{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.07820617381880747,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.692034435272216,
      "epoch": 0.00011172310545543924,
      "grad_norm": 16.25,
      "learning_rate": 2e-06,
      "loss": 10.6039,
      "mean_token_accuracy": 0.0,
      "num_tokens": 4250.0,
      "step": 5
    },
    {
      "entropy": 10.692024230957031,
      "epoch": 0.00022344621091087847,
      "grad_norm": 17.125,
      "learning_rate": 4.5e-06,
      "loss": 10.5084,
      "mean_token_accuracy": 0.0,
      "num_tokens": 8228.0,
      "step": 10
    },
    {
      "entropy": 10.691641712188721,
      "epoch": 0.0003351693163663177,
      "grad_norm": 13.125,
      "learning_rate": 7e-06,
      "loss": 10.2399,
      "mean_token_accuracy": 0.01585084570106119,
      "num_tokens": 12209.0,
      "step": 15
    },
    {
      "entropy": 10.6864333152771,
      "epoch": 0.00044689242182175694,
      "grad_norm": 8.875,
      "learning_rate": 9.5e-06,
      "loss": 9.9231,
      "mean_token_accuracy": 0.04045895729213953,
      "num_tokens": 16225.0,
      "step": 20
    },
    {
      "entropy": 10.658738136291504,
      "epoch": 0.0005586155272771962,
      "grad_norm": 5.71875,
      "learning_rate": 1.2e-05,
      "loss": 9.6193,
      "mean_token_accuracy": 0.05384383201599121,
      "num_tokens": 20212.0,
      "step": 25
    },
    {
      "entropy": 10.617362594604492,
      "epoch": 0.0006703386327326354,
      "grad_norm": 4.46875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 9.3987,
      "mean_token_accuracy": 0.06703231446444988,
      "num_tokens": 24445.0,
      "step": 30
    },
    {
      "entropy": 10.580356121063232,
      "epoch": 0.0007820617381880747,
      "grad_norm": 4.09375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 9.2682,
      "mean_token_accuracy": 0.06737326122820378,
      "num_tokens": 28365.0,
      "step": 35
    },
    {
      "entropy": 10.578466510772705,
      "epoch": 0.0008937848436435139,
      "grad_norm": 3.9375,
      "learning_rate": 1.95e-05,
      "loss": 9.3685,
      "mean_token_accuracy": 0.05515027567744255,
      "num_tokens": 33055.0,
      "step": 40
    },
    {
      "entropy": 10.575786018371582,
      "epoch": 0.0010055079490989532,
      "grad_norm": 3.5625,
      "learning_rate": 2.2e-05,
      "loss": 9.2537,
      "mean_token_accuracy": 0.0627420760691166,
      "num_tokens": 37599.0,
      "step": 45
    },
    {
      "entropy": 10.472075748443604,
      "epoch": 0.0011172310545543925,
      "grad_norm": 3.5,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.1143,
      "mean_token_accuracy": 0.07111909165978432,
      "num_tokens": 41934.0,
      "step": 50
    },
    {
      "entropy": 10.315015411376953,
      "epoch": 0.0012289541600098315,
      "grad_norm": 3.21875,
      "learning_rate": 2.7e-05,
      "loss": 9.0802,
      "mean_token_accuracy": 0.06700236462056637,
      "num_tokens": 46178.0,
      "step": 55
    },
    {
      "entropy": 10.27945785522461,
      "epoch": 0.0013406772654652708,
      "grad_norm": 3.546875,
      "learning_rate": 2.95e-05,
      "loss": 8.9726,
      "mean_token_accuracy": 0.0730497345328331,
      "num_tokens": 50513.0,
      "step": 60
    },
    {
      "entropy": 10.237419605255127,
      "epoch": 0.0014524003709207101,
      "grad_norm": 2.859375,
      "learning_rate": 3.2e-05,
      "loss": 8.9616,
      "mean_token_accuracy": 0.07224201261997223,
      "num_tokens": 54924.0,
      "step": 65
    },
    {
      "entropy": 10.185122776031495,
      "epoch": 0.0015641234763761494,
      "grad_norm": 3.125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 8.8671,
      "mean_token_accuracy": 0.06654836386442184,
      "num_tokens": 59083.0,
      "step": 70
    },
    {
      "entropy": 10.172991943359374,
      "epoch": 0.0016758465818315885,
      "grad_norm": 3.9375,
      "learning_rate": 3.7e-05,
      "loss": 8.7827,
      "mean_token_accuracy": 0.06427431814372539,
      "num_tokens": 63324.0,
      "step": 75
    },
    {
      "entropy": 10.14193344116211,
      "epoch": 0.0017875696872870278,
      "grad_norm": 2.640625,
      "learning_rate": 3.95e-05,
      "loss": 8.8298,
      "mean_token_accuracy": 0.07168133780360222,
      "num_tokens": 67738.0,
      "step": 80
    },
    {
      "entropy": 10.173482418060303,
      "epoch": 0.001899292792742467,
      "grad_norm": 2.359375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 8.7488,
      "mean_token_accuracy": 0.0716201275587082,
      "num_tokens": 72305.0,
      "step": 85
    },
    {
      "entropy": 10.12889585494995,
      "epoch": 0.0020110158981979064,
      "grad_norm": 3.046875,
      "learning_rate": 4.45e-05,
      "loss": 8.5249,
      "mean_token_accuracy": 0.0720902569591999,
      "num_tokens": 76579.0,
      "step": 90
    },
    {
      "entropy": 10.03916826248169,
      "epoch": 0.0021227390036533456,
      "grad_norm": 2.203125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.5006,
      "mean_token_accuracy": 0.07380848824977874,
      "num_tokens": 80812.0,
      "step": 95
    },
    {
      "entropy": 9.937397384643555,
      "epoch": 0.002234462109108785,
      "grad_norm": 1.953125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.4135,
      "mean_token_accuracy": 0.07211422435939312,
      "num_tokens": 85090.0,
      "step": 100
    },
    {
      "entropy": 9.917206573486329,
      "epoch": 0.0023461852145642242,
      "grad_norm": 2.53125,
      "learning_rate": 5.2e-05,
      "loss": 8.344,
      "mean_token_accuracy": 0.07151557318866253,
      "num_tokens": 89578.0,
      "step": 105
    },
    {
      "entropy": 9.800644588470458,
      "epoch": 0.002457908320019663,
      "grad_norm": 1.9140625,
      "learning_rate": 5.45e-05,
      "loss": 8.2336,
      "mean_token_accuracy": 0.07592308931052685,
      "num_tokens": 94117.0,
      "step": 110
    },
    {
      "entropy": 9.670217800140382,
      "epoch": 0.0025696314254751024,
      "grad_norm": 2.1875,
      "learning_rate": 5.7e-05,
      "loss": 8.0337,
      "mean_token_accuracy": 0.08074008114635944,
      "num_tokens": 98082.0,
      "step": 115
    },
    {
      "entropy": 9.533781147003173,
      "epoch": 0.0026813545309305417,
      "grad_norm": 1.875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.0226,
      "mean_token_accuracy": 0.07543974481523037,
      "num_tokens": 102327.0,
      "step": 120
    },
    {
      "entropy": 9.384562969207764,
      "epoch": 0.002793077636385981,
      "grad_norm": 1.9296875,
      "learning_rate": 6.2e-05,
      "loss": 7.8842,
      "mean_token_accuracy": 0.07998752668499946,
      "num_tokens": 106567.0,
      "step": 125
    },
    {
      "entropy": 9.304523754119874,
      "epoch": 0.0029048007418414202,
      "grad_norm": 1.9140625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.7414,
      "mean_token_accuracy": 0.07796499505639076,
      "num_tokens": 110654.0,
      "step": 130
    },
    {
      "entropy": 9.14549970626831,
      "epoch": 0.0030165238472968595,
      "grad_norm": 2.140625,
      "learning_rate": 6.7e-05,
      "loss": 7.7396,
      "mean_token_accuracy": 0.0759395856410265,
      "num_tokens": 114679.0,
      "step": 135
    },
    {
      "entropy": 8.918598842620849,
      "epoch": 0.003128246952752299,
      "grad_norm": 1.7578125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.7208,
      "mean_token_accuracy": 0.0734277281910181,
      "num_tokens": 118817.0,
      "step": 140
    },
    {
      "entropy": 8.880987071990967,
      "epoch": 0.003239970058207738,
      "grad_norm": 1.640625,
      "learning_rate": 7.2e-05,
      "loss": 7.63,
      "mean_token_accuracy": 0.08075562492012978,
      "num_tokens": 123188.0,
      "step": 145
    },
    {
      "entropy": 8.665369319915772,
      "epoch": 0.003351693163663177,
      "grad_norm": 1.96875,
      "learning_rate": 7.45e-05,
      "loss": 7.571,
      "mean_token_accuracy": 0.0803100574761629,
      "num_tokens": 127767.0,
      "step": 150
    },
    {
      "entropy": 8.561956405639648,
      "epoch": 0.0034634162691186163,
      "grad_norm": 1.8125,
      "learning_rate": 7.7e-05,
      "loss": 7.3994,
      "mean_token_accuracy": 0.0842840738594532,
      "num_tokens": 131837.0,
      "step": 155
    },
    {
      "entropy": 8.48619909286499,
      "epoch": 0.0035751393745740555,
      "grad_norm": 1.515625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.4235,
      "mean_token_accuracy": 0.08158716633915901,
      "num_tokens": 136247.0,
      "step": 160
    },
    {
      "entropy": 8.290351867675781,
      "epoch": 0.003686862480029495,
      "grad_norm": 2.203125,
      "learning_rate": 8.2e-05,
      "loss": 7.264,
      "mean_token_accuracy": 0.0918136142194271,
      "num_tokens": 140170.0,
      "step": 165
    },
    {
      "entropy": 8.1635573387146,
      "epoch": 0.003798585585484934,
      "grad_norm": 1.546875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.3583,
      "mean_token_accuracy": 0.09022970125079155,
      "num_tokens": 144139.0,
      "step": 170
    },
    {
      "entropy": 8.112592220306396,
      "epoch": 0.003910308690940373,
      "grad_norm": 1.5859375,
      "learning_rate": 8.7e-05,
      "loss": 7.3255,
      "mean_token_accuracy": 0.08751063793897629,
      "num_tokens": 148792.0,
      "step": 175
    },
    {
      "entropy": 8.065679502487182,
      "epoch": 0.004022031796395813,
      "grad_norm": 2.1875,
      "learning_rate": 8.95e-05,
      "loss": 7.2901,
      "mean_token_accuracy": 0.08218263126909733,
      "num_tokens": 152844.0,
      "step": 180
    },
    {
      "entropy": 8.083935976028442,
      "epoch": 0.0041337549018512516,
      "grad_norm": 1.7265625,
      "learning_rate": 9.2e-05,
      "loss": 7.3158,
      "mean_token_accuracy": 0.08304233588278294,
      "num_tokens": 157366.0,
      "step": 185
    },
    {
      "entropy": 7.946479368209839,
      "epoch": 0.004245478007306691,
      "grad_norm": 2.015625,
      "learning_rate": 9.45e-05,
      "loss": 7.1574,
      "mean_token_accuracy": 0.093728469312191,
      "num_tokens": 161348.0,
      "step": 190
    },
    {
      "entropy": 7.886520957946777,
      "epoch": 0.00435720111276213,
      "grad_norm": 1.8046875,
      "learning_rate": 9.7e-05,
      "loss": 7.1876,
      "mean_token_accuracy": 0.10094638615846634,
      "num_tokens": 165647.0,
      "step": 195
    },
    {
      "entropy": 7.8108118057250975,
      "epoch": 0.00446892421821757,
      "grad_norm": 1.8515625,
      "learning_rate": 9.95e-05,
      "loss": 7.1137,
      "mean_token_accuracy": 0.09259389042854309,
      "num_tokens": 169521.0,
      "step": 200
    },
    {
      "entropy": 7.7959239959716795,
      "epoch": 0.004580647323673009,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000102,
      "loss": 7.1787,
      "mean_token_accuracy": 0.09544989094138145,
      "num_tokens": 173466.0,
      "step": 205
    },
    {
      "entropy": 7.768160486221314,
      "epoch": 0.0046923704291284484,
      "grad_norm": 1.734375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.1178,
      "mean_token_accuracy": 0.08695701211690902,
      "num_tokens": 177663.0,
      "step": 210
    },
    {
      "entropy": 7.66025218963623,
      "epoch": 0.004804093534583887,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000107,
      "loss": 7.1533,
      "mean_token_accuracy": 0.08611816093325615,
      "num_tokens": 181778.0,
      "step": 215
    },
    {
      "entropy": 7.759128427505493,
      "epoch": 0.004915816640039326,
      "grad_norm": 2.046875,
      "learning_rate": 0.0001095,
      "loss": 7.1497,
      "mean_token_accuracy": 0.09380927979946137,
      "num_tokens": 185525.0,
      "step": 220
    },
    {
      "entropy": 7.712249279022217,
      "epoch": 0.005027539745494766,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000112,
      "loss": 7.0063,
      "mean_token_accuracy": 0.10432608500123024,
      "num_tokens": 189418.0,
      "step": 225
    },
    {
      "entropy": 7.597976541519165,
      "epoch": 0.005139262850950205,
      "grad_norm": 2.0625,
      "learning_rate": 0.0001145,
      "loss": 7.0884,
      "mean_token_accuracy": 0.10287595391273499,
      "num_tokens": 193494.0,
      "step": 230
    },
    {
      "entropy": 7.700113201141358,
      "epoch": 0.0052509859564056445,
      "grad_norm": 1.53125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1323,
      "mean_token_accuracy": 0.09622926861047745,
      "num_tokens": 198018.0,
      "step": 235
    },
    {
      "entropy": 7.695586109161377,
      "epoch": 0.005362709061861083,
      "grad_norm": 1.78125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.0591,
      "mean_token_accuracy": 0.09503638371825218,
      "num_tokens": 202296.0,
      "step": 240
    },
    {
      "entropy": 7.5984455108642575,
      "epoch": 0.005474432167316523,
      "grad_norm": 1.515625,
      "learning_rate": 0.000122,
      "loss": 7.0644,
      "mean_token_accuracy": 0.09454682543873787,
      "num_tokens": 206694.0,
      "step": 245
    },
    {
      "entropy": 7.6198304176330565,
      "epoch": 0.005586155272771962,
      "grad_norm": 2.03125,
      "learning_rate": 0.0001245,
      "loss": 7.0863,
      "mean_token_accuracy": 0.10327245965600014,
      "num_tokens": 210810.0,
      "step": 250
    },
    {
      "entropy": 7.671643877029419,
      "epoch": 0.005697878378227402,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000127,
      "loss": 7.0545,
      "mean_token_accuracy": 0.09972267299890518,
      "num_tokens": 215044.0,
      "step": 255
    },
    {
      "entropy": 7.580405235290527,
      "epoch": 0.0058096014836828405,
      "grad_norm": 1.734375,
      "learning_rate": 0.0001295,
      "loss": 7.1836,
      "mean_token_accuracy": 0.09796730950474739,
      "num_tokens": 219235.0,
      "step": 260
    },
    {
      "entropy": 7.618637609481811,
      "epoch": 0.005921324589138279,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000132,
      "loss": 6.9848,
      "mean_token_accuracy": 0.10349723398685455,
      "num_tokens": 223639.0,
      "step": 265
    },
    {
      "entropy": 7.4419373035430905,
      "epoch": 0.006033047694593719,
      "grad_norm": 1.75,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0238,
      "mean_token_accuracy": 0.10185197666287422,
      "num_tokens": 227873.0,
      "step": 270
    },
    {
      "entropy": 7.483202838897705,
      "epoch": 0.006144770800049158,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.9487,
      "mean_token_accuracy": 0.11078577488660812,
      "num_tokens": 232147.0,
      "step": 275
    },
    {
      "entropy": 7.483991575241089,
      "epoch": 0.006256493905504598,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0001395,
      "loss": 7.0367,
      "mean_token_accuracy": 0.1014625109732151,
      "num_tokens": 236456.0,
      "step": 280
    },
    {
      "entropy": 7.457619524002075,
      "epoch": 0.0063682170109600365,
      "grad_norm": 1.46875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0498,
      "mean_token_accuracy": 0.10427944287657738,
      "num_tokens": 241039.0,
      "step": 285
    },
    {
      "entropy": 7.737891387939453,
      "epoch": 0.006479940116415476,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0001445,
      "loss": 7.097,
      "mean_token_accuracy": 0.09794913604855537,
      "num_tokens": 245132.0,
      "step": 290
    },
    {
      "entropy": 7.422194004058838,
      "epoch": 0.006591663221870915,
      "grad_norm": 1.71875,
      "learning_rate": 0.000147,
      "loss": 7.0166,
      "mean_token_accuracy": 0.1079220362007618,
      "num_tokens": 249152.0,
      "step": 295
    },
    {
      "entropy": 7.482765865325928,
      "epoch": 0.006703386327326354,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0001495,
      "loss": 6.9395,
      "mean_token_accuracy": 0.10692858770489692,
      "num_tokens": 253439.0,
      "step": 300
    },
    {
      "entropy": 7.519821882247925,
      "epoch": 0.006815109432781794,
      "grad_norm": 1.78125,
      "learning_rate": 0.000152,
      "loss": 6.9877,
      "mean_token_accuracy": 0.10241441354155541,
      "num_tokens": 258066.0,
      "step": 305
    },
    {
      "entropy": 7.4823202133178714,
      "epoch": 0.0069268325382372325,
      "grad_norm": 1.890625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9284,
      "mean_token_accuracy": 0.1084162712097168,
      "num_tokens": 261954.0,
      "step": 310
    },
    {
      "entropy": 7.320609712600708,
      "epoch": 0.007038555643692672,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000157,
      "loss": 6.9852,
      "mean_token_accuracy": 0.10961857140064239,
      "num_tokens": 266650.0,
      "step": 315
    },
    {
      "entropy": 7.539988470077515,
      "epoch": 0.007150278749148111,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0001595,
      "loss": 7.0467,
      "mean_token_accuracy": 0.10235239341855049,
      "num_tokens": 271069.0,
      "step": 320
    },
    {
      "entropy": 7.4312175750732425,
      "epoch": 0.007262001854603551,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000162,
      "loss": 6.9195,
      "mean_token_accuracy": 0.10588012635707855,
      "num_tokens": 275084.0,
      "step": 325
    },
    {
      "entropy": 7.405316781997681,
      "epoch": 0.00737372496005899,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.9708,
      "mean_token_accuracy": 0.11100029796361924,
      "num_tokens": 279721.0,
      "step": 330
    },
    {
      "entropy": 7.45182785987854,
      "epoch": 0.007485448065514429,
      "grad_norm": 1.546875,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9745,
      "mean_token_accuracy": 0.11023727878928184,
      "num_tokens": 284317.0,
      "step": 335
    },
    {
      "entropy": 7.376049518585205,
      "epoch": 0.007597171170969868,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9627,
      "mean_token_accuracy": 0.11432994231581688,
      "num_tokens": 288870.0,
      "step": 340
    },
    {
      "entropy": 7.399447011947632,
      "epoch": 0.007708894276425307,
      "grad_norm": 1.65625,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8883,
      "mean_token_accuracy": 0.11889545023441314,
      "num_tokens": 292996.0,
      "step": 345
    },
    {
      "entropy": 7.417701625823975,
      "epoch": 0.007820617381880746,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.8748,
      "mean_token_accuracy": 0.11174583658576012,
      "num_tokens": 297238.0,
      "step": 350
    },
    {
      "entropy": 7.387735509872437,
      "epoch": 0.007932340487336187,
      "grad_norm": 1.890625,
      "learning_rate": 0.000177,
      "loss": 6.8957,
      "mean_token_accuracy": 0.1190306767821312,
      "num_tokens": 301453.0,
      "step": 355
    },
    {
      "entropy": 7.298361539840698,
      "epoch": 0.008044063592791625,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0001795,
      "loss": 6.8362,
      "mean_token_accuracy": 0.11696951016783715,
      "num_tokens": 305949.0,
      "step": 360
    },
    {
      "entropy": 7.339731693267822,
      "epoch": 0.008155786698247064,
      "grad_norm": 1.734375,
      "learning_rate": 0.000182,
      "loss": 6.8502,
      "mean_token_accuracy": 0.12272856459021568,
      "num_tokens": 310097.0,
      "step": 365
    },
    {
      "entropy": 7.378586864471435,
      "epoch": 0.008267509803702503,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0001845,
      "loss": 6.8746,
      "mean_token_accuracy": 0.1073436789214611,
      "num_tokens": 314567.0,
      "step": 370
    },
    {
      "entropy": 7.258294773101807,
      "epoch": 0.008379232909157944,
      "grad_norm": 1.828125,
      "learning_rate": 0.000187,
      "loss": 6.9539,
      "mean_token_accuracy": 0.1148001030087471,
      "num_tokens": 319166.0,
      "step": 375
    },
    {
      "entropy": 7.30250129699707,
      "epoch": 0.008490956014613383,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0001895,
      "loss": 6.8336,
      "mean_token_accuracy": 0.11580941900610923,
      "num_tokens": 323682.0,
      "step": 380
    },
    {
      "entropy": 7.259005832672119,
      "epoch": 0.008602679120068821,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000192,
      "loss": 6.7239,
      "mean_token_accuracy": 0.12464385628700256,
      "num_tokens": 327994.0,
      "step": 385
    },
    {
      "entropy": 7.200961637496948,
      "epoch": 0.00871440222552426,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0001945,
      "loss": 6.8242,
      "mean_token_accuracy": 0.11740145906805992,
      "num_tokens": 332026.0,
      "step": 390
    },
    {
      "entropy": 7.481966304779053,
      "epoch": 0.0088261253309797,
      "grad_norm": 3.078125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8603,
      "mean_token_accuracy": 0.12176617532968521,
      "num_tokens": 336552.0,
      "step": 395
    },
    {
      "entropy": 7.178363084793091,
      "epoch": 0.00893784843643514,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8005,
      "mean_token_accuracy": 0.11857323870062828,
      "num_tokens": 340643.0,
      "step": 400
    },
    {
      "entropy": 7.28194899559021,
      "epoch": 0.009049571541890579,
      "grad_norm": 1.65625,
      "learning_rate": 0.000202,
      "loss": 6.7174,
      "mean_token_accuracy": 0.12609147280454636,
      "num_tokens": 344886.0,
      "step": 405
    },
    {
      "entropy": 7.182888126373291,
      "epoch": 0.009161294647346017,
      "grad_norm": 1.8125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7831,
      "mean_token_accuracy": 0.12051703482866287,
      "num_tokens": 349115.0,
      "step": 410
    },
    {
      "entropy": 7.436312055587768,
      "epoch": 0.009273017752801456,
      "grad_norm": 1.71875,
      "learning_rate": 0.000207,
      "loss": 6.7448,
      "mean_token_accuracy": 0.12795047089457512,
      "num_tokens": 353368.0,
      "step": 415
    },
    {
      "entropy": 7.033882093429566,
      "epoch": 0.009384740858256897,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002095,
      "loss": 6.7951,
      "mean_token_accuracy": 0.11617444008588791,
      "num_tokens": 357382.0,
      "step": 420
    },
    {
      "entropy": 7.290721368789673,
      "epoch": 0.009496463963712336,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000212,
      "loss": 6.7152,
      "mean_token_accuracy": 0.12455029636621476,
      "num_tokens": 361542.0,
      "step": 425
    },
    {
      "entropy": 7.23753170967102,
      "epoch": 0.009608187069167775,
      "grad_norm": 1.5,
      "learning_rate": 0.0002145,
      "loss": 6.7497,
      "mean_token_accuracy": 0.12337451875209808,
      "num_tokens": 366006.0,
      "step": 430
    },
    {
      "entropy": 7.083537578582764,
      "epoch": 0.009719910174623213,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6902,
      "mean_token_accuracy": 0.1326639696955681,
      "num_tokens": 370021.0,
      "step": 435
    },
    {
      "entropy": 7.248496341705322,
      "epoch": 0.009831633280078652,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002195,
      "loss": 6.805,
      "mean_token_accuracy": 0.12277387753129006,
      "num_tokens": 374434.0,
      "step": 440
    },
    {
      "entropy": 7.251986265182495,
      "epoch": 0.009943356385534093,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000222,
      "loss": 6.8176,
      "mean_token_accuracy": 0.11638971790671349,
      "num_tokens": 378634.0,
      "step": 445
    },
    {
      "entropy": 7.187275457382202,
      "epoch": 0.010055079490989532,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002245,
      "loss": 6.6952,
      "mean_token_accuracy": 0.12703127264976502,
      "num_tokens": 382904.0,
      "step": 450
    },
    {
      "entropy": 7.244022417068481,
      "epoch": 0.01016680259644497,
      "grad_norm": 1.65625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6803,
      "mean_token_accuracy": 0.12515532448887826,
      "num_tokens": 386970.0,
      "step": 455
    },
    {
      "entropy": 7.082424831390381,
      "epoch": 0.01027852570190041,
      "grad_norm": 1.59375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6572,
      "mean_token_accuracy": 0.13228979632258414,
      "num_tokens": 391043.0,
      "step": 460
    },
    {
      "entropy": 7.1885302543640135,
      "epoch": 0.01039024880735585,
      "grad_norm": 1.390625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.6428,
      "mean_token_accuracy": 0.13703735694289207,
      "num_tokens": 395413.0,
      "step": 465
    },
    {
      "entropy": 7.166730117797852,
      "epoch": 0.010501971912811289,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7767,
      "mean_token_accuracy": 0.12424237206578255,
      "num_tokens": 399821.0,
      "step": 470
    },
    {
      "entropy": 7.0755476474761965,
      "epoch": 0.010613695018266728,
      "grad_norm": 1.59375,
      "learning_rate": 0.000237,
      "loss": 6.5815,
      "mean_token_accuracy": 0.13624327406287193,
      "num_tokens": 404043.0,
      "step": 475
    },
    {
      "entropy": 7.202919864654541,
      "epoch": 0.010725418123722167,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0002395,
      "loss": 6.7277,
      "mean_token_accuracy": 0.1173846572637558,
      "num_tokens": 408339.0,
      "step": 480
    },
    {
      "entropy": 7.004957008361816,
      "epoch": 0.010837141229177606,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000242,
      "loss": 6.6233,
      "mean_token_accuracy": 0.1330234356224537,
      "num_tokens": 412384.0,
      "step": 485
    },
    {
      "entropy": 7.170296764373779,
      "epoch": 0.010948864334633046,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002445,
      "loss": 6.7419,
      "mean_token_accuracy": 0.122147136926651,
      "num_tokens": 416891.0,
      "step": 490
    },
    {
      "entropy": 7.086709833145141,
      "epoch": 0.011060587440088485,
      "grad_norm": 1.6875,
      "learning_rate": 0.000247,
      "loss": 6.6048,
      "mean_token_accuracy": 0.13360075876116753,
      "num_tokens": 420980.0,
      "step": 495
    },
    {
      "entropy": 6.981275844573974,
      "epoch": 0.011172310545543924,
      "grad_norm": 1.921875,
      "learning_rate": 0.0002495,
      "loss": 6.6585,
      "mean_token_accuracy": 0.12793864756822587,
      "num_tokens": 425454.0,
      "step": 500
    },
    {
      "entropy": 7.063224363327026,
      "epoch": 0.011284033650999363,
      "grad_norm": 1.5,
      "learning_rate": 0.000252,
      "loss": 6.6792,
      "mean_token_accuracy": 0.12707910239696502,
      "num_tokens": 430087.0,
      "step": 505
    },
    {
      "entropy": 7.057527685165406,
      "epoch": 0.011395756756454803,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002545,
      "loss": 6.6015,
      "mean_token_accuracy": 0.13783879056572915,
      "num_tokens": 434062.0,
      "step": 510
    },
    {
      "entropy": 6.918734359741211,
      "epoch": 0.011507479861910242,
      "grad_norm": 1.6875,
      "learning_rate": 0.000257,
      "loss": 6.5727,
      "mean_token_accuracy": 0.1411380872130394,
      "num_tokens": 437882.0,
      "step": 515
    },
    {
      "entropy": 7.062899255752564,
      "epoch": 0.011619202967365681,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002595,
      "loss": 6.6023,
      "mean_token_accuracy": 0.13055641129612922,
      "num_tokens": 442248.0,
      "step": 520
    },
    {
      "entropy": 7.028418397903442,
      "epoch": 0.01173092607282112,
      "grad_norm": 1.84375,
      "learning_rate": 0.000262,
      "loss": 6.5713,
      "mean_token_accuracy": 0.13718463256955146,
      "num_tokens": 446492.0,
      "step": 525
    },
    {
      "entropy": 6.917926502227783,
      "epoch": 0.011842649178276559,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6019,
      "mean_token_accuracy": 0.13765231221914292,
      "num_tokens": 450359.0,
      "step": 530
    },
    {
      "entropy": 7.029662656784057,
      "epoch": 0.011954372283732,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6264,
      "mean_token_accuracy": 0.12923550009727477,
      "num_tokens": 454508.0,
      "step": 535
    },
    {
      "entropy": 7.076862907409668,
      "epoch": 0.012066095389187438,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6009,
      "mean_token_accuracy": 0.13477242439985276,
      "num_tokens": 459028.0,
      "step": 540
    },
    {
      "entropy": 6.926972484588623,
      "epoch": 0.012177818494642877,
      "grad_norm": 1.703125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5556,
      "mean_token_accuracy": 0.1286642447113991,
      "num_tokens": 462911.0,
      "step": 545
    },
    {
      "entropy": 6.978501319885254,
      "epoch": 0.012289541600098316,
      "grad_norm": 1.625,
      "learning_rate": 0.0002745,
      "loss": 6.5891,
      "mean_token_accuracy": 0.13105733469128608,
      "num_tokens": 467129.0,
      "step": 550
    },
    {
      "entropy": 7.02958312034607,
      "epoch": 0.012401264705553756,
      "grad_norm": 1.6875,
      "learning_rate": 0.000277,
      "loss": 6.4952,
      "mean_token_accuracy": 0.13785838335752487,
      "num_tokens": 471324.0,
      "step": 555
    },
    {
      "entropy": 6.861265659332275,
      "epoch": 0.012512987811009195,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002795,
      "loss": 6.6661,
      "mean_token_accuracy": 0.1318783350288868,
      "num_tokens": 475761.0,
      "step": 560
    },
    {
      "entropy": 6.988262891769409,
      "epoch": 0.012624710916464634,
      "grad_norm": 1.515625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.4608,
      "mean_token_accuracy": 0.1332152046263218,
      "num_tokens": 479532.0,
      "step": 565
    },
    {
      "entropy": 6.955751943588257,
      "epoch": 0.012736434021920073,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002845,
      "loss": 6.6147,
      "mean_token_accuracy": 0.1329347386956215,
      "num_tokens": 483844.0,
      "step": 570
    },
    {
      "entropy": 7.067316389083862,
      "epoch": 0.012848157127375512,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000287,
      "loss": 6.6526,
      "mean_token_accuracy": 0.12774468660354615,
      "num_tokens": 488402.0,
      "step": 575
    },
    {
      "entropy": 6.788709592819214,
      "epoch": 0.012959880232830952,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002895,
      "loss": 6.3961,
      "mean_token_accuracy": 0.13899129256606102,
      "num_tokens": 492473.0,
      "step": 580
    },
    {
      "entropy": 6.816238927841186,
      "epoch": 0.013071603338286391,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000292,
      "loss": 6.5152,
      "mean_token_accuracy": 0.1357610009610653,
      "num_tokens": 496830.0,
      "step": 585
    },
    {
      "entropy": 7.027258968353271,
      "epoch": 0.01318332644374183,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002945,
      "loss": 6.5867,
      "mean_token_accuracy": 0.13579039499163628,
      "num_tokens": 500996.0,
      "step": 590
    },
    {
      "entropy": 6.797857189178467,
      "epoch": 0.013295049549197269,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000297,
      "loss": 6.5446,
      "mean_token_accuracy": 0.13526690602302552,
      "num_tokens": 505174.0,
      "step": 595
    },
    {
      "entropy": 6.864916706085205,
      "epoch": 0.013406772654652708,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002995,
      "loss": 6.548,
      "mean_token_accuracy": 0.1401192255318165,
      "num_tokens": 509425.0,
      "step": 600
    },
    {
      "entropy": 6.837484741210938,
      "epoch": 0.013518495760108148,
      "grad_norm": 1.859375,
      "learning_rate": 0.000302,
      "loss": 6.4098,
      "mean_token_accuracy": 0.14489975348114967,
      "num_tokens": 513271.0,
      "step": 605
    },
    {
      "entropy": 6.8677136421203615,
      "epoch": 0.013630218865563587,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003045,
      "loss": 6.6053,
      "mean_token_accuracy": 0.14355736672878266,
      "num_tokens": 517879.0,
      "step": 610
    },
    {
      "entropy": 6.855578374862671,
      "epoch": 0.013741941971019026,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000307,
      "loss": 6.5708,
      "mean_token_accuracy": 0.13215027675032615,
      "num_tokens": 522321.0,
      "step": 615
    },
    {
      "entropy": 6.7266088962554935,
      "epoch": 0.013853665076474465,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003095,
      "loss": 6.3619,
      "mean_token_accuracy": 0.1518048860132694,
      "num_tokens": 525884.0,
      "step": 620
    },
    {
      "entropy": 6.896519756317138,
      "epoch": 0.013965388181929906,
      "grad_norm": 1.640625,
      "learning_rate": 0.000312,
      "loss": 6.4982,
      "mean_token_accuracy": 0.1376729391515255,
      "num_tokens": 530373.0,
      "step": 625
    },
    {
      "entropy": 6.760099697113037,
      "epoch": 0.014077111287385344,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003145,
      "loss": 6.3254,
      "mean_token_accuracy": 0.1519668973982334,
      "num_tokens": 534571.0,
      "step": 630
    },
    {
      "entropy": 6.845727109909058,
      "epoch": 0.014188834392840783,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000317,
      "loss": 6.496,
      "mean_token_accuracy": 0.1373538002371788,
      "num_tokens": 538938.0,
      "step": 635
    },
    {
      "entropy": 6.782427406311035,
      "epoch": 0.014300557498296222,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003195,
      "loss": 6.4534,
      "mean_token_accuracy": 0.15009932219982147,
      "num_tokens": 542977.0,
      "step": 640
    },
    {
      "entropy": 6.756459665298462,
      "epoch": 0.014412280603751661,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000322,
      "loss": 6.4253,
      "mean_token_accuracy": 0.14535169675946236,
      "num_tokens": 547005.0,
      "step": 645
    },
    {
      "entropy": 6.681589221954345,
      "epoch": 0.014524003709207102,
      "grad_norm": 1.53125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.468,
      "mean_token_accuracy": 0.1390581026673317,
      "num_tokens": 551361.0,
      "step": 650
    },
    {
      "entropy": 6.782884311676026,
      "epoch": 0.01463572681466254,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.4134,
      "mean_token_accuracy": 0.13999306932091712,
      "num_tokens": 555574.0,
      "step": 655
    },
    {
      "entropy": 6.632973098754883,
      "epoch": 0.01474744992011798,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.462,
      "mean_token_accuracy": 0.14115875586867332,
      "num_tokens": 560368.0,
      "step": 660
    },
    {
      "entropy": 6.901856851577759,
      "epoch": 0.014859173025573418,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4939,
      "mean_token_accuracy": 0.13747724890708923,
      "num_tokens": 564884.0,
      "step": 665
    },
    {
      "entropy": 6.668564367294311,
      "epoch": 0.014970896131028859,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4878,
      "mean_token_accuracy": 0.1415390208363533,
      "num_tokens": 569494.0,
      "step": 670
    },
    {
      "entropy": 6.764341497421265,
      "epoch": 0.015082619236484298,
      "grad_norm": 1.46875,
      "learning_rate": 0.000337,
      "loss": 6.506,
      "mean_token_accuracy": 0.1322600431740284,
      "num_tokens": 574161.0,
      "step": 675
    },
    {
      "entropy": 6.753030443191529,
      "epoch": 0.015194342341939737,
      "grad_norm": 1.828125,
      "learning_rate": 0.0003395,
      "loss": 6.4248,
      "mean_token_accuracy": 0.14356208443641663,
      "num_tokens": 578510.0,
      "step": 680
    },
    {
      "entropy": 6.674463367462158,
      "epoch": 0.015306065447395175,
      "grad_norm": 1.578125,
      "learning_rate": 0.000342,
      "loss": 6.4388,
      "mean_token_accuracy": 0.1411791481077671,
      "num_tokens": 582695.0,
      "step": 685
    },
    {
      "entropy": 6.776868438720703,
      "epoch": 0.015417788552850614,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4011,
      "mean_token_accuracy": 0.1373072750866413,
      "num_tokens": 586622.0,
      "step": 690
    },
    {
      "entropy": 6.6929234027862545,
      "epoch": 0.015529511658306055,
      "grad_norm": 2.09375,
      "learning_rate": 0.000347,
      "loss": 6.4018,
      "mean_token_accuracy": 0.14830828532576562,
      "num_tokens": 591704.0,
      "step": 695
    },
    {
      "entropy": 6.481786489486694,
      "epoch": 0.015641234763761492,
      "grad_norm": 1.859375,
      "learning_rate": 0.0003495,
      "loss": 6.2342,
      "mean_token_accuracy": 0.15492519065737725,
      "num_tokens": 596029.0,
      "step": 700
    },
    {
      "entropy": 6.6203367710113525,
      "epoch": 0.015752957869216933,
      "grad_norm": 1.609375,
      "learning_rate": 0.000352,
      "loss": 6.391,
      "mean_token_accuracy": 0.14638458639383317,
      "num_tokens": 600115.0,
      "step": 705
    },
    {
      "entropy": 6.719356822967529,
      "epoch": 0.015864680974672373,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003545,
      "loss": 6.3892,
      "mean_token_accuracy": 0.14423182979226112,
      "num_tokens": 604488.0,
      "step": 710
    },
    {
      "entropy": 6.73731255531311,
      "epoch": 0.01597640408012781,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000357,
      "loss": 6.4097,
      "mean_token_accuracy": 0.1440168045461178,
      "num_tokens": 608776.0,
      "step": 715
    },
    {
      "entropy": 6.5224494457244875,
      "epoch": 0.01608812718558325,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003595,
      "loss": 6.4047,
      "mean_token_accuracy": 0.14601838812232018,
      "num_tokens": 612771.0,
      "step": 720
    },
    {
      "entropy": 6.620055294036865,
      "epoch": 0.01619985029103869,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000362,
      "loss": 6.3194,
      "mean_token_accuracy": 0.15284316167235373,
      "num_tokens": 616947.0,
      "step": 725
    },
    {
      "entropy": 6.584221696853637,
      "epoch": 0.01631157339649413,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003645,
      "loss": 6.4028,
      "mean_token_accuracy": 0.14883917048573495,
      "num_tokens": 621064.0,
      "step": 730
    },
    {
      "entropy": 6.636436128616333,
      "epoch": 0.01642329650194957,
      "grad_norm": 1.6875,
      "learning_rate": 0.000367,
      "loss": 6.3876,
      "mean_token_accuracy": 0.1397322230041027,
      "num_tokens": 625349.0,
      "step": 735
    },
    {
      "entropy": 6.571576929092407,
      "epoch": 0.016535019607405006,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003695,
      "loss": 6.3184,
      "mean_token_accuracy": 0.15138321444392205,
      "num_tokens": 629754.0,
      "step": 740
    },
    {
      "entropy": 6.620907211303711,
      "epoch": 0.016646742712860447,
      "grad_norm": 1.578125,
      "learning_rate": 0.000372,
      "loss": 6.3518,
      "mean_token_accuracy": 0.14085613414645196,
      "num_tokens": 634111.0,
      "step": 745
    },
    {
      "entropy": 6.450802040100098,
      "epoch": 0.016758465818315887,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003745,
      "loss": 6.2157,
      "mean_token_accuracy": 0.16091271564364434,
      "num_tokens": 638527.0,
      "step": 750
    },
    {
      "entropy": 6.442024230957031,
      "epoch": 0.016870188923771325,
      "grad_norm": 1.796875,
      "learning_rate": 0.000377,
      "loss": 6.2322,
      "mean_token_accuracy": 0.1605646252632141,
      "num_tokens": 642530.0,
      "step": 755
    },
    {
      "entropy": 6.5630576610565186,
      "epoch": 0.016981912029226765,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003795,
      "loss": 6.3145,
      "mean_token_accuracy": 0.15974024683237076,
      "num_tokens": 646707.0,
      "step": 760
    },
    {
      "entropy": 6.436832427978516,
      "epoch": 0.017093635134682202,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000382,
      "loss": 6.3549,
      "mean_token_accuracy": 0.15013475716114044,
      "num_tokens": 651326.0,
      "step": 765
    },
    {
      "entropy": 6.54889612197876,
      "epoch": 0.017205358240137643,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003845,
      "loss": 6.4225,
      "mean_token_accuracy": 0.1450472079217434,
      "num_tokens": 655651.0,
      "step": 770
    },
    {
      "entropy": 6.598716402053833,
      "epoch": 0.017317081345593083,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3493,
      "mean_token_accuracy": 0.15063260793685912,
      "num_tokens": 659868.0,
      "step": 775
    },
    {
      "entropy": 6.613773393630981,
      "epoch": 0.01742880445104852,
      "grad_norm": 1.609375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.415,
      "mean_token_accuracy": 0.14450336694717408,
      "num_tokens": 664404.0,
      "step": 780
    },
    {
      "entropy": 6.383858156204224,
      "epoch": 0.01754052755650396,
      "grad_norm": 1.640625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2007,
      "mean_token_accuracy": 0.15557888001203538,
      "num_tokens": 668543.0,
      "step": 785
    },
    {
      "entropy": 6.472045183181763,
      "epoch": 0.0176522506619594,
      "grad_norm": 1.578125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.203,
      "mean_token_accuracy": 0.1600804843008518,
      "num_tokens": 672637.0,
      "step": 790
    },
    {
      "entropy": 6.512582492828369,
      "epoch": 0.01776397376741484,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.2935,
      "mean_token_accuracy": 0.15304349958896638,
      "num_tokens": 676929.0,
      "step": 795
    },
    {
      "entropy": 6.386533451080322,
      "epoch": 0.01787569687287028,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003995,
      "loss": 6.3513,
      "mean_token_accuracy": 0.15382541790604592,
      "num_tokens": 681287.0,
      "step": 800
    },
    {
      "entropy": 6.473147869110107,
      "epoch": 0.017987419978325717,
      "grad_norm": 1.546875,
      "learning_rate": 0.000402,
      "loss": 6.2439,
      "mean_token_accuracy": 0.15017452612519264,
      "num_tokens": 685543.0,
      "step": 805
    },
    {
      "entropy": 6.567328834533692,
      "epoch": 0.018099143083781157,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004045,
      "loss": 6.2352,
      "mean_token_accuracy": 0.14613626077771186,
      "num_tokens": 689479.0,
      "step": 810
    },
    {
      "entropy": 6.370502090454101,
      "epoch": 0.018210866189236594,
      "grad_norm": 1.4375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.1598,
      "mean_token_accuracy": 0.16322519332170488,
      "num_tokens": 693564.0,
      "step": 815
    },
    {
      "entropy": 6.502647733688354,
      "epoch": 0.018322589294692035,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004095,
      "loss": 6.2557,
      "mean_token_accuracy": 0.1501608557999134,
      "num_tokens": 697719.0,
      "step": 820
    },
    {
      "entropy": 6.316255664825439,
      "epoch": 0.018434312400147475,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000412,
      "loss": 6.4171,
      "mean_token_accuracy": 0.15079098790884018,
      "num_tokens": 701937.0,
      "step": 825
    },
    {
      "entropy": 6.541279220581055,
      "epoch": 0.018546035505602913,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004145,
      "loss": 6.3393,
      "mean_token_accuracy": 0.14786507934331894,
      "num_tokens": 706457.0,
      "step": 830
    },
    {
      "entropy": 6.479259777069092,
      "epoch": 0.018657758611058353,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000417,
      "loss": 6.3349,
      "mean_token_accuracy": 0.15288059860467912,
      "num_tokens": 710538.0,
      "step": 835
    },
    {
      "entropy": 6.2955930709838865,
      "epoch": 0.018769481716513794,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004195,
      "loss": 6.222,
      "mean_token_accuracy": 0.15921739637851715,
      "num_tokens": 714663.0,
      "step": 840
    },
    {
      "entropy": 6.474664878845215,
      "epoch": 0.01888120482196923,
      "grad_norm": 1.515625,
      "learning_rate": 0.000422,
      "loss": 6.2263,
      "mean_token_accuracy": 0.15058805868029596,
      "num_tokens": 718828.0,
      "step": 845
    },
    {
      "entropy": 6.231873464584351,
      "epoch": 0.01899292792742467,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004245,
      "loss": 6.2178,
      "mean_token_accuracy": 0.15931910127401352,
      "num_tokens": 723155.0,
      "step": 850
    },
    {
      "entropy": 6.52202787399292,
      "epoch": 0.01910465103288011,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000427,
      "loss": 6.2688,
      "mean_token_accuracy": 0.157795250415802,
      "num_tokens": 727605.0,
      "step": 855
    },
    {
      "entropy": 6.3318055152893065,
      "epoch": 0.01921637413833555,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004295,
      "loss": 6.2337,
      "mean_token_accuracy": 0.15254703313112258,
      "num_tokens": 731510.0,
      "step": 860
    },
    {
      "entropy": 6.418648195266724,
      "epoch": 0.01932809724379099,
      "grad_norm": 1.359375,
      "learning_rate": 0.000432,
      "loss": 6.467,
      "mean_token_accuracy": 0.13948606923222542,
      "num_tokens": 736048.0,
      "step": 865
    },
    {
      "entropy": 6.461650276184082,
      "epoch": 0.019439820349246427,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004345,
      "loss": 6.2286,
      "mean_token_accuracy": 0.1571047306060791,
      "num_tokens": 740714.0,
      "step": 870
    },
    {
      "entropy": 6.340702772140503,
      "epoch": 0.019551543454701868,
      "grad_norm": 1.671875,
      "learning_rate": 0.000437,
      "loss": 6.3055,
      "mean_token_accuracy": 0.1546173483133316,
      "num_tokens": 745167.0,
      "step": 875
    },
    {
      "entropy": 6.3855945587158205,
      "epoch": 0.019663266560157305,
      "grad_norm": 1.375,
      "learning_rate": 0.0004395,
      "loss": 6.1172,
      "mean_token_accuracy": 0.16043180972337723,
      "num_tokens": 749401.0,
      "step": 880
    },
    {
      "entropy": 6.357611846923828,
      "epoch": 0.019774989665612745,
      "grad_norm": 1.453125,
      "learning_rate": 0.000442,
      "loss": 6.1698,
      "mean_token_accuracy": 0.16577810943126678,
      "num_tokens": 753554.0,
      "step": 885
    },
    {
      "entropy": 6.337348079681396,
      "epoch": 0.019886712771068186,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004445,
      "loss": 6.2451,
      "mean_token_accuracy": 0.15055422931909562,
      "num_tokens": 757744.0,
      "step": 890
    },
    {
      "entropy": 6.3222455978393555,
      "epoch": 0.019998435876523623,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000447,
      "loss": 6.1133,
      "mean_token_accuracy": 0.15861648172140122,
      "num_tokens": 761858.0,
      "step": 895
    },
    {
      "entropy": 6.288092851638794,
      "epoch": 0.020110158981979064,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.1429,
      "mean_token_accuracy": 0.16834202855825425,
      "num_tokens": 765852.0,
      "step": 900
    },
    {
      "entropy": 6.2346821308135985,
      "epoch": 0.0202218820874345,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1649,
      "mean_token_accuracy": 0.1537023276090622,
      "num_tokens": 770274.0,
      "step": 905
    },
    {
      "entropy": 6.315171623229981,
      "epoch": 0.02033360519288994,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2175,
      "mean_token_accuracy": 0.1551866799592972,
      "num_tokens": 774775.0,
      "step": 910
    },
    {
      "entropy": 6.4123328685760494,
      "epoch": 0.020445328298345382,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2295,
      "mean_token_accuracy": 0.16605380773544312,
      "num_tokens": 779374.0,
      "step": 915
    },
    {
      "entropy": 6.239690780639648,
      "epoch": 0.02055705140380082,
      "grad_norm": 1.75,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2391,
      "mean_token_accuracy": 0.1497175522148609,
      "num_tokens": 783537.0,
      "step": 920
    },
    {
      "entropy": 6.284845304489136,
      "epoch": 0.02066877450925626,
      "grad_norm": 1.40625,
      "learning_rate": 0.000462,
      "loss": 6.3045,
      "mean_token_accuracy": 0.15461198166012763,
      "num_tokens": 787562.0,
      "step": 925
    },
    {
      "entropy": 6.4740015983581545,
      "epoch": 0.0207804976147117,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004645,
      "loss": 6.3361,
      "mean_token_accuracy": 0.14666738733649254,
      "num_tokens": 792303.0,
      "step": 930
    },
    {
      "entropy": 6.263271856307983,
      "epoch": 0.020892220720167137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000467,
      "loss": 6.1427,
      "mean_token_accuracy": 0.16184145957231522,
      "num_tokens": 796402.0,
      "step": 935
    },
    {
      "entropy": 6.380016136169433,
      "epoch": 0.021003943825622578,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004695,
      "loss": 6.2087,
      "mean_token_accuracy": 0.1643604129552841,
      "num_tokens": 800004.0,
      "step": 940
    },
    {
      "entropy": 6.157491159439087,
      "epoch": 0.021115666931078015,
      "grad_norm": 1.484375,
      "learning_rate": 0.000472,
      "loss": 6.0888,
      "mean_token_accuracy": 0.1660729944705963,
      "num_tokens": 803782.0,
      "step": 945
    },
    {
      "entropy": 6.282210779190064,
      "epoch": 0.021227390036533456,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004745,
      "loss": 6.1556,
      "mean_token_accuracy": 0.16121889799833297,
      "num_tokens": 808013.0,
      "step": 950
    },
    {
      "entropy": 6.159784889221191,
      "epoch": 0.021339113141988896,
      "grad_norm": 1.4375,
      "learning_rate": 0.000477,
      "loss": 6.1523,
      "mean_token_accuracy": 0.1605899378657341,
      "num_tokens": 812130.0,
      "step": 955
    },
    {
      "entropy": 6.311656665802002,
      "epoch": 0.021450836247444333,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004795,
      "loss": 6.1163,
      "mean_token_accuracy": 0.16060882434248924,
      "num_tokens": 816246.0,
      "step": 960
    },
    {
      "entropy": 6.1442811489105225,
      "epoch": 0.021562559352899774,
      "grad_norm": 1.46875,
      "learning_rate": 0.000482,
      "loss": 6.1295,
      "mean_token_accuracy": 0.15766306221485138,
      "num_tokens": 820628.0,
      "step": 965
    },
    {
      "entropy": 6.282287883758545,
      "epoch": 0.02167428245835521,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004845,
      "loss": 6.1044,
      "mean_token_accuracy": 0.16587121784687042,
      "num_tokens": 824864.0,
      "step": 970
    },
    {
      "entropy": 6.298398399353028,
      "epoch": 0.02178600556381065,
      "grad_norm": 1.640625,
      "learning_rate": 0.000487,
      "loss": 6.1733,
      "mean_token_accuracy": 0.1572745218873024,
      "num_tokens": 828887.0,
      "step": 975
    },
    {
      "entropy": 6.119819927215576,
      "epoch": 0.021897728669266092,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004895,
      "loss": 6.0463,
      "mean_token_accuracy": 0.16708520650863648,
      "num_tokens": 832784.0,
      "step": 980
    },
    {
      "entropy": 6.1874528408050535,
      "epoch": 0.02200945177472153,
      "grad_norm": 1.40625,
      "learning_rate": 0.000492,
      "loss": 6.0893,
      "mean_token_accuracy": 0.15289360731840135,
      "num_tokens": 837408.0,
      "step": 985
    },
    {
      "entropy": 6.194441604614258,
      "epoch": 0.02212117488017697,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004945,
      "loss": 6.1339,
      "mean_token_accuracy": 0.16136935949325562,
      "num_tokens": 841536.0,
      "step": 990
    },
    {
      "entropy": 6.170191717147827,
      "epoch": 0.022232897985632407,
      "grad_norm": 1.484375,
      "learning_rate": 0.000497,
      "loss": 6.0497,
      "mean_token_accuracy": 0.16993979513645172,
      "num_tokens": 846044.0,
      "step": 995
    },
    {
      "entropy": 6.106483221054077,
      "epoch": 0.022344621091087848,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004995,
      "loss": 6.0682,
      "mean_token_accuracy": 0.16326213181018828,
      "num_tokens": 850254.0,
      "step": 1000
    },
    {
      "entropy": 6.112219524383545,
      "epoch": 0.022456344196543288,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000499998026082006,
      "loss": 6.0539,
      "mean_token_accuracy": 0.16882037818431855,
      "num_tokens": 854548.0,
      "step": 1005
    },
    {
      "entropy": 6.08699860572815,
      "epoch": 0.022568067301998725,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004999900070995136,
      "loss": 6.2141,
      "mean_token_accuracy": 0.1636434242129326,
      "num_tokens": 858837.0,
      "step": 1010
    },
    {
      "entropy": 6.176929616928101,
      "epoch": 0.022679790407454166,
      "grad_norm": 1.5,
      "learning_rate": 0.0004999758199023239,
      "loss": 6.0198,
      "mean_token_accuracy": 0.15852618217468262,
      "num_tokens": 862772.0,
      "step": 1015
    },
    {
      "entropy": 6.200733661651611,
      "epoch": 0.022791513512909607,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999554648793858,
      "loss": 6.098,
      "mean_token_accuracy": 0.1593088060617447,
      "num_tokens": 866909.0,
      "step": 1020
    },
    {
      "entropy": 6.177529811859131,
      "epoch": 0.022903236618365044,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004999289425887425,
      "loss": 6.2321,
      "mean_token_accuracy": 0.1547963500022888,
      "num_tokens": 871127.0,
      "step": 1025
    },
    {
      "entropy": 6.2023234367370605,
      "epoch": 0.023014959723820484,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004998962537575161,
      "loss": 6.1609,
      "mean_token_accuracy": 0.16332077234983444,
      "num_tokens": 875292.0,
      "step": 1030
    },
    {
      "entropy": 6.16438570022583,
      "epoch": 0.02312668282927592,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004998573992818874,
      "loss": 6.0549,
      "mean_token_accuracy": 0.16016169339418412,
      "num_tokens": 879286.0,
      "step": 1035
    },
    {
      "entropy": 6.2433970928192135,
      "epoch": 0.023238405934731362,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004998123802270715,
      "loss": 6.0876,
      "mean_token_accuracy": 0.17015287280082703,
      "num_tokens": 883490.0,
      "step": 1040
    },
    {
      "entropy": 6.14397964477539,
      "epoch": 0.023350129040186803,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004997611978272886,
      "loss": 6.0609,
      "mean_token_accuracy": 0.16376899629831315,
      "num_tokens": 887786.0,
      "step": 1045
    },
    {
      "entropy": 6.1847413063049315,
      "epoch": 0.02346185214564224,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004997038534857298,
      "loss": 6.0817,
      "mean_token_accuracy": 0.17141184508800505,
      "num_tokens": 892035.0,
      "step": 1050
    },
    {
      "entropy": 6.1673887252807615,
      "epoch": 0.02357357525109768,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004996403487745194,
      "loss": 6.0719,
      "mean_token_accuracy": 0.16404020488262178,
      "num_tokens": 896099.0,
      "step": 1055
    },
    {
      "entropy": 6.215155744552613,
      "epoch": 0.023685298356553117,
      "grad_norm": 1.578125,
      "learning_rate": 0.000499570685434671,
      "loss": 5.9941,
      "mean_token_accuracy": 0.17437216117978097,
      "num_tokens": 900448.0,
      "step": 1060
    },
    {
      "entropy": 6.082410001754761,
      "epoch": 0.023797021462008558,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004994948653760405,
      "loss": 6.0422,
      "mean_token_accuracy": 0.16581026017665862,
      "num_tokens": 904538.0,
      "step": 1065
    },
    {
      "entropy": 6.211330890655518,
      "epoch": 0.023908744567464,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004994128906772729,
      "loss": 6.2105,
      "mean_token_accuracy": 0.15310067683458328,
      "num_tokens": 908680.0,
      "step": 1070
    },
    {
      "entropy": 6.132528066635132,
      "epoch": 0.024020467672919436,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000499324763585746,
      "loss": 6.1633,
      "mean_token_accuracy": 0.154509137570858,
      "num_tokens": 912868.0,
      "step": 1075
    },
    {
      "entropy": 6.17259407043457,
      "epoch": 0.024132190778374876,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004992304865175085,
      "loss": 6.192,
      "mean_token_accuracy": 0.16236126869916917,
      "num_tokens": 917312.0,
      "step": 1080
    },
    {
      "entropy": 6.335460948944092,
      "epoch": 0.024243913883830313,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991300620572138,
      "loss": 6.1838,
      "mean_token_accuracy": 0.16260358244180678,
      "num_tokens": 921977.0,
      "step": 1085
    },
    {
      "entropy": 6.128043842315674,
      "epoch": 0.024355636989285754,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004990234929580494,
      "loss": 6.1085,
      "mean_token_accuracy": 0.16749032437801362,
      "num_tokens": 925964.0,
      "step": 1090
    },
    {
      "entropy": 6.223902463912964,
      "epoch": 0.024467360094741195,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004989107821416609,
      "loss": 5.9766,
      "mean_token_accuracy": 0.1657112054526806,
      "num_tokens": 930077.0,
      "step": 1095
    },
    {
      "entropy": 6.065625762939453,
      "epoch": 0.02457908320019663,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987919326980723,
      "loss": 6.0066,
      "mean_token_accuracy": 0.1741485133767128,
      "num_tokens": 934547.0,
      "step": 1100
    },
    {
      "entropy": 6.170744371414185,
      "epoch": 0.024690806305652072,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004986669478856011,
      "loss": 6.0122,
      "mean_token_accuracy": 0.17556725889444352,
      "num_tokens": 938583.0,
      "step": 1105
    },
    {
      "entropy": 6.233295917510986,
      "epoch": 0.024802529411107513,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004985358311307688,
      "loss": 6.0892,
      "mean_token_accuracy": 0.16876054555177689,
      "num_tokens": 943007.0,
      "step": 1110
    },
    {
      "entropy": 6.1148707389831545,
      "epoch": 0.02491425251656295,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004983985860282081,
      "loss": 6.089,
      "mean_token_accuracy": 0.16639824211597443,
      "num_tokens": 947281.0,
      "step": 1115
    },
    {
      "entropy": 6.048860692977906,
      "epoch": 0.02502597562201839,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004982552163405623,
      "loss": 5.952,
      "mean_token_accuracy": 0.17678665965795518,
      "num_tokens": 951443.0,
      "step": 1120
    },
    {
      "entropy": 6.026111602783203,
      "epoch": 0.025137698727473828,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004981057259983839,
      "loss": 6.0601,
      "mean_token_accuracy": 0.16274188607931137,
      "num_tokens": 955131.0,
      "step": 1125
    },
    {
      "entropy": 6.128307962417603,
      "epoch": 0.02524942183292927,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004979501191000262,
      "loss": 6.0103,
      "mean_token_accuracy": 0.16576859056949617,
      "num_tokens": 959408.0,
      "step": 1130
    },
    {
      "entropy": 5.998388624191284,
      "epoch": 0.02536114493838471,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004977883999115311,
      "loss": 6.0554,
      "mean_token_accuracy": 0.16911191642284393,
      "num_tokens": 963657.0,
      "step": 1135
    },
    {
      "entropy": 6.107620096206665,
      "epoch": 0.025472868043840146,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004976205728665113,
      "loss": 6.0781,
      "mean_token_accuracy": 0.1575411856174469,
      "num_tokens": 967973.0,
      "step": 1140
    },
    {
      "entropy": 6.082073783874511,
      "epoch": 0.025584591149295587,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974466425660307,
      "loss": 5.8803,
      "mean_token_accuracy": 0.18416397273540497,
      "num_tokens": 971957.0,
      "step": 1145
    },
    {
      "entropy": 5.985405111312867,
      "epoch": 0.025696314254751024,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004972666137784759,
      "loss": 5.8823,
      "mean_token_accuracy": 0.1817352905869484,
      "num_tokens": 975679.0,
      "step": 1150
    },
    {
      "entropy": 6.04285044670105,
      "epoch": 0.025808037360206464,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004970804914394271,
      "loss": 6.1732,
      "mean_token_accuracy": 0.15990505665540694,
      "num_tokens": 979806.0,
      "step": 1155
    },
    {
      "entropy": 6.167063045501709,
      "epoch": 0.025919760465661905,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004968882806515225,
      "loss": 6.0201,
      "mean_token_accuracy": 0.1767844155430794,
      "num_tokens": 984221.0,
      "step": 1160
    },
    {
      "entropy": 6.029861211776733,
      "epoch": 0.026031483571117342,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004966899866843177,
      "loss": 5.9773,
      "mean_token_accuracy": 0.16563293039798738,
      "num_tokens": 988323.0,
      "step": 1165
    },
    {
      "entropy": 6.0827898502349855,
      "epoch": 0.026143206676572783,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000496485614974142,
      "loss": 6.038,
      "mean_token_accuracy": 0.1660424754023552,
      "num_tokens": 992643.0,
      "step": 1170
    },
    {
      "entropy": 5.991740322113037,
      "epoch": 0.02625492978202822,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004962751711239492,
      "loss": 5.8579,
      "mean_token_accuracy": 0.1826507866382599,
      "num_tokens": 996994.0,
      "step": 1175
    },
    {
      "entropy": 5.962495994567871,
      "epoch": 0.02636665288748366,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004960586609031636,
      "loss": 6.0109,
      "mean_token_accuracy": 0.17780142575502395,
      "num_tokens": 1001345.0,
      "step": 1180
    },
    {
      "entropy": 6.076908683776855,
      "epoch": 0.0264783759929391,
      "grad_norm": 1.375,
      "learning_rate": 0.0004958360902475224,
      "loss": 5.9154,
      "mean_token_accuracy": 0.17819885313510894,
      "num_tokens": 1005839.0,
      "step": 1185
    },
    {
      "entropy": 6.0804304599761965,
      "epoch": 0.026590099098394538,
      "grad_norm": 1.375,
      "learning_rate": 0.0004956074652589125,
      "loss": 6.0655,
      "mean_token_accuracy": 0.1642119474709034,
      "num_tokens": 1010698.0,
      "step": 1190
    },
    {
      "entropy": 6.093788766860962,
      "epoch": 0.02670182220384998,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004953727922052035,
      "loss": 6.0287,
      "mean_token_accuracy": 0.1619931563735008,
      "num_tokens": 1014982.0,
      "step": 1195
    },
    {
      "entropy": 5.86436014175415,
      "epoch": 0.026813545309305416,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004951320775200756,
      "loss": 5.9534,
      "mean_token_accuracy": 0.18322545439004898,
      "num_tokens": 1019235.0,
      "step": 1200
    },
    {
      "entropy": 6.108639049530029,
      "epoch": 0.026925268414760856,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004948853278028436,
      "loss": 6.0397,
      "mean_token_accuracy": 0.1752349078655243,
      "num_tokens": 1023855.0,
      "step": 1205
    },
    {
      "entropy": 6.110106515884399,
      "epoch": 0.027036991520216297,
      "grad_norm": 1.375,
      "learning_rate": 0.0004946325498182755,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17471726685762407,
      "num_tokens": 1028464.0,
      "step": 1210
    },
    {
      "entropy": 5.944786262512207,
      "epoch": 0.027148714625671734,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004943737504964076,
      "loss": 5.9103,
      "mean_token_accuracy": 0.18010717928409575,
      "num_tokens": 1033110.0,
      "step": 1215
    },
    {
      "entropy": 5.954358720779419,
      "epoch": 0.027260437731127175,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000494108936932354,
      "loss": 6.0102,
      "mean_token_accuracy": 0.16354090273380278,
      "num_tokens": 1037470.0,
      "step": 1220
    },
    {
      "entropy": 6.164341068267822,
      "epoch": 0.027372160836582615,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004938381163861124,
      "loss": 6.0126,
      "mean_token_accuracy": 0.16803643256425857,
      "num_tokens": 1041662.0,
      "step": 1225
    },
    {
      "entropy": 6.096904563903808,
      "epoch": 0.027483883942038052,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004935612962823645,
      "loss": 5.92,
      "mean_token_accuracy": 0.17529453635215758,
      "num_tokens": 1046041.0,
      "step": 1230
    },
    {
      "entropy": 5.942513656616211,
      "epoch": 0.027595607047493493,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004932784842102739,
      "loss": 5.9454,
      "mean_token_accuracy": 0.1719459719955921,
      "num_tokens": 1049893.0,
      "step": 1235
    },
    {
      "entropy": 6.186945247650146,
      "epoch": 0.02770733015294893,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004929896879232758,
      "loss": 6.1042,
      "mean_token_accuracy": 0.15955148935317992,
      "num_tokens": 1054048.0,
      "step": 1240
    },
    {
      "entropy": 6.0783411979675295,
      "epoch": 0.02781905325840437,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004926949153388668,
      "loss": 5.9941,
      "mean_token_accuracy": 0.16463877633213997,
      "num_tokens": 1058176.0,
      "step": 1245
    },
    {
      "entropy": 6.107759857177735,
      "epoch": 0.02793077636385981,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004923941745383859,
      "loss": 5.9543,
      "mean_token_accuracy": 0.16195504367351532,
      "num_tokens": 1062031.0,
      "step": 1250
    },
    {
      "entropy": 5.933470392227173,
      "epoch": 0.02804249946931525,
      "grad_norm": 1.40625,
      "learning_rate": 0.000492087473766794,
      "loss": 5.9244,
      "mean_token_accuracy": 0.17964119464159012,
      "num_tokens": 1066705.0,
      "step": 1255
    },
    {
      "entropy": 6.079675674438477,
      "epoch": 0.02815422257477069,
      "grad_norm": 1.390625,
      "learning_rate": 0.000491774821432448,
      "loss": 5.9448,
      "mean_token_accuracy": 0.18084608763456345,
      "num_tokens": 1070960.0,
      "step": 1260
    },
    {
      "entropy": 6.124145603179931,
      "epoch": 0.028265945680226126,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004914562261068693,
      "loss": 6.0945,
      "mean_token_accuracy": 0.15580475330352783,
      "num_tokens": 1075576.0,
      "step": 1265
    },
    {
      "entropy": 6.012061548233032,
      "epoch": 0.028377668785681567,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004911316965245098,
      "loss": 5.8603,
      "mean_token_accuracy": 0.16867389604449273,
      "num_tokens": 1079538.0,
      "step": 1270
    },
    {
      "entropy": 5.922763729095459,
      "epoch": 0.028489391891137007,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000490801241582512,
      "loss": 5.9411,
      "mean_token_accuracy": 0.1767105147242546,
      "num_tokens": 1083296.0,
      "step": 1275
    },
    {
      "entropy": 5.973857307434082,
      "epoch": 0.028601114996592444,
      "grad_norm": 1.4375,
      "learning_rate": 0.000490464870340465,
      "loss": 5.9737,
      "mean_token_accuracy": 0.17723702639341354,
      "num_tokens": 1087688.0,
      "step": 1280
    },
    {
      "entropy": 6.082131719589233,
      "epoch": 0.028712838102047885,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004901225920201563,
      "loss": 5.9207,
      "mean_token_accuracy": 0.17827319502830505,
      "num_tokens": 1091988.0,
      "step": 1285
    },
    {
      "entropy": 6.049498653411865,
      "epoch": 0.028824561207503322,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000489774416005319,
      "loss": 5.9449,
      "mean_token_accuracy": 0.17476076930761336,
      "num_tokens": 1095949.0,
      "step": 1290
    },
    {
      "entropy": 5.945538663864136,
      "epoch": 0.028936284312958763,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004894203518413742,
      "loss": 5.8761,
      "mean_token_accuracy": 0.17729702293872834,
      "num_tokens": 1099795.0,
      "step": 1295
    },
    {
      "entropy": 5.840633201599121,
      "epoch": 0.029048007418414203,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004890604092351701,
      "loss": 5.7914,
      "mean_token_accuracy": 0.18237984031438828,
      "num_tokens": 1103894.0,
      "step": 1300
    },
    {
      "entropy": 6.022551822662353,
      "epoch": 0.02915973052386964,
      "grad_norm": 1.46875,
      "learning_rate": 0.000488694598054715,
      "loss": 6.0166,
      "mean_token_accuracy": 0.1801459826529026,
      "num_tokens": 1108382.0,
      "step": 1305
    },
    {
      "entropy": 5.830238723754883,
      "epoch": 0.02927145362932508,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004883229283289071,
      "loss": 5.772,
      "mean_token_accuracy": 0.1923062965273857,
      "num_tokens": 1112841.0,
      "step": 1310
    },
    {
      "entropy": 6.063643503189087,
      "epoch": 0.02938317673478052,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048794541024725993,
      "loss": 6.0292,
      "mean_token_accuracy": 0.16797914803028108,
      "num_tokens": 1117319.0,
      "step": 1315
    },
    {
      "entropy": 6.104041051864624,
      "epoch": 0.02949489984023596,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004875620541596221,
      "loss": 5.976,
      "mean_token_accuracy": 0.1713699594140053,
      "num_tokens": 1121371.0,
      "step": 1320
    },
    {
      "entropy": 5.860101127624512,
      "epoch": 0.0296066229456914,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048717287057589454,
      "loss": 5.8614,
      "mean_token_accuracy": 0.17473118454217912,
      "num_tokens": 1125496.0,
      "step": 1325
    },
    {
      "entropy": 6.027236604690552,
      "epoch": 0.029718346051146836,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004867778701657417,
      "loss": 5.9316,
      "mean_token_accuracy": 0.17779530882835387,
      "num_tokens": 1129929.0,
      "step": 1330
    },
    {
      "entropy": 5.962581586837769,
      "epoch": 0.029830069156602277,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048637706375829955,
      "loss": 5.8692,
      "mean_token_accuracy": 0.18104994744062425,
      "num_tokens": 1133982.0,
      "step": 1335
    },
    {
      "entropy": 5.87730565071106,
      "epoch": 0.029941792262057718,
      "grad_norm": 1.484375,
      "learning_rate": 0.000485970462341878,
      "loss": 5.8023,
      "mean_token_accuracy": 0.17854024320840836,
      "num_tokens": 1138270.0,
      "step": 1340
    },
    {
      "entropy": 6.029954719543457,
      "epoch": 0.030053515367513155,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048555807706366044,
      "loss": 5.9336,
      "mean_token_accuracy": 0.17757426649332048,
      "num_tokens": 1142197.0,
      "step": 1345
    },
    {
      "entropy": 5.986188697814941,
      "epoch": 0.030165238472968595,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048513991922939756,
      "loss": 5.9806,
      "mean_token_accuracy": 0.17655822336673738,
      "num_tokens": 1146439.0,
      "step": 1350
    },
    {
      "entropy": 5.9268670082092285,
      "epoch": 0.030276961578424032,
      "grad_norm": 1.375,
      "learning_rate": 0.00048471600030309744,
      "loss": 5.9419,
      "mean_token_accuracy": 0.1782355859875679,
      "num_tokens": 1150809.0,
      "step": 1355
    },
    {
      "entropy": 6.041894149780274,
      "epoch": 0.030388684683879473,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048428633190671186,
      "loss": 6.0754,
      "mean_token_accuracy": 0.16220128163695335,
      "num_tokens": 1155132.0,
      "step": 1360
    },
    {
      "entropy": 6.196034908294678,
      "epoch": 0.030500407789334914,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004838509258198167,
      "loss": 6.0108,
      "mean_token_accuracy": 0.16138193309307097,
      "num_tokens": 1159694.0,
      "step": 1365
    },
    {
      "entropy": 5.919046926498413,
      "epoch": 0.03061213089479035,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048340979397929,
      "loss": 5.9393,
      "mean_token_accuracy": 0.16888010501861572,
      "num_tokens": 1164070.0,
      "step": 1370
    },
    {
      "entropy": 5.922135591506958,
      "epoch": 0.03072385400024579,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048296294847898386,
      "loss": 5.7818,
      "mean_token_accuracy": 0.1902960404753685,
      "num_tokens": 1168097.0,
      "step": 1375
    },
    {
      "entropy": 5.983519697189331,
      "epoch": 0.03083557710570123,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004825104015693934,
      "loss": 5.9272,
      "mean_token_accuracy": 0.1677720718085766,
      "num_tokens": 1172375.0,
      "step": 1380
    },
    {
      "entropy": 5.890913963317871,
      "epoch": 0.03094730021115667,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004820521656573208,
      "loss": 5.9675,
      "mean_token_accuracy": 0.16840586364269255,
      "num_tokens": 1176628.0,
      "step": 1385
    },
    {
      "entropy": 5.930252313613892,
      "epoch": 0.03105902331661211,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048158825330553505,
      "loss": 5.9445,
      "mean_token_accuracy": 0.16537177711725234,
      "num_tokens": 1180500.0,
      "step": 1390
    },
    {
      "entropy": 5.924455261230468,
      "epoch": 0.031170746422067547,
      "grad_norm": 1.5,
      "learning_rate": 0.00048111867723242763,
      "loss": 5.9551,
      "mean_token_accuracy": 0.17216705679893493,
      "num_tokens": 1184883.0,
      "step": 1395
    },
    {
      "entropy": 6.142921876907349,
      "epoch": 0.031282469527522984,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004806434503116637,
      "loss": 5.9368,
      "mean_token_accuracy": 0.1701400026679039,
      "num_tokens": 1188723.0,
      "step": 1400
    },
    {
      "entropy": 6.0046467781066895,
      "epoch": 0.03139419263297843,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004801625855718296,
      "loss": 6.0485,
      "mean_token_accuracy": 0.1668444462120533,
      "num_tokens": 1193318.0,
      "step": 1405
    },
    {
      "entropy": 6.041379928588867,
      "epoch": 0.031505915738433865,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047967609619607477,
      "loss": 5.9308,
      "mean_token_accuracy": 0.16749139726161957,
      "num_tokens": 1197663.0,
      "step": 1410
    },
    {
      "entropy": 5.913714504241943,
      "epoch": 0.0316176388438893,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004791839955217513,
      "loss": 5.7884,
      "mean_token_accuracy": 0.18735090345144273,
      "num_tokens": 1201949.0,
      "step": 1415
    },
    {
      "entropy": 6.008267402648926,
      "epoch": 0.031729361949344746,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047868629704004786,
      "loss": 6.0329,
      "mean_token_accuracy": 0.16291153579950332,
      "num_tokens": 1206360.0,
      "step": 1420
    },
    {
      "entropy": 5.969942045211792,
      "epoch": 0.03184108505480018,
      "grad_norm": 1.5,
      "learning_rate": 0.00047818301439561965,
      "loss": 5.8897,
      "mean_token_accuracy": 0.1797643095254898,
      "num_tokens": 1210738.0,
      "step": 1425
    },
    {
      "entropy": 5.823257780075073,
      "epoch": 0.03195280816025562,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047767416138621454,
      "loss": 5.8901,
      "mean_token_accuracy": 0.17570878565311432,
      "num_tokens": 1214722.0,
      "step": 1430
    },
    {
      "entropy": 6.028759336471557,
      "epoch": 0.032064531265711065,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000477159751962295,
      "loss": 5.9947,
      "mean_token_accuracy": 0.17196338325738908,
      "num_tokens": 1218811.0,
      "step": 1435
    },
    {
      "entropy": 6.077054071426391,
      "epoch": 0.0321762543711665,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047663980022665507,
      "loss": 5.9492,
      "mean_token_accuracy": 0.1680076539516449,
      "num_tokens": 1223089.0,
      "step": 1440
    },
    {
      "entropy": 5.646877670288086,
      "epoch": 0.03228797747662194,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047611432043403437,
      "loss": 5.7538,
      "mean_token_accuracy": 0.18203833252191542,
      "num_tokens": 1226970.0,
      "step": 1445
    },
    {
      "entropy": 5.893757677078247,
      "epoch": 0.03239970058207738,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004755833269907267,
      "loss": 5.8528,
      "mean_token_accuracy": 0.17612469345331191,
      "num_tokens": 1230734.0,
      "step": 1450
    },
    {
      "entropy": 5.9495970726013185,
      "epoch": 0.03251142368753282,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004750468344541857,
      "loss": 5.8897,
      "mean_token_accuracy": 0.16842119693756102,
      "num_tokens": 1235363.0,
      "step": 1455
    },
    {
      "entropy": 5.881724309921265,
      "epoch": 0.03262314679298826,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047450485753262525,
      "loss": 5.739,
      "mean_token_accuracy": 0.17950696647167205,
      "num_tokens": 1239826.0,
      "step": 1460
    },
    {
      "entropy": 5.911339473724365,
      "epoch": 0.032734869898443694,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047395741108461633,
      "loss": 5.9883,
      "mean_token_accuracy": 0.16461902111768723,
      "num_tokens": 1244067.0,
      "step": 1465
    },
    {
      "entropy": 5.944296741485596,
      "epoch": 0.03284659300389914,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047340451011867985,
      "loss": 5.7769,
      "mean_token_accuracy": 0.18020680397748948,
      "num_tokens": 1248307.0,
      "step": 1470
    },
    {
      "entropy": 5.810064172744751,
      "epoch": 0.032958316109354575,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047284616979287515,
      "loss": 5.8516,
      "mean_token_accuracy": 0.1767789214849472,
      "num_tokens": 1252504.0,
      "step": 1475
    },
    {
      "entropy": 5.784211921691894,
      "epoch": 0.03307003921481001,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047228240541438433,
      "loss": 5.8911,
      "mean_token_accuracy": 0.17289571911096574,
      "num_tokens": 1256823.0,
      "step": 1480
    },
    {
      "entropy": 5.894031572341919,
      "epoch": 0.03318176232026546,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047171323243909257,
      "loss": 5.8256,
      "mean_token_accuracy": 0.17667611539363862,
      "num_tokens": 1261236.0,
      "step": 1485
    },
    {
      "entropy": 5.957602310180664,
      "epoch": 0.033293485425720894,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047113866647116457,
      "loss": 5.836,
      "mean_token_accuracy": 0.18172842562198638,
      "num_tokens": 1265360.0,
      "step": 1490
    },
    {
      "entropy": 5.874336624145508,
      "epoch": 0.03340520853117633,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004705587232626164,
      "loss": 5.8134,
      "mean_token_accuracy": 0.18341318368911744,
      "num_tokens": 1269683.0,
      "step": 1495
    },
    {
      "entropy": 5.891045808792114,
      "epoch": 0.033516931636631775,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046997341871288424,
      "loss": 5.8988,
      "mean_token_accuracy": 0.1770934522151947,
      "num_tokens": 1273843.0,
      "step": 1500
    },
    {
      "entropy": 5.960714864730835,
      "epoch": 0.03362865474208721,
      "grad_norm": 1.25,
      "learning_rate": 0.0004693827688683879,
      "loss": 5.8822,
      "mean_token_accuracy": 0.18434463143348695,
      "num_tokens": 1278299.0,
      "step": 1505
    },
    {
      "entropy": 5.963765954971313,
      "epoch": 0.03374037784754265,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004687867899220914,
      "loss": 5.7639,
      "mean_token_accuracy": 0.18203867971897125,
      "num_tokens": 1282541.0,
      "step": 1510
    },
    {
      "entropy": 5.752007484436035,
      "epoch": 0.033852100952998086,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046818549821305846,
      "loss": 5.7004,
      "mean_token_accuracy": 0.18360009491443635,
      "num_tokens": 1286318.0,
      "step": 1515
    },
    {
      "entropy": 5.6970837116241455,
      "epoch": 0.03396382405845353,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046757891022600494,
      "loss": 5.8109,
      "mean_token_accuracy": 0.18161729574203492,
      "num_tokens": 1290713.0,
      "step": 1520
    },
    {
      "entropy": 6.068321418762207,
      "epoch": 0.03407554716390897,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004669670425908471,
      "loss": 5.7778,
      "mean_token_accuracy": 0.17242596298456192,
      "num_tokens": 1294820.0,
      "step": 1525
    },
    {
      "entropy": 5.848509216308594,
      "epoch": 0.034187270269364405,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004663499120822451,
      "loss": 5.7949,
      "mean_token_accuracy": 0.17674396187067032,
      "num_tokens": 1299232.0,
      "step": 1530
    },
    {
      "entropy": 5.775924873352051,
      "epoch": 0.03429899337481985,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004657275356191437,
      "loss": 5.7961,
      "mean_token_accuracy": 0.18478017151355744,
      "num_tokens": 1303287.0,
      "step": 1535
    },
    {
      "entropy": 5.834440803527832,
      "epoch": 0.034410716480275286,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046509993026430804,
      "loss": 5.8024,
      "mean_token_accuracy": 0.18893743604421614,
      "num_tokens": 1307563.0,
      "step": 1540
    },
    {
      "entropy": 5.639309597015381,
      "epoch": 0.03452243958573072,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004644671132238558,
      "loss": 5.7013,
      "mean_token_accuracy": 0.18279279619455338,
      "num_tokens": 1311484.0,
      "step": 1545
    },
    {
      "entropy": 6.000847482681275,
      "epoch": 0.03463416269118617,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046382910184678585,
      "loss": 5.9404,
      "mean_token_accuracy": 0.17179854363203048,
      "num_tokens": 1316093.0,
      "step": 1550
    },
    {
      "entropy": 5.936440086364746,
      "epoch": 0.034745885796641604,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004631859136245025,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17483644485473632,
      "num_tokens": 1321001.0,
      "step": 1555
    },
    {
      "entropy": 5.812693357467651,
      "epoch": 0.03485760890209704,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004625375661903357,
      "loss": 5.7145,
      "mean_token_accuracy": 0.18931587934494018,
      "num_tokens": 1325069.0,
      "step": 1560
    },
    {
      "entropy": 5.8191815376281735,
      "epoch": 0.034969332007552485,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046188407731905787,
      "loss": 5.6621,
      "mean_token_accuracy": 0.175711190700531,
      "num_tokens": 1328887.0,
      "step": 1565
    },
    {
      "entropy": 5.904877519607544,
      "epoch": 0.03508105511300792,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046122546492639643,
      "loss": 5.9677,
      "mean_token_accuracy": 0.170508536696434,
      "num_tokens": 1333395.0,
      "step": 1570
    },
    {
      "entropy": 6.098869657516479,
      "epoch": 0.03519277821846336,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000460561747068543,
      "loss": 6.0506,
      "mean_token_accuracy": 0.16190842911601067,
      "num_tokens": 1337457.0,
      "step": 1575
    },
    {
      "entropy": 5.817472219467163,
      "epoch": 0.0353045013239188,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004598929419416578,
      "loss": 5.6626,
      "mean_token_accuracy": 0.18710167407989503,
      "num_tokens": 1341768.0,
      "step": 1580
    },
    {
      "entropy": 5.893204212188721,
      "epoch": 0.03541622442937424,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045921906788137123,
      "loss": 5.7164,
      "mean_token_accuracy": 0.19159406274557114,
      "num_tokens": 1346191.0,
      "step": 1585
    },
    {
      "entropy": 5.7586006164550785,
      "epoch": 0.03552794753482968,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045854014336228115,
      "loss": 5.7747,
      "mean_token_accuracy": 0.18223194628953934,
      "num_tokens": 1350228.0,
      "step": 1590
    },
    {
      "entropy": 5.8750648021698,
      "epoch": 0.035639670640285115,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045785618699744615,
      "loss": 5.8952,
      "mean_token_accuracy": 0.1681994952261448,
      "num_tokens": 1354722.0,
      "step": 1595
    },
    {
      "entropy": 5.9419848918914795,
      "epoch": 0.03575139374574056,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045716721753787543,
      "loss": 5.8673,
      "mean_token_accuracy": 0.1742061510682106,
      "num_tokens": 1358985.0,
      "step": 1600
    },
    {
      "entropy": 5.980447912216187,
      "epoch": 0.035863116851195996,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004564732538720148,
      "loss": 5.7788,
      "mean_token_accuracy": 0.17965463250875474,
      "num_tokens": 1363281.0,
      "step": 1605
    },
    {
      "entropy": 5.878961086273193,
      "epoch": 0.03597483995665143,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045577431502522877,
      "loss": 5.8718,
      "mean_token_accuracy": 0.1725798264145851,
      "num_tokens": 1367777.0,
      "step": 1610
    },
    {
      "entropy": 5.8722765922546385,
      "epoch": 0.03608656306210688,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004550704201592787,
      "loss": 5.9285,
      "mean_token_accuracy": 0.1609680324792862,
      "num_tokens": 1372537.0,
      "step": 1615
    },
    {
      "entropy": 5.778023958206177,
      "epoch": 0.036198286167562314,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004543615885717981,
      "loss": 5.6864,
      "mean_token_accuracy": 0.18914838135242462,
      "num_tokens": 1377341.0,
      "step": 1620
    },
    {
      "entropy": 5.8790647983551025,
      "epoch": 0.03631000927301775,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045364783969576296,
      "loss": 5.8012,
      "mean_token_accuracy": 0.17915536761283873,
      "num_tokens": 1381480.0,
      "step": 1625
    },
    {
      "entropy": 5.8201066017150875,
      "epoch": 0.03642173237847319,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004529291930989592,
      "loss": 5.7437,
      "mean_token_accuracy": 0.16947454288601876,
      "num_tokens": 1386057.0,
      "step": 1630
    },
    {
      "entropy": 5.769669675827027,
      "epoch": 0.03653345548392863,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004522056684834464,
      "loss": 5.7184,
      "mean_token_accuracy": 0.18831546157598494,
      "num_tokens": 1390405.0,
      "step": 1635
    },
    {
      "entropy": 5.873564338684082,
      "epoch": 0.03664517858938407,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004514772856850173,
      "loss": 5.7209,
      "mean_token_accuracy": 0.18445121347904206,
      "num_tokens": 1394972.0,
      "step": 1640
    },
    {
      "entropy": 5.762715196609497,
      "epoch": 0.03675690169483951,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004507440646726542,
      "loss": 5.6785,
      "mean_token_accuracy": 0.18628646433353424,
      "num_tokens": 1399270.0,
      "step": 1645
    },
    {
      "entropy": 5.652107191085816,
      "epoch": 0.03686862480029495,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004500060255479818,
      "loss": 5.6945,
      "mean_token_accuracy": 0.19375374019145966,
      "num_tokens": 1403506.0,
      "step": 1650
    },
    {
      "entropy": 5.74113359451294,
      "epoch": 0.03698034790575039,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004492631885447151,
      "loss": 5.8009,
      "mean_token_accuracy": 0.17932575345039367,
      "num_tokens": 1407493.0,
      "step": 1655
    },
    {
      "entropy": 5.981715536117553,
      "epoch": 0.037092071011205825,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044851557402810616,
      "loss": 5.7784,
      "mean_token_accuracy": 0.17953559160232543,
      "num_tokens": 1411871.0,
      "step": 1660
    },
    {
      "entropy": 5.7225306034088135,
      "epoch": 0.03720379411666127,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044776320249438444,
      "loss": 5.739,
      "mean_token_accuracy": 0.17465803027153015,
      "num_tokens": 1415951.0,
      "step": 1665
    },
    {
      "entropy": 5.740856504440307,
      "epoch": 0.037315517222116706,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044700609457019565,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1844460114836693,
      "num_tokens": 1420394.0,
      "step": 1670
    },
    {
      "entropy": 5.932256269454956,
      "epoch": 0.037427240327572144,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004462442710120359,
      "loss": 5.8594,
      "mean_token_accuracy": 0.17785218358039856,
      "num_tokens": 1424921.0,
      "step": 1675
    },
    {
      "entropy": 5.777454376220703,
      "epoch": 0.03753896343302759,
      "grad_norm": 1.421875,
      "learning_rate": 0.000445477752705683,
      "loss": 5.7103,
      "mean_token_accuracy": 0.1833161398768425,
      "num_tokens": 1428638.0,
      "step": 1680
    },
    {
      "entropy": 5.9318077087402346,
      "epoch": 0.037650686538483025,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044470656066562336,
      "loss": 5.7964,
      "mean_token_accuracy": 0.17520285099744798,
      "num_tokens": 1432780.0,
      "step": 1685
    },
    {
      "entropy": 5.804677724838257,
      "epoch": 0.03776240964393846,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004439307160344765,
      "loss": 5.7169,
      "mean_token_accuracy": 0.18259765058755875,
      "num_tokens": 1437178.0,
      "step": 1690
    },
    {
      "entropy": 5.928979682922363,
      "epoch": 0.0378741327493939,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00044315024008241473,
      "loss": 5.7682,
      "mean_token_accuracy": 0.17713233679533005,
      "num_tokens": 1441040.0,
      "step": 1695
    },
    {
      "entropy": 5.898324060440063,
      "epoch": 0.03798585585484934,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004423651542065806,
      "loss": 5.7677,
      "mean_token_accuracy": 0.17906588166952134,
      "num_tokens": 1445530.0,
      "step": 1700
    },
    {
      "entropy": 5.927756214141846,
      "epoch": 0.03809757896030478,
      "grad_norm": 1.375,
      "learning_rate": 0.00044157547993050006,
      "loss": 5.8536,
      "mean_token_accuracy": 0.1840515822172165,
      "num_tokens": 1449623.0,
      "step": 1705
    },
    {
      "entropy": 5.821538782119751,
      "epoch": 0.03820930206576022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044078123890349227,
      "loss": 5.6882,
      "mean_token_accuracy": 0.1885247677564621,
      "num_tokens": 1454167.0,
      "step": 1710
    },
    {
      "entropy": 5.715845966339112,
      "epoch": 0.03832102517121566,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043998245290007606,
      "loss": 5.6951,
      "mean_token_accuracy": 0.1866699069738388,
      "num_tokens": 1458666.0,
      "step": 1715
    },
    {
      "entropy": 5.808634424209595,
      "epoch": 0.0384327482766711,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043917914381937323,
      "loss": 5.7194,
      "mean_token_accuracy": 0.1920030653476715,
      "num_tokens": 1463253.0,
      "step": 1720
    },
    {
      "entropy": 5.779128122329712,
      "epoch": 0.038544471382126536,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043837133368450815,
      "loss": 5.7039,
      "mean_token_accuracy": 0.17917657792568206,
      "num_tokens": 1467487.0,
      "step": 1725
    },
    {
      "entropy": 5.916869115829468,
      "epoch": 0.03865619448758198,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004375590446420037,
      "loss": 5.7899,
      "mean_token_accuracy": 0.17375679612159728,
      "num_tokens": 1471408.0,
      "step": 1730
    },
    {
      "entropy": 5.81466851234436,
      "epoch": 0.03876791759303742,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004367422989611743,
      "loss": 5.5575,
      "mean_token_accuracy": 0.20052251368761062,
      "num_tokens": 1475296.0,
      "step": 1735
    },
    {
      "entropy": 5.8537859439849855,
      "epoch": 0.038879640698492854,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004359211190335153,
      "loss": 5.7213,
      "mean_token_accuracy": 0.17733037620782852,
      "num_tokens": 1479593.0,
      "step": 1740
    },
    {
      "entropy": 5.804961729049682,
      "epoch": 0.0389913638039483,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043509552737208923,
      "loss": 5.644,
      "mean_token_accuracy": 0.19278691858053207,
      "num_tokens": 1483587.0,
      "step": 1745
    },
    {
      "entropy": 5.787391281127929,
      "epoch": 0.039103086909403735,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043426554661090853,
      "loss": 5.7231,
      "mean_token_accuracy": 0.18333306312561035,
      "num_tokens": 1488152.0,
      "step": 1750
    },
    {
      "entropy": 5.771423101425171,
      "epoch": 0.03921481001485917,
      "grad_norm": 1.484375,
      "learning_rate": 0.00043343119950431516,
      "loss": 5.7442,
      "mean_token_accuracy": 0.17990843206644058,
      "num_tokens": 1492511.0,
      "step": 1755
    },
    {
      "entropy": 5.883594274520874,
      "epoch": 0.03932653312031461,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043259250892635644,
      "loss": 5.686,
      "mean_token_accuracy": 0.1835270807147026,
      "num_tokens": 1496634.0,
      "step": 1760
    },
    {
      "entropy": 5.877381658554077,
      "epoch": 0.03943825622577005,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004317494978701582,
      "loss": 5.7195,
      "mean_token_accuracy": 0.18765006512403487,
      "num_tokens": 1500904.0,
      "step": 1765
    },
    {
      "entropy": 5.817433214187622,
      "epoch": 0.03954997933122549,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004309021894472943,
      "loss": 5.641,
      "mean_token_accuracy": 0.19316388219594954,
      "num_tokens": 1504612.0,
      "step": 1770
    },
    {
      "entropy": 5.7231629371643065,
      "epoch": 0.03966170243668093,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004300506068871534,
      "loss": 5.6771,
      "mean_token_accuracy": 0.18761029839515686,
      "num_tokens": 1508867.0,
      "step": 1775
    },
    {
      "entropy": 5.717128753662109,
      "epoch": 0.03977342554213637,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00042919477353630135,
      "loss": 5.4243,
      "mean_token_accuracy": 0.21169085800647736,
      "num_tokens": 1512355.0,
      "step": 1780
    },
    {
      "entropy": 5.604061508178711,
      "epoch": 0.03988514864759181,
      "grad_norm": 1.40625,
      "learning_rate": 0.000428334712857842,
      "loss": 5.6445,
      "mean_token_accuracy": 0.1938673362135887,
      "num_tokens": 1516609.0,
      "step": 1785
    },
    {
      "entropy": 5.717974853515625,
      "epoch": 0.039996871753047246,
      "grad_norm": 1.375,
      "learning_rate": 0.00042747044843077304,
      "loss": 5.8014,
      "mean_token_accuracy": 0.1799890711903572,
      "num_tokens": 1520945.0,
      "step": 1790
    },
    {
      "entropy": 5.886101341247558,
      "epoch": 0.04010859485850269,
      "grad_norm": 1.40625,
      "learning_rate": 0.00042660200394934047,
      "loss": 5.8089,
      "mean_token_accuracy": 0.1731388159096241,
      "num_tokens": 1525136.0,
      "step": 1795
    },
    {
      "entropy": 5.8484992504119875,
      "epoch": 0.04022031796395813,
      "grad_norm": 1.484375,
      "learning_rate": 0.00042572940322238844,
      "loss": 5.7564,
      "mean_token_accuracy": 0.17428047806024552,
      "num_tokens": 1529088.0,
      "step": 1800
    },
    {
      "entropy": 5.745491313934326,
      "epoch": 0.040332041069413564,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00042485267017270664,
      "loss": 5.7456,
      "mean_token_accuracy": 0.1762354075908661,
      "num_tokens": 1533923.0,
      "step": 1805
    },
    {
      "entropy": 5.763177680969238,
      "epoch": 0.040443764174869,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004239718288363745,
      "loss": 5.6023,
      "mean_token_accuracy": 0.19920098632574082,
      "num_tokens": 1538031.0,
      "step": 1810
    },
    {
      "entropy": 5.743065643310547,
      "epoch": 0.040555487280324445,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004230869033621023,
      "loss": 5.7189,
      "mean_token_accuracy": 0.17735423594713212,
      "num_tokens": 1542217.0,
      "step": 1815
    },
    {
      "entropy": 5.834994745254517,
      "epoch": 0.04066721038577988,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004221979180105688,
      "loss": 5.6623,
      "mean_token_accuracy": 0.18281597793102264,
      "num_tokens": 1546102.0,
      "step": 1820
    },
    {
      "entropy": 5.805409002304077,
      "epoch": 0.04077893349123532,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042130489715375645,
      "loss": 5.632,
      "mean_token_accuracy": 0.1972121849656105,
      "num_tokens": 1550292.0,
      "step": 1825
    },
    {
      "entropy": 5.683276176452637,
      "epoch": 0.040890656596690764,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00042040786527428335,
      "loss": 5.6792,
      "mean_token_accuracy": 0.183244788646698,
      "num_tokens": 1554231.0,
      "step": 1830
    },
    {
      "entropy": 5.609442949295044,
      "epoch": 0.0410023797021462,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004195068469647315,
      "loss": 5.6571,
      "mean_token_accuracy": 0.18493013978004455,
      "num_tokens": 1558101.0,
      "step": 1835
    },
    {
      "entropy": 5.659271860122681,
      "epoch": 0.04111410280760164,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041860186692697297,
      "loss": 5.6225,
      "mean_token_accuracy": 0.19041107445955277,
      "num_tokens": 1562735.0,
      "step": 1840
    },
    {
      "entropy": 5.774989414215088,
      "epoch": 0.04122582591305708,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041769294997149264,
      "loss": 5.5778,
      "mean_token_accuracy": 0.20120995193719865,
      "num_tokens": 1566953.0,
      "step": 1845
    },
    {
      "entropy": 5.70683217048645,
      "epoch": 0.04133754901851252,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004167801210167081,
      "loss": 5.615,
      "mean_token_accuracy": 0.19226820766925812,
      "num_tokens": 1571389.0,
      "step": 1850
    },
    {
      "entropy": 5.650122833251953,
      "epoch": 0.041449272123967956,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004158634050882861,
      "loss": 5.7396,
      "mean_token_accuracy": 0.1780322551727295,
      "num_tokens": 1575737.0,
      "step": 1855
    },
    {
      "entropy": 5.853532600402832,
      "epoch": 0.0415609952294234,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004149428273184569,
      "loss": 5.7569,
      "mean_token_accuracy": 0.18002261966466904,
      "num_tokens": 1580371.0,
      "step": 1860
    },
    {
      "entropy": 5.816565465927124,
      "epoch": 0.04167271833487884,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004140184129453253,
      "loss": 5.7605,
      "mean_token_accuracy": 0.17988699525594712,
      "num_tokens": 1584569.0,
      "step": 1865
    },
    {
      "entropy": 5.734528732299805,
      "epoch": 0.041784441440334275,
      "grad_norm": 1.296875,
      "learning_rate": 0.000413090187312178,
      "loss": 5.6903,
      "mean_token_accuracy": 0.19021912217140197,
      "num_tokens": 1588869.0,
      "step": 1870
    },
    {
      "entropy": 5.85476336479187,
      "epoch": 0.04189616454578971,
      "grad_norm": 1.375,
      "learning_rate": 0.0004121581758667898,
      "loss": 5.7116,
      "mean_token_accuracy": 0.1797335773706436,
      "num_tokens": 1593254.0,
      "step": 1875
    },
    {
      "entropy": 5.764283180236816,
      "epoch": 0.042007887651245156,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041122240416072533,
      "loss": 5.6083,
      "mean_token_accuracy": 0.1957382544875145,
      "num_tokens": 1597506.0,
      "step": 1880
    },
    {
      "entropy": 5.680334758758545,
      "epoch": 0.04211961075670059,
      "grad_norm": 1.375,
      "learning_rate": 0.0004102828978486385,
      "loss": 5.7782,
      "mean_token_accuracy": 0.18792699724435807,
      "num_tokens": 1601930.0,
      "step": 1885
    },
    {
      "entropy": 5.850901412963867,
      "epoch": 0.04223133386215603,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004093396826875695,
      "loss": 5.7666,
      "mean_token_accuracy": 0.1885473236441612,
      "num_tokens": 1606634.0,
      "step": 1890
    },
    {
      "entropy": 5.900485515594482,
      "epoch": 0.042343056967611474,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00040839278453623837,
      "loss": 5.6675,
      "mean_token_accuracy": 0.1905714750289917,
      "num_tokens": 1610457.0,
      "step": 1895
    },
    {
      "entropy": 5.699015712738037,
      "epoch": 0.04245478007306691,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004074422293543363,
      "loss": 5.5776,
      "mean_token_accuracy": 0.19200229197740554,
      "num_tokens": 1614315.0,
      "step": 1900
    },
    {
      "entropy": 5.615668153762817,
      "epoch": 0.04256650317852235,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004064880432018137,
      "loss": 5.7362,
      "mean_token_accuracy": 0.18060311526060105,
      "num_tokens": 1618613.0,
      "step": 1905
    },
    {
      "entropy": 5.884034299850464,
      "epoch": 0.04267822628397779,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00040553025223816615,
      "loss": 5.6203,
      "mean_token_accuracy": 0.18918341398239136,
      "num_tokens": 1622874.0,
      "step": 1910
    },
    {
      "entropy": 5.8201299667358395,
      "epoch": 0.04278994938943323,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040456888272171653,
      "loss": 5.5972,
      "mean_token_accuracy": 0.18606456071138383,
      "num_tokens": 1626799.0,
      "step": 1915
    },
    {
      "entropy": 5.58936185836792,
      "epoch": 0.04290167249488867,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040360396100889577,
      "loss": 5.6497,
      "mean_token_accuracy": 0.19720107764005662,
      "num_tokens": 1631082.0,
      "step": 1920
    },
    {
      "entropy": 5.884630441665649,
      "epoch": 0.043013395600344104,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004026355135535202,
      "loss": 5.6225,
      "mean_token_accuracy": 0.19111260622739792,
      "num_tokens": 1635075.0,
      "step": 1925
    },
    {
      "entropy": 5.834641075134277,
      "epoch": 0.04312511870579955,
      "grad_norm": 1.34375,
      "learning_rate": 0.000401663566906066,
      "loss": 5.6622,
      "mean_token_accuracy": 0.18582070171833037,
      "num_tokens": 1639663.0,
      "step": 1930
    },
    {
      "entropy": 5.706852054595947,
      "epoch": 0.043236841811254985,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040068814771294134,
      "loss": 5.8219,
      "mean_token_accuracy": 0.1685791566967964,
      "num_tokens": 1644285.0,
      "step": 1935
    },
    {
      "entropy": 5.769976663589477,
      "epoch": 0.04334856491671042,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003997092827157562,
      "loss": 5.6525,
      "mean_token_accuracy": 0.1879347950220108,
      "num_tokens": 1648542.0,
      "step": 1940
    },
    {
      "entropy": 5.851756381988525,
      "epoch": 0.043460288022165866,
      "grad_norm": 1.34375,
      "learning_rate": 0.000398726998750589,
      "loss": 5.7242,
      "mean_token_accuracy": 0.18249233961105346,
      "num_tokens": 1652982.0,
      "step": 1945
    },
    {
      "entropy": 5.843126392364502,
      "epoch": 0.0435720111276213,
      "grad_norm": 1.421875,
      "learning_rate": 0.00039774132274725076,
      "loss": 5.588,
      "mean_token_accuracy": 0.1961601123213768,
      "num_tokens": 1657111.0,
      "step": 1950
    },
    {
      "entropy": 5.722390270233154,
      "epoch": 0.04368373423307674,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039675228172854707,
      "loss": 5.55,
      "mean_token_accuracy": 0.2040131390094757,
      "num_tokens": 1661541.0,
      "step": 1955
    },
    {
      "entropy": 5.790045356750488,
      "epoch": 0.043795457338532184,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003957599028095371,
      "loss": 5.7184,
      "mean_token_accuracy": 0.18772038519382478,
      "num_tokens": 1665606.0,
      "step": 1960
    },
    {
      "entropy": 5.723642206192016,
      "epoch": 0.04390718044398762,
      "grad_norm": 1.546875,
      "learning_rate": 0.00039476421319679017,
      "loss": 5.5654,
      "mean_token_accuracy": 0.19957753419876098,
      "num_tokens": 1670110.0,
      "step": 1965
    },
    {
      "entropy": 5.727631521224976,
      "epoch": 0.04401890354944306,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039376524018764,
      "loss": 5.7035,
      "mean_token_accuracy": 0.18647708594799042,
      "num_tokens": 1674724.0,
      "step": 1970
    },
    {
      "entropy": 5.6798807144165036,
      "epoch": 0.0441306266548985,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039276301116943616,
      "loss": 5.5037,
      "mean_token_accuracy": 0.204180046916008,
      "num_tokens": 1679022.0,
      "step": 1975
    },
    {
      "entropy": 5.791165924072265,
      "epoch": 0.04424234976035394,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003917575536187936,
      "loss": 5.6215,
      "mean_token_accuracy": 0.18042138814926148,
      "num_tokens": 1683192.0,
      "step": 1980
    },
    {
      "entropy": 5.757804203033447,
      "epoch": 0.04435407286580938,
      "grad_norm": 1.421875,
      "learning_rate": 0.00039074889510083894,
      "loss": 5.5671,
      "mean_token_accuracy": 0.19676164537668228,
      "num_tokens": 1687442.0,
      "step": 1985
    },
    {
      "entropy": 5.7241490364074705,
      "epoch": 0.044465795971264814,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038973706326845495,
      "loss": 5.6109,
      "mean_token_accuracy": 0.19887208938598633,
      "num_tokens": 1691528.0,
      "step": 1990
    },
    {
      "entropy": 5.706973314285278,
      "epoch": 0.04457751907672026,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003887220858615225,
      "loss": 5.7249,
      "mean_token_accuracy": 0.18426525145769118,
      "num_tokens": 1695936.0,
      "step": 1995
    },
    {
      "entropy": 5.724706077575684,
      "epoch": 0.044689242182175695,
      "grad_norm": 1.25,
      "learning_rate": 0.0003877039907061597,
      "loss": 5.6597,
      "mean_token_accuracy": 0.18481703847646713,
      "num_tokens": 1700729.0,
      "step": 2000
    },
    {
      "entropy": 5.661561107635498,
      "epoch": 0.04480096528763113,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003866828057139598,
      "loss": 5.6118,
      "mean_token_accuracy": 0.19120104908943175,
      "num_tokens": 1704962.0,
      "step": 2005
    },
    {
      "entropy": 5.825762557983398,
      "epoch": 0.044912688393086576,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00038565855888122503,
      "loss": 5.7476,
      "mean_token_accuracy": 0.18562862128019333,
      "num_tokens": 1709408.0,
      "step": 2010
    },
    {
      "entropy": 5.708235836029052,
      "epoch": 0.045024411498542014,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038463127828819975,
      "loss": 5.555,
      "mean_token_accuracy": 0.2008541688323021,
      "num_tokens": 1713894.0,
      "step": 2015
    },
    {
      "entropy": 5.700809955596924,
      "epoch": 0.04513613460399745,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038360099209830043,
      "loss": 5.5303,
      "mean_token_accuracy": 0.1931731179356575,
      "num_tokens": 1718477.0,
      "step": 2020
    },
    {
      "entropy": 5.6299162864685055,
      "epoch": 0.045247857709452895,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003825677285573433,
      "loss": 5.5068,
      "mean_token_accuracy": 0.20083038061857222,
      "num_tokens": 1722438.0,
      "step": 2025
    },
    {
      "entropy": 5.582123613357544,
      "epoch": 0.04535958081490833,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038153151599277027,
      "loss": 5.6107,
      "mean_token_accuracy": 0.18591762483119964,
      "num_tokens": 1726360.0,
      "step": 2030
    },
    {
      "entropy": 5.639498996734619,
      "epoch": 0.04547130392036377,
      "grad_norm": 1.375,
      "learning_rate": 0.0003804923828128723,
      "loss": 5.5953,
      "mean_token_accuracy": 0.1981348842382431,
      "num_tokens": 1730550.0,
      "step": 2035
    },
    {
      "entropy": 5.841020154953003,
      "epoch": 0.04558302702581921,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003794503575060104,
      "loss": 5.6303,
      "mean_token_accuracy": 0.19250714033842087,
      "num_tokens": 1734685.0,
      "step": 2040
    },
    {
      "entropy": 5.816658306121826,
      "epoch": 0.04569475013127465,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037840546863983484,
      "loss": 5.6727,
      "mean_token_accuracy": 0.18337119072675706,
      "num_tokens": 1738714.0,
      "step": 2045
    },
    {
      "entropy": 5.6304834365844725,
      "epoch": 0.04580647323673009,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003773577448605015,
      "loss": 5.6487,
      "mean_token_accuracy": 0.18985280990600586,
      "num_tokens": 1742714.0,
      "step": 2050
    },
    {
      "entropy": 5.6639081001281735,
      "epoch": 0.045918196342185524,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003763072148918872,
      "loss": 5.5674,
      "mean_token_accuracy": 0.19057373851537704,
      "num_tokens": 1747053.0,
      "step": 2055
    },
    {
      "entropy": 5.765899181365967,
      "epoch": 0.04602991944764097,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003752539075348017,
      "loss": 5.6464,
      "mean_token_accuracy": 0.18599009662866592,
      "num_tokens": 1750766.0,
      "step": 2060
    },
    {
      "entropy": 5.633810901641846,
      "epoch": 0.046141642553096406,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037419785166619817,
      "loss": 5.5613,
      "mean_token_accuracy": 0.1995617777109146,
      "num_tokens": 1754835.0,
      "step": 2065
    },
    {
      "entropy": 5.703394365310669,
      "epoch": 0.04625336565855184,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003731390762383818,
      "loss": 5.6528,
      "mean_token_accuracy": 0.19336580187082292,
      "num_tokens": 1759088.0,
      "step": 2070
    },
    {
      "entropy": 5.638295269012451,
      "epoch": 0.04636508876400729,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003720776102782158,
      "loss": 5.4773,
      "mean_token_accuracy": 0.19688699394464493,
      "num_tokens": 1763493.0,
      "step": 2075
    },
    {
      "entropy": 5.639464426040649,
      "epoch": 0.046476811869462724,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037101348288632555,
      "loss": 5.6322,
      "mean_token_accuracy": 0.19464462846517563,
      "num_tokens": 1767859.0,
      "step": 2080
    },
    {
      "entropy": 5.74103889465332,
      "epoch": 0.04658853497491816,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003699467232363012,
      "loss": 5.7186,
      "mean_token_accuracy": 0.18696200251579284,
      "num_tokens": 1772448.0,
      "step": 2085
    },
    {
      "entropy": 5.743357467651367,
      "epoch": 0.046700258080373605,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003688773605738973,
      "loss": 5.5853,
      "mean_token_accuracy": 0.19198814928531646,
      "num_tokens": 1776420.0,
      "step": 2090
    },
    {
      "entropy": 5.695070838928222,
      "epoch": 0.04681198118582904,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036780542421623134,
      "loss": 5.4846,
      "mean_token_accuracy": 0.19678378999233245,
      "num_tokens": 1780503.0,
      "step": 2095
    },
    {
      "entropy": 5.665157270431519,
      "epoch": 0.04692370429128448,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003667309435509802,
      "loss": 5.5423,
      "mean_token_accuracy": 0.188940691947937,
      "num_tokens": 1784730.0,
      "step": 2100
    },
    {
      "entropy": 5.633287715911865,
      "epoch": 0.047035427396739916,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003656539480355741,
      "loss": 5.5475,
      "mean_token_accuracy": 0.19798828661441803,
      "num_tokens": 1789317.0,
      "step": 2105
    },
    {
      "entropy": 5.651819515228271,
      "epoch": 0.04714715050219536,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003645744671963891,
      "loss": 5.7429,
      "mean_token_accuracy": 0.18275324553251265,
      "num_tokens": 1793431.0,
      "step": 2110
    },
    {
      "entropy": 5.714201354980469,
      "epoch": 0.0472588736076508,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003634925306279376,
      "loss": 5.6055,
      "mean_token_accuracy": 0.18972734063863755,
      "num_tokens": 1797546.0,
      "step": 2115
    },
    {
      "entropy": 5.736758518218994,
      "epoch": 0.047370596713106235,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003624081679920574,
      "loss": 5.6071,
      "mean_token_accuracy": 0.18960150480270385,
      "num_tokens": 1801995.0,
      "step": 2120
    },
    {
      "entropy": 5.62506251335144,
      "epoch": 0.04748231981856168,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003613214090170977,
      "loss": 5.551,
      "mean_token_accuracy": 0.19271351546049117,
      "num_tokens": 1805957.0,
      "step": 2125
    },
    {
      "entropy": 5.496391439437867,
      "epoch": 0.047594042924017116,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003602322834971048,
      "loss": 5.5688,
      "mean_token_accuracy": 0.20091331452131272,
      "num_tokens": 1810216.0,
      "step": 2130
    },
    {
      "entropy": 5.57668776512146,
      "epoch": 0.04770576602947255,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003591408212910051,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1942553251981735,
      "num_tokens": 1814727.0,
      "step": 2135
    },
    {
      "entropy": 5.792499685287476,
      "epoch": 0.047817489134928,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003580470523217863,
      "loss": 5.5754,
      "mean_token_accuracy": 0.19351515620946885,
      "num_tokens": 1819076.0,
      "step": 2140
    },
    {
      "entropy": 5.799184322357178,
      "epoch": 0.047929212240383434,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003569510065756771,
      "loss": 5.6495,
      "mean_token_accuracy": 0.18864707350730897,
      "num_tokens": 1823289.0,
      "step": 2145
    },
    {
      "entropy": 5.420970630645752,
      "epoch": 0.04804093534583887,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003558527141013254,
      "loss": 5.4771,
      "mean_token_accuracy": 0.2005689561367035,
      "num_tokens": 1827301.0,
      "step": 2150
    },
    {
      "entropy": 5.653721857070923,
      "epoch": 0.048152658451294315,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003547522050089742,
      "loss": 5.6337,
      "mean_token_accuracy": 0.1886797547340393,
      "num_tokens": 1831300.0,
      "step": 2155
    },
    {
      "entropy": 5.715897750854492,
      "epoch": 0.04826438155674975,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035364950946963606,
      "loss": 5.4187,
      "mean_token_accuracy": 0.20591014325618745,
      "num_tokens": 1835445.0,
      "step": 2160
    },
    {
      "entropy": 5.697785568237305,
      "epoch": 0.04837610466220519,
      "grad_norm": 1.375,
      "learning_rate": 0.0003525446577142663,
      "loss": 5.6007,
      "mean_token_accuracy": 0.18906023651361464,
      "num_tokens": 1839773.0,
      "step": 2165
    },
    {
      "entropy": 5.764755582809448,
      "epoch": 0.04848782776766063,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035143768003293395,
      "loss": 5.5597,
      "mean_token_accuracy": 0.19898805320262908,
      "num_tokens": 1844011.0,
      "step": 2170
    },
    {
      "entropy": 5.754788970947265,
      "epoch": 0.04859955087311607,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003503286067739913,
      "loss": 5.6625,
      "mean_token_accuracy": 0.18419571667909623,
      "num_tokens": 1848063.0,
      "step": 2175
    },
    {
      "entropy": 5.781708574295044,
      "epoch": 0.04871127397857151,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034921746834324193,
      "loss": 5.7336,
      "mean_token_accuracy": 0.18319133371114732,
      "num_tokens": 1852403.0,
      "step": 2180
    },
    {
      "entropy": 5.65818943977356,
      "epoch": 0.048822997084026945,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003481042952031072,
      "loss": 5.5805,
      "mean_token_accuracy": 0.1911611184477806,
      "num_tokens": 1856680.0,
      "step": 2185
    },
    {
      "entropy": 5.830749273300171,
      "epoch": 0.04893472018948239,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003469891178717911,
      "loss": 5.7408,
      "mean_token_accuracy": 0.18272164165973664,
      "num_tokens": 1861166.0,
      "step": 2190
    },
    {
      "entropy": 5.721550703048706,
      "epoch": 0.049046443294937826,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003458719669224436,
      "loss": 5.5687,
      "mean_token_accuracy": 0.19427022337913513,
      "num_tokens": 1865334.0,
      "step": 2195
    },
    {
      "entropy": 5.672908163070678,
      "epoch": 0.04915816640039326,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003447528729823221,
      "loss": 5.5333,
      "mean_token_accuracy": 0.1998296007514,
      "num_tokens": 1869391.0,
      "step": 2200
    },
    {
      "entropy": 5.675740146636963,
      "epoch": 0.04926988950584871,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003436318667319525,
      "loss": 5.6662,
      "mean_token_accuracy": 0.18038018196821212,
      "num_tokens": 1873811.0,
      "step": 2205
    },
    {
      "entropy": 5.786928987503051,
      "epoch": 0.049381612611304145,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034250897890428716,
      "loss": 5.63,
      "mean_token_accuracy": 0.19128481298685074,
      "num_tokens": 1878269.0,
      "step": 2210
    },
    {
      "entropy": 5.77509298324585,
      "epoch": 0.04949333571675958,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003413842402838633,
      "loss": 5.7361,
      "mean_token_accuracy": 0.17976828813552856,
      "num_tokens": 1882800.0,
      "step": 2215
    },
    {
      "entropy": 5.720989131927491,
      "epoch": 0.049605058822215026,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034025768170595834,
      "loss": 5.6285,
      "mean_token_accuracy": 0.18693387508392334,
      "num_tokens": 1886968.0,
      "step": 2220
    },
    {
      "entropy": 5.70858998298645,
      "epoch": 0.04971678192767046,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003391293340557446,
      "loss": 5.6753,
      "mean_token_accuracy": 0.18215965777635573,
      "num_tokens": 1891158.0,
      "step": 2225
    },
    {
      "entropy": 5.543679046630859,
      "epoch": 0.0498285050331259,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003379992282674431,
      "loss": 5.3801,
      "mean_token_accuracy": 0.2130328193306923,
      "num_tokens": 1895431.0,
      "step": 2230
    },
    {
      "entropy": 5.666289806365967,
      "epoch": 0.04994022813858134,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003368673953234749,
      "loss": 5.5843,
      "mean_token_accuracy": 0.19352493286132813,
      "num_tokens": 1899750.0,
      "step": 2235
    },
    {
      "entropy": 5.670787143707275,
      "epoch": 0.05005195124403678,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033573386625361176,
      "loss": 5.4846,
      "mean_token_accuracy": 0.19894968271255492,
      "num_tokens": 1904076.0,
      "step": 2240
    },
    {
      "entropy": 5.586210012435913,
      "epoch": 0.05016367434949222,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033459867213412567,
      "loss": 5.6071,
      "mean_token_accuracy": 0.19177706241607667,
      "num_tokens": 1908859.0,
      "step": 2245
    },
    {
      "entropy": 5.4901470184326175,
      "epoch": 0.050275397454947655,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000333461844086937,
      "loss": 5.4025,
      "mean_token_accuracy": 0.20574979335069657,
      "num_tokens": 1913049.0,
      "step": 2250
    },
    {
      "entropy": 5.612983560562133,
      "epoch": 0.0503871205604031,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033232341327876097,
      "loss": 5.6216,
      "mean_token_accuracy": 0.18734033405780792,
      "num_tokens": 1917247.0,
      "step": 2255
    },
    {
      "entropy": 5.75532636642456,
      "epoch": 0.05049884366585854,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003311834109202531,
      "loss": 5.5987,
      "mean_token_accuracy": 0.1979236125946045,
      "num_tokens": 1920958.0,
      "step": 2260
    },
    {
      "entropy": 5.552858877182007,
      "epoch": 0.050610566771313974,
      "grad_norm": 1.375,
      "learning_rate": 0.00033004186826515416,
      "loss": 5.5027,
      "mean_token_accuracy": 0.19557077586650848,
      "num_tokens": 1925184.0,
      "step": 2265
    },
    {
      "entropy": 5.63628921508789,
      "epoch": 0.05072228987676942,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003288988166094324,
      "loss": 5.4803,
      "mean_token_accuracy": 0.1963364765048027,
      "num_tokens": 1929392.0,
      "step": 2270
    },
    {
      "entropy": 5.666030311584473,
      "epoch": 0.050834012982224855,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032775428729042656,
      "loss": 5.5791,
      "mean_token_accuracy": 0.20280695855617523,
      "num_tokens": 1933535.0,
      "step": 2275
    },
    {
      "entropy": 5.651852464675903,
      "epoch": 0.05094573608768029,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000326608311685986,
      "loss": 5.4906,
      "mean_token_accuracy": 0.19224414229393005,
      "num_tokens": 1937830.0,
      "step": 2280
    },
    {
      "entropy": 5.550401639938355,
      "epoch": 0.05105745919313573,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003254609212136108,
      "loss": 5.4906,
      "mean_token_accuracy": 0.20217254310846328,
      "num_tokens": 1941521.0,
      "step": 2285
    },
    {
      "entropy": 5.583224821090698,
      "epoch": 0.05116918229859117,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032431214732959036,
      "loss": 5.527,
      "mean_token_accuracy": 0.20249635577201844,
      "num_tokens": 1945662.0,
      "step": 2290
    },
    {
      "entropy": 5.619274425506592,
      "epoch": 0.05128090540404661,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000323162021528141,
      "loss": 5.3316,
      "mean_token_accuracy": 0.19660361409187316,
      "num_tokens": 1949529.0,
      "step": 2295
    },
    {
      "entropy": 5.52543077468872,
      "epoch": 0.05139262850950205,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00032201057534054264,
      "loss": 5.3834,
      "mean_token_accuracy": 0.21025778502225875,
      "num_tokens": 1953541.0,
      "step": 2300
    },
    {
      "entropy": 5.502612018585205,
      "epoch": 0.05150435161495749,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032085784033427414,
      "loss": 5.4665,
      "mean_token_accuracy": 0.19938438087701799,
      "num_tokens": 1957863.0,
      "step": 2305
    },
    {
      "entropy": 5.612597036361694,
      "epoch": 0.05161607472041293,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003197038481121478,
      "loss": 5.491,
      "mean_token_accuracy": 0.19755251854658126,
      "num_tokens": 1962085.0,
      "step": 2310
    },
    {
      "entropy": 5.710309934616089,
      "epoch": 0.051727797825868366,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003185486303114436,
      "loss": 5.4584,
      "mean_token_accuracy": 0.20661440938711167,
      "num_tokens": 1966707.0,
      "step": 2315
    },
    {
      "entropy": 5.611154699325562,
      "epoch": 0.05183952093132381,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003173922186030409,
      "loss": 5.5871,
      "mean_token_accuracy": 0.18784187585115433,
      "num_tokens": 1970461.0,
      "step": 2320
    },
    {
      "entropy": 5.628113603591919,
      "epoch": 0.05195124403677925,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000316234644690551,
      "loss": 5.5058,
      "mean_token_accuracy": 0.19585638344287873,
      "num_tokens": 1974611.0,
      "step": 2325
    },
    {
      "entropy": 5.521657705307007,
      "epoch": 0.052062967142234684,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003150759403094473,
      "loss": 5.3594,
      "mean_token_accuracy": 0.21671968698501587,
      "num_tokens": 1978644.0,
      "step": 2330
    },
    {
      "entropy": 5.493622589111328,
      "epoch": 0.05217469024769013,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031391613722619587,
      "loss": 5.3684,
      "mean_token_accuracy": 0.20494887232780457,
      "num_tokens": 1983045.0,
      "step": 2335
    },
    {
      "entropy": 5.592142248153687,
      "epoch": 0.052286413353145565,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000312755267237384,
      "loss": 5.4356,
      "mean_token_accuracy": 0.19318086951971053,
      "num_tokens": 1987046.0,
      "step": 2340
    },
    {
      "entropy": 5.719300317764282,
      "epoch": 0.052398136458601,
      "grad_norm": 1.375,
      "learning_rate": 0.0003115933621688488,
      "loss": 5.7009,
      "mean_token_accuracy": 0.18828426897525788,
      "num_tokens": 1992201.0,
      "step": 2345
    },
    {
      "entropy": 5.798365688323974,
      "epoch": 0.05250985956405644,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031043045387480487,
      "loss": 5.525,
      "mean_token_accuracy": 0.19892065227031708,
      "num_tokens": 1996124.0,
      "step": 2350
    },
    {
      "entropy": 5.615959596633911,
      "epoch": 0.052621582669511884,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003092665742369703,
      "loss": 5.5271,
      "mean_token_accuracy": 0.1865956410765648,
      "num_tokens": 2000602.0,
      "step": 2355
    },
    {
      "entropy": 5.663578510284424,
      "epoch": 0.05273330577496732,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030810175516369343,
      "loss": 5.5947,
      "mean_token_accuracy": 0.18611914217472075,
      "num_tokens": 2005538.0,
      "step": 2360
    },
    {
      "entropy": 5.671791028976441,
      "epoch": 0.05284502888042276,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003069360285890775,
      "loss": 5.5075,
      "mean_token_accuracy": 0.19378492385149002,
      "num_tokens": 2010158.0,
      "step": 2365
    },
    {
      "entropy": 5.567385244369507,
      "epoch": 0.0529567519858782,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00030576942647210547,
      "loss": 5.4607,
      "mean_token_accuracy": 0.20393914580345154,
      "num_tokens": 2013993.0,
      "step": 2370
    },
    {
      "entropy": 5.586142444610596,
      "epoch": 0.05306847509133364,
      "grad_norm": 1.375,
      "learning_rate": 0.00030460198079576355,
      "loss": 5.5127,
      "mean_token_accuracy": 0.19779219925403596,
      "num_tokens": 2018312.0,
      "step": 2375
    },
    {
      "entropy": 5.7746532440185545,
      "epoch": 0.053180198196789076,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003034337235661648,
      "loss": 5.6479,
      "mean_token_accuracy": 0.18985842019319535,
      "num_tokens": 2022957.0,
      "step": 2380
    },
    {
      "entropy": 5.652853965759277,
      "epoch": 0.05329192130224452,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003022646868116714,
      "loss": 5.5748,
      "mean_token_accuracy": 0.1886435940861702,
      "num_tokens": 2027546.0,
      "step": 2385
    },
    {
      "entropy": 5.62093677520752,
      "epoch": 0.05340364440769996,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003010949025820163,
      "loss": 5.3653,
      "mean_token_accuracy": 0.19798131585121154,
      "num_tokens": 2031370.0,
      "step": 2390
    },
    {
      "entropy": 5.60447211265564,
      "epoch": 0.053515367513155394,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002999244029474252,
      "loss": 5.5451,
      "mean_token_accuracy": 0.19064643681049348,
      "num_tokens": 2036314.0,
      "step": 2395
    },
    {
      "entropy": 5.684001779556274,
      "epoch": 0.05362709061861083,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029875321999773684,
      "loss": 5.5181,
      "mean_token_accuracy": 0.19244933873414993,
      "num_tokens": 2040819.0,
      "step": 2400
    },
    {
      "entropy": 5.607943344116211,
      "epoch": 0.053738813724066276,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029758138584152333,
      "loss": 5.4871,
      "mean_token_accuracy": 0.2002251401543617,
      "num_tokens": 2045230.0,
      "step": 2405
    },
    {
      "entropy": 5.654977750778198,
      "epoch": 0.05385053682952171,
      "grad_norm": 1.5,
      "learning_rate": 0.0002964089326052102,
      "loss": 5.5119,
      "mean_token_accuracy": 0.2028552696108818,
      "num_tokens": 2049688.0,
      "step": 2410
    },
    {
      "entropy": 5.570733118057251,
      "epoch": 0.05396225993497715,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002952358924321949,
      "loss": 5.4707,
      "mean_token_accuracy": 0.20194039940834047,
      "num_tokens": 2053811.0,
      "step": 2415
    },
    {
      "entropy": 5.535793685913086,
      "epoch": 0.054073983040432594,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029406229748196657,
      "loss": 5.5214,
      "mean_token_accuracy": 0.1908762738108635,
      "num_tokens": 2058018.0,
      "step": 2420
    },
    {
      "entropy": 5.634843015670777,
      "epoch": 0.05418570614588803,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002928881799292235,
      "loss": 5.5395,
      "mean_token_accuracy": 0.18729523271322251,
      "num_tokens": 2062258.0,
      "step": 2425
    },
    {
      "entropy": 5.616613721847534,
      "epoch": 0.05429742925134347,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029171357196299154,
      "loss": 5.5708,
      "mean_token_accuracy": 0.18940937966108323,
      "num_tokens": 2066465.0,
      "step": 2430
    },
    {
      "entropy": 5.5524194717407225,
      "epoch": 0.05440915235679891,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002905385057857414,
      "loss": 5.4176,
      "mean_token_accuracy": 0.20998302698135377,
      "num_tokens": 2070679.0,
      "step": 2435
    },
    {
      "entropy": 5.582164573669433,
      "epoch": 0.05452087546225435,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002893630136125058,
      "loss": 5.4877,
      "mean_token_accuracy": 0.19743818342685698,
      "num_tokens": 2075443.0,
      "step": 2440
    },
    {
      "entropy": 5.582848262786865,
      "epoch": 0.054632598567709786,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002881871276699967,
      "loss": 5.4972,
      "mean_token_accuracy": 0.18803117871284486,
      "num_tokens": 2079726.0,
      "step": 2445
    },
    {
      "entropy": 5.485451030731201,
      "epoch": 0.05474432167316523,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028701088019572114,
      "loss": 5.432,
      "mean_token_accuracy": 0.19790401607751845,
      "num_tokens": 2083868.0,
      "step": 2450
    },
    {
      "entropy": 5.5643316268920895,
      "epoch": 0.05485604477862067,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002858343034370977,
      "loss": 5.4652,
      "mean_token_accuracy": 0.19667036533355714,
      "num_tokens": 2087722.0,
      "step": 2455
    },
    {
      "entropy": 5.629821634292602,
      "epoch": 0.054967767884076105,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028465742965057267,
      "loss": 5.3771,
      "mean_token_accuracy": 0.21432919651269913,
      "num_tokens": 2092293.0,
      "step": 2460
    },
    {
      "entropy": 5.548560333251953,
      "epoch": 0.05507949098953154,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028348029110073533,
      "loss": 5.5023,
      "mean_token_accuracy": 0.2013688862323761,
      "num_tokens": 2096715.0,
      "step": 2465
    },
    {
      "entropy": 5.694032478332519,
      "epoch": 0.055191214094986986,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028230292005943365,
      "loss": 5.5699,
      "mean_token_accuracy": 0.1971820428967476,
      "num_tokens": 2100810.0,
      "step": 2470
    },
    {
      "entropy": 5.648245429992675,
      "epoch": 0.05530293720044242,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028112534880488945,
      "loss": 5.4822,
      "mean_token_accuracy": 0.18982537537813188,
      "num_tokens": 2105125.0,
      "step": 2475
    },
    {
      "entropy": 5.604338979721069,
      "epoch": 0.05541466030589786,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002799476096208137,
      "loss": 5.5598,
      "mean_token_accuracy": 0.19004852026700975,
      "num_tokens": 2109218.0,
      "step": 2480
    },
    {
      "entropy": 5.593366622924805,
      "epoch": 0.055526383411353304,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027876973479552087,
      "loss": 5.4758,
      "mean_token_accuracy": 0.19235745221376419,
      "num_tokens": 2113361.0,
      "step": 2485
    },
    {
      "entropy": 5.575938034057617,
      "epoch": 0.05563810651680874,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027759175662104424,
      "loss": 5.4216,
      "mean_token_accuracy": 0.2013538047671318,
      "num_tokens": 2117457.0,
      "step": 2490
    },
    {
      "entropy": 5.648002290725708,
      "epoch": 0.05574982962226418,
      "grad_norm": 1.5,
      "learning_rate": 0.0002764137073922508,
      "loss": 5.4526,
      "mean_token_accuracy": 0.2085655465722084,
      "num_tokens": 2121198.0,
      "step": 2495
    },
    {
      "entropy": 5.62782769203186,
      "epoch": 0.05586155272771962,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027523561940595505,
      "loss": 5.5748,
      "mean_token_accuracy": 0.19579420536756514,
      "num_tokens": 2125742.0,
      "step": 2500
    },
    {
      "entropy": 5.591427230834961,
      "epoch": 0.05597327583317506,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002740575249600342,
      "loss": 5.4562,
      "mean_token_accuracy": 0.19945420026779176,
      "num_tokens": 2129851.0,
      "step": 2505
    },
    {
      "entropy": 5.560900545120239,
      "epoch": 0.0560849989386305,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027287945635254263,
      "loss": 5.2606,
      "mean_token_accuracy": 0.21626812666654588,
      "num_tokens": 2133814.0,
      "step": 2510
    },
    {
      "entropy": 5.648103046417236,
      "epoch": 0.05619672204408594,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027170144588082635,
      "loss": 5.5301,
      "mean_token_accuracy": 0.20282886773347855,
      "num_tokens": 2138191.0,
      "step": 2515
    },
    {
      "entropy": 5.618161869049072,
      "epoch": 0.05630844514954138,
      "grad_norm": 1.375,
      "learning_rate": 0.00027052352584063763,
      "loss": 5.5811,
      "mean_token_accuracy": 0.19190315306186675,
      "num_tokens": 2142144.0,
      "step": 2520
    },
    {
      "entropy": 5.691967725753784,
      "epoch": 0.056420168254996815,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026934572852524907,
      "loss": 5.4425,
      "mean_token_accuracy": 0.19738697111606598,
      "num_tokens": 2146308.0,
      "step": 2525
    },
    {
      "entropy": 5.726579999923706,
      "epoch": 0.05653189136045225,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026816808622456937,
      "loss": 5.5617,
      "mean_token_accuracy": 0.18826697170734405,
      "num_tokens": 2150810.0,
      "step": 2530
    },
    {
      "entropy": 5.6576322555542,
      "epoch": 0.056643614465907696,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002669906312242569,
      "loss": 5.5467,
      "mean_token_accuracy": 0.18115203231573104,
      "num_tokens": 2154865.0,
      "step": 2535
    },
    {
      "entropy": 5.487837648391723,
      "epoch": 0.05675533757136313,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026581339580483525,
      "loss": 5.4071,
      "mean_token_accuracy": 0.20803821682929993,
      "num_tokens": 2159045.0,
      "step": 2540
    },
    {
      "entropy": 5.549320459365845,
      "epoch": 0.05686706067681857,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002646364122408082,
      "loss": 5.4352,
      "mean_token_accuracy": 0.20072186142206191,
      "num_tokens": 2163151.0,
      "step": 2545
    },
    {
      "entropy": 5.503663015365601,
      "epoch": 0.056978783782274015,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002634597127997749,
      "loss": 5.5074,
      "mean_token_accuracy": 0.20518606156110764,
      "num_tokens": 2167099.0,
      "step": 2550
    },
    {
      "entropy": 5.607196474075318,
      "epoch": 0.05709050688772945,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002622833297415445,
      "loss": 5.5578,
      "mean_token_accuracy": 0.18713489323854446,
      "num_tokens": 2171299.0,
      "step": 2555
    },
    {
      "entropy": 5.678381443023682,
      "epoch": 0.05720222999318489,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002611072953172531,
      "loss": 5.5743,
      "mean_token_accuracy": 0.19254956394433975,
      "num_tokens": 2175491.0,
      "step": 2560
    },
    {
      "entropy": 5.659079456329346,
      "epoch": 0.05731395309864033,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025993164176847845,
      "loss": 5.3514,
      "mean_token_accuracy": 0.19644243270158768,
      "num_tokens": 2179677.0,
      "step": 2565
    },
    {
      "entropy": 5.631671285629272,
      "epoch": 0.05742567620409577,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002587564013263564,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1980074465274811,
      "num_tokens": 2184038.0,
      "step": 2570
    },
    {
      "entropy": 5.626058292388916,
      "epoch": 0.05753739930955121,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002575816062106974,
      "loss": 5.4796,
      "mean_token_accuracy": 0.19614142179489136,
      "num_tokens": 2188278.0,
      "step": 2575
    },
    {
      "entropy": 5.6068353176116945,
      "epoch": 0.057649122415006644,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025640728862910293,
      "loss": 5.6207,
      "mean_token_accuracy": 0.1917508587241173,
      "num_tokens": 2192402.0,
      "step": 2580
    },
    {
      "entropy": 5.462264633178711,
      "epoch": 0.05776084552046209,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025523348077608285,
      "loss": 5.3489,
      "mean_token_accuracy": 0.20899475514888763,
      "num_tokens": 2196631.0,
      "step": 2585
    },
    {
      "entropy": 5.59301986694336,
      "epoch": 0.057872568625917525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025406021483217225,
      "loss": 5.4389,
      "mean_token_accuracy": 0.20359549969434737,
      "num_tokens": 2201016.0,
      "step": 2590
    },
    {
      "entropy": 5.567645835876465,
      "epoch": 0.05798429173137296,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025288752296304963,
      "loss": 5.3945,
      "mean_token_accuracy": 0.21075689047574997,
      "num_tokens": 2205091.0,
      "step": 2595
    },
    {
      "entropy": 5.641946744918823,
      "epoch": 0.05809601483682841,
      "grad_norm": 1.359375,
      "learning_rate": 0.000251715437318655,
      "loss": 5.4691,
      "mean_token_accuracy": 0.19774311035871506,
      "num_tokens": 2209525.0,
      "step": 2600
    },
    {
      "entropy": 5.582809400558472,
      "epoch": 0.058207737942283844,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002505439900323084,
      "loss": 5.42,
      "mean_token_accuracy": 0.20146647095680237,
      "num_tokens": 2214093.0,
      "step": 2605
    },
    {
      "entropy": 5.5769775867462155,
      "epoch": 0.05831946104773928,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024937321321982894,
      "loss": 5.487,
      "mean_token_accuracy": 0.19734087586402893,
      "num_tokens": 2218619.0,
      "step": 2610
    },
    {
      "entropy": 5.571866989135742,
      "epoch": 0.058431184153194725,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024820313897865433,
      "loss": 5.3724,
      "mean_token_accuracy": 0.20756171494722367,
      "num_tokens": 2222953.0,
      "step": 2615
    },
    {
      "entropy": 5.536918497085571,
      "epoch": 0.05854290725865016,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00024703379938696105,
      "loss": 5.4559,
      "mean_token_accuracy": 0.197811096906662,
      "num_tokens": 2227162.0,
      "step": 2620
    },
    {
      "entropy": 5.575473737716675,
      "epoch": 0.0586546303641056,
      "grad_norm": 1.296875,
      "learning_rate": 0.00024586522650278447,
      "loss": 5.42,
      "mean_token_accuracy": 0.1999548003077507,
      "num_tokens": 2231333.0,
      "step": 2625
    },
    {
      "entropy": 5.648802614212036,
      "epoch": 0.05876635346956104,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024469745236314064,
      "loss": 5.5065,
      "mean_token_accuracy": 0.18912915736436844,
      "num_tokens": 2235160.0,
      "step": 2630
    },
    {
      "entropy": 5.6090099811553955,
      "epoch": 0.05887807657501648,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00024353050898314767,
      "loss": 5.5155,
      "mean_token_accuracy": 0.19891011267900466,
      "num_tokens": 2239699.0,
      "step": 2635
    },
    {
      "entropy": 5.730726051330566,
      "epoch": 0.05898979968047192,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024236442835514743,
      "loss": 5.4919,
      "mean_token_accuracy": 0.19372162222862244,
      "num_tokens": 2244146.0,
      "step": 2640
    },
    {
      "entropy": 5.6738152503967285,
      "epoch": 0.059101522785927355,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024119924244782965,
      "loss": 5.5331,
      "mean_token_accuracy": 0.19131239801645278,
      "num_tokens": 2248190.0,
      "step": 2645
    },
    {
      "entropy": 5.542245864868164,
      "epoch": 0.0592132458913828,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024003498320535462,
      "loss": 5.3719,
      "mean_token_accuracy": 0.21510188281536102,
      "num_tokens": 2252389.0,
      "step": 2650
    },
    {
      "entropy": 5.411032438278198,
      "epoch": 0.059324968996838236,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00023887168254647727,
      "loss": 5.3938,
      "mean_token_accuracy": 0.21622349321842194,
      "num_tokens": 2256430.0,
      "step": 2655
    },
    {
      "entropy": 5.474547338485718,
      "epoch": 0.05943669210229367,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023770937236367308,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1985387310385704,
      "num_tokens": 2260632.0,
      "step": 2660
    },
    {
      "entropy": 5.540546131134033,
      "epoch": 0.05954841520774912,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023654808452226278,
      "loss": 5.3951,
      "mean_token_accuracy": 0.20164178013801576,
      "num_tokens": 2264593.0,
      "step": 2665
    },
    {
      "entropy": 5.556380558013916,
      "epoch": 0.059660138313204554,
      "grad_norm": 1.421875,
      "learning_rate": 0.00023538785085953912,
      "loss": 5.3618,
      "mean_token_accuracy": 0.20641586035490037,
      "num_tokens": 2268994.0,
      "step": 2670
    },
    {
      "entropy": 5.585520076751709,
      "epoch": 0.05977186141865999,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023422870318389404,
      "loss": 5.4611,
      "mean_token_accuracy": 0.2091505765914917,
      "num_tokens": 2273355.0,
      "step": 2675
    },
    {
      "entropy": 5.48478889465332,
      "epoch": 0.059883584524115435,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002330706732739468,
      "loss": 5.4803,
      "mean_token_accuracy": 0.20879880338907242,
      "num_tokens": 2277618.0,
      "step": 2680
    },
    {
      "entropy": 5.571560049057007,
      "epoch": 0.05999530762957087,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023191379287767211,
      "loss": 5.4958,
      "mean_token_accuracy": 0.19661557972431182,
      "num_tokens": 2281751.0,
      "step": 2685
    },
    {
      "entropy": 5.614952898025512,
      "epoch": 0.06010703073502631,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002307580937115305,
      "loss": 5.5625,
      "mean_token_accuracy": 0.18459359258413316,
      "num_tokens": 2286581.0,
      "step": 2690
    },
    {
      "entropy": 5.566756916046143,
      "epoch": 0.06021875384048175,
      "grad_norm": 1.375,
      "learning_rate": 0.00022960360745959846,
      "loss": 5.4223,
      "mean_token_accuracy": 0.20224524140357972,
      "num_tokens": 2290733.0,
      "step": 2695
    },
    {
      "entropy": 5.630762529373169,
      "epoch": 0.06033047694593719,
      "grad_norm": 1.34375,
      "learning_rate": 0.00022845036577269972,
      "loss": 5.371,
      "mean_token_accuracy": 0.2050411343574524,
      "num_tokens": 2295150.0,
      "step": 2700
    },
    {
      "entropy": 5.594348001480102,
      "epoch": 0.06044220005139263,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00022729840026753777,
      "loss": 5.4489,
      "mean_token_accuracy": 0.20504524558782578,
      "num_tokens": 2299461.0,
      "step": 2705
    },
    {
      "entropy": 5.425336313247681,
      "epoch": 0.060553923156848065,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002261477425258287,
      "loss": 5.2897,
      "mean_token_accuracy": 0.21590072959661483,
      "num_tokens": 2303981.0,
      "step": 2710
    },
    {
      "entropy": 5.718755674362183,
      "epoch": 0.06066564626230351,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002249984240934358,
      "loss": 5.6543,
      "mean_token_accuracy": 0.18148277401924134,
      "num_tokens": 2308514.0,
      "step": 2715
    },
    {
      "entropy": 5.696688175201416,
      "epoch": 0.060777369367758946,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00022385047647950464,
      "loss": 5.5257,
      "mean_token_accuracy": 0.19327573478221893,
      "num_tokens": 2312803.0,
      "step": 2720
    },
    {
      "entropy": 5.5735913753509525,
      "epoch": 0.06088909247321438,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002227039311555986,
      "loss": 5.442,
      "mean_token_accuracy": 0.20054861307144164,
      "num_tokens": 2317136.0,
      "step": 2725
    },
    {
      "entropy": 5.624294853210449,
      "epoch": 0.06100081557866983,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002215588195548372,
      "loss": 5.4348,
      "mean_token_accuracy": 0.20062965750694275,
      "num_tokens": 2321064.0,
      "step": 2730
    },
    {
      "entropy": 5.593418312072754,
      "epoch": 0.061112538684125264,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022041517307103337,
      "loss": 5.374,
      "mean_token_accuracy": 0.20263641327619553,
      "num_tokens": 2324971.0,
      "step": 2735
    },
    {
      "entropy": 5.5768012523651125,
      "epoch": 0.0612242617895807,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002192730230578331,
      "loss": 5.481,
      "mean_token_accuracy": 0.19864298403263092,
      "num_tokens": 2329433.0,
      "step": 2740
    },
    {
      "entropy": 5.505443143844604,
      "epoch": 0.061335984895036146,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002181324008278559,
      "loss": 5.3796,
      "mean_token_accuracy": 0.21429672539234162,
      "num_tokens": 2333493.0,
      "step": 2745
    },
    {
      "entropy": 5.426680946350098,
      "epoch": 0.06144770800049158,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021699333765183655,
      "loss": 5.2397,
      "mean_token_accuracy": 0.21503425091505052,
      "num_tokens": 2337663.0,
      "step": 2750
    },
    {
      "entropy": 5.513641881942749,
      "epoch": 0.06155943110594702,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002158558647577673,
      "loss": 5.2693,
      "mean_token_accuracy": 0.2166907772421837,
      "num_tokens": 2342291.0,
      "step": 2755
    },
    {
      "entropy": 5.546995401382446,
      "epoch": 0.06167115421140246,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00021472001333004215,
      "loss": 5.4227,
      "mean_token_accuracy": 0.2089788571000099,
      "num_tokens": 2346660.0,
      "step": 2760
    },
    {
      "entropy": 5.567002630233764,
      "epoch": 0.0617828773168579,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00021358581450860186,
      "loss": 5.383,
      "mean_token_accuracy": 0.19727942496538162,
      "num_tokens": 2350605.0,
      "step": 2765
    },
    {
      "entropy": 5.573090124130249,
      "epoch": 0.06189460042231334,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002124532993880799,
      "loss": 5.3777,
      "mean_token_accuracy": 0.2063875451683998,
      "num_tokens": 2355125.0,
      "step": 2770
    },
    {
      "entropy": 5.603260803222656,
      "epoch": 0.062006323527768775,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00021132249901695044,
      "loss": 5.5299,
      "mean_token_accuracy": 0.18890753388404846,
      "num_tokens": 2359058.0,
      "step": 2775
    },
    {
      "entropy": 5.577706432342529,
      "epoch": 0.06211804663322422,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00021019344439667705,
      "loss": 5.4785,
      "mean_token_accuracy": 0.2110677868127823,
      "num_tokens": 2363326.0,
      "step": 2780
    },
    {
      "entropy": 5.587746477127075,
      "epoch": 0.062229769738679656,
      "grad_norm": 1.453125,
      "learning_rate": 0.00020906616648086213,
      "loss": 5.2759,
      "mean_token_accuracy": 0.21217510849237442,
      "num_tokens": 2367241.0,
      "step": 2785
    },
    {
      "entropy": 5.68141770362854,
      "epoch": 0.062341492844135094,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020794069617439942,
      "loss": 5.4647,
      "mean_token_accuracy": 0.1912406340241432,
      "num_tokens": 2371660.0,
      "step": 2790
    },
    {
      "entropy": 5.618228816986084,
      "epoch": 0.06245321594959054,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00020681706433262593,
      "loss": 5.5128,
      "mean_token_accuracy": 0.19959444403648377,
      "num_tokens": 2376083.0,
      "step": 2795
    },
    {
      "entropy": 5.53573784828186,
      "epoch": 0.06256493905504597,
      "grad_norm": 1.4375,
      "learning_rate": 0.00020569530176047602,
      "loss": 5.3969,
      "mean_token_accuracy": 0.20906035751104354,
      "num_tokens": 2380006.0,
      "step": 2800
    },
    {
      "entropy": 5.543073892593384,
      "epoch": 0.06267666216050141,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002045754392116374,
      "loss": 5.4287,
      "mean_token_accuracy": 0.21249081641435624,
      "num_tokens": 2384382.0,
      "step": 2805
    },
    {
      "entropy": 5.537890958786011,
      "epoch": 0.06278838526595686,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020345750738770757,
      "loss": 5.3584,
      "mean_token_accuracy": 0.20751726925373076,
      "num_tokens": 2388550.0,
      "step": 2810
    },
    {
      "entropy": 5.501313591003418,
      "epoch": 0.06290010837141229,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00020234153693735214,
      "loss": 5.3933,
      "mean_token_accuracy": 0.20292940586805344,
      "num_tokens": 2392189.0,
      "step": 2815
    },
    {
      "entropy": 5.453502511978149,
      "epoch": 0.06301183147686773,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002012275584554647,
      "loss": 5.3101,
      "mean_token_accuracy": 0.2178672045469284,
      "num_tokens": 2396016.0,
      "step": 2820
    },
    {
      "entropy": 5.507620620727539,
      "epoch": 0.06312355458232317,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00020011560248232803,
      "loss": 5.4116,
      "mean_token_accuracy": 0.19674886614084244,
      "num_tokens": 2400053.0,
      "step": 2825
    },
    {
      "entropy": 5.616665172576904,
      "epoch": 0.0632352776877786,
      "grad_norm": 1.46875,
      "learning_rate": 0.00019900569950277692,
      "loss": 5.511,
      "mean_token_accuracy": 0.1812056928873062,
      "num_tokens": 2404164.0,
      "step": 2830
    },
    {
      "entropy": 5.551771831512451,
      "epoch": 0.06334700079323405,
      "grad_norm": 1.265625,
      "learning_rate": 0.00019789787994536228,
      "loss": 5.4838,
      "mean_token_accuracy": 0.2024390071630478,
      "num_tokens": 2408477.0,
      "step": 2835
    },
    {
      "entropy": 5.6671067237854,
      "epoch": 0.06345872389868949,
      "grad_norm": 1.453125,
      "learning_rate": 0.00019679217418151667,
      "loss": 5.532,
      "mean_token_accuracy": 0.18993941694498062,
      "num_tokens": 2413299.0,
      "step": 2840
    },
    {
      "entropy": 5.541928005218506,
      "epoch": 0.06357044700414492,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00019568861252472236,
      "loss": 5.2577,
      "mean_token_accuracy": 0.21702981293201445,
      "num_tokens": 2417668.0,
      "step": 2845
    },
    {
      "entropy": 5.578932189941407,
      "epoch": 0.06368217010960037,
      "grad_norm": 1.296875,
      "learning_rate": 0.00019458722522967952,
      "loss": 5.4111,
      "mean_token_accuracy": 0.2051840081810951,
      "num_tokens": 2422531.0,
      "step": 2850
    },
    {
      "entropy": 5.458756589889527,
      "epoch": 0.06379389321505581,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00019348804249147723,
      "loss": 5.1974,
      "mean_token_accuracy": 0.22611149847507478,
      "num_tokens": 2426958.0,
      "step": 2855
    },
    {
      "entropy": 5.442702007293701,
      "epoch": 0.06390561632051124,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0001923910944447655,
      "loss": 5.3937,
      "mean_token_accuracy": 0.2129739284515381,
      "num_tokens": 2431011.0,
      "step": 2860
    },
    {
      "entropy": 5.441421890258789,
      "epoch": 0.06401733942596669,
      "grad_norm": 1.46875,
      "learning_rate": 0.00019129641116292928,
      "loss": 5.3534,
      "mean_token_accuracy": 0.20348150730133058,
      "num_tokens": 2435050.0,
      "step": 2865
    },
    {
      "entropy": 5.646132755279541,
      "epoch": 0.06412906253142213,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00019020402265726343,
      "loss": 5.5116,
      "mean_token_accuracy": 0.19797807186841965,
      "num_tokens": 2440017.0,
      "step": 2870
    },
    {
      "entropy": 5.582180166244507,
      "epoch": 0.06424078563687756,
      "grad_norm": 1.453125,
      "learning_rate": 0.0001891139588761509,
      "loss": 5.3964,
      "mean_token_accuracy": 0.21056918501853944,
      "num_tokens": 2444177.0,
      "step": 2875
    },
    {
      "entropy": 5.549179029464722,
      "epoch": 0.064352508742333,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00018802624970424076,
      "loss": 5.5072,
      "mean_token_accuracy": 0.1970616325736046,
      "num_tokens": 2448287.0,
      "step": 2880
    },
    {
      "entropy": 5.567583322525024,
      "epoch": 0.06446423184778845,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00018694092496162945,
      "loss": 5.4531,
      "mean_token_accuracy": 0.20046740174293518,
      "num_tokens": 2452901.0,
      "step": 2885
    },
    {
      "entropy": 5.523518848419189,
      "epoch": 0.06457595495324388,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00018585801440304306,
      "loss": 5.29,
      "mean_token_accuracy": 0.22705289870500564,
      "num_tokens": 2457333.0,
      "step": 2890
    },
    {
      "entropy": 5.709904718399048,
      "epoch": 0.06468767805869932,
      "grad_norm": 1.328125,
      "learning_rate": 0.00018477754771702165,
      "loss": 5.4641,
      "mean_token_accuracy": 0.19585065692663192,
      "num_tokens": 2461502.0,
      "step": 2895
    },
    {
      "entropy": 5.63380274772644,
      "epoch": 0.06479940116415477,
      "grad_norm": 1.390625,
      "learning_rate": 0.00018369955452510506,
      "loss": 5.567,
      "mean_token_accuracy": 0.1850685879588127,
      "num_tokens": 2465742.0,
      "step": 2900
    },
    {
      "entropy": 5.575063753128052,
      "epoch": 0.0649111242696102,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0001826240643810212,
      "loss": 5.4862,
      "mean_token_accuracy": 0.1963152125477791,
      "num_tokens": 2469768.0,
      "step": 2905
    },
    {
      "entropy": 5.492411041259766,
      "epoch": 0.06502284737506564,
      "grad_norm": 1.390625,
      "learning_rate": 0.0001815511067698758,
      "loss": 5.4261,
      "mean_token_accuracy": 0.20247098952531814,
      "num_tokens": 2473867.0,
      "step": 2910
    },
    {
      "entropy": 5.508422088623047,
      "epoch": 0.06513457048052107,
      "grad_norm": 1.40625,
      "learning_rate": 0.0001804807111073436,
      "loss": 5.3556,
      "mean_token_accuracy": 0.20493078380823135,
      "num_tokens": 2477783.0,
      "step": 2915
    },
    {
      "entropy": 5.5772562503814695,
      "epoch": 0.06524629358597651,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0001794129067388625,
      "loss": 5.5152,
      "mean_token_accuracy": 0.1926484927535057,
      "num_tokens": 2482135.0,
      "step": 2920
    },
    {
      "entropy": 5.566284275054931,
      "epoch": 0.06535801669143196,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00017834772293882868,
      "loss": 5.455,
      "mean_token_accuracy": 0.19815311580896378,
      "num_tokens": 2486171.0,
      "step": 2925
    },
    {
      "entropy": 5.457381963729858,
      "epoch": 0.06546973979688739,
      "grad_norm": 1.421875,
      "learning_rate": 0.000177285188909794,
      "loss": 5.2721,
      "mean_token_accuracy": 0.21690144687891005,
      "num_tokens": 2490372.0,
      "step": 2930
    },
    {
      "entropy": 5.511955738067627,
      "epoch": 0.06558146290234283,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001762253337816656,
      "loss": 5.2055,
      "mean_token_accuracy": 0.22533015310764312,
      "num_tokens": 2494662.0,
      "step": 2935
    },
    {
      "entropy": 5.456213045120239,
      "epoch": 0.06569318600779828,
      "grad_norm": 1.546875,
      "learning_rate": 0.00017516818661090738,
      "loss": 5.3538,
      "mean_token_accuracy": 0.21707223057746888,
      "num_tokens": 2498164.0,
      "step": 2940
    },
    {
      "entropy": 5.53596453666687,
      "epoch": 0.0658049091132537,
      "grad_norm": 1.328125,
      "learning_rate": 0.0001741137763797428,
      "loss": 5.4243,
      "mean_token_accuracy": 0.1941936880350113,
      "num_tokens": 2502335.0,
      "step": 2945
    },
    {
      "entropy": 5.64856538772583,
      "epoch": 0.06591663221870915,
      "grad_norm": 1.46875,
      "learning_rate": 0.00017306213199536115,
      "loss": 5.5751,
      "mean_token_accuracy": 0.1955207407474518,
      "num_tokens": 2506792.0,
      "step": 2950
    },
    {
      "entropy": 5.600952863693237,
      "epoch": 0.0660283553241646,
      "grad_norm": 1.25,
      "learning_rate": 0.0001720132822891243,
      "loss": 5.3522,
      "mean_token_accuracy": 0.21116236299276353,
      "num_tokens": 2510863.0,
      "step": 2955
    },
    {
      "entropy": 5.678525066375732,
      "epoch": 0.06614007842962003,
      "grad_norm": 1.40625,
      "learning_rate": 0.0001709672560157769,
      "loss": 5.4757,
      "mean_token_accuracy": 0.19187746047973633,
      "num_tokens": 2515220.0,
      "step": 2960
    },
    {
      "entropy": 5.567672872543335,
      "epoch": 0.06625180153507547,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00016992408185265758,
      "loss": 5.3598,
      "mean_token_accuracy": 0.1987830951809883,
      "num_tokens": 2519501.0,
      "step": 2965
    },
    {
      "entropy": 5.565345335006714,
      "epoch": 0.06636352464053091,
      "grad_norm": 1.453125,
      "learning_rate": 0.00016888378839891298,
      "loss": 5.4164,
      "mean_token_accuracy": 0.20309228003025054,
      "num_tokens": 2523838.0,
      "step": 2970
    },
    {
      "entropy": 5.483179807662964,
      "epoch": 0.06647524774598634,
      "grad_norm": 1.375,
      "learning_rate": 0.0001678464041747137,
      "loss": 5.3708,
      "mean_token_accuracy": 0.20249621272087098,
      "num_tokens": 2528480.0,
      "step": 2975
    },
    {
      "entropy": 5.53153920173645,
      "epoch": 0.06658697085144179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00016681195762047223,
      "loss": 5.4239,
      "mean_token_accuracy": 0.1959275111556053,
      "num_tokens": 2532596.0,
      "step": 2980
    },
    {
      "entropy": 5.562540483474732,
      "epoch": 0.06669869395689723,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00016578047709606337,
      "loss": 5.3401,
      "mean_token_accuracy": 0.20386775434017182,
      "num_tokens": 2536980.0,
      "step": 2985
    },
    {
      "entropy": 5.630472135543823,
      "epoch": 0.06681041706235266,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00016475199088004678,
      "loss": 5.4943,
      "mean_token_accuracy": 0.19038521200418473,
      "num_tokens": 2540729.0,
      "step": 2990
    },
    {
      "entropy": 5.634172344207764,
      "epoch": 0.0669221401678081,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00016372652716889163,
      "loss": 5.5563,
      "mean_token_accuracy": 0.1887037053704262,
      "num_tokens": 2545164.0,
      "step": 2995
    },
    {
      "entropy": 5.562699651718139,
      "epoch": 0.06703386327326355,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001627041140762035,
      "loss": 5.2977,
      "mean_token_accuracy": 0.21568341553211212,
      "num_tokens": 2549182.0,
      "step": 3000
    },
    {
      "entropy": 5.533506679534912,
      "epoch": 0.06714558637871898,
      "grad_norm": 1.359375,
      "learning_rate": 0.00016168477963195382,
      "loss": 5.4094,
      "mean_token_accuracy": 0.2054365247488022,
      "num_tokens": 2553673.0,
      "step": 3005
    },
    {
      "entropy": 5.587810325622558,
      "epoch": 0.06725730948417442,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001606685517817114,
      "loss": 5.4964,
      "mean_token_accuracy": 0.20472187250852586,
      "num_tokens": 2558494.0,
      "step": 3010
    },
    {
      "entropy": 5.576572275161743,
      "epoch": 0.06736903258962987,
      "grad_norm": 1.453125,
      "learning_rate": 0.00015965545838587592,
      "loss": 5.3804,
      "mean_token_accuracy": 0.19869962185621262,
      "num_tokens": 2562488.0,
      "step": 3015
    },
    {
      "entropy": 5.548072576522827,
      "epoch": 0.0674807556950853,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00015864552721891467,
      "loss": 5.3049,
      "mean_token_accuracy": 0.20522964745759964,
      "num_tokens": 2566404.0,
      "step": 3020
    },
    {
      "entropy": 5.403257369995117,
      "epoch": 0.06759247880054074,
      "grad_norm": 1.34375,
      "learning_rate": 0.00015763878596860076,
      "loss": 5.2853,
      "mean_token_accuracy": 0.21069473326206206,
      "num_tokens": 2571000.0,
      "step": 3025
    },
    {
      "entropy": 5.545701313018799,
      "epoch": 0.06770420190599617,
      "grad_norm": 1.203125,
      "learning_rate": 0.00015663526223525412,
      "loss": 5.4689,
      "mean_token_accuracy": 0.1939232811331749,
      "num_tokens": 2575795.0,
      "step": 3030
    },
    {
      "entropy": 5.481280088424683,
      "epoch": 0.06781592501145162,
      "grad_norm": 1.359375,
      "learning_rate": 0.0001556349835309848,
      "loss": 5.2921,
      "mean_token_accuracy": 0.21501746624708176,
      "num_tokens": 2579885.0,
      "step": 3035
    },
    {
      "entropy": 5.55908842086792,
      "epoch": 0.06792764811690706,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001546379772789389,
      "loss": 5.3067,
      "mean_token_accuracy": 0.2067527338862419,
      "num_tokens": 2583951.0,
      "step": 3040
    },
    {
      "entropy": 5.420316076278686,
      "epoch": 0.06803937122236249,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00015364427081254622,
      "loss": 5.2772,
      "mean_token_accuracy": 0.21368366628885269,
      "num_tokens": 2587881.0,
      "step": 3045
    },
    {
      "entropy": 5.505311679840088,
      "epoch": 0.06815109432781793,
      "grad_norm": 1.34375,
      "learning_rate": 0.00015265389137477165,
      "loss": 5.3664,
      "mean_token_accuracy": 0.21869818568229676,
      "num_tokens": 2592123.0,
      "step": 3050
    },
    {
      "entropy": 5.624979543685913,
      "epoch": 0.06826281743327338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00015166686611736786,
      "loss": 5.536,
      "mean_token_accuracy": 0.19260048121213913,
      "num_tokens": 2596315.0,
      "step": 3055
    },
    {
      "entropy": 5.6317283630371096,
      "epoch": 0.06837454053872881,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00015068322210013064,
      "loss": 5.3305,
      "mean_token_accuracy": 0.20026403814554214,
      "num_tokens": 2600472.0,
      "step": 3060
    },
    {
      "entropy": 5.474189901351929,
      "epoch": 0.06848626364418425,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001497029862901578,
      "loss": 5.2144,
      "mean_token_accuracy": 0.22472612112760543,
      "num_tokens": 2604310.0,
      "step": 3065
    },
    {
      "entropy": 5.531822681427002,
      "epoch": 0.0685979867496397,
      "grad_norm": 1.3125,
      "learning_rate": 0.00014872618556110905,
      "loss": 5.447,
      "mean_token_accuracy": 0.1974410429596901,
      "num_tokens": 2608771.0,
      "step": 3070
    },
    {
      "entropy": 5.451178169250488,
      "epoch": 0.06870970985509513,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00014775284669246992,
      "loss": 5.3274,
      "mean_token_accuracy": 0.21316346824169158,
      "num_tokens": 2613299.0,
      "step": 3075
    },
    {
      "entropy": 5.607516241073609,
      "epoch": 0.06882143296055057,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00014678299636881716,
      "loss": 5.3839,
      "mean_token_accuracy": 0.20785063207149507,
      "num_tokens": 2617778.0,
      "step": 3080
    },
    {
      "entropy": 5.493184614181518,
      "epoch": 0.06893315606600602,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001458166611790873,
      "loss": 5.333,
      "mean_token_accuracy": 0.21613772660493852,
      "num_tokens": 2621636.0,
      "step": 3085
    },
    {
      "entropy": 5.521860265731812,
      "epoch": 0.06904487917146145,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00014485386761584773,
      "loss": 5.4048,
      "mean_token_accuracy": 0.19960226416587828,
      "num_tokens": 2625923.0,
      "step": 3090
    },
    {
      "entropy": 5.569070720672608,
      "epoch": 0.06915660227691689,
      "grad_norm": 1.40625,
      "learning_rate": 0.00014389464207457042,
      "loss": 5.481,
      "mean_token_accuracy": 0.1899915635585785,
      "num_tokens": 2630282.0,
      "step": 3095
    },
    {
      "entropy": 5.449570608139038,
      "epoch": 0.06926832538237233,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00014293901085290795,
      "loss": 5.268,
      "mean_token_accuracy": 0.2146723374724388,
      "num_tokens": 2634376.0,
      "step": 3100
    },
    {
      "entropy": 5.508600187301636,
      "epoch": 0.06938004848782776,
      "grad_norm": 1.46875,
      "learning_rate": 0.00014198700014997307,
      "loss": 5.3895,
      "mean_token_accuracy": 0.203936605155468,
      "num_tokens": 2638486.0,
      "step": 3105
    },
    {
      "entropy": 5.609038972854615,
      "epoch": 0.06949177159328321,
      "grad_norm": 1.328125,
      "learning_rate": 0.00014103863606562016,
      "loss": 5.4556,
      "mean_token_accuracy": 0.18898191303014755,
      "num_tokens": 2643023.0,
      "step": 3110
    },
    {
      "entropy": 5.556278228759766,
      "epoch": 0.06960349469873865,
      "grad_norm": 1.359375,
      "learning_rate": 0.00014009394459972964,
      "loss": 5.368,
      "mean_token_accuracy": 0.2124757096171379,
      "num_tokens": 2646977.0,
      "step": 3115
    },
    {
      "entropy": 5.629961919784546,
      "epoch": 0.06971521780419408,
      "grad_norm": 1.46875,
      "learning_rate": 0.00013915295165149513,
      "loss": 5.3671,
      "mean_token_accuracy": 0.21044702976942062,
      "num_tokens": 2651296.0,
      "step": 3120
    },
    {
      "entropy": 5.551492357254029,
      "epoch": 0.06982694090964953,
      "grad_norm": 1.359375,
      "learning_rate": 0.00013821568301871384,
      "loss": 5.4196,
      "mean_token_accuracy": 0.20507991760969163,
      "num_tokens": 2655511.0,
      "step": 3125
    },
    {
      "entropy": 5.543411779403686,
      "epoch": 0.06993866401510497,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00013728216439707862,
      "loss": 5.3965,
      "mean_token_accuracy": 0.20317531377077103,
      "num_tokens": 2660457.0,
      "step": 3130
    },
    {
      "entropy": 5.5066040515899655,
      "epoch": 0.0700503871205604,
      "grad_norm": 1.40625,
      "learning_rate": 0.00013635242137947419,
      "loss": 5.3101,
      "mean_token_accuracy": 0.2202605977654457,
      "num_tokens": 2664675.0,
      "step": 3135
    },
    {
      "entropy": 5.551229429244995,
      "epoch": 0.07016211022601584,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00013542647945527498,
      "loss": 5.3026,
      "mean_token_accuracy": 0.22133247405290604,
      "num_tokens": 2668809.0,
      "step": 3140
    },
    {
      "entropy": 5.533741140365601,
      "epoch": 0.07027383333147127,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0001345043640096465,
      "loss": 5.3337,
      "mean_token_accuracy": 0.20904677212238312,
      "num_tokens": 2672974.0,
      "step": 3145
    },
    {
      "entropy": 5.553868627548217,
      "epoch": 0.07038555643692672,
      "grad_norm": 1.296875,
      "learning_rate": 0.00013358610032284957,
      "loss": 5.3573,
      "mean_token_accuracy": 0.2051636353135109,
      "num_tokens": 2677633.0,
      "step": 3150
    },
    {
      "entropy": 5.5611162185668945,
      "epoch": 0.07049727954238216,
      "grad_norm": 1.40625,
      "learning_rate": 0.000132671713569547,
      "loss": 5.2561,
      "mean_token_accuracy": 0.20905781239271165,
      "num_tokens": 2681398.0,
      "step": 3155
    },
    {
      "entropy": 5.614703702926636,
      "epoch": 0.0706090026478376,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0001317612288181136,
      "loss": 5.4679,
      "mean_token_accuracy": 0.1954231396317482,
      "num_tokens": 2685785.0,
      "step": 3160
    },
    {
      "entropy": 5.500372457504272,
      "epoch": 0.07072072575329304,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00013085467102994864,
      "loss": 5.3446,
      "mean_token_accuracy": 0.20620400309562684,
      "num_tokens": 2689804.0,
      "step": 3165
    },
    {
      "entropy": 5.6220344543457035,
      "epoch": 0.07083244885874848,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00012995206505879198,
      "loss": 5.5199,
      "mean_token_accuracy": 0.19301192164421083,
      "num_tokens": 2694632.0,
      "step": 3170
    },
    {
      "entropy": 5.4843446731567385,
      "epoch": 0.07094417196420391,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0001290534356500421,
      "loss": 5.1761,
      "mean_token_accuracy": 0.2106317937374115,
      "num_tokens": 2698723.0,
      "step": 3175
    },
    {
      "entropy": 5.394832372665405,
      "epoch": 0.07105589506965936,
      "grad_norm": 1.390625,
      "learning_rate": 0.00012815880744007827,
      "loss": 5.1025,
      "mean_token_accuracy": 0.22589420527219772,
      "num_tokens": 2702746.0,
      "step": 3180
    },
    {
      "entropy": 5.52421875,
      "epoch": 0.0711676181751148,
      "grad_norm": 1.46875,
      "learning_rate": 0.00012726820495558483,
      "loss": 5.3383,
      "mean_token_accuracy": 0.20268831700086593,
      "num_tokens": 2707037.0,
      "step": 3185
    },
    {
      "entropy": 5.565604829788208,
      "epoch": 0.07127934128057023,
      "grad_norm": 1.359375,
      "learning_rate": 0.00012638165261287868,
      "loss": 5.483,
      "mean_token_accuracy": 0.19443231523036958,
      "num_tokens": 2711529.0,
      "step": 3190
    },
    {
      "entropy": 5.4620850563049315,
      "epoch": 0.07139106438602567,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001254991747172402,
      "loss": 5.341,
      "mean_token_accuracy": 0.2169657066464424,
      "num_tokens": 2715804.0,
      "step": 3195
    },
    {
      "entropy": 5.413955783843994,
      "epoch": 0.07150278749148112,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00012462079546224662,
      "loss": 5.2217,
      "mean_token_accuracy": 0.2313481852412224,
      "num_tokens": 2719902.0,
      "step": 3200
    },
    {
      "entropy": 5.447470235824585,
      "epoch": 0.07161451059693655,
      "grad_norm": 1.34375,
      "learning_rate": 0.00012374653892910896,
      "loss": 5.1621,
      "mean_token_accuracy": 0.22463787198066712,
      "num_tokens": 2723673.0,
      "step": 3205
    },
    {
      "entropy": 5.533495712280273,
      "epoch": 0.07172623370239199,
      "grad_norm": 1.421875,
      "learning_rate": 0.00012287642908601166,
      "loss": 5.4301,
      "mean_token_accuracy": 0.19854580461978913,
      "num_tokens": 2727634.0,
      "step": 3210
    },
    {
      "entropy": 5.44229474067688,
      "epoch": 0.07183795680784744,
      "grad_norm": 1.3125,
      "learning_rate": 0.00012201048978745569,
      "loss": 5.1981,
      "mean_token_accuracy": 0.2350959450006485,
      "num_tokens": 2731931.0,
      "step": 3215
    },
    {
      "entropy": 5.457131242752075,
      "epoch": 0.07194967991330287,
      "grad_norm": 1.46875,
      "learning_rate": 0.00012114874477360427,
      "loss": 5.2268,
      "mean_token_accuracy": 0.21981479823589326,
      "num_tokens": 2736212.0,
      "step": 3220
    },
    {
      "entropy": 5.552663373947143,
      "epoch": 0.07206140301875831,
      "grad_norm": 1.421875,
      "learning_rate": 0.00012029121766963236,
      "loss": 5.4433,
      "mean_token_accuracy": 0.19506857991218568,
      "num_tokens": 2740696.0,
      "step": 3225
    },
    {
      "entropy": 5.478365278244018,
      "epoch": 0.07217312612421375,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00011943793198507858,
      "loss": 5.3454,
      "mean_token_accuracy": 0.21160352677106858,
      "num_tokens": 2744695.0,
      "step": 3230
    },
    {
      "entropy": 5.507468414306641,
      "epoch": 0.07228484922966918,
      "grad_norm": 1.5,
      "learning_rate": 0.00011858891111320104,
      "loss": 5.3719,
      "mean_token_accuracy": 0.20120661407709123,
      "num_tokens": 2748855.0,
      "step": 3235
    },
    {
      "entropy": 5.471977281570434,
      "epoch": 0.07239657233512463,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001177441783303359,
      "loss": 5.2239,
      "mean_token_accuracy": 0.20875055342912674,
      "num_tokens": 2753099.0,
      "step": 3240
    },
    {
      "entropy": 5.5574604034423825,
      "epoch": 0.07250829544058007,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00011690375679525896,
      "loss": 5.3847,
      "mean_token_accuracy": 0.1993442580103874,
      "num_tokens": 2757290.0,
      "step": 3245
    },
    {
      "entropy": 5.566975259780884,
      "epoch": 0.0726200185460355,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00011606766954855124,
      "loss": 5.4143,
      "mean_token_accuracy": 0.2066384956240654,
      "num_tokens": 2761044.0,
      "step": 3250
    },
    {
      "entropy": 5.588833665847778,
      "epoch": 0.07273174165149095,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00011523593951196702,
      "loss": 5.4072,
      "mean_token_accuracy": 0.20698025375604628,
      "num_tokens": 2765353.0,
      "step": 3255
    },
    {
      "entropy": 5.5550299167633055,
      "epoch": 0.07284346475694638,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00011440858948780523,
      "loss": 5.3039,
      "mean_token_accuracy": 0.20871637314558028,
      "num_tokens": 2769530.0,
      "step": 3260
    },
    {
      "entropy": 5.61817889213562,
      "epoch": 0.07295518786240182,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00011358564215828484,
      "loss": 5.3615,
      "mean_token_accuracy": 0.2078099176287651,
      "num_tokens": 2773876.0,
      "step": 3265
    },
    {
      "entropy": 5.656089353561401,
      "epoch": 0.07306691096785727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00011276712008492254,
      "loss": 5.5268,
      "mean_token_accuracy": 0.19516537785530091,
      "num_tokens": 2778648.0,
      "step": 3270
    },
    {
      "entropy": 5.457943773269653,
      "epoch": 0.0731786340733127,
      "grad_norm": 1.484375,
      "learning_rate": 0.00011195304570791451,
      "loss": 5.2562,
      "mean_token_accuracy": 0.22165347635746002,
      "num_tokens": 2782651.0,
      "step": 3275
    },
    {
      "entropy": 5.558605003356933,
      "epoch": 0.07329035717876814,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00011114344134552094,
      "loss": 5.394,
      "mean_token_accuracy": 0.20966942757368087,
      "num_tokens": 2786665.0,
      "step": 3280
    },
    {
      "entropy": 5.511362552642822,
      "epoch": 0.07340208028422358,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001103383291934545,
      "loss": 5.3282,
      "mean_token_accuracy": 0.21039337515830994,
      "num_tokens": 2791037.0,
      "step": 3285
    },
    {
      "entropy": 5.5587486743927,
      "epoch": 0.07351380338967901,
      "grad_norm": 1.40625,
      "learning_rate": 0.00010953773132427141,
      "loss": 5.3747,
      "mean_token_accuracy": 0.20165727883577347,
      "num_tokens": 2795026.0,
      "step": 3290
    },
    {
      "entropy": 5.415568113327026,
      "epoch": 0.07362552649513446,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00010874166968676677,
      "loss": 5.1669,
      "mean_token_accuracy": 0.23009648472070693,
      "num_tokens": 2798998.0,
      "step": 3295
    },
    {
      "entropy": 5.434786701202393,
      "epoch": 0.0737372496005899,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00010795016610537251,
      "loss": 5.284,
      "mean_token_accuracy": 0.21612590700387954,
      "num_tokens": 2803676.0,
      "step": 3300
    },
    {
      "entropy": 5.459165620803833,
      "epoch": 0.07384897270604533,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00010716324227955904,
      "loss": 5.2361,
      "mean_token_accuracy": 0.22047095149755477,
      "num_tokens": 2808227.0,
      "step": 3305
    },
    {
      "entropy": 5.41816725730896,
      "epoch": 0.07396069581150078,
      "grad_norm": 1.3125,
      "learning_rate": 0.0001063809197832406,
      "loss": 5.3362,
      "mean_token_accuracy": 0.219697405397892,
      "num_tokens": 2812340.0,
      "step": 3310
    },
    {
      "entropy": 5.436846685409546,
      "epoch": 0.07407241891695622,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00010560322006418368,
      "loss": 5.2255,
      "mean_token_accuracy": 0.2216632604598999,
      "num_tokens": 2816329.0,
      "step": 3315
    },
    {
      "entropy": 5.657596206665039,
      "epoch": 0.07418414202241165,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00010483016444341887,
      "loss": 5.539,
      "mean_token_accuracy": 0.18904027342796326,
      "num_tokens": 2820759.0,
      "step": 3320
    },
    {
      "entropy": 5.437835741043091,
      "epoch": 0.0742958651278671,
      "grad_norm": 1.28125,
      "learning_rate": 0.00010406177411465654,
      "loss": 5.1291,
      "mean_token_accuracy": 0.2266139045357704,
      "num_tokens": 2825463.0,
      "step": 3325
    },
    {
      "entropy": 5.542234277725219,
      "epoch": 0.07440758823332254,
      "grad_norm": 1.3125,
      "learning_rate": 0.00010329807014370562,
      "loss": 5.3376,
      "mean_token_accuracy": 0.20447732508182526,
      "num_tokens": 2829453.0,
      "step": 3330
    },
    {
      "entropy": 5.503271675109863,
      "epoch": 0.07451931133877797,
      "grad_norm": 1.390625,
      "learning_rate": 0.00010253907346789632,
      "loss": 5.3448,
      "mean_token_accuracy": 0.21958455294370652,
      "num_tokens": 2833526.0,
      "step": 3335
    },
    {
      "entropy": 5.57898850440979,
      "epoch": 0.07463103444423341,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00010178480489550596,
      "loss": 5.4124,
      "mean_token_accuracy": 0.21505914777517318,
      "num_tokens": 2837634.0,
      "step": 3340
    },
    {
      "entropy": 5.527571487426758,
      "epoch": 0.07474275754968886,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00010103528510518836,
      "loss": 5.2429,
      "mean_token_accuracy": 0.21626538634300232,
      "num_tokens": 2842126.0,
      "step": 3345
    },
    {
      "entropy": 5.479366874694824,
      "epoch": 0.07485448065514429,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001002905346454073,
      "loss": 5.2928,
      "mean_token_accuracy": 0.2194333478808403,
      "num_tokens": 2846518.0,
      "step": 3350
    },
    {
      "entropy": 5.518194770812988,
      "epoch": 0.07496620376059973,
      "grad_norm": 1.4140625,
      "learning_rate": 9.955057393387285e-05,
      "loss": 5.3799,
      "mean_token_accuracy": 0.20560420155525208,
      "num_tokens": 2851045.0,
      "step": 3355
    },
    {
      "entropy": 5.531414890289307,
      "epoch": 0.07507792686605518,
      "grad_norm": 1.40625,
      "learning_rate": 9.88154232569816e-05,
      "loss": 5.3578,
      "mean_token_accuracy": 0.20951698422431947,
      "num_tokens": 2855235.0,
      "step": 3360
    },
    {
      "entropy": 5.576348972320557,
      "epoch": 0.0751896499715106,
      "grad_norm": 1.328125,
      "learning_rate": 9.808510276926075e-05,
      "loss": 5.292,
      "mean_token_accuracy": 0.20938485413789748,
      "num_tokens": 2859507.0,
      "step": 3365
    },
    {
      "entropy": 5.586604976654053,
      "epoch": 0.07530137307696605,
      "grad_norm": 1.40625,
      "learning_rate": 9.735963249281549e-05,
      "loss": 5.4149,
      "mean_token_accuracy": 0.1966403216123581,
      "num_tokens": 2863833.0,
      "step": 3370
    },
    {
      "entropy": 5.621471929550171,
      "epoch": 0.0754130961824215,
      "grad_norm": 1.4375,
      "learning_rate": 9.663903231677974e-05,
      "loss": 5.5328,
      "mean_token_accuracy": 0.18283545225858688,
      "num_tokens": 2868540.0,
      "step": 3375
    },
    {
      "entropy": 5.59570198059082,
      "epoch": 0.07552481928787692,
      "grad_norm": 1.4609375,
      "learning_rate": 9.592332199677145e-05,
      "loss": 5.2865,
      "mean_token_accuracy": 0.20847386866807938,
      "num_tokens": 2872407.0,
      "step": 3380
    },
    {
      "entropy": 5.6583184719085695,
      "epoch": 0.07563654239333237,
      "grad_norm": 1.4375,
      "learning_rate": 9.521252115435061e-05,
      "loss": 5.3678,
      "mean_token_accuracy": 0.20249831974506377,
      "num_tokens": 2876428.0,
      "step": 3385
    },
    {
      "entropy": 5.654592895507813,
      "epoch": 0.0757482654987878,
      "grad_norm": 1.21875,
      "learning_rate": 9.450664927648126e-05,
      "loss": 5.4111,
      "mean_token_accuracy": 0.201308237016201,
      "num_tokens": 2880536.0,
      "step": 3390
    },
    {
      "entropy": 5.58061580657959,
      "epoch": 0.07585998860424324,
      "grad_norm": 1.3671875,
      "learning_rate": 9.380572571499758e-05,
      "loss": 5.3291,
      "mean_token_accuracy": 0.21148840487003326,
      "num_tokens": 2884607.0,
      "step": 3395
    },
    {
      "entropy": 5.424838495254517,
      "epoch": 0.07597171170969869,
      "grad_norm": 1.5078125,
      "learning_rate": 9.310976968607307e-05,
      "loss": 5.3244,
      "mean_token_accuracy": 0.21160314083099366,
      "num_tokens": 2888996.0,
      "step": 3400
    },
    {
      "entropy": 5.4981842041015625,
      "epoch": 0.07608343481515412,
      "grad_norm": 1.3125,
      "learning_rate": 9.241880026969381e-05,
      "loss": 5.2962,
      "mean_token_accuracy": 0.21013507694005967,
      "num_tokens": 2893526.0,
      "step": 3405
    },
    {
      "entropy": 5.546281337738037,
      "epoch": 0.07619515792060956,
      "grad_norm": 1.2578125,
      "learning_rate": 9.173283640913537e-05,
      "loss": 5.4652,
      "mean_token_accuracy": 0.2015298828482628,
      "num_tokens": 2898170.0,
      "step": 3410
    },
    {
      "entropy": 5.572264862060547,
      "epoch": 0.076306881026065,
      "grad_norm": 1.2890625,
      "learning_rate": 9.10518969104436e-05,
      "loss": 5.3884,
      "mean_token_accuracy": 0.20227549523115157,
      "num_tokens": 2902694.0,
      "step": 3415
    },
    {
      "entropy": 5.649074268341065,
      "epoch": 0.07641860413152043,
      "grad_norm": 1.3671875,
      "learning_rate": 9.037600044191868e-05,
      "loss": 5.4148,
      "mean_token_accuracy": 0.20267524868249892,
      "num_tokens": 2906529.0,
      "step": 3420
    },
    {
      "entropy": 5.653632545471192,
      "epoch": 0.07653032723697588,
      "grad_norm": 1.4921875,
      "learning_rate": 8.970516553360383e-05,
      "loss": 5.4751,
      "mean_token_accuracy": 0.1963999703526497,
      "num_tokens": 2910643.0,
      "step": 3425
    },
    {
      "entropy": 5.542692184448242,
      "epoch": 0.07664205034243132,
      "grad_norm": 1.3984375,
      "learning_rate": 8.903941057677692e-05,
      "loss": 5.2885,
      "mean_token_accuracy": 0.20829614251852036,
      "num_tokens": 2914755.0,
      "step": 3430
    },
    {
      "entropy": 5.541115999221802,
      "epoch": 0.07675377344788675,
      "grad_norm": 1.421875,
      "learning_rate": 8.837875382344635e-05,
      "loss": 5.3152,
      "mean_token_accuracy": 0.21061815321445465,
      "num_tokens": 2919054.0,
      "step": 3435
    },
    {
      "entropy": 5.378523445129394,
      "epoch": 0.0768654965533422,
      "grad_norm": 1.375,
      "learning_rate": 8.772321338585076e-05,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2321953296661377,
      "num_tokens": 2922887.0,
      "step": 3440
    },
    {
      "entropy": 5.501231241226196,
      "epoch": 0.07697721965879764,
      "grad_norm": 1.3125,
      "learning_rate": 8.707280723596242e-05,
      "loss": 5.3906,
      "mean_token_accuracy": 0.21212373673915863,
      "num_tokens": 2927587.0,
      "step": 3445
    },
    {
      "entropy": 5.4999463081359865,
      "epoch": 0.07708894276425307,
      "grad_norm": 1.59375,
      "learning_rate": 8.64275532049944e-05,
      "loss": 5.3351,
      "mean_token_accuracy": 0.2148713067173958,
      "num_tokens": 2931485.0,
      "step": 3450
    },
    {
      "entropy": 5.441826915740966,
      "epoch": 0.07720066586970852,
      "grad_norm": 1.40625,
      "learning_rate": 8.578746898291198e-05,
      "loss": 5.2734,
      "mean_token_accuracy": 0.22465914189815522,
      "num_tokens": 2935592.0,
      "step": 3455
    },
    {
      "entropy": 5.529314613342285,
      "epoch": 0.07731238897516396,
      "grad_norm": 1.390625,
      "learning_rate": 8.515257211794742e-05,
      "loss": 5.3514,
      "mean_token_accuracy": 0.20800221264362334,
      "num_tokens": 2939758.0,
      "step": 3460
    },
    {
      "entropy": 5.458008670806885,
      "epoch": 0.07742411208061939,
      "grad_norm": 1.5859375,
      "learning_rate": 8.452288001611896e-05,
      "loss": 5.2006,
      "mean_token_accuracy": 0.2216313362121582,
      "num_tokens": 2943927.0,
      "step": 3465
    },
    {
      "entropy": 5.538556623458862,
      "epoch": 0.07753583518607483,
      "grad_norm": 1.3515625,
      "learning_rate": 8.389840994075379e-05,
      "loss": 5.3277,
      "mean_token_accuracy": 0.20335785448551177,
      "num_tokens": 2948310.0,
      "step": 3470
    },
    {
      "entropy": 5.543921613693238,
      "epoch": 0.07764755829153028,
      "grad_norm": 1.5,
      "learning_rate": 8.327917901201435e-05,
      "loss": 5.3341,
      "mean_token_accuracy": 0.20996950566768646,
      "num_tokens": 2952647.0,
      "step": 3475
    },
    {
      "entropy": 5.49157567024231,
      "epoch": 0.07775928139698571,
      "grad_norm": 1.375,
      "learning_rate": 8.266520420642931e-05,
      "loss": 5.3147,
      "mean_token_accuracy": 0.2074354499578476,
      "num_tokens": 2956940.0,
      "step": 3480
    },
    {
      "entropy": 5.464511346817017,
      "epoch": 0.07787100450244115,
      "grad_norm": 1.2734375,
      "learning_rate": 8.205650235642828e-05,
      "loss": 5.1543,
      "mean_token_accuracy": 0.21790542602539062,
      "num_tokens": 2960954.0,
      "step": 3485
    },
    {
      "entropy": 5.560961675643921,
      "epoch": 0.0779827276078966,
      "grad_norm": 1.5390625,
      "learning_rate": 8.145309014987978e-05,
      "loss": 5.4677,
      "mean_token_accuracy": 0.206323279440403,
      "num_tokens": 2965161.0,
      "step": 3490
    },
    {
      "entropy": 5.579499959945679,
      "epoch": 0.07809445071335203,
      "grad_norm": 1.5390625,
      "learning_rate": 8.085498412963437e-05,
      "loss": 5.3248,
      "mean_token_accuracy": 0.21123748123645783,
      "num_tokens": 2969071.0,
      "step": 3495
    },
    {
      "entropy": 5.539063549041748,
      "epoch": 0.07820617381880747,
      "grad_norm": 1.3046875,
      "learning_rate": 8.026220069307078e-05,
      "loss": 5.3398,
      "mean_token_accuracy": 0.20731965601444244,
      "num_tokens": 2973593.0,
      "step": 3500
    }
  ],
  "logging_steps": 5,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 4310012067840000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}