,
      "mean_token_accuracy": 0.16802308410406114,
      "num_tokens": 3040900.0,
      "step": 1640
    },
    {
      "entropy": 5.917369699478149,
      "epoch": 1.3765690376569037,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961703427891154,
      "loss": 5.6341,
      "mean_token_accuracy": 0.16583087146282197,
      "num_tokens": 3050477.0,
      "step": 1645
    },
    {
      "entropy": 5.9194018840789795,
      "epoch": 1.3807531380753137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961108177873466,
      "loss": 5.7088,
      "mean_token_accuracy": 0.1571326047182083,
      "num_tokens": 3059743.0,
      "step": 1650
    },
    {
      "entropy": 5.843382024765015,
      "epoch": 1.3849372384937237,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004960508377750638,
      "loss": 5.5053,
      "mean_token_accuracy": 0.16926025450229645,
      "num_tokens": 3067801.0,
      "step": 1655
    },
    {
      "entropy": 5.825117397308349,
      "epoch": 1.389121338912134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959904028756961,
      "loss": 5.6868,
      "mean_token_accuracy": 0.156205852329731,
      "num_tokens": 3077404.0,
      "step": 1660
    },
    {
      "entropy": 5.940052938461304,
      "epoch": 1.393305439330544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959295132136091,
      "loss": 5.7066,
      "mean_token_accuracy": 0.15989601761102676,
      "num_tokens": 3087130.0,
      "step": 1665
    },
    {
      "entropy": 5.978323650360108,
      "epoch": 1.397489539748954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004958681689141036,
      "loss": 5.7219,
      "mean_token_accuracy": 0.1560539923608303,
      "num_tokens": 3096038.0,
      "step": 1670
    },
    {
      "entropy": 5.858187627792359,
      "epoch": 1.401673640167364,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004958063701034167,
      "loss": 5.5735,
      "mean_token_accuracy": 0.16297442018985747,
      "num_tokens": 3105668.0,
      "step": 1675
    },
    {
      "entropy": 5.857545661926269,
      "epoch": 1.4058577405857742,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00049574411690872,
      "loss": 5.5643,
      "mean_token_accuracy": 0.15730429887771608,
      "num_tokens": 3114020.0,
      "step": 1680
    },
    {
      "entropy": 5.732494449615478,
      "epoch": 1.4100418410041842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956814094581208,
      "loss": 5.602,
      "mean_token_accuracy": 0.16669758707284926,
      "num_tokens": 3124096.0,
      "step": 1685
    },
    {
      "entropy": 5.935622072219848,
      "epoch": 1.4142259414225942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495618247880661,
      "loss": 5.6677,
      "mean_token_accuracy": 0.16170076429843902,
      "num_tokens": 3132806.0,
      "step": 1690
    },
    {
      "entropy": 5.8426909923553465,
      "epoch": 1.4184100418410042,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955546323063168,
      "loss": 5.5876,
      "mean_token_accuracy": 0.17422180920839309,
      "num_tokens": 3142001.0,
      "step": 1695
    },
    {
      "entropy": 5.8333580017089846,
      "epoch": 1.4225941422594142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954905628659989,
      "loss": 5.5666,
      "mean_token_accuracy": 0.16144170612096786,
      "num_tokens": 3151658.0,
      "step": 1700
    },
    {
      "entropy": 5.846378564834595,
      "epoch": 1.4267782426778242,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004954260396915521,
      "loss": 5.591,
      "mean_token_accuracy": 0.1657014384865761,
      "num_tokens": 3159573.0,
      "step": 1705
    },
    {
      "entropy": 5.669739723205566,
      "epoch": 1.4309623430962344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004953610629157544,
      "loss": 5.4647,
      "mean_token_accuracy": 0.17399210333824158,
      "num_tokens": 3168312.0,
      "step": 1710
    },
    {
      "entropy": 5.927958869934082,
      "epoch": 1.4351464435146444,
      "grad_norm": 1.25,
      "learning_rate": 0.0004952956326723177,
      "loss": 5.5834,
      "mean_token_accuracy": 0.17099785804748535,
      "num_tokens": 3176811.0,
      "step": 1715
    },
    {
      "entropy": 5.8302654266357425,
      "epoch": 1.4393305439330544,
      "grad_norm": 1.109375,
      "learning_rate": 0.000495229749095887,
      "loss": 5.6826,
      "mean_token_accuracy": 0.15766623616218567,
      "num_tokens": 3186599.0,
      "step": 1720
    },
    {
      "entropy": 5.99775767326355,
      "epoch": 1.4435146443514644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00049516341232204,
      "loss": 5.7751,
      "mean_token_accuracy": 0.15921125411987305,
      "num_tokens": 3195905.0,
      "step": 1725
    },
    {
      "entropy": 5.9371254444122314,
      "epoch": 1.4476987447698746,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000495096622487287,
      "loss": 5.7281,
      "mean_token_accuracy": 0.15743090063333512,
      "num_tokens": 3205132.0,
      "step": 1730
    },
    {
      "entropy": 5.935694646835327,
      "epoch": 1.4518828451882846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000495029379729071,
      "loss": 5.7187,
      "mean_token_accuracy": 0.15283734053373338,
      "num_tokens": 3214362.0,
      "step": 1735
    },
    {
      "entropy": 5.887416410446167,
      "epoch": 1.4560669456066946,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949616841857666,
      "loss": 5.7136,
      "mean_token_accuracy": 0.15853825807571412,
      "num_tokens": 3224248.0,
      "step": 1740
    },
    {
      "entropy": 5.913235187530518,
      "epoch": 1.4602510460251046,
      "grad_norm": 1.125,
      "learning_rate": 0.0004948935359966806,
      "loss": 5.5927,
      "mean_token_accuracy": 0.16305283904075624,
      "num_tokens": 3233416.0,
      "step": 1745
    },
    {
      "entropy": 5.725965213775635,
      "epoch": 1.4644351464435146,
      "grad_norm": 1.25,
      "learning_rate": 0.0004948249353020506,
      "loss": 5.5373,
      "mean_token_accuracy": 0.1675902485847473,
      "num_tokens": 3242354.0,
      "step": 1750
    },
    {
      "entropy": 5.873963499069214,
      "epoch": 1.4686192468619246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004947558822430461,
      "loss": 5.6023,
      "mean_token_accuracy": 0.1656225636601448,
      "num_tokens": 3251026.0,
      "step": 1755
    },
    {
      "entropy": 5.83093957901001,
      "epoch": 1.4728033472803348,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004946863769617672,
      "loss": 5.5668,
      "mean_token_accuracy": 0.1644663542509079,
      "num_tokens": 3259429.0,
      "step": 1760
    },
    {
      "entropy": 5.83068356513977,
      "epoch": 1.4769874476987448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004946164196012445,
      "loss": 5.6217,
      "mean_token_accuracy": 0.16164354085922242,
      "num_tokens": 3269305.0,
      "step": 1765
    },
    {
      "entropy": 5.877197027206421,
      "epoch": 1.4811715481171548,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004945460103054389,
      "loss": 5.6843,
      "mean_token_accuracy": 0.15862189531326293,
      "num_tokens": 3278846.0,
      "step": 1770
    },
    {
      "entropy": 5.899416828155518,
      "epoch": 1.4853556485355648,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004944751492192416,
      "loss": 5.611,
      "mean_token_accuracy": 0.16517737805843352,
      "num_tokens": 3287200.0,
      "step": 1775
    },
    {
      "entropy": 5.9020170211792,
      "epoch": 1.489539748953975,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004944038364884734,
      "loss": 5.6415,
      "mean_token_accuracy": 0.15735085159540177,
      "num_tokens": 3296651.0,
      "step": 1780
    },
    {
      "entropy": 5.842372989654541,
      "epoch": 1.493723849372385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943320722598842,
      "loss": 5.524,
      "mean_token_accuracy": 0.17010669261217118,
      "num_tokens": 3304887.0,
      "step": 1785
    },
    {
      "entropy": 5.83018159866333,
      "epoch": 1.497907949790795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004942598566811532,
      "loss": 5.543,
      "mean_token_accuracy": 0.1716176301240921,
      "num_tokens": 3313925.0,
      "step": 1790
    },
    {
      "entropy": 5.847691488265991,
      "epoch": 1.502092050209205,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004941871899008887,
      "loss": 5.5298,
      "mean_token_accuracy": 0.16718751192092896,
      "num_tokens": 3322943.0,
      "step": 1795
    },
    {
      "entropy": 5.730998134613037,
      "epoch": 1.506276150627615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494114072068627,
      "loss": 5.5467,
      "mean_token_accuracy": 0.16466595083475113,
      "num_tokens": 3332742.0,
      "step": 1800
    },
    {
      "entropy": 5.813080930709839,
      "epoch": 1.510460251046025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940405033348332,
      "loss": 5.6352,
      "mean_token_accuracy": 0.16255978345870972,
      "num_tokens": 3342228.0,
      "step": 1805
    },
    {
      "entropy": 5.814670515060425,
      "epoch": 1.514644351464435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939664838508996,
      "loss": 5.6105,
      "mean_token_accuracy": 0.1650616854429245,
      "num_tokens": 3351854.0,
      "step": 1810
    },
    {
      "entropy": 5.786225271224976,
      "epoch": 1.5188284518828452,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004938920137691465,
      "loss": 5.586,
      "mean_token_accuracy": 0.16562919616699218,
      "num_tokens": 3360809.0,
      "step": 1815
    },
    {
      "entropy": 5.833091688156128,
      "epoch": 1.5230125523012552,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938170932428214,
      "loss": 5.6233,
      "mean_token_accuracy": 0.1586744472384453,
      "num_tokens": 3371033.0,
      "step": 1820
    },
    {
      "entropy": 5.90294919013977,
      "epoch": 1.5271966527196654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004937417224260984,
      "loss": 5.4947,
      "mean_token_accuracy": 0.17257912904024125,
      "num_tokens": 3379943.0,
      "step": 1825
    },
    {
      "entropy": 5.752410268783569,
      "epoch": 1.5313807531380754,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004936659014740791,
      "loss": 5.6229,
      "mean_token_accuracy": 0.16704918593168258,
      "num_tokens": 3389316.0,
      "step": 1830
    },
    {
      "entropy": 5.85987696647644,
      "epoch": 1.5355648535564854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004935896305427904,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16151361167430878,
      "num_tokens": 3398914.0,
      "step": 1835
    },
    {
      "entropy": 5.975553369522094,
      "epoch": 1.5397489539748954,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004935129097891857,
      "loss": 5.6227,
      "mean_token_accuracy": 0.16726224571466447,
      "num_tokens": 3407967.0,
      "step": 1840
    },
    {
      "entropy": 5.9012768268585205,
      "epoch": 1.5439330543933054,
      "grad_norm": 1.25,
      "learning_rate": 0.000493435739371144,
      "loss": 5.6502,
      "mean_token_accuracy": 0.1688822716474533,
      "num_tokens": 3416530.0,
      "step": 1845
    },
    {
      "entropy": 5.824185705184936,
      "epoch": 1.5481171548117154,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004933581194474695,
      "loss": 5.5298,
      "mean_token_accuracy": 0.1656462609767914,
      "num_tokens": 3425762.0,
      "step": 1850
    },
    {
      "entropy": 5.774158477783203,
      "epoch": 1.5523012552301254,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004932800501778918,
      "loss": 5.6079,
      "mean_token_accuracy": 0.1647947371006012,
      "num_tokens": 3434799.0,
      "step": 1855
    },
    {
      "entropy": 5.902548933029175,
      "epoch": 1.5564853556485354,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004932015317230646,
      "loss": 5.6221,
      "mean_token_accuracy": 0.1627484917640686,
      "num_tokens": 3442986.0,
      "step": 1860
    },
    {
      "entropy": 5.7775719165802,
      "epoch": 1.5606694560669456,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004931225642445665,
      "loss": 5.6517,
      "mean_token_accuracy": 0.16315302401781082,
      "num_tokens": 3452914.0,
      "step": 1865
    },
    {
      "entropy": 5.883452749252319,
      "epoch": 1.5648535564853556,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004930431479048998,
      "loss": 5.5036,
      "mean_token_accuracy": 0.16831693202257156,
      "num_tokens": 3462093.0,
      "step": 1870
    },
    {
      "entropy": 5.7762549877166744,
      "epoch": 1.5690376569037658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004929632828674904,
      "loss": 5.6812,
      "mean_token_accuracy": 0.15995465219020844,
      "num_tokens": 3471259.0,
      "step": 1875
    },
    {
      "entropy": 5.880064201354981,
      "epoch": 1.5732217573221758,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004928829692966879,
      "loss": 5.6262,
      "mean_token_accuracy": 0.17060019671916962,
      "num_tokens": 3480071.0,
      "step": 1880
    },
    {
      "entropy": 5.924511051177978,
      "epoch": 1.5774058577405858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928022073577646,
      "loss": 5.6393,
      "mean_token_accuracy": 0.17406027764081955,
      "num_tokens": 3489075.0,
      "step": 1885
    },
    {
      "entropy": 5.866768455505371,
      "epoch": 1.5815899581589958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004927209972169157,
      "loss": 5.475,
      "mean_token_accuracy": 0.17666401118040084,
      "num_tokens": 3498495.0,
      "step": 1890
    },
    {
      "entropy": 5.721832513809204,
      "epoch": 1.5857740585774058,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004926393390412586,
      "loss": 5.6119,
      "mean_token_accuracy": 0.16383255571126937,
      "num_tokens": 3508590.0,
      "step": 1895
    },
    {
      "entropy": 5.934581184387207,
      "epoch": 1.5899581589958158,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004925572329988325,
      "loss": 5.6307,
      "mean_token_accuracy": 0.16441386789083481,
      "num_tokens": 3518397.0,
      "step": 1900
    },
    {
      "entropy": 5.789706468582153,
      "epoch": 1.5941422594142258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004924746792585987,
      "loss": 5.608,
      "mean_token_accuracy": 0.1638333410024643,
      "num_tokens": 3526620.0,
      "step": 1905
    },
    {
      "entropy": 5.928402853012085,
      "epoch": 1.5983263598326358,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004923916779904394,
      "loss": 5.6147,
      "mean_token_accuracy": 0.16687250435352324,
      "num_tokens": 3535534.0,
      "step": 1910
    },
    {
      "entropy": 5.735380840301514,
      "epoch": 1.602510460251046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004923082293651578,
      "loss": 5.5745,
      "mean_token_accuracy": 0.16308026015758514,
      "num_tokens": 3544771.0,
      "step": 1915
    },
    {
      "entropy": 5.762562894821167,
      "epoch": 1.606694560669456,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000492224333554478,
      "loss": 5.5449,
      "mean_token_accuracy": 0.1586107388138771,
      "num_tokens": 3553970.0,
      "step": 1920
    },
    {
      "entropy": 5.760586595535278,
      "epoch": 1.6108786610878663,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004921399907310437,
      "loss": 5.4547,
      "mean_token_accuracy": 0.16664935201406478,
      "num_tokens": 3563116.0,
      "step": 1925
    },
    {
      "entropy": 5.777907133102417,
      "epoch": 1.6150627615062763,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920552010684191,
      "loss": 5.6021,
      "mean_token_accuracy": 0.16663048416376114,
      "num_tokens": 3571605.0,
      "step": 1930
    },
    {
      "entropy": 5.784053897857666,
      "epoch": 1.6192468619246863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004919699647410878,
      "loss": 5.6637,
      "mean_token_accuracy": 0.15977739542722702,
      "num_tokens": 3581837.0,
      "step": 1935
    },
    {
      "entropy": 5.804090404510498,
      "epoch": 1.6234309623430963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004918842819244523,
      "loss": 5.6153,
      "mean_token_accuracy": 0.1600283220410347,
      "num_tokens": 3590853.0,
      "step": 1940
    },
    {
      "entropy": 5.812429189682007,
      "epoch": 1.6276150627615062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917981527948342,
      "loss": 5.5107,
      "mean_token_accuracy": 0.16755159795284272,
      "num_tokens": 3599708.0,
      "step": 1945
    },
    {
      "entropy": 5.736073541641235,
      "epoch": 1.6317991631799162,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917115775294731,
      "loss": 5.538,
      "mean_token_accuracy": 0.16955381333827974,
      "num_tokens": 3608634.0,
      "step": 1950
    },
    {
      "entropy": 5.845906925201416,
      "epoch": 1.6359832635983262,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004916245563065273,
      "loss": 5.6043,
      "mean_token_accuracy": 0.16488172560930253,
      "num_tokens": 3618776.0,
      "step": 1955
    },
    {
      "entropy": 5.776925754547119,
      "epoch": 1.6401673640167362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915370893050722,
      "loss": 5.6009,
      "mean_token_accuracy": 0.1616799533367157,
      "num_tokens": 3628595.0,
      "step": 1960
    },
    {
      "entropy": 5.918238115310669,
      "epoch": 1.6443514644351465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004914491767051012,
      "loss": 5.7267,
      "mean_token_accuracy": 0.15665314570069314,
      "num_tokens": 3638133.0,
      "step": 1965
    },
    {
      "entropy": 5.785878467559814,
      "epoch": 1.6485355648535565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004913608186875241,
      "loss": 5.5189,
      "mean_token_accuracy": 0.17225270271301268,
      "num_tokens": 3646679.0,
      "step": 1970
    },
    {
      "entropy": 5.835364151000976,
      "epoch": 1.6527196652719667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004912720154341674,
      "loss": 5.6365,
      "mean_token_accuracy": 0.16081463247537614,
      "num_tokens": 3656131.0,
      "step": 1975
    },
    {
      "entropy": 5.834744548797607,
      "epoch": 1.6569037656903767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004911827671277739,
      "loss": 5.5349,
      "mean_token_accuracy": 0.16812530755996705,
      "num_tokens": 3665187.0,
      "step": 1980
    },
    {
      "entropy": 5.789411354064941,
      "epoch": 1.6610878661087867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004910930739520025,
      "loss": 5.6057,
      "mean_token_accuracy": 0.16710528880357742,
      "num_tokens": 3674649.0,
      "step": 1985
    },
    {
      "entropy": 5.87918701171875,
      "epoch": 1.6652719665271967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004910029360914271,
      "loss": 5.6179,
      "mean_token_accuracy": 0.16003223210573198,
      "num_tokens": 3684680.0,
      "step": 1990
    },
    {
      "entropy": 5.727990341186524,
      "epoch": 1.6694560669456067,
      "grad_norm": 1.140625,
      "learning_rate": 0.000490912353731537,
      "loss": 5.562,
      "mean_token_accuracy": 0.16933501809835433,
      "num_tokens": 3693560.0,
      "step": 1995
    },
    {
      "entropy": 5.739227724075318,
      "epoch": 1.6736401673640167,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004908213270587363,
      "loss": 5.554,
      "mean_token_accuracy": 0.16775781214237212,
      "num_tokens": 3702632.0,
      "step": 2000
    },
    {
      "epoch": 1.6736401673640167,
      "eval_entropy": 5.588579570832672,
      "eval_loss": 5.659614562988281,
      "eval_mean_token_accuracy": 0.16905502562081642,
      "eval_num_tokens": 3702632.0,
      "eval_runtime": 2.2974,
      "eval_samples_per_second": 1700.65,
      "eval_steps_per_second": 212.853,
      "step": 2000
    },
    {
      "entropy": 5.733205652236938,
      "epoch": 1.6778242677824267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004907298562603431,
      "loss": 5.4419,
      "mean_token_accuracy": 0.179792420566082,
      "num_tokens": 3711022.0,
      "step": 2005
    },
    {
      "entropy": 5.765319538116455,
      "epoch": 1.6820083682008367,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004906379415245896,
      "loss": 5.5635,
      "mean_token_accuracy": 0.17425089180469513,
      "num_tokens": 3719778.0,
      "step": 2010
    },
    {
      "entropy": 5.72466812133789,
      "epoch": 1.6861924686192469,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004905455830406215,
      "loss": 5.541,
      "mean_token_accuracy": 0.16978248208761215,
      "num_tokens": 3728773.0,
      "step": 2015
    },
    {
      "entropy": 5.749242210388184,
      "epoch": 1.6903765690376569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004904527809984978,
      "loss": 5.5758,
      "mean_token_accuracy": 0.17408233433961867,
      "num_tokens": 3737781.0,
      "step": 2020
    },
    {
      "entropy": 5.907482528686524,
      "epoch": 1.694560669456067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004903595355891902,
      "loss": 5.6325,
      "mean_token_accuracy": 0.15966468602418898,
      "num_tokens": 3748010.0,
      "step": 2025
    },
    {
      "entropy": 5.757623910903931,
      "epoch": 1.698744769874477,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004902658470045826,
      "loss": 5.6104,
      "mean_token_accuracy": 0.16668835431337356,
      "num_tokens": 3757894.0,
      "step": 2030
    },
    {
      "entropy": 5.913280773162842,
      "epoch": 1.702928870292887,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901717154374713,
      "loss": 5.7001,
      "mean_token_accuracy": 0.1627259835600853,
      "num_tokens": 3768760.0,
      "step": 2035
    },
    {
      "entropy": 5.737261056900024,
      "epoch": 1.707112970711297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004900771410815638,
      "loss": 5.5313,
      "mean_token_accuracy": 0.17250364124774933,
      "num_tokens": 3778273.0,
      "step": 2040
    },
    {
      "entropy": 5.72805814743042,
      "epoch": 1.711297071129707,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004899821241314788,
      "loss": 5.5411,
      "mean_token_accuracy": 0.1667292982339859,
      "num_tokens": 3787012.0,
      "step": 2045
    },
    {
      "entropy": 5.802943515777588,
      "epoch": 1.715481171548117,
      "grad_norm": 1.1875,
      "learning_rate": 0.000489886664782746,
      "loss": 5.495,
      "mean_token_accuracy": 0.17157293409109114,
      "num_tokens": 3795144.0,
      "step": 2050
    },
    {
      "entropy": 5.84842586517334,
      "epoch": 1.719665271966527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004897907632318055,
      "loss": 5.7066,
      "mean_token_accuracy": 0.16221378445625306,
      "num_tokens": 3804534.0,
      "step": 2055
    },
    {
      "entropy": 5.764593982696534,
      "epoch": 1.723849372384937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896944196760071,
      "loss": 5.4946,
      "mean_token_accuracy": 0.17056726664304733,
      "num_tokens": 3813757.0,
      "step": 2060
    },
    {
      "entropy": 5.6439008712768555,
      "epoch": 1.7280334728033473,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895976343136106,
      "loss": 5.5233,
      "mean_token_accuracy": 0.1781034454703331,
      "num_tokens": 3823157.0,
      "step": 2065
    },
    {
      "entropy": 5.68931622505188,
      "epoch": 1.7322175732217573,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895004073437844,
      "loss": 5.4059,
      "mean_token_accuracy": 0.1745637461543083,
      "num_tokens": 3831661.0,
      "step": 2070
    },
    {
      "entropy": 5.6525579452514645,
      "epoch": 1.7364016736401675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004894027389666061,
      "loss": 5.4101,
      "mean_token_accuracy": 0.17606336176395415,
      "num_tokens": 3840591.0,
      "step": 2075
    },
    {
      "entropy": 5.792623472213745,
      "epoch": 1.7405857740585775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893046293830616,
      "loss": 5.5278,
      "mean_token_accuracy": 0.16255860030651093,
      "num_tokens": 3850094.0,
      "step": 2080
    },
    {
      "entropy": 5.742485284805298,
      "epoch": 1.7447698744769875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892060787950447,
      "loss": 5.6075,
      "mean_token_accuracy": 0.16815222799777985,
      "num_tokens": 3858973.0,
      "step": 2085
    },
    {
      "entropy": 5.63767294883728,
      "epoch": 1.7489539748953975,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004891070874053564,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1753972977399826,
      "num_tokens": 3868315.0,
      "step": 2090
    },
    {
      "entropy": 5.736858320236206,
      "epoch": 1.7531380753138075,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004890076554177051,
      "loss": 5.5362,
      "mean_token_accuracy": 0.16723127663135529,
      "num_tokens": 3877605.0,
      "step": 2095
    },
    {
      "entropy": 5.74579815864563,
      "epoch": 1.7573221757322175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889077830367061,
      "loss": 5.5509,
      "mean_token_accuracy": 0.171517913043499,
      "num_tokens": 3886998.0,
      "step": 2100
    },
    {
      "entropy": 5.767471218109131,
      "epoch": 1.7615062761506275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004888074704678806,
      "loss": 5.5082,
      "mean_token_accuracy": 0.17135567963123322,
      "num_tokens": 3896018.0,
      "step": 2105
    },
    {
      "entropy": 5.735943555831909,
      "epoch": 1.7656903765690377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004887067179176558,
      "loss": 5.4913,
      "mean_token_accuracy": 0.17377638518810273,
      "num_tokens": 3904957.0,
      "step": 2110
    },
    {
      "entropy": 5.741572475433349,
      "epoch": 1.7698744769874477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886055255933639,
      "loss": 5.6059,
      "mean_token_accuracy": 0.16303271129727365,
      "num_tokens": 3913427.0,
      "step": 2115
    },
    {
      "entropy": 5.74202823638916,
      "epoch": 1.7740585774058577,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004885038937032429,
      "loss": 5.5313,
      "mean_token_accuracy": 0.16900549381971358,
      "num_tokens": 3923134.0,
      "step": 2120
    },
    {
      "entropy": 5.753329849243164,
      "epoch": 1.778242677824268,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004884018224564345,
      "loss": 5.4325,
      "mean_token_accuracy": 0.17378120124340057,
      "num_tokens": 3931867.0,
      "step": 2125
    },
    {
      "entropy": 5.744208335876465,
      "epoch": 1.782426778242678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004882993120629852,
      "loss": 5.5386,
      "mean_token_accuracy": 0.1757911577820778,
      "num_tokens": 3941057.0,
      "step": 2130
    },
    {
      "entropy": 5.839227104187012,
      "epoch": 1.786610878661088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004881963627338449,
      "loss": 5.6526,
      "mean_token_accuracy": 0.1612655907869339,
      "num_tokens": 3950511.0,
      "step": 2135
    },
    {
      "entropy": 5.7714259147644045,
      "epoch": 1.790794979079498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048809297468086654,
      "loss": 5.5708,
      "mean_token_accuracy": 0.17301054149866105,
      "num_tokens": 3959769.0,
      "step": 2140
    },
    {
      "entropy": 5.754560470581055,
      "epoch": 1.794979079497908,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004879891481168063,
      "loss": 5.4885,
      "mean_token_accuracy": 0.17386019825935364,
      "num_tokens": 3968462.0,
      "step": 2145
    },
    {
      "entropy": 5.694684267044067,
      "epoch": 1.799163179916318,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048788488325532235,
      "loss": 5.525,
      "mean_token_accuracy": 0.17304963022470474,
      "num_tokens": 3977464.0,
      "step": 2150
    },
    {
      "entropy": 5.735943078994751,
      "epoch": 1.803347280334728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004877801803109753,
      "loss": 5.449,
      "mean_token_accuracy": 0.176297365128994,
      "num_tokens": 3986303.0,
      "step": 2155
    },
    {
      "entropy": 5.791642427444458,
      "epoch": 1.8075313807531381,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048767503949922683,
      "loss": 5.4987,
      "mean_token_accuracy": 0.173785038292408,
      "num_tokens": 3994970.0,
      "step": 2160
    },
    {
      "entropy": 5.815527820587159,
      "epoch": 1.8117154811715481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004875694610364398,
      "loss": 5.5822,
      "mean_token_accuracy": 0.1655818536877632,
      "num_tokens": 4004679.0,
      "step": 2165
    },
    {
      "entropy": 5.808756685256958,
      "epoch": 1.8158995815899581,
      "grad_norm": 1.125,
      "learning_rate": 0.00048746344513987775,
      "loss": 5.5967,
      "mean_token_accuracy": 0.1684105336666107,
      "num_tokens": 4013516.0,
      "step": 2170
    },
    {
      "entropy": 5.781185150146484,
      "epoch": 1.8200836820083683,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004873569920277043,
      "loss": 5.466,
      "mean_token_accuracy": 0.17886558920145035,
      "num_tokens": 4022335.0,
      "step": 2175
    },
    {
      "entropy": 5.631121635437012,
      "epoch": 1.8242677824267783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048725010191898296,
      "loss": 5.4482,
      "mean_token_accuracy": 0.17270184606313704,
      "num_tokens": 4031295.0,
      "step": 2180
    },
    {
      "entropy": 5.809648323059082,
      "epoch": 1.8284518828451883,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048714277503367627,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1627314418554306,
      "num_tokens": 4040378.0,
      "step": 2185
    },
    {
      "entropy": 5.838449716567993,
      "epoch": 1.8326359832635983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048703501159264576,
      "loss": 5.5254,
      "mean_token_accuracy": 0.16325923651456833,
      "num_tokens": 4050217.0,
      "step": 2190
    },
    {
      "entropy": 5.614929103851319,
      "epoch": 1.8368200836820083,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869268118176512,
      "loss": 5.3981,
      "mean_token_accuracy": 0.16747482419013976,
      "num_tokens": 4059088.0,
      "step": 2195
    },
    {
      "entropy": 5.792963457107544,
      "epoch": 1.8410041841004183,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048681817593135046,
      "loss": 5.5137,
      "mean_token_accuracy": 0.1806218609213829,
      "num_tokens": 4067609.0,
      "step": 2200
    },
    {
      "entropy": 5.725982236862182,
      "epoch": 1.8451882845188283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048670910415729855,
      "loss": 5.4925,
      "mean_token_accuracy": 0.17104198634624482,
      "num_tokens": 4076763.0,
      "step": 2205
    },
    {
      "entropy": 5.796593093872071,
      "epoch": 1.8493723849372385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048659959671994786,
      "loss": 5.6414,
      "mean_token_accuracy": 0.16381513625383376,
      "num_tokens": 4086552.0,
      "step": 2210
    },
    {
      "entropy": 5.615056848526001,
      "epoch": 1.8535564853556485,
      "grad_norm": 1.0,
      "learning_rate": 0.00048648965384464693,
      "loss": 5.4132,
      "mean_token_accuracy": 0.17252946645021439,
      "num_tokens": 4096200.0,
      "step": 2215
    },
    {
      "entropy": 5.649881219863891,
      "epoch": 1.8577405857740585,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048637927575764057,
      "loss": 5.4644,
      "mean_token_accuracy": 0.17159941494464875,
      "num_tokens": 4105080.0,
      "step": 2220
    },
    {
      "entropy": 5.693881797790527,
      "epoch": 1.8619246861924688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048626846268606913,
      "loss": 5.3967,
      "mean_token_accuracy": 0.1740979105234146,
      "num_tokens": 4113972.0,
      "step": 2225
    },
    {
      "entropy": 5.787312078475952,
      "epoch": 1.8661087866108788,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048615721485796803,
      "loss": 5.6322,
      "mean_token_accuracy": 0.16151763051748275,
      "num_tokens": 4123615.0,
      "step": 2230
    },
    {
      "entropy": 5.778605556488037,
      "epoch": 1.8702928870292888,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004860455325022676,
      "loss": 5.4188,
      "mean_token_accuracy": 0.17042351365089417,
      "num_tokens": 4132128.0,
      "step": 2235
    },
    {
      "entropy": 5.806691884994507,
      "epoch": 1.8744769874476988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000485933415848792,
      "loss": 5.6195,
      "mean_token_accuracy": 0.16596710234880446,
      "num_tokens": 4141575.0,
      "step": 2240
    },
    {
      "entropy": 5.74137282371521,
      "epoch": 1.8786610878661087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004858208651282593,
      "loss": 5.4838,
      "mean_token_accuracy": 0.16864698082208635,
      "num_tokens": 4150372.0,
      "step": 2245
    },
    {
      "entropy": 5.717060470581055,
      "epoch": 1.8828451882845187,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000485707880572281,
      "loss": 5.5708,
      "mean_token_accuracy": 0.17850400805473327,
      "num_tokens": 4159007.0,
      "step": 2250
    },
    {
      "entropy": 5.713363838195801,
      "epoch": 1.8870292887029287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004855944624133608,
      "loss": 5.606,
      "mean_token_accuracy": 0.16776737868785857,
      "num_tokens": 4168646.0,
      "step": 2255
    },
    {
      "entropy": 5.706373119354248,
      "epoch": 1.891213389121339,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048548061088489523,
      "loss": 5.5191,
      "mean_token_accuracy": 0.16962461769580842,
      "num_tokens": 4177979.0,
      "step": 2260
    },
    {
      "entropy": 5.747662687301636,
      "epoch": 1.895397489539749,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048536632622117246,
      "loss": 5.4532,
      "mean_token_accuracy": 0.1754816398024559,
      "num_tokens": 4186781.0,
      "step": 2265
    },
    {
      "entropy": 5.674567317962646,
      "epoch": 1.899581589958159,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048525160865737186,
      "loss": 5.5135,
      "mean_token_accuracy": 0.17038754671812056,
      "num_tokens": 4195661.0,
      "step": 2270
    },
    {
      "entropy": 5.711926651000977,
      "epoch": 1.9037656903765692,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048513645842956384,
      "loss": 5.4171,
      "mean_token_accuracy": 0.17328426092863083,
      "num_tokens": 4203902.0,
      "step": 2275
    },
    {
      "entropy": 5.65511646270752,
      "epoch": 1.9079497907949792,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048502087577470895,
      "loss": 5.4801,
      "mean_token_accuracy": 0.1693297103047371,
      "num_tokens": 4213663.0,
      "step": 2280
    },
    {
      "entropy": 5.7536944389343265,
      "epoch": 1.9121338912133892,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048490486093065774,
      "loss": 5.5155,
      "mean_token_accuracy": 0.17362722307443618,
      "num_tokens": 4222152.0,
      "step": 2285
    },
    {
      "entropy": 5.7442272186279295,
      "epoch": 1.9163179916317992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048478841413615017,
      "loss": 5.5351,
      "mean_token_accuracy": 0.17005095928907393,
      "num_tokens": 4232347.0,
      "step": 2290
    },
    {
      "entropy": 5.7341259002685545,
      "epoch": 1.9205020920502092,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048467153563081487,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17492114007472992,
      "num_tokens": 4240607.0,
      "step": 2295
    },
    {
      "entropy": 5.781266546249389,
      "epoch": 1.9246861924686192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845542256551692,
      "loss": 5.59,
      "mean_token_accuracy": 0.16862932443618775,
      "num_tokens": 4251131.0,
      "step": 2300
    },
    {
      "entropy": 5.73166675567627,
      "epoch": 1.9288702928870292,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004844364844506181,
      "loss": 5.5218,
      "mean_token_accuracy": 0.17353187799453734,
      "num_tokens": 4259955.0,
      "step": 2305
    },
    {
      "entropy": 5.801835727691651,
      "epoch": 1.9330543933054394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004843183122594541,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1740906909108162,
      "num_tokens": 4268686.0,
      "step": 2310
    },
    {
      "entropy": 5.691273832321167,
      "epoch": 1.9372384937238494,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004841997093248566,
      "loss": 5.4908,
      "mean_token_accuracy": 0.17442866414785385,
      "num_tokens": 4277377.0,
      "step": 2315
    },
    {
      "entropy": 5.642513227462769,
      "epoch": 1.9414225941422594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004840806758908913,
      "loss": 5.4783,
      "mean_token_accuracy": 0.1721930593252182,
      "num_tokens": 4287037.0,
      "step": 2320
    },
    {
      "entropy": 5.82609486579895,
      "epoch": 1.9456066945606696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048396121220250983,
      "loss": 5.6235,
      "mean_token_accuracy": 0.15890175402164458,
      "num_tokens": 4296923.0,
      "step": 2325
    },
    {
      "entropy": 5.645568037033081,
      "epoch": 1.9497907949790796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838413185055494,
      "loss": 5.3844,
      "mean_token_accuracy": 0.1826627552509308,
      "num_tokens": 4305873.0,
      "step": 2330
    },
    {
      "entropy": 5.69848518371582,
      "epoch": 1.9539748953974896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004837209950467318,
      "loss": 5.5444,
      "mean_token_accuracy": 0.16894280910491943,
      "num_tokens": 4314912.0,
      "step": 2335
    },
    {
      "entropy": 5.754768705368042,
      "epoch": 1.9581589958158996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004836002420736635,
      "loss": 5.5253,
      "mean_token_accuracy": 0.16813971996307372,
      "num_tokens": 4324606.0,
      "step": 2340
    },
    {
      "entropy": 5.638680171966553,
      "epoch": 1.9623430962343096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004834790598348347,
      "loss": 5.403,
      "mean_token_accuracy": 0.18076600432395934,
      "num_tokens": 4333076.0,
      "step": 2345
    },
    {
      "entropy": 5.783349657058716,
      "epoch": 1.9665271966527196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004833574485796188,
      "loss": 5.5358,
      "mean_token_accuracy": 0.17332175523042678,
      "num_tokens": 4341893.0,
      "step": 2350
    },
    {
      "entropy": 5.636692905426026,
      "epoch": 1.9707112970711296,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048323540855827244,
      "loss": 5.5143,
      "mean_token_accuracy": 0.1677964821457863,
      "num_tokens": 4352303.0,
      "step": 2355
    },
    {
      "entropy": 5.820708417892456,
      "epoch": 1.9748953974895398,
      "grad_norm": 1.125,
      "learning_rate": 0.0004831129400219341,
      "loss": 5.4715,
      "mean_token_accuracy": 0.16791264712810516,
      "num_tokens": 4361166.0,
      "step": 2360
    },
    {
      "entropy": 5.729510307312012,
      "epoch": 1.9790794979079498,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048299004322262457,
      "loss": 5.5136,
      "mean_token_accuracy": 0.17077897489070892,
      "num_tokens": 4369999.0,
      "step": 2365
    },
    {
      "entropy": 5.690209627151489,
      "epoch": 1.9832635983263598,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048286671841324565,
      "loss": 5.4774,
      "mean_token_accuracy": 0.17686031758785248,
      "num_tokens": 4379373.0,
      "step": 2370
    },
    {
      "entropy": 5.721845436096191,
      "epoch": 1.98744769874477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004827429658475799,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1777857258915901,
      "num_tokens": 4389035.0,
      "step": 2375
    },
    {
      "entropy": 5.727389669418335,
      "epoch": 1.99163179916318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048261878578029026,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1699129104614258,
      "num_tokens": 4399366.0,
      "step": 2380
    },
    {
      "entropy": 5.701954936981201,
      "epoch": 1.99581589958159,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004824941784669193,
      "loss": 5.4662,
      "mean_token_accuracy": 0.17321588844060898,
      "num_tokens": 4408183.0,
      "step": 2385
    },
    {
      "entropy": 5.705373239517212,
      "epoch": 2.0,
      "grad_norm": 1.125,
      "learning_rate": 0.0004823691441638889,
      "loss": 5.5908,
      "mean_token_accuracy": 0.15743454173207283,
      "num_tokens": 4418644.0,
      "step": 2390
    },
    {
      "entropy": 5.676972484588623,
      "epoch": 2.00418410041841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004822436831284995,
      "loss": 5.2176,
      "mean_token_accuracy": 0.18418388664722443,
      "num_tokens": 4427828.0,
      "step": 2395
    },
    {
      "entropy": 5.63955602645874,
      "epoch": 2.00836820083682,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004821177956189297,
      "loss": 5.1468,
      "mean_token_accuracy": 0.18794921636581421,
      "num_tokens": 4436194.0,
      "step": 2400
    },
    {
      "entropy": 5.572412300109863,
      "epoch": 2.01255230125523,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004819914818942359,
      "loss": 5.2065,
      "mean_token_accuracy": 0.19006218761205673,
      "num_tokens": 4444468.0,
      "step": 2405
    },
    {
      "entropy": 5.723620510101318,
      "epoch": 2.01673640167364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048186474221435127,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1960455819964409,
      "num_tokens": 4452760.0,
      "step": 2410
    },
    {
      "entropy": 5.634727859497071,
      "epoch": 2.02092050209205,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048173757684008583,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19306624829769134,
      "num_tokens": 4462154.0,
      "step": 2415
    },
    {
      "entropy": 5.666573143005371,
      "epoch": 2.0251046025104604,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004816099860331255,
      "loss": 5.2761,
      "mean_token_accuracy": 0.18292328268289565,
      "num_tokens": 4471987.0,
      "step": 2420
    },
    {
      "entropy": 5.602439880371094,
      "epoch": 2.0292887029288704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048148197005603145,
      "loss": 5.1894,
      "mean_token_accuracy": 0.18424307256937028,
      "num_tokens": 4481645.0,
      "step": 2425
    },
    {
      "entropy": 5.563078451156616,
      "epoch": 2.0334728033472804,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048135352917224005,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19173673838376998,
      "num_tokens": 4491166.0,
      "step": 2430
    },
    {
      "entropy": 5.69947361946106,
      "epoch": 2.0376569037656904,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048122466364606214,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18710510134696962,
      "num_tokens": 4500159.0,
      "step": 2435
    },
    {
      "entropy": 5.537801504135132,
      "epoch": 2.0418410041841004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048109537374268216,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1943747103214264,
      "num_tokens": 4509737.0,
      "step": 2440
    },
    {
      "entropy": 5.414461278915406,
      "epoch": 2.0460251046025104,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048096565972815793,
      "loss": 5.068,
      "mean_token_accuracy": 0.19497113823890685,
      "num_tokens": 4518379.0,
      "step": 2445
    },
    {
      "entropy": 5.585501289367675,
      "epoch": 2.0502092050209204,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004808355218694199,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19233749955892562,
      "num_tokens": 4527885.0,
      "step": 2450
    },
    {
      "entropy": 5.642935276031494,
      "epoch": 2.0543933054393304,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048070496043427116,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18954924792051314,
      "num_tokens": 4537894.0,
      "step": 2455
    },
    {
      "entropy": 5.6098230361938475,
      "epoch": 2.0585774058577404,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048057397569138596,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19356006979942322,
      "num_tokens": 4546462.0,
      "step": 2460
    },
    {
      "entropy": 5.581704616546631,
      "epoch": 2.062761506276151,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004804425679103099,
      "loss": 5.1662,
      "mean_token_accuracy": 0.18637972325086594,
      "num_tokens": 4555970.0,
      "step": 2465
    },
    {
      "entropy": 5.614779233932495,
      "epoch": 2.066945606694561,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004803107373614589,
      "loss": 5.2263,
      "mean_token_accuracy": 0.18251074105501175,
      "num_tokens": 4565799.0,
      "step": 2470
    },
    {
      "entropy": 5.591456508636474,
      "epoch": 2.071129707112971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004801784843161194,
      "loss": 5.2588,
      "mean_token_accuracy": 0.18001774698495865,
      "num_tokens": 4575571.0,
      "step": 2475
    },
    {
      "entropy": 5.648578453063965,
      "epoch": 2.075313807531381,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004800458090464465,
      "loss": 5.1558,
      "mean_token_accuracy": 0.19157422780990602,
      "num_tokens": 4584263.0,
      "step": 2480
    },
    {
      "entropy": 5.549553346633911,
      "epoch": 2.079497907949791,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047991271182546476,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19961231499910354,
      "num_tokens": 4592979.0,
      "step": 2485
    },
    {
      "entropy": 5.545702266693115,
      "epoch": 2.083682008368201,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047977919292706696,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1967400923371315,
      "num_tokens": 4602042.0,
      "step": 2490
    },
    {
      "entropy": 5.575488901138305,
      "epoch": 2.087866108786611,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004796452526260134,
      "loss": 5.2003,
      "mean_token_accuracy": 0.19144488722085953,
      "num_tokens": 4611197.0,
      "step": 2495
    },
    {
      "entropy": 5.467491197586059,
      "epoch": 2.092050209205021,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047951089119793164,
      "loss": 5.1512,
      "mean_token_accuracy": 0.18873886466026307,
      "num_tokens": 4621612.0,
      "step": 2500
    },
    {
      "epoch": 2.092050209205021,
      "eval_entropy": 5.31002169103954,
      "eval_loss": 5.553900241851807,
      "eval_mean_token_accuracy": 0.1790137020813907,
      "eval_num_tokens": 4621612.0,
      "eval_runtime": 2.0824,
      "eval_samples_per_second": 1876.19,
      "eval_steps_per_second": 234.824,
      "step": 2500
    },
    {
      "entropy": 5.466035270690918,
      "epoch": 2.096234309623431,
      "grad_norm": 1.15625,
      "learning_rate": 0.000479376108919316,
      "loss": 5.0416,
      "mean_token_accuracy": 0.2022264838218689,
      "num_tokens": 4630198.0,
      "step": 2505
    },
    {
      "entropy": 5.5752606868743895,
      "epoch": 2.100418410041841,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004792409060675267,
      "loss": 5.1309,
      "mean_token_accuracy": 0.18947514593601228,
      "num_tokens": 4638729.0,
      "step": 2510
    },
    {
      "entropy": 5.543802404403687,
      "epoch": 2.104602510460251,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004791052829207895,
      "loss": 5.281,
      "mean_token_accuracy": 0.18752007484436034,
      "num_tokens": 4648759.0,
      "step": 2515
    },
    {
      "entropy": 5.539657258987427,
      "epoch": 2.1087866108786613,
      "grad_norm": 1.125,
      "learning_rate": 0.000478969239758195,
      "loss": 5.2249,
      "mean_token_accuracy": 0.18912642896175386,
      "num_tokens": 4657744.0,
      "step": 2520
    },
    {
      "entropy": 5.583879613876343,
      "epoch": 2.1129707112970713,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047883277685969815,
      "loss": 5.1756,
      "mean_token_accuracy": 0.19520145207643508,
      "num_tokens": 4666154.0,
      "step": 2525
    },
    {
      "entropy": 5.605193853378296,
      "epoch": 2.1171548117154813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047869589450611773,
      "loss": 5.1974,
      "mean_token_accuracy": 0.18547933399677277,
      "num_tokens": 4675677.0,
      "step": 2530
    },
    {
      "entropy": 5.5078535079956055,
      "epoch": 2.1213389121338913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004785585929791355,
      "loss": 5.2027,
      "mean_token_accuracy": 0.18887757360935212,
      "num_tokens": 4684733.0,
      "step": 2535
    },
    {
      "entropy": 5.568530702590943,
      "epoch": 2.1255230125523012,
      "grad_norm": 1.21875,
      "learning_rate": 0.000478420872561296,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19232923835515975,
      "num_tokens": 4694155.0,
      "step": 2540
    },
    {
      "entropy": 5.501318836212159,
      "epoch": 2.1297071129707112,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047828273353600577,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1967197060585022,
      "num_tokens": 4703286.0,
      "step": 2545
    },
    {
      "entropy": 5.541385889053345,
      "epoch": 2.1338912133891212,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047814417618753256,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18664800822734834,
      "num_tokens": 4711666.0,
      "step": 2550
    },
    {
      "entropy": 5.5342179298400875,
      "epoch": 2.1380753138075312,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047800520080100526,
      "loss": 5.1789,
      "mean_token_accuracy": 0.18600201457738877,
      "num_tokens": 4721644.0,
      "step": 2555
    },
    {
      "entropy": 5.53387713432312,
      "epoch": 2.1422594142259412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047786580766241265,
      "loss": 5.1499,
      "mean_token_accuracy": 0.19156892448663712,
      "num_tokens": 4730885.0,
      "step": 2560
    },
    {
      "entropy": 5.480189800262451,
      "epoch": 2.1464435146443517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004777259970586036,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1930587887763977,
      "num_tokens": 4740871.0,
      "step": 2565
    },
    {
      "entropy": 5.532391023635864,
      "epoch": 2.1506276150627617,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004775857692772859,
      "loss": 5.117,
      "mean_token_accuracy": 0.1901528760790825,
      "num_tokens": 4750603.0,
      "step": 2570
    },
    {
      "entropy": 5.529564762115479,
      "epoch": 2.1548117154811717,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004774451246070257,
      "loss": 5.2211,
      "mean_token_accuracy": 0.1945394530892372,
      "num_tokens": 4760681.0,
      "step": 2575
    },
    {
      "entropy": 5.55622820854187,
      "epoch": 2.1589958158995817,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004773040633372471,
      "loss": 5.2048,
      "mean_token_accuracy": 0.19357844293117524,
      "num_tokens": 4770237.0,
      "step": 2580
    },
    {
      "entropy": 5.567328310012817,
      "epoch": 2.1631799163179917,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047716258575823147,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19111842960119246,
      "num_tokens": 4779022.0,
      "step": 2585
    },
    {
      "entropy": 5.59231653213501,
      "epoch": 2.1673640167364017,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477020692161117,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1882817640900612,
      "num_tokens": 4788644.0,
      "step": 2590
    },
    {
      "entropy": 5.468843793869018,
      "epoch": 2.1715481171548117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004768783828378978,
      "loss": 5.248,
      "mean_token_accuracy": 0.18198808878660203,
      "num_tokens": 4798628.0,
      "step": 2595
    },
    {
      "entropy": 5.709801912307739,
      "epoch": 2.1757322175732217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004767356580814237,
      "loss": 5.3477,
      "mean_token_accuracy": 0.18395230323076248,
      "num_tokens": 4807727.0,
      "step": 2600
    },
    {
      "entropy": 5.627778053283691,
      "epoch": 2.1799163179916317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004765925181853993,
      "loss": 5.2205,
      "mean_token_accuracy": 0.18238564878702163,
      "num_tokens": 4817896.0,
      "step": 2605
    },
    {
      "entropy": 5.565152502059936,
      "epoch": 2.1841004184100417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004764489634443834,
      "loss": 5.1815,
      "mean_token_accuracy": 0.18810514807701112,
      "num_tokens": 4827135.0,
      "step": 2610
    },
    {
      "entropy": 5.571887063980102,
      "epoch": 2.1882845188284517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047630499415378863,
      "loss": 5.2606,
      "mean_token_accuracy": 0.185622002184391,
      "num_tokens": 4837554.0,
      "step": 2615
    },
    {
      "entropy": 5.454201078414917,
      "epoch": 2.192468619246862,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004761606106098806,
      "loss": 5.1374,
      "mean_token_accuracy": 0.19262371063232422,
      "num_tokens": 4846739.0,
      "step": 2620
    },
    {
      "entropy": 5.549982786178589,
      "epoch": 2.196652719665272,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047601581310977747,
      "loss": 5.2102,
      "mean_token_accuracy": 0.1892186775803566,
      "num_tokens": 4855707.0,
      "step": 2625
    },
    {
      "entropy": 5.534137153625489,
      "epoch": 2.200836820083682,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004758706019514493,
      "loss": 5.1952,
      "mean_token_accuracy": 0.19004165083169938,
      "num_tokens": 4865098.0,
      "step": 2630
    },
    {
      "entropy": 5.605195140838623,
      "epoch": 2.205020920502092,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004757249774337172,
      "loss": 5.2084,
      "mean_token_accuracy": 0.19359517991542816,
      "num_tokens": 4874426.0,
      "step": 2635
    },
    {
      "entropy": 5.523797559738159,
      "epoch": 2.209205020920502,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047557893985625293,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1928726240992546,
      "num_tokens": 4883240.0,
      "step": 2640
    },
    {
      "entropy": 5.478337240219116,
      "epoch": 2.213389121338912,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004754324895195786,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1926814079284668,
      "num_tokens": 4891989.0,
      "step": 2645
    },
    {
      "entropy": 5.5565934658050535,
      "epoch": 2.217573221757322,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004752856267250652,
      "loss": 5.2583,
      "mean_token_accuracy": 0.18897372037172316,
      "num_tokens": 4901546.0,
      "step": 2650
    },
    {
      "entropy": 5.444426965713501,
      "epoch": 2.221757322175732,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047513835177493286,
      "loss": 5.0804,
      "mean_token_accuracy": 0.19561816602945328,
      "num_tokens": 4910222.0,
      "step": 2655
    },
    {
      "entropy": 5.5684233665466305,
      "epoch": 2.225941422594142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004749906649722498,
      "loss": 5.3037,
      "mean_token_accuracy": 0.1823274552822113,
      "num_tokens": 4920184.0,
      "step": 2660
    },
    {
      "entropy": 5.530097723007202,
      "epoch": 2.2301255230125525,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004748425666209316,
      "loss": 5.1993,
      "mean_token_accuracy": 0.1915793240070343,
      "num_tokens": 4930014.0,
      "step": 2665
    },
    {
      "entropy": 5.52888674736023,
      "epoch": 2.2343096234309625,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004746940570257409,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19633713066577912,
      "num_tokens": 4939556.0,
      "step": 2670
    },
    {
      "entropy": 5.456666135787964,
      "epoch": 2.2384937238493725,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745451364922867,
      "loss": 5.1854,
      "mean_token_accuracy": 0.18619713932275772,
      "num_tokens": 4949066.0,
      "step": 2675
    },
    {
      "entropy": 5.60462965965271,
      "epoch": 2.2426778242677825,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004743958053270233,
      "loss": 5.1857,
      "mean_token_accuracy": 0.19684374779462815,
      "num_tokens": 4957881.0,
      "step": 2680
    },
    {
      "entropy": 5.516653776168823,
      "epoch": 2.2468619246861925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047424606383725037,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1889663055539131,
      "num_tokens": 4967105.0,
      "step": 2685
    },
    {
      "entropy": 5.500734186172485,
      "epoch": 2.2510460251046025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004740959123311118,
      "loss": 5.1201,
      "mean_token_accuracy": 0.19694772958755494,
      "num_tokens": 4977676.0,
      "step": 2690
    },
    {
      "entropy": 5.448371934890747,
      "epoch": 2.2552301255230125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047394535111759526,
      "loss": 5.1571,
      "mean_token_accuracy": 0.19497282058000565,
      "num_tokens": 4986163.0,
      "step": 2695
    },
    {
      "entropy": 5.516038846969605,
      "epoch": 2.2594142259414225,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004737943805065314,
      "loss": 5.1569,
      "mean_token_accuracy": 0.18557292073965073,
      "num_tokens": 4995179.0,
      "step": 2700
    },
    {
      "entropy": 5.57070837020874,
      "epoch": 2.2635983263598325,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047364300080859363,
      "loss": 5.2456,
      "mean_token_accuracy": 0.18458279073238373,
      "num_tokens": 5003856.0,
      "step": 2705
    },
    {
      "entropy": 5.54024863243103,
      "epoch": 2.2677824267782425,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047349121233529695,
      "loss": 5.2841,
      "mean_token_accuracy": 0.18307160437107087,
      "num_tokens": 5013506.0,
      "step": 2710
    },
    {
      "entropy": 5.517574882507324,
      "epoch": 2.2719665271966525,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004733390153989976,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19041822850704193,
      "num_tokens": 5022159.0,
      "step": 2715
    },
    {
      "entropy": 5.544815492630005,
      "epoch": 2.276150627615063,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047318641031289245,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1895567938685417,
      "num_tokens": 5031869.0,
      "step": 2720
    },
    {
      "entropy": 5.4250811576843265,
      "epoch": 2.280334728033473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730333973910183,
      "loss": 5.031,
      "mean_token_accuracy": 0.20200449526309966,
      "num_tokens": 5039904.0,
      "step": 2725
    },
    {
      "entropy": 5.529689121246338,
      "epoch": 2.284518828451883,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004728799769482511,
      "loss": 5.3029,
      "mean_token_accuracy": 0.18466683924198152,
      "num_tokens": 5049144.0,
      "step": 2730
    },
    {
      "entropy": 5.519205617904663,
      "epoch": 2.288702928870293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047272614930030533,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19491563737392426,
      "num_tokens": 5058302.0,
      "step": 2735
    },
    {
      "entropy": 5.497249746322632,
      "epoch": 2.292887029288703,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047257191476373375,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19153977185487747,
      "num_tokens": 5067443.0,
      "step": 2740
    },
    {
      "entropy": 5.462966060638427,
      "epoch": 2.297071129707113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047241727365592603,
      "loss": 5.2445,
      "mean_token_accuracy": 0.1829087108373642,
      "num_tokens": 5076415.0,
      "step": 2745
    },
    {
      "entropy": 5.544723129272461,
      "epoch": 2.301255230125523,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004722622262951089,
      "loss": 5.2121,
      "mean_token_accuracy": 0.18837485909461976,
      "num_tokens": 5085882.0,
      "step": 2750
    },
    {
      "entropy": 5.561155891418457,
      "epoch": 2.305439330543933,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004721067730003448,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1986195296049118,
      "num_tokens": 5094685.0,
      "step": 2755
    },
    {
      "entropy": 5.522579193115234,
      "epoch": 2.309623430962343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004719509140915317,
      "loss": 5.1946,
      "mean_token_accuracy": 0.1885368198156357,
      "num_tokens": 5103885.0,
      "step": 2760
    },
    {
      "entropy": 5.523617172241211,
      "epoch": 2.3138075313807533,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047179464988940205,
      "loss": 5.1232,
      "mean_token_accuracy": 0.20171012580394745,
      "num_tokens": 5112164.0,
      "step": 2765
    },
    {
      "entropy": 5.397410917282104,
      "epoch": 2.3179916317991633,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047163798071552253,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1930816426873207,
      "num_tokens": 5121913.0,
      "step": 2770
    },
    {
      "entropy": 5.499872779846191,
      "epoch": 2.3221757322175733,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004714809068922931,
      "loss": 5.0993,
      "mean_token_accuracy": 0.20085703879594802,
      "num_tokens": 5130447.0,
      "step": 2775
    },
    {
      "entropy": 5.530040597915649,
      "epoch": 2.3263598326359833,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004713234287429465,
      "loss": 5.2262,
      "mean_token_accuracy": 0.18991127461194993,
      "num_tokens": 5139863.0,
      "step": 2780
    },
    {
      "entropy": 5.390472364425659,
      "epoch": 2.3305439330543933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004711655465915473,
      "loss": 5.1751,
      "mean_token_accuracy": 0.1947562500834465,
      "num_tokens": 5149945.0,
      "step": 2785
    },
    {
      "entropy": 5.560927772521973,
      "epoch": 2.3347280334728033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004710072607629917,
      "loss": 5.1167,
      "mean_token_accuracy": 0.18760036081075668,
      "num_tokens": 5160293.0,
      "step": 2790
    },
    {
      "entropy": 5.460022878646851,
      "epoch": 2.3389121338912133,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047084857158300646,
      "loss": 5.1393,
      "mean_token_accuracy": 0.1829672560095787,
      "num_tokens": 5169896.0,
      "step": 2795
    },
    {
      "entropy": 5.503097057342529,
      "epoch": 2.3430962343096233,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047068947937814844,
      "loss": 5.1958,
      "mean_token_accuracy": 0.18684695065021514,
      "num_tokens": 5178898.0,
      "step": 2800
    },
    {
      "entropy": 5.561424970626831,
      "epoch": 2.3472803347280333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047052998447580374,
      "loss": 5.2588,
      "mean_token_accuracy": 0.18708261400461196,
      "num_tokens": 5188733.0,
      "step": 2805
    },
    {
      "entropy": 5.4390233039855955,
      "epoch": 2.3514644351464433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047037008720418724,
      "loss": 5.1555,
      "mean_token_accuracy": 0.19035712778568267,
      "num_tokens": 5198117.0,
      "step": 2810
    },
    {
      "entropy": 5.538933753967285,
      "epoch": 2.3556485355648533,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004702097878923419,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1884738877415657,
      "num_tokens": 5206791.0,
      "step": 2815
    },
    {
      "entropy": 5.4940228939056395,
      "epoch": 2.3598326359832638,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004700490868701377,
      "loss": 5.1058,
      "mean_token_accuracy": 0.19394215941429138,
      "num_tokens": 5216427.0,
      "step": 2820
    },
    {
      "entropy": 5.3847336769104,
      "epoch": 2.3640167364016738,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004698879844682718,
      "loss": 5.0899,
      "mean_token_accuracy": 0.1956726461648941,
      "num_tokens": 5225983.0,
      "step": 2825
    },
    {
      "entropy": 5.439257955551147,
      "epoch": 2.3682008368200838,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046972648101826695,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19315690398216248,
      "num_tokens": 5234547.0,
      "step": 2830
    },
    {
      "entropy": 5.596295785903931,
      "epoch": 2.3723849372384938,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004695645768524712,
      "loss": 5.2238,
      "mean_token_accuracy": 0.19835204333066941,
      "num_tokens": 5243281.0,
      "step": 2835
    },
    {
      "entropy": 5.490299701690674,
      "epoch": 2.3765690376569037,
      "grad_norm": 1.125,
      "learning_rate": 0.0004694022723040573,
      "loss": 5.1261,
      "mean_token_accuracy": 0.18668307811021806,
      "num_tokens": 5252318.0,
      "step": 2840
    },
    {
      "entropy": 5.515432500839234,
      "epoch": 2.3807531380753137,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046923956770702206,
      "loss": 5.204,
      "mean_token_accuracy": 0.1940845176577568,
      "num_tokens": 5261781.0,
      "step": 2845
    },
    {
      "entropy": 5.573498725891113,
      "epoch": 2.3849372384937237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004690764633961852,
      "loss": 5.11,
      "mean_token_accuracy": 0.19316617548465728,
      "num_tokens": 5270409.0,
      "step": 2850
    },
    {
      "entropy": 5.364716100692749,
      "epoch": 2.3891213389121337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004689129597071893,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20516035258769988,
      "num_tokens": 5278766.0,
      "step": 2855
    },
    {
      "entropy": 5.515507650375366,
      "epoch": 2.393305439330544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046874905697649857,
      "loss": 5.2754,
      "mean_token_accuracy": 0.1816343769431114,
      "num_tokens": 5287572.0,
      "step": 2860
    },
    {
      "entropy": 5.501306343078613,
      "epoch": 2.397489539748954,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046858475554139857,
      "loss": 5.1833,
      "mean_token_accuracy": 0.19312700480222703,
      "num_tokens": 5297348.0,
      "step": 2865
    },
    {
      "entropy": 5.549457693099976,
      "epoch": 2.401673640167364,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004684200557399952,
      "loss": 5.1952,
      "mean_token_accuracy": 0.1903080239892006,
      "num_tokens": 5305748.0,
      "step": 2870
    },
    {
      "entropy": 5.492236948013305,
      "epoch": 2.405857740585774,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046825495791121414,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19962307959794998,
      "num_tokens": 5314482.0,
      "step": 2875
    },
    {
      "entropy": 5.50344614982605,
      "epoch": 2.410041841004184,
      "grad_norm": 1.0,
      "learning_rate": 0.00046808946239480017,
      "loss": 5.2372,
      "mean_token_accuracy": 0.1866622343659401,
      "num_tokens": 5325161.0,
      "step": 2880
    },
    {
      "entropy": 5.4480592727661135,
      "epoch": 2.414225941422594,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004679235695313165,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18587664663791656,
      "num_tokens": 5335262.0,
      "step": 2885
    },
    {
      "entropy": 5.471191024780273,
      "epoch": 2.418410041841004,
      "grad_norm": 1.25,
      "learning_rate": 0.00046775727966214393,
      "loss": 5.1161,
      "mean_token_accuracy": 0.18551510572433472,
      "num_tokens": 5345145.0,
      "step": 2890
    },
    {
      "entropy": 5.438928413391113,
      "epoch": 2.422594142259414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004675905931294803,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19111968725919723,
      "num_tokens": 5354715.0,
      "step": 2895
    },
    {
      "entropy": 5.437091255187989,
      "epoch": 2.426778242677824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004674235102763398,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1983390673995018,
      "num_tokens": 5364180.0,
      "step": 2900
    },
    {
      "entropy": 5.454319906234741,
      "epoch": 2.430962343096234,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046725603144655195,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19619197398424149,
      "num_tokens": 5372919.0,
      "step": 2905
    },
    {
      "entropy": 5.462763833999634,
      "epoch": 2.435146443514644,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004670881569847612,
      "loss": 5.104,
      "mean_token_accuracy": 0.20203736573457717,
      "num_tokens": 5383002.0,
      "step": 2910
    },
    {
      "entropy": 5.55969090461731,
      "epoch": 2.439330543933054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046691988723642636,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18525985181331633,
      "num_tokens": 5393622.0,
      "step": 2915
    },
    {
      "entropy": 5.5230185985565186,
      "epoch": 2.4435146443514646,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004667512225478195,
      "loss": 5.1987,
      "mean_token_accuracy": 0.18882321268320085,
      "num_tokens": 5403106.0,
      "step": 2920
    },
    {
      "entropy": 5.513156461715698,
      "epoch": 2.4476987447698746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046658216326602526,
      "loss": 5.1784,
      "mean_token_accuracy": 0.1932978704571724,
      "num_tokens": 5412596.0,
      "step": 2925
    },
    {
      "entropy": 5.55132269859314,
      "epoch": 2.4518828451882846,
      "grad_norm": 1.125,
      "learning_rate": 0.0004664127097389406,
      "loss": 5.2235,
      "mean_token_accuracy": 0.19170084595680237,
      "num_tokens": 5422301.0,
      "step": 2930
    },
    {
      "entropy": 5.533843612670898,
      "epoch": 2.4560669456066946,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046624286231527367,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1933741256594658,
      "num_tokens": 5432178.0,
      "step": 2935
    },
    {
      "entropy": 5.372134637832642,
      "epoch": 2.4602510460251046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004660726213445431,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19677756428718568,
      "num_tokens": 5441162.0,
      "step": 2940
    },
    {
      "entropy": 5.468031597137451,
      "epoch": 2.4644351464435146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004659019871770774,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20391910523176193,
      "num_tokens": 5450566.0,
      "step": 2945
    },
    {
      "entropy": 5.5154331684112545,
      "epoch": 2.4686192468619246,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046573096016401435,
      "loss": 5.144,
      "mean_token_accuracy": 0.20277026444673538,
      "num_tokens": 5460289.0,
      "step": 2950
    },
    {
      "entropy": 5.427037620544434,
      "epoch": 2.4728033472803346,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004655595406572999,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19660421907901765,
      "num_tokens": 5469581.0,
      "step": 2955
    },
    {
      "entropy": 5.518569612503052,
      "epoch": 2.476987447698745,
      "grad_norm": 1.0625,
      "learning_rate": 0.000465387729009688,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19369943886995317,
      "num_tokens": 5478005.0,
      "step": 2960
    },
    {
      "entropy": 5.485342979431152,
      "epoch": 2.481171548117155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046521552557473927,
      "loss": 5.2092,
      "mean_token_accuracy": 0.19067464172840118,
      "num_tokens": 5486963.0,
      "step": 2965
    },
    {
      "entropy": 5.396170234680175,
      "epoch": 2.485355648535565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046504293070682073,
      "loss": 5.0971,
      "mean_token_accuracy": 0.19292196929454802,
      "num_tokens": 5496368.0,
      "step": 2970
    },
    {
      "entropy": 5.4054327487945555,
      "epoch": 2.489539748953975,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046486994476110493,
      "loss": 5.0894,
      "mean_token_accuracy": 0.19642373919487,
      "num_tokens": 5505706.0,
      "step": 2975
    },
    {
      "entropy": 5.394309616088867,
      "epoch": 2.493723849372385,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046469656809356916,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20244767069816588,
      "num_tokens": 5513945.0,
      "step": 2980
    },
    {
      "entropy": 5.3953783988952635,
      "epoch": 2.497907949790795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046452280106099454,
      "loss": 5.1325,
      "mean_token_accuracy": 0.188616843521595,
      "num_tokens": 5523831.0,
      "step": 2985
    },
    {
      "entropy": 5.527229118347168,
      "epoch": 2.502092050209205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004643486440209659,
      "loss": 5.2699,
      "mean_token_accuracy": 0.18961481899023055,
      "num_tokens": 5532705.0,
      "step": 2990
    },
    {
      "entropy": 5.525477695465088,
      "epoch": 2.506276150627615,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004641740973318702,
      "loss": 5.1864,
      "mean_token_accuracy": 0.19227942377328872,
      "num_tokens": 5542793.0,
      "step": 2995
    },
    {
      "entropy": 5.5265655517578125,
      "epoch": 2.510460251046025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004639991613528967,
      "loss": 5.186,
      "mean_token_accuracy": 0.19159698039293288,
      "num_tokens": 5551938.0,
      "step": 3000
    },
    {
      "epoch": 2.510460251046025,
      "eval_entropy": 5.213770697697052,
      "eval_loss": 5.475037574768066,
      "eval_mean_token_accuracy": 0.1852224213823463,
      "eval_num_tokens": 5551938.0,
      "eval_runtime": 2.2909,
      "eval_samples_per_second": 1705.412,
      "eval_steps_per_second": 213.449,
      "step": 3000
    },
    {
      "entropy": 5.462968111038208,
      "epoch": 2.514644351464435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046382383644403543,
      "loss": 5.2129,
      "mean_token_accuracy": 0.18845061659812928,
      "num_tokens": 5561374.0,
      "step": 3005
    },
    {
      "entropy": 5.3695238590240475,
      "epoch": 2.518828451882845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046364812296607676,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19389346539974212,
      "num_tokens": 5570459.0,
      "step": 3010
    },
    {
      "entropy": 5.408000421524048,
      "epoch": 2.523012552301255,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046347202128061094,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1907093733549118,
      "num_tokens": 5579692.0,
      "step": 3015
    },
    {
      "entropy": 5.527827787399292,
      "epoch": 2.5271966527196654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004632955317500268,
      "loss": 5.1324,
      "mean_token_accuracy": 0.19062095284461975,
      "num_tokens": 5588338.0,
      "step": 3020
    },
    {
      "entropy": 5.4718592166900635,
      "epoch": 2.5313807531380754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046311865473751154,
      "loss": 5.22,
      "mean_token_accuracy": 0.19249685406684874,
      "num_tokens": 5598589.0,
      "step": 3025
    },
    {
      "entropy": 5.462061405181885,
      "epoch": 2.5355648535564854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004629413906070495,
      "loss": 5.1907,
      "mean_token_accuracy": 0.19549299776554108,
      "num_tokens": 5608127.0,
      "step": 3030
    },
    {
      "entropy": 5.548971509933471,
      "epoch": 2.5397489539748954,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046276373972342177,
      "loss": 5.235,
      "mean_token_accuracy": 0.19585680067539216,
      "num_tokens": 5617728.0,
      "step": 3035
    },
    {
      "entropy": 5.430024433135986,
      "epoch": 2.5439330543933054,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004625857024522054,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1904665008187294,
      "num_tokens": 5626823.0,
      "step": 3040
    },
    {
      "entropy": 5.46235408782959,
      "epoch": 2.5481171548117154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046240727915977236,
      "loss": 5.158,
      "mean_token_accuracy": 0.19430482536554336,
      "num_tokens": 5636284.0,
      "step": 3045
    },
    {
      "entropy": 5.555997657775879,
      "epoch": 2.5523012552301254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004622284702132893,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1841801658272743,
      "num_tokens": 5645721.0,
      "step": 3050
    },
    {
      "entropy": 5.413286399841309,
      "epoch": 2.5564853556485354,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046204927598071595,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19346165359020234,
      "num_tokens": 5654734.0,
      "step": 3055
    },
    {
      "entropy": 5.527040958404541,
      "epoch": 2.560669456066946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046186969683080547,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20113472640514374,
      "num_tokens": 5664010.0,
      "step": 3060
    },
    {
      "entropy": 5.474124813079834,
      "epoch": 2.564853556485356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046168973313310294,
      "loss": 5.17,
      "mean_token_accuracy": 0.18840867280960083,
      "num_tokens": 5673028.0,
      "step": 3065
    },
    {
      "entropy": 5.43609094619751,
      "epoch": 2.569037656903766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004615093852579445,
      "loss": 5.1336,
      "mean_token_accuracy": 0.19433332085609437,
      "num_tokens": 5682517.0,
      "step": 3070
    },
    {
      "entropy": 5.3480634689331055,
      "epoch": 2.573221757322176,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046132865357645717,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20391639918088914,
      "num_tokens": 5690920.0,
      "step": 3075
    },
    {
      "entropy": 5.414696264266968,
      "epoch": 2.577405857740586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046114753846055774,
      "loss": 5.143,
      "mean_token_accuracy": 0.19163515716791152,
      "num_tokens": 5700336.0,
      "step": 3080
    },
    {
      "entropy": 5.515250396728516,
      "epoch": 2.581589958158996,
      "grad_norm": 1.03125,
      "learning_rate": 0.000460966040282952,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18952076584100724,
      "num_tokens": 5711085.0,
      "step": 3085
    },
    {
      "entropy": 5.500215530395508,
      "epoch": 2.585774058577406,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046078415941713384,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20170718282461167,
      "num_tokens": 5719560.0,
      "step": 3090
    },
    {
      "entropy": 5.378318500518799,
      "epoch": 2.589958158995816,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004606018962373852,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20389935672283171,
      "num_tokens": 5727865.0,
      "step": 3095
    },
    {
      "entropy": 5.4493427753448485,
      "epoch": 2.594142259414226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004604192511187741,
      "loss": 5.1844,
      "mean_token_accuracy": 0.1872186169028282,
      "num_tokens": 5737705.0,
      "step": 3100
    },
    {
      "entropy": 5.503284740447998,
      "epoch": 2.598326359832636,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046023622443715483,
      "loss": 5.2045,
      "mean_token_accuracy": 0.1855546861886978,
      "num_tokens": 5747057.0,
      "step": 3105
    },
    {
      "entropy": 5.434207344055176,
      "epoch": 2.602510460251046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046005281656916707,
      "loss": 5.156,
      "mean_token_accuracy": 0.19481234550476073,
      "num_tokens": 5755793.0,
      "step": 3110
    },
    {
      "entropy": 5.47066297531128,
      "epoch": 2.606694560669456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004598690278922346,
      "loss": 5.1581,
      "mean_token_accuracy": 0.1878283441066742,
      "num_tokens": 5765385.0,
      "step": 3115
    },
    {
      "entropy": 5.33690767288208,
      "epoch": 2.6108786610878663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045968485878456497,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19226381778717042,
      "num_tokens": 5773947.0,
      "step": 3120
    },
    {
      "entropy": 5.443906211853028,
      "epoch": 2.6150627615062763,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004595003096251486,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19814347773790358,
      "num_tokens": 5782700.0,
      "step": 3125
    },
    {
      "entropy": 5.493331956863403,
      "epoch": 2.6192468619246863,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045931538079375803,
      "loss": 5.2399,
      "mean_token_accuracy": 0.18730711191892624,
      "num_tokens": 5792452.0,
      "step": 3130
    },
    {
      "entropy": 5.486668348312378,
      "epoch": 2.6234309623430963,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045913007267094704,
      "loss": 5.1706,
      "mean_token_accuracy": 0.19296336323022842,
      "num_tokens": 5801270.0,
      "step": 3135
    },
    {
      "entropy": 5.532931137084961,
      "epoch": 2.6276150627615062,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004589443856380499,
      "loss": 5.211,
      "mean_token_accuracy": 0.19481545835733413,
      "num_tokens": 5810704.0,
      "step": 3140
    },
    {
      "entropy": 5.481737565994263,
      "epoch": 2.6317991631799162,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004587583200771807,
      "loss": 5.1578,
      "mean_token_accuracy": 0.1936139225959778,
      "num_tokens": 5819690.0,
      "step": 3145
    },
    {
      "entropy": 5.37274956703186,
      "epoch": 2.6359832635983262,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004585718763712325,
      "loss": 5.0867,
      "mean_token_accuracy": 0.19192241877317429,
      "num_tokens": 5828912.0,
      "step": 3150
    },
    {
      "entropy": 5.31561918258667,
      "epoch": 2.6401673640167362,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004583850549038764,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19629174768924712,
      "num_tokens": 5836914.0,
      "step": 3155
    },
    {
      "entropy": 5.444818162918091,
      "epoch": 2.6443514644351467,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045819785605956105,
      "loss": 5.1609,
      "mean_token_accuracy": 0.19689084142446517,
      "num_tokens": 5846268.0,
      "step": 3160
    },
    {
      "entropy": 5.400204181671143,
      "epoch": 2.6485355648535567,
      "grad_norm": 1.125,
      "learning_rate": 0.0004580102802235115,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1971372663974762,
      "num_tokens": 5855029.0,
      "step": 3165
    },
    {
      "entropy": 5.49607892036438,
      "epoch": 2.6527196652719667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004578223277817286,
      "loss": 5.1717,
      "mean_token_accuracy": 0.18653464615345,
      "num_tokens": 5865049.0,
      "step": 3170
    },
    {
      "entropy": 5.5427673816680905,
      "epoch": 2.6569037656903767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045763399912098835,
      "loss": 5.1744,
      "mean_token_accuracy": 0.18617971241474152,
      "num_tokens": 5873593.0,
      "step": 3175
    },
    {
      "entropy": 5.361571311950684,
      "epoch": 2.6610878661087867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045744529462884086,
      "loss": 5.0781,
      "mean_token_accuracy": 0.19762487709522247,
      "num_tokens": 5882562.0,
      "step": 3180
    },
    {
      "entropy": 5.429044437408447,
      "epoch": 2.6652719665271967,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045725621469360964,
      "loss": 5.2668,
      "mean_token_accuracy": 0.18453310579061508,
      "num_tokens": 5892516.0,
      "step": 3185
    },
    {
      "entropy": 5.4789777278900145,
      "epoch": 2.6694560669456067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004570667597043907,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19438394159078598,
      "num_tokens": 5901494.0,
      "step": 3190
    },
    {
      "entropy": 5.456045532226563,
      "epoch": 2.6736401673640167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000456876930051052,
      "loss": 5.1421,
      "mean_token_accuracy": 0.19510222375392913,
      "num_tokens": 5910750.0,
      "step": 3195
    },
    {
      "entropy": 5.541888809204101,
      "epoch": 2.6778242677824267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004566867261242325,
      "loss": 5.183,
      "mean_token_accuracy": 0.19630327373743056,
      "num_tokens": 5920277.0,
      "step": 3200
    },
    {
      "entropy": 5.410405969619751,
      "epoch": 2.6820083682008367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045649614831534127,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1965371772646904,
      "num_tokens": 5929880.0,
      "step": 3205
    },
    {
      "entropy": 5.527627420425415,
      "epoch": 2.6861924686192467,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045630519701655677,
      "loss": 5.321,
      "mean_token_accuracy": 0.18892762362957,
      "num_tokens": 5939538.0,
      "step": 3210
    },
    {
      "entropy": 5.448320198059082,
      "epoch": 2.6903765690376567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045611387262082604,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19353785663843154,
      "num_tokens": 5948747.0,
      "step": 3215
    },
    {
      "entropy": 5.4134666442871096,
      "epoch": 2.694560669456067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045592217552186407,
      "loss": 5.1274,
      "mean_token_accuracy": 0.19890922158956528,
      "num_tokens": 5958209.0,
      "step": 3220
    },
    {
      "entropy": 5.474023628234863,
      "epoch": 2.698744769874477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004557301061141525,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19570582509040832,
      "num_tokens": 5967184.0,
      "step": 3225
    },
    {
      "entropy": 5.3882392883300785,
      "epoch": 2.702928870292887,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045553766479293945,
      "loss": 5.1882,
      "mean_token_accuracy": 0.19017804712057113,
      "num_tokens": 5976006.0,
      "step": 3230
    },
    {
      "entropy": 5.338523197174072,
      "epoch": 2.707112970711297,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045534485195423806,
      "loss": 5.032,
      "mean_token_accuracy": 0.19907400161027908,
      "num_tokens": 5984671.0,
      "step": 3235
    },
    {
      "entropy": 5.536633491516113,
      "epoch": 2.711297071129707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045515166799482633,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1948863998055458,
      "num_tokens": 5993190.0,
      "step": 3240
    },
    {
      "entropy": 5.443432140350342,
      "epoch": 2.715481171548117,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045495811331224564,
      "loss": 5.1621,
      "mean_token_accuracy": 0.19129905104637146,
      "num_tokens": 6004080.0,
      "step": 3245
    },
    {
      "entropy": 5.404340171813965,
      "epoch": 2.719665271966527,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004547641883048005,
      "loss": 5.1551,
      "mean_token_accuracy": 0.19158968031406404,
      "num_tokens": 6013212.0,
      "step": 3250
    },
    {
      "entropy": 5.437668323516846,
      "epoch": 2.723849372384937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004545698933715573,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19823037832975388,
      "num_tokens": 6023444.0,
      "step": 3255
    },
    {
      "entropy": 5.4280712604522705,
      "epoch": 2.7280334728033475,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045437522891234386,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19958617091178893,
      "num_tokens": 6032217.0,
      "step": 3260
    },
    {
      "entropy": 5.404166889190674,
      "epoch": 2.7322175732217575,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045418019532774827,
      "loss": 5.0588,
      "mean_token_accuracy": 0.19934537559747695,
      "num_tokens": 6040246.0,
      "step": 3265
    },
    {
      "entropy": 5.41451325416565,
      "epoch": 2.7364016736401675,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045398479301911836,
      "loss": 5.173,
      "mean_token_accuracy": 0.1900328889489174,
      "num_tokens": 6049447.0,
      "step": 3270
    },
    {
      "entropy": 5.332231903076172,
      "epoch": 2.7405857740585775,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045378902238856064,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19907548427581787,
      "num_tokens": 6058294.0,
      "step": 3275
    },
    {
      "entropy": 5.449809646606445,
      "epoch": 2.7447698744769875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004535928838389396,
      "loss": 5.1878,
      "mean_token_accuracy": 0.19494476914405823,
      "num_tokens": 6067237.0,
      "step": 3280
    },
    {
      "entropy": 5.51179518699646,
      "epoch": 2.7489539748953975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004533963777738768,
      "loss": 5.1898,
      "mean_token_accuracy": 0.19057921171188355,
      "num_tokens": 6075599.0,
      "step": 3285
    },
    {
      "entropy": 5.460715055465698,
      "epoch": 2.7531380753138075,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045319950459775024,
      "loss": 5.1629,
      "mean_token_accuracy": 0.18881246596574783,
      "num_tokens": 6085131.0,
      "step": 3290
    },
    {
      "entropy": 5.487667608261108,
      "epoch": 2.7573221757322175,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004530022647156931,
      "loss": 5.1314,
      "mean_token_accuracy": 0.19969360232353212,
      "num_tokens": 6094745.0,
      "step": 3295
    },
    {
      "entropy": 5.3998754024505615,
      "epoch": 2.7615062761506275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004528046585335935,
      "loss": 5.1953,
      "mean_token_accuracy": 0.18483933806419373,
      "num_tokens": 6104182.0,
      "step": 3300
    },
    {
      "entropy": 5.410027885437012,
      "epoch": 2.7656903765690375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045260668645809304,
      "loss": 5.0517,
      "mean_token_accuracy": 0.2025168314576149,
      "num_tokens": 6113370.0,
      "step": 3305
    },
    {
      "entropy": 5.356280279159546,
      "epoch": 2.7698744769874475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004524083488965866,
      "loss": 5.1293,
      "mean_token_accuracy": 0.19631304293870927,
      "num_tokens": 6122202.0,
      "step": 3310
    },
    {
      "entropy": 5.374682331085205,
      "epoch": 2.7740585774058575,
      "grad_norm": 1.125,
      "learning_rate": 0.0004522096462572208,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1911054402589798,
      "num_tokens": 6131081.0,
      "step": 3315
    },
    {
      "entropy": 5.586391258239746,
      "epoch": 2.778242677824268,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004520105789488939,
      "loss": 5.2668,
      "mean_token_accuracy": 0.18820440918207168,
      "num_tokens": 6141707.0,
      "step": 3320
    },
    {
      "entropy": 5.4600382328033445,
      "epoch": 2.782426778242678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045181114738125427,
      "loss": 5.1162,
      "mean_token_accuracy": 0.19284327626228331,
      "num_tokens": 6151651.0,
      "step": 3325
    },
    {
      "entropy": 5.379871606826782,
      "epoch": 2.786610878661088,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045161135196470015,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19958383738994598,
      "num_tokens": 6159928.0,
      "step": 3330
    },
    {
      "entropy": 5.37453989982605,
      "epoch": 2.790794979079498,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004514111931103783,
      "loss": 5.1734,
      "mean_token_accuracy": 0.19033128768205643,
      "num_tokens": 6169115.0,
      "step": 3335
    },
    {
      "entropy": 5.4737201690673825,
      "epoch": 2.794979079497908,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045121067123018353,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1947602093219757,
      "num_tokens": 6177995.0,
      "step": 3340
    },
    {
      "entropy": 5.435278463363647,
      "epoch": 2.799163179916318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045100978673675755,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19039879590272904,
      "num_tokens": 6188081.0,
      "step": 3345
    },
    {
      "entropy": 5.50736927986145,
      "epoch": 2.803347280334728,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045080854004348836,
      "loss": 5.1291,
      "mean_token_accuracy": 0.1885787844657898,
      "num_tokens": 6198010.0,
      "step": 3350
    },
    {
      "entropy": 5.363803195953369,
      "epoch": 2.8075313807531384,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045060693156450933,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19714236706495286,
      "num_tokens": 6207164.0,
      "step": 3355
    },
    {
      "entropy": 5.4973828315734865,
      "epoch": 2.8117154811715483,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004504049617146983,
      "loss": 5.2401,
      "mean_token_accuracy": 0.18754298090934754,
      "num_tokens": 6216512.0,
      "step": 3360
    },
    {
      "entropy": 5.427261829376221,
      "epoch": 2.8158995815899583,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045020263090967674,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1960294485092163,
      "num_tokens": 6226196.0,
      "step": 3365
    },
    {
      "entropy": 5.4154823303222654,
      "epoch": 2.8200836820083683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004499999395658088,
      "loss": 5.0659,
      "mean_token_accuracy": 0.19427476078271866,
      "num_tokens": 6234820.0,
      "step": 3370
    },
    {
      "entropy": 5.381675148010254,
      "epoch": 2.8242677824267783,
      "grad_norm": 1.09375,
      "learning_rate": 0.000449796888100201,
      "loss": 5.1385,
      "mean_token_accuracy": 0.19204012900590897,
      "num_tokens": 6243813.0,
      "step": 3375
    },
    {
      "entropy": 5.518025684356689,
      "epoch": 2.8284518828451883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044959347693070026,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1914212316274643,
      "num_tokens": 6253376.0,
      "step": 3380
    },
    {
      "entropy": 5.555794095993042,
      "epoch": 2.8326359832635983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004493897064758943,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1916716992855072,
      "num_tokens": 6262640.0,
      "step": 3385
    },
    {
      "entropy": 5.519548797607422,
      "epoch": 2.8368200836820083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004491855771551098,
      "loss": 5.2022,
      "mean_token_accuracy": 0.1959479421377182,
      "num_tokens": 6270908.0,
      "step": 3390
    },
    {
      "entropy": 5.38501124382019,
      "epoch": 2.8410041841004183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004489810893884121,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19345175325870514,
      "num_tokens": 6280361.0,
      "step": 3395
    },
    {
      "entropy": 5.398903322219849,
      "epoch": 2.8451882845188283,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004487762435966042,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19799116104841233,
      "num_tokens": 6288832.0,
      "step": 3400
    },
    {
      "entropy": 5.410068607330322,
      "epoch": 2.8493723849372383,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044857104020122585,
      "loss": 5.1068,
      "mean_token_accuracy": 0.19581186324357985,
      "num_tokens": 6298553.0,
      "step": 3405
    },
    {
      "entropy": 5.341890954971314,
      "epoch": 2.8535564853556483,
      "grad_norm": 1.125,
      "learning_rate": 0.0004483654796245526,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19626759737730026,
      "num_tokens": 6307692.0,
      "step": 3410
    },
    {
      "entropy": 5.482541465759278,
      "epoch": 2.8577405857740583,
      "grad_norm": 1.0625,
      "learning_rate": 0.000448159562289595,
      "loss": 5.1455,
      "mean_token_accuracy": 0.19332865476608277,
      "num_tokens": 6316729.0,
      "step": 3415
    },
    {
      "entropy": 5.4931943893432615,
      "epoch": 2.8619246861924688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044795328862009794,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18952576369047164,
      "num_tokens": 6326210.0,
      "step": 3420
    },
    {
      "entropy": 5.321086454391479,
      "epoch": 2.8661087866108788,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044774665904053943,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1981959745287895,
      "num_tokens": 6335156.0,
      "step": 3425
    },
    {
      "entropy": 5.361148929595947,
      "epoch": 2.8702928870292888,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004475396739761299,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20167842358350754,
      "num_tokens": 6344264.0,
      "step": 3430
    },
    {
      "entropy": 5.422211742401123,
      "epoch": 2.8744769874476988,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044733233385281117,
      "loss": 5.093,
      "mean_token_accuracy": 0.200849287211895,
      "num_tokens": 6352997.0,
      "step": 3435
    },
    {
      "entropy": 5.363988924026489,
      "epoch": 2.8786610878661087,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044712463909725626,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1967838779091835,
      "num_tokens": 6362312.0,
      "step": 3440
    },
    {
      "entropy": 5.528323650360107,
      "epoch": 2.8828451882845187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044691659013686727,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19050872921943665,
      "num_tokens": 6372107.0,
      "step": 3445
    },
    {
      "entropy": 5.387990474700928,
      "epoch": 2.8870292887029287,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004467081873997756,
      "loss": 5.0526,
      "mean_token_accuracy": 0.20210649967193603,
      "num_tokens": 6381306.0,
      "step": 3450
    },
    {
      "entropy": 5.337385892868042,
      "epoch": 2.891213389121339,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044649943131484056,
      "loss": 5.0115,
      "mean_token_accuracy": 0.19747427701950074,
      "num_tokens": 6390404.0,
      "step": 3455
    },
    {
      "entropy": 5.489330101013183,
      "epoch": 2.895397489539749,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044629032231164874,
      "loss": 5.2212,
      "mean_token_accuracy": 0.1843263640999794,
      "num_tokens": 6400366.0,
      "step": 3460
    },
    {
      "entropy": 5.408727359771729,
      "epoch": 2.899581589958159,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004460808608205127,
      "loss": 5.0302,
      "mean_token_accuracy": 0.1996001347899437,
      "num_tokens": 6408955.0,
      "step": 3465
    },
    {
      "entropy": 5.262295055389404,
      "epoch": 2.903765690376569,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004458710472724706,
      "loss": 5.0918,
      "mean_token_accuracy": 0.19133829474449157,
      "num_tokens": 6417172.0,
      "step": 3470
    },
    {
      "entropy": 5.39056544303894,
      "epoch": 2.907949790794979,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044566088209928485,
      "loss": 5.1034,
      "mean_token_accuracy": 0.19498197138309478,
      "num_tokens": 6425898.0,
      "step": 3475
    },
    {
      "entropy": 5.371723222732544,
      "epoch": 2.912133891213389,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044545036573344175,
      "loss": 5.1406,
      "mean_token_accuracy": 0.19467767626047133,
      "num_tokens": 6435940.0,
      "step": 3480
    },
    {
      "entropy": 5.376276540756225,
      "epoch": 2.916317991631799,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044523949860815015,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20263846069574357,
      "num_tokens": 6445119.0,
      "step": 3485
    },
    {
      "entropy": 5.384667587280274,
      "epoch": 2.920502092050209,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004450282811573405,
      "loss": 5.1117,
      "mean_token_accuracy": 0.19618954360485077,
      "num_tokens": 6453986.0,
      "step": 3490
    },
    {
      "entropy": 5.376170206069946,
      "epoch": 2.924686192468619,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044481671381566446,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19363461285829545,
      "num_tokens": 6463658.0,
      "step": 3495
    },
    {
      "entropy": 5.379592990875244,
      "epoch": 2.928870292887029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004446047970184937,
      "loss": 5.1255,
      "mean_token_accuracy": 0.19629420936107636,
      "num_tokens": 6472370.0,
      "step": 3500
    },
    {
      "epoch": 2.928870292887029,
      "eval_entropy": 5.151964174945418,
      "eval_loss": 5.394122123718262,
      "eval_mean_token_accuracy": 0.1907018796066565,
      "eval_num_tokens": 6472370.0,
      "eval_runtime": 2.0851,
      "eval_samples_per_second": 1873.805,
      "eval_steps_per_second": 234.525,
      "step": 3500
    },
    {
      "entropy": 5.438113689422607,
      "epoch": 2.933054393305439,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044439253120191877,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1858174741268158,
      "num_tokens": 6482642.0,
      "step": 3505
    },
    {
      "entropy": 5.365430307388306,
      "epoch": 2.937238493723849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044417991680274874,
      "loss": 5.1493,
      "mean_token_accuracy": 0.1935872480273247,
      "num_tokens": 6491818.0,
      "step": 3510
    },
    {
      "entropy": 5.419385147094727,
      "epoch": 2.941422594142259,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004439669542585097,
      "loss": 5.1454,
      "mean_token_accuracy": 0.18960228860378264,
      "num_tokens": 6500908.0,
      "step": 3515
    },
    {
      "entropy": 5.352400112152099,
      "epoch": 2.9456066945606696,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004437536440074444,
      "loss": 5.1151,
      "mean_token_accuracy": 0.19481378197669982,
      "num_tokens": 6510182.0,
      "step": 3520
    },
    {
      "entropy": 5.35415678024292,
      "epoch": 2.9497907949790796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004435399864885111,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1997893199324608,
      "num_tokens": 6519051.0,
      "step": 3525
    },
    {
      "entropy": 5.330616283416748,
      "epoch": 2.9539748953974896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004433259821413825,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20871360450983048,
      "num_tokens": 6527949.0,
      "step": 3530
    },
    {
      "entropy": 5.506001710891724,
      "epoch": 2.9581589958158996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044311163140644524,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1850181758403778,
      "num_tokens": 6537227.0,
      "step": 3535
    },
    {
      "entropy": 5.38893723487854,
      "epoch": 2.9623430962343096,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044289693472479867,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1902851089835167,
      "num_tokens": 6546938.0,
      "step": 3540
    },
    {
      "entropy": 5.363958740234375,
      "epoch": 2.9665271966527196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044268189253825397,
      "loss": 5.1071,
      "mean_token_accuracy": 0.19246285557746887,
      "num_tokens": 6555750.0,
      "step": 3545
    },
    {
      "entropy": 5.310831022262573,
      "epoch": 2.9707112970711296,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004424665052893334,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19983767420053483,
      "num_tokens": 6565680.0,
      "step": 3550
    },
    {
      "entropy": 5.266908311843872,
      "epoch": 2.97489539748954,
      "grad_norm": 1.125,
      "learning_rate": 0.00044225077342126935,
      "loss": 5.014,
      "mean_token_accuracy": 0.20443485230207442,
      "num_tokens": 6574595.0,
      "step": 3555
    },
    {
      "entropy": 5.436822366714478,
      "epoch": 2.97907949790795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004420346973780033,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19291843771934508,
      "num_tokens": 6583333.0,
      "step": 3560
    },
    {
      "entropy": 5.461638736724853,
      "epoch": 2.98326359832636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000441818277604185,
      "loss": 5.1758,
      "mean_token_accuracy": 0.2010513201355934,
      "num_tokens": 6592252.0,
      "step": 3565
    },
    {
      "entropy": 5.345468807220459,
      "epoch": 2.98744769874477,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004416015145451715,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20407403260469437,
      "num_tokens": 6600675.0,
      "step": 3570
    },
    {
      "entropy": 5.34960904121399,
      "epoch": 2.99163179916318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004413844086470264,
      "loss": 5.0959,
      "mean_token_accuracy": 0.19900625348091125,
      "num_tokens": 6609722.0,
      "step": 3575
    },
    {
      "entropy": 5.422952175140381,
      "epoch": 2.99581589958159,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004411669603565187,
      "loss": 5.0922,
      "mean_token_accuracy": 0.20171553194522857,
      "num_tokens": 6618415.0,
      "step": 3580
    },
    {
      "entropy": 5.397150373458862,
      "epoch": 3.0,
      "grad_norm": 1.09375,
      "learning_rate": 0.000440949170121122,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19668337553739548,
      "num_tokens": 6627966.0,
      "step": 3585
    },
    {
      "entropy": 5.417689371109009,
      "epoch": 3.00418410041841,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044073103838901373,
      "loss": 4.7726,
      "mean_token_accuracy": 0.22511388063430787,
      "num_tokens": 6636820.0,
      "step": 3590
    },
    {
      "entropy": 5.302271938323974,
      "epoch": 3.00836820083682,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044051256560907374,
      "loss": 4.8061,
      "mean_token_accuracy": 0.21354635208845138,
      "num_tokens": 6645792.0,
      "step": 3595
    },
    {
      "entropy": 5.42719087600708,
      "epoch": 3.01255230125523,
      "grad_norm": 1.25,
      "learning_rate": 0.00044029375223088394,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21108814477920532,
      "num_tokens": 6654850.0,
      "step": 3600
    },
    {
      "entropy": 5.293935489654541,
      "epoch": 3.01673640167364,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044007459870472706,
      "loss": 4.8389,
      "mean_token_accuracy": 0.21405893117189406,
      "num_tokens": 6664524.0,
      "step": 3605
    },
    {
      "entropy": 5.289966678619384,
      "epoch": 3.02092050209205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004398551054815859,
      "loss": 4.7767,
      "mean_token_accuracy": 0.21690000742673873,
      "num_tokens": 6674330.0,
      "step": 3610
    },
    {
      "entropy": 5.350748205184937,
      "epoch": 3.0251046025104604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004396352730131422,
      "loss": 4.8596,
      "mean_token_accuracy": 0.21818140000104905,
      "num_tokens": 6683121.0,
      "step": 3615
    },
    {
      "entropy": 5.360624027252197,
      "epoch": 3.0292887029288704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004394151017517758,
      "loss": 4.816,
      "mean_token_accuracy": 0.2204051747918129,
      "num_tokens": 6692566.0,
      "step": 3620
    },
    {
      "entropy": 5.379842805862427,
      "epoch": 3.0334728033472804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004391945921505638,
      "loss": 4.8994,
      "mean_token_accuracy": 0.20829516500234604,
      "num_tokens": 6701577.0,
      "step": 3625
    },
    {
      "entropy": 5.25770583152771,
      "epoch": 3.0376569037656904,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043897374466327944,
      "loss": 4.7579,
      "mean_token_accuracy": 0.22355487197637558,
      "num_tokens": 6710672.0,
      "step": 3630
    },
    {
      "entropy": 5.287049102783203,
      "epoch": 3.0418410041841004,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043875255974439145,
      "loss": 4.7783,
      "mean_token_accuracy": 0.21751174479722976,
      "num_tokens": 6719412.0,
      "step": 3635
    },
    {
      "entropy": 5.277928113937378,
      "epoch": 3.0460251046025104,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004385310378490627,
      "loss": 4.7796,
      "mean_token_accuracy": 0.21473814100027083,
      "num_tokens": 6729169.0,
      "step": 3640
    },
    {
      "entropy": 5.305811214447021,
      "epoch": 3.0502092050209204,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043830917943314977,
      "loss": 4.7221,
      "mean_token_accuracy": 0.2207148939371109,
      "num_tokens": 6738549.0,
      "step": 3645
    },
    {
      "entropy": 5.262117481231689,
      "epoch": 3.0543933054393304,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043808698495320154,
      "loss": 4.7783,
      "mean_token_accuracy": 0.22020159810781478,
      "num_tokens": 6747909.0,
      "step": 3650
    },
    {
      "entropy": 5.2283553123474125,
      "epoch": 3.0585774058577404,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043786445486645846,
      "loss": 4.835,
      "mean_token_accuracy": 0.21614689081907273,
      "num_tokens": 6757125.0,
      "step": 3655
    },
    {
      "entropy": 5.379630041122437,
      "epoch": 3.062761506276151,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004376415896308518,
      "loss": 4.8186,
      "mean_token_accuracy": 0.22030971944332123,
      "num_tokens": 6765774.0,
      "step": 3660
    },
    {
      "entropy": 5.341221857070923,
      "epoch": 3.066945606694561,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043741838970500234,
      "loss": 4.874,
      "mean_token_accuracy": 0.21571923047304153,
      "num_tokens": 6774980.0,
      "step": 3665
    },
    {
      "entropy": 5.3157493591308596,
      "epoch": 3.071129707112971,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004371948555482196,
      "loss": 4.8663,
      "mean_token_accuracy": 0.21201944798231126,
      "num_tokens": 6784187.0,
      "step": 3670
    },
    {
      "entropy": 5.28399658203125,
      "epoch": 3.075313807531381,
      "grad_norm": 1.15625,
      "learning_rate": 0.000436970987620501,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21151311695575714,
      "num_tokens": 6792953.0,
      "step": 3675
    },
    {
      "entropy": 5.325482034683228,
      "epoch": 3.079497907949791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043674678638253066,
      "loss": 4.7583,
      "mean_token_accuracy": 0.22190648466348648,
      "num_tokens": 6803004.0,
      "step": 3680
    },
    {
      "entropy": 5.262072944641114,
      "epoch": 3.083682008368201,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004365222522956788,
      "loss": 4.8085,
      "mean_token_accuracy": 0.215910604596138,
      "num_tokens": 6811883.0,
      "step": 3685
    },
    {
      "entropy": 5.257794713973999,
      "epoch": 3.087866108786611,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004362973858220004,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2243594378232956,
      "num_tokens": 6820327.0,
      "step": 3690
    },
    {
      "entropy": 5.31516466140747,
      "epoch": 3.092050209205021,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043607218742423447,
      "loss": 4.8363,
      "mean_token_accuracy": 0.21631665378808976,
      "num_tokens": 6829412.0,
      "step": 3695
    },
    {
      "entropy": 5.25584397315979,
      "epoch": 3.096234309623431,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043584665756580323,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2158111184835434,
      "num_tokens": 6838997.0,
      "step": 3700
    },
    {
      "entropy": 5.361407375335693,
      "epoch": 3.100418410041841,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004356207967108106,
      "loss": 4.8709,
      "mean_token_accuracy": 0.20751748085021973,
      "num_tokens": 6848957.0,
      "step": 3705
    },
    {
      "entropy": 5.308306169509888,
      "epoch": 3.104602510460251,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004353946053240422,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21284400522708893,
      "num_tokens": 6859688.0,
      "step": 3710
    },
    {
      "entropy": 5.336264944076538,
      "epoch": 3.1087866108786613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004351680838709635,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2013743668794632,
      "num_tokens": 6869149.0,
      "step": 3715
    },
    {
      "entropy": 5.385293483734131,
      "epoch": 3.1129707112970713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043494123281771915,
      "loss": 4.909,
      "mean_token_accuracy": 0.20783170610666274,
      "num_tokens": 6879432.0,
      "step": 3720
    },
    {
      "entropy": 5.26263394355774,
      "epoch": 3.1171548117154813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043471405263113214,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21832842081785203,
      "num_tokens": 6888398.0,
      "step": 3725
    },
    {
      "entropy": 5.298207092285156,
      "epoch": 3.1213389121338913,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004344865437787027,
      "loss": 4.7927,
      "mean_token_accuracy": 0.21798287332057953,
      "num_tokens": 6898082.0,
      "step": 3730
    },
    {
      "entropy": 5.222813844680786,
      "epoch": 3.1255230125523012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043425870672860774,
      "loss": 4.7684,
      "mean_token_accuracy": 0.21932897120714187,
      "num_tokens": 6907693.0,
      "step": 3735
    },
    {
      "entropy": 5.389808130264282,
      "epoch": 3.1297071129707112,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004340305419496989,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20066292434930802,
      "num_tokens": 6917095.0,
      "step": 3740
    },
    {
      "entropy": 5.317314767837525,
      "epoch": 3.1338912133891212,
      "grad_norm": 1.125,
      "learning_rate": 0.0004338020499115029,
      "loss": 4.9058,
      "mean_token_accuracy": 0.20590936839580537,
      "num_tokens": 6926952.0,
      "step": 3745
    },
    {
      "entropy": 5.308876848220825,
      "epoch": 3.1380753138075312,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004335732310842196,
      "loss": 4.8358,
      "mean_token_accuracy": 0.21241551339626313,
      "num_tokens": 6936704.0,
      "step": 3750
    },
    {
      "entropy": 5.27081069946289,
      "epoch": 3.1422594142259412,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004333440859387212,
      "loss": 4.8492,
      "mean_token_accuracy": 0.21257602721452712,
      "num_tokens": 6946614.0,
      "step": 3755
    },
    {
      "entropy": 5.350625896453858,
      "epoch": 3.1464435146443517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043311461494655165,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2187810242176056,
      "num_tokens": 6955980.0,
      "step": 3760
    },
    {
      "entropy": 5.359692811965942,
      "epoch": 3.1506276150627617,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004328848185799254,
      "loss": 4.8457,
      "mean_token_accuracy": 0.21638285219669343,
      "num_tokens": 6964108.0,
      "step": 3765
    },
    {
      "entropy": 5.319613885879517,
      "epoch": 3.1548117154811717,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004326546973117264,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22057590186595916,
      "num_tokens": 6972895.0,
      "step": 3770
    },
    {
      "entropy": 5.270555591583252,
      "epoch": 3.1589958158995817,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043242425161550723,
      "loss": 4.858,
      "mean_token_accuracy": 0.2126601293683052,
      "num_tokens": 6982222.0,
      "step": 3775
    },
    {
      "entropy": 5.32768440246582,
      "epoch": 3.1631799163179917,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043219348196548807,
      "loss": 4.8495,
      "mean_token_accuracy": 0.21487165689468385,
      "num_tokens": 6990820.0,
      "step": 3780
    },
    {
      "entropy": 5.277839851379395,
      "epoch": 3.1673640167364017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004319623888365557,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21117942333221434,
      "num_tokens": 7000846.0,
      "step": 3785
    },
    {
      "entropy": 5.257175159454346,
      "epoch": 3.1715481171548117,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043173097270426275,
      "loss": 4.8065,
      "mean_token_accuracy": 0.21191611289978027,
      "num_tokens": 7009086.0,
      "step": 3790
    },
    {
      "entropy": 5.408314180374146,
      "epoch": 3.1757322175732217,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043149923404482634,
      "loss": 4.8774,
      "mean_token_accuracy": 0.21050115823745727,
      "num_tokens": 7017752.0,
      "step": 3795
    },
    {
      "entropy": 5.266625022888183,
      "epoch": 3.1799163179916317,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043126717333512733,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22228188961744308,
      "num_tokens": 7026535.0,
      "step": 3800
    },
    {
      "entropy": 5.29256157875061,
      "epoch": 3.1841004184100417,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004310347910527094,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21733937710523604,
      "num_tokens": 7035462.0,
      "step": 3805
    },
    {
      "entropy": 5.292649555206299,
      "epoch": 3.1882845188284517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004308020876757778,
      "loss": 4.8296,
      "mean_token_accuracy": 0.21759999841451644,
      "num_tokens": 7043978.0,
      "step": 3810
    },
    {
      "entropy": 5.2791407108306885,
      "epoch": 3.192468619246862,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000430569063683199,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22045563608407975,
      "num_tokens": 7053757.0,
      "step": 3815
    },
    {
      "entropy": 5.215965270996094,
      "epoch": 3.196652719665272,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004303357195544986,
      "loss": 4.7617,
      "mean_token_accuracy": 0.2122887060046196,
      "num_tokens": 7062491.0,
      "step": 3820
    },
    {
      "entropy": 5.24834885597229,
      "epoch": 3.200836820083682,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004301020557698615,
      "loss": 4.8122,
      "mean_token_accuracy": 0.21571405231952667,
      "num_tokens": 7072157.0,
      "step": 3825
    },
    {
      "entropy": 5.243205547332764,
      "epoch": 3.205020920502092,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042986807281013013,
      "loss": 4.771,
      "mean_token_accuracy": 0.21973754316568375,
      "num_tokens": 7080631.0,
      "step": 3830
    },
    {
      "entropy": 5.338223552703857,
      "epoch": 3.209205020920502,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004296337711568037,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2220018669962883,
      "num_tokens": 7089944.0,
      "step": 3835
    },
    {
      "entropy": 5.306629037857055,
      "epoch": 3.213389121338912,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042939915129203745,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21429578810930253,
      "num_tokens": 7099190.0,
      "step": 3840
    },
    {
      "entropy": 5.15996675491333,
      "epoch": 3.217573221757322,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004291642136986413,
      "loss": 4.8527,
      "mean_token_accuracy": 0.20626747012138366,
      "num_tokens": 7107944.0,
      "step": 3845
    },
    {
      "entropy": 5.237857961654663,
      "epoch": 3.221757322175732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000428928958860079,
      "loss": 4.7044,
      "mean_token_accuracy": 0.21445917338132858,
      "num_tokens": 7116953.0,
      "step": 3850
    },
    {
      "entropy": 5.225046586990357,
      "epoch": 3.225941422594142,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004286933872604673,
      "loss": 4.7621,
      "mean_token_accuracy": 0.22059133797883987,
      "num_tokens": 7126236.0,
      "step": 3855
    },
    {
      "entropy": 5.29739933013916,
      "epoch": 3.2301255230125525,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004284574993845746,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2139418512582779,
      "num_tokens": 7136286.0,
      "step": 3860
    },
    {
      "entropy": 5.284384250640869,
      "epoch": 3.2343096234309625,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042822129571782025,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2139872580766678,
      "num_tokens": 7145551.0,
      "step": 3865
    },
    {
      "entropy": 5.313933515548706,
      "epoch": 3.2384937238493725,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042798477674627343,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2191818654537201,
      "num_tokens": 7155486.0,
      "step": 3870
    },
    {
      "entropy": 5.2608051776885985,
      "epoch": 3.2426778242677825,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042774794295665213,
      "loss": 4.8475,
      "mean_token_accuracy": 0.21361929476261138,
      "num_tokens": 7163984.0,
      "step": 3875
    },
    {
      "entropy": 5.26825442314148,
      "epoch": 3.2468619246861925,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004275107948363223,
      "loss": 4.7978,
      "mean_token_accuracy": 0.21317970901727676,
      "num_tokens": 7172854.0,
      "step": 3880
    },
    {
      "entropy": 5.197178030014038,
      "epoch": 3.2510460251046025,
      "grad_norm": 1.25,
      "learning_rate": 0.0004272733328732965,
      "loss": 4.8196,
      "mean_token_accuracy": 0.20986296534538268,
      "num_tokens": 7182202.0,
      "step": 3885
    },
    {
      "entropy": 5.286785459518432,
      "epoch": 3.2552301255230125,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042703555755623337,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2052895739674568,
      "num_tokens": 7192062.0,
      "step": 3890
    },
    {
      "entropy": 5.4119648933410645,
      "epoch": 3.2594142259414225,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004267974693744362,
      "loss": 4.9313,
      "mean_token_accuracy": 0.20530452877283095,
      "num_tokens": 7201766.0,
      "step": 3895
    },
    {
      "entropy": 5.283129405975342,
      "epoch": 3.2635983263598325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004265590688178523,
      "loss": 4.8391,
      "mean_token_accuracy": 0.20876989662647247,
      "num_tokens": 7211361.0,
      "step": 3900
    },
    {
      "entropy": 5.275751447677612,
      "epoch": 3.2677824267782425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004263203563770716,
      "loss": 4.8111,
      "mean_token_accuracy": 0.21774289458990098,
      "num_tokens": 7220221.0,
      "step": 3905
    },
    {
      "entropy": 5.2844874382019045,
      "epoch": 3.2719665271966525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000426081332543326,
      "loss": 4.9261,
      "mean_token_accuracy": 0.20752736777067185,
      "num_tokens": 7230134.0,
      "step": 3910
    },
    {
      "entropy": 5.192396020889282,
      "epoch": 3.276150627615063,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004258419978084881,
      "loss": 4.7405,
      "mean_token_accuracy": 0.22151417583227156,
      "num_tokens": 7238713.0,
      "step": 3915
    },
    {
      "entropy": 5.34380407333374,
      "epoch": 3.280334728033473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004256023526650702,
      "loss": 4.8633,
      "mean_token_accuracy": 0.21645605862140654,
      "num_tokens": 7247603.0,
      "step": 3920
    },
    {
      "entropy": 5.252472591400147,
      "epoch": 3.284518828451883,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004253623976062236,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21033538281917571,
      "num_tokens": 7257452.0,
      "step": 3925
    },
    {
      "entropy": 5.377154922485351,
      "epoch": 3.288702928870293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004251221331257372,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21271614581346512,
      "num_tokens": 7266401.0,
      "step": 3930
    },
    {
      "entropy": 5.290256547927856,
      "epoch": 3.292887029288703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004248815597180366,
      "loss": 4.841,
      "mean_token_accuracy": 0.2185330480337143,
      "num_tokens": 7275836.0,
      "step": 3935
    },
    {
      "entropy": 5.27263913154602,
      "epoch": 3.297071129707113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042464067787818333,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2124941200017929,
      "num_tokens": 7284854.0,
      "step": 3940
    },
    {
      "entropy": 5.299465465545654,
      "epoch": 3.301255230125523,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004243994881018735,
      "loss": 4.8587,
      "mean_token_accuracy": 0.21343783140182496,
      "num_tokens": 7293421.0,
      "step": 3945
    },
    {
      "entropy": 5.272430753707885,
      "epoch": 3.305439330543933,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004241579908854367,
      "loss": 4.8021,
      "mean_token_accuracy": 0.2127429038286209,
      "num_tokens": 7302184.0,
      "step": 3950
    },
    {
      "entropy": 5.383147096633911,
      "epoch": 3.309623430962343,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042391618672583546,
      "loss": 4.9469,
      "mean_token_accuracy": 0.209491965174675,
      "num_tokens": 7311596.0,
      "step": 3955
    },
    {
      "entropy": 5.273970890045166,
      "epoch": 3.3138075313807533,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042367407612066383,
      "loss": 4.7863,
      "mean_token_accuracy": 0.21876903474330903,
      "num_tokens": 7319866.0,
      "step": 3960
    },
    {
      "entropy": 5.2668585777282715,
      "epoch": 3.3179916317991633,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004234316595681468,
      "loss": 4.8559,
      "mean_token_accuracy": 0.21466988176107407,
      "num_tokens": 7329327.0,
      "step": 3965
    },
    {
      "entropy": 5.3854817867279055,
      "epoch": 3.3221757322175733,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004231889375671382,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20021829903125762,
      "num_tokens": 7339717.0,
      "step": 3970
    },
    {
      "entropy": 5.314059782028198,
      "epoch": 3.3263598326359833,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042294591061712124,
      "loss": 4.8441,
      "mean_token_accuracy": 0.21653733253479004,
      "num_tokens": 7348850.0,
      "step": 3975
    },
    {
      "entropy": 5.2205463409423825,
      "epoch": 3.3305439330543933,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042270257921820623,
      "loss": 4.789,
      "mean_token_accuracy": 0.2239212676882744,
      "num_tokens": 7357081.0,
      "step": 3980
    },
    {
      "entropy": 5.1775803565979,
      "epoch": 3.3347280334728033,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042245894387112987,
      "loss": 4.8069,
      "mean_token_accuracy": 0.2220182329416275,
      "num_tokens": 7365762.0,
      "step": 3985
    },
    {
      "entropy": 5.281820964813233,
      "epoch": 3.3389121338912133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004222150050772548,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2108008921146393,
      "num_tokens": 7374870.0,
      "step": 3990
    },
    {
      "entropy": 5.243097877502441,
      "epoch": 3.3430962343096233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004219707633385677,
      "loss": 4.7353,
      "mean_token_accuracy": 0.2225353464484215,
      "num_tokens": 7383836.0,
      "step": 3995
    },
    {
      "entropy": 5.2761904239654545,
      "epoch": 3.3472803347280333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042172621915767886,
      "loss": 4.8533,
      "mean_token_accuracy": 0.21824376583099364,
      "num_tokens": 7394059.0,
      "step": 4000
    },
    {
      "epoch": 3.3472803347280333,
      "eval_entropy": 5.010007921659629,
      "eval_loss": 5.385568141937256,
      "eval_mean_token_accuracy": 0.19328919281623114,
      "eval_num_tokens": 7394059.0,
      "eval_runtime": 2.2664,
      "eval_samples_per_second": 1723.889,
      "eval_steps_per_second": 215.762,
      "step": 4000
    }
  ],
  "logging_steps": 5,
  "max_steps": 11950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1595941224284160.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}