{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.601828438047073,
  "eval_steps": 3000,
  "global_step": 72000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.34880576133728,
      "epoch": 0.00038902937171756465,
      "grad_norm": 1.0703125,
      "learning_rate": 2e-06,
      "loss": 7.0252,
      "mean_token_accuracy": 0.10256106480956077,
      "num_tokens": 9243.0,
      "step": 5
    },
    {
      "entropy": 7.446655797958374,
      "epoch": 0.0007780587434351293,
      "grad_norm": 1.1328125,
      "learning_rate": 4.5e-06,
      "loss": 7.1577,
      "mean_token_accuracy": 0.0929850086569786,
      "num_tokens": 18424.0,
      "step": 10
    },
    {
      "entropy": 7.352576017379761,
      "epoch": 0.001167088115152694,
      "grad_norm": 1.0703125,
      "learning_rate": 7e-06,
      "loss": 7.0451,
      "mean_token_accuracy": 0.0964082270860672,
      "num_tokens": 27006.0,
      "step": 15
    },
    {
      "entropy": 7.343383884429931,
      "epoch": 0.0015561174868702586,
      "grad_norm": 1.078125,
      "learning_rate": 9.5e-06,
      "loss": 7.0864,
      "mean_token_accuracy": 0.0972217358648777,
      "num_tokens": 35577.0,
      "step": 20
    },
    {
      "entropy": 7.42727689743042,
      "epoch": 0.0019451468585878235,
      "grad_norm": 0.96875,
      "learning_rate": 1.2e-05,
      "loss": 7.0596,
      "mean_token_accuracy": 0.09545032754540443,
      "num_tokens": 43706.0,
      "step": 25
    },
    {
      "entropy": 7.413392829895019,
      "epoch": 0.002334176230305388,
      "grad_norm": 1.0859375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 7.0333,
      "mean_token_accuracy": 0.09936482384800911,
      "num_tokens": 52571.0,
      "step": 30
    },
    {
      "entropy": 7.465293312072754,
      "epoch": 0.0027232056020229526,
      "grad_norm": 0.984375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.0159,
      "mean_token_accuracy": 0.09479603916406631,
      "num_tokens": 61900.0,
      "step": 35
    },
    {
      "entropy": 7.500431871414184,
      "epoch": 0.0031122349737405172,
      "grad_norm": 1.0234375,
      "learning_rate": 1.95e-05,
      "loss": 7.0403,
      "mean_token_accuracy": 0.0956958994269371,
      "num_tokens": 71234.0,
      "step": 40
    },
    {
      "entropy": 7.569026231765747,
      "epoch": 0.003501264345458082,
      "grad_norm": 1.140625,
      "learning_rate": 2.2e-05,
      "loss": 7.1142,
      "mean_token_accuracy": 0.09231705740094184,
      "num_tokens": 79610.0,
      "step": 45
    },
    {
      "entropy": 7.532996749877929,
      "epoch": 0.003890293717175647,
      "grad_norm": 1.09375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.9294,
      "mean_token_accuracy": 0.1006363645195961,
      "num_tokens": 87527.0,
      "step": 50
    },
    {
      "entropy": 7.560778427124023,
      "epoch": 0.004279323088893211,
      "grad_norm": 1.125,
      "learning_rate": 2.7e-05,
      "loss": 7.0757,
      "mean_token_accuracy": 0.09589893892407417,
      "num_tokens": 95512.0,
      "step": 55
    },
    {
      "entropy": 7.501343441009522,
      "epoch": 0.004668352460610776,
      "grad_norm": 0.99609375,
      "learning_rate": 2.95e-05,
      "loss": 7.0864,
      "mean_token_accuracy": 0.09744687899947166,
      "num_tokens": 104395.0,
      "step": 60
    },
    {
      "entropy": 7.478601932525635,
      "epoch": 0.0050573818323283405,
      "grad_norm": 0.9609375,
      "learning_rate": 3.2e-05,
      "loss": 6.9791,
      "mean_token_accuracy": 0.10184086710214615,
      "num_tokens": 112562.0,
      "step": 65
    },
    {
      "entropy": 7.405571556091308,
      "epoch": 0.005446411204045905,
      "grad_norm": 1.1015625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.86,
      "mean_token_accuracy": 0.09663667231798172,
      "num_tokens": 121962.0,
      "step": 70
    },
    {
      "entropy": 7.51507978439331,
      "epoch": 0.00583544057576347,
      "grad_norm": 1.0234375,
      "learning_rate": 3.7e-05,
      "loss": 7.126,
      "mean_token_accuracy": 0.09231036975979805,
      "num_tokens": 130714.0,
      "step": 75
    },
    {
      "entropy": 7.520365238189697,
      "epoch": 0.0062244699474810344,
      "grad_norm": 0.95703125,
      "learning_rate": 3.95e-05,
      "loss": 6.9241,
      "mean_token_accuracy": 0.10009283497929573,
      "num_tokens": 139674.0,
      "step": 80
    },
    {
      "entropy": 7.47720103263855,
      "epoch": 0.006613499319198599,
      "grad_norm": 1.03125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.9803,
      "mean_token_accuracy": 0.09953748062252998,
      "num_tokens": 148654.0,
      "step": 85
    },
    {
      "entropy": 7.46751070022583,
      "epoch": 0.007002528690916164,
      "grad_norm": 1.09375,
      "learning_rate": 4.45e-05,
      "loss": 6.8751,
      "mean_token_accuracy": 0.10261099562048911,
      "num_tokens": 156180.0,
      "step": 90
    },
    {
      "entropy": 7.383714389801026,
      "epoch": 0.007391558062633729,
      "grad_norm": 0.94140625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.9565,
      "mean_token_accuracy": 0.09331886917352676,
      "num_tokens": 165608.0,
      "step": 95
    },
    {
      "entropy": 7.432891321182251,
      "epoch": 0.007780587434351294,
      "grad_norm": 0.9921875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.936,
      "mean_token_accuracy": 0.10059967562556267,
      "num_tokens": 173871.0,
      "step": 100
    },
    {
      "entropy": 7.461685752868652,
      "epoch": 0.008169616806068859,
      "grad_norm": 1.0390625,
      "learning_rate": 5.2e-05,
      "loss": 7.0248,
      "mean_token_accuracy": 0.09571366757154465,
      "num_tokens": 182356.0,
      "step": 105
    },
    {
      "entropy": 7.499224996566772,
      "epoch": 0.008558646177786422,
      "grad_norm": 0.98828125,
      "learning_rate": 5.45e-05,
      "loss": 7.0179,
      "mean_token_accuracy": 0.09466839730739593,
      "num_tokens": 190361.0,
      "step": 110
    },
    {
      "entropy": 7.435716581344605,
      "epoch": 0.008947675549503988,
      "grad_norm": 1.0234375,
      "learning_rate": 5.7e-05,
      "loss": 6.9827,
      "mean_token_accuracy": 0.0984235167503357,
      "num_tokens": 199253.0,
      "step": 115
    },
    {
      "entropy": 7.429847049713135,
      "epoch": 0.009336704921221552,
      "grad_norm": 1.109375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.9557,
      "mean_token_accuracy": 0.09870526567101479,
      "num_tokens": 208181.0,
      "step": 120
    },
    {
      "entropy": 7.515597820281982,
      "epoch": 0.009725734292939117,
      "grad_norm": 0.9921875,
      "learning_rate": 6.2e-05,
      "loss": 6.9546,
      "mean_token_accuracy": 0.09591879844665527,
      "num_tokens": 216898.0,
      "step": 125
    },
    {
      "entropy": 7.421197032928466,
      "epoch": 0.010114763664656681,
      "grad_norm": 1.0625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.0039,
      "mean_token_accuracy": 0.09688223078846932,
      "num_tokens": 226051.0,
      "step": 130
    },
    {
      "entropy": 7.492958879470825,
      "epoch": 0.010503793036374246,
      "grad_norm": 1.015625,
      "learning_rate": 6.7e-05,
      "loss": 7.042,
      "mean_token_accuracy": 0.09841900020837784,
      "num_tokens": 235018.0,
      "step": 135
    },
    {
      "entropy": 7.432940196990967,
      "epoch": 0.01089282240809181,
      "grad_norm": 1.0546875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.9703,
      "mean_token_accuracy": 0.09938597083091735,
      "num_tokens": 244268.0,
      "step": 140
    },
    {
      "entropy": 7.446199369430542,
      "epoch": 0.011281851779809376,
      "grad_norm": 0.98046875,
      "learning_rate": 7.2e-05,
      "loss": 6.9842,
      "mean_token_accuracy": 0.09659678265452384,
      "num_tokens": 253545.0,
      "step": 145
    },
    {
      "entropy": 7.365108060836792,
      "epoch": 0.01167088115152694,
      "grad_norm": 1.1015625,
      "learning_rate": 7.45e-05,
      "loss": 6.8048,
      "mean_token_accuracy": 0.10624724552035332,
      "num_tokens": 261767.0,
      "step": 150
    },
    {
      "entropy": 7.435336446762085,
      "epoch": 0.012059910523244505,
      "grad_norm": 0.9609375,
      "learning_rate": 7.7e-05,
      "loss": 6.9655,
      "mean_token_accuracy": 0.10120068341493607,
      "num_tokens": 270372.0,
      "step": 155
    },
    {
      "entropy": 7.463044929504394,
      "epoch": 0.012448939894962069,
      "grad_norm": 1.046875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.026,
      "mean_token_accuracy": 0.09834155589342117,
      "num_tokens": 279726.0,
      "step": 160
    },
    {
      "entropy": 7.4832648754119875,
      "epoch": 0.012837969266679634,
      "grad_norm": 1.109375,
      "learning_rate": 8.2e-05,
      "loss": 6.9905,
      "mean_token_accuracy": 0.10089058354496956,
      "num_tokens": 288955.0,
      "step": 165
    },
    {
      "entropy": 7.449814510345459,
      "epoch": 0.013226998638397198,
      "grad_norm": 1.1171875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.9218,
      "mean_token_accuracy": 0.09420814290642739,
      "num_tokens": 297898.0,
      "step": 170
    },
    {
      "entropy": 7.396070957183838,
      "epoch": 0.013616028010114764,
      "grad_norm": 1.046875,
      "learning_rate": 8.7e-05,
      "loss": 6.9188,
      "mean_token_accuracy": 0.10343953669071197,
      "num_tokens": 306189.0,
      "step": 175
    },
    {
      "entropy": 7.427816915512085,
      "epoch": 0.014005057381832327,
      "grad_norm": 1.0703125,
      "learning_rate": 8.95e-05,
      "loss": 6.9462,
      "mean_token_accuracy": 0.09905305579304695,
      "num_tokens": 314881.0,
      "step": 180
    },
    {
      "entropy": 7.363733434677124,
      "epoch": 0.014394086753549893,
      "grad_norm": 1.1796875,
      "learning_rate": 9.2e-05,
      "loss": 6.9415,
      "mean_token_accuracy": 0.09854480773210525,
      "num_tokens": 323357.0,
      "step": 185
    },
    {
      "entropy": 7.500346422195435,
      "epoch": 0.014783116125267459,
      "grad_norm": 1.046875,
      "learning_rate": 9.45e-05,
      "loss": 6.9926,
      "mean_token_accuracy": 0.09660296440124512,
      "num_tokens": 331799.0,
      "step": 190
    },
    {
      "entropy": 7.395345211029053,
      "epoch": 0.015172145496985022,
      "grad_norm": 1.125,
      "learning_rate": 9.7e-05,
      "loss": 6.904,
      "mean_token_accuracy": 0.10265823528170585,
      "num_tokens": 340320.0,
      "step": 195
    },
    {
      "entropy": 7.404849243164063,
      "epoch": 0.015561174868702588,
      "grad_norm": 1.171875,
      "learning_rate": 9.95e-05,
      "loss": 6.8857,
      "mean_token_accuracy": 0.1047187753021717,
      "num_tokens": 348990.0,
      "step": 200
    },
    {
      "entropy": 7.39699239730835,
      "epoch": 0.01595020424042015,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000102,
      "loss": 6.9065,
      "mean_token_accuracy": 0.09710254445672035,
      "num_tokens": 357661.0,
      "step": 205
    },
    {
      "entropy": 7.321025276184082,
      "epoch": 0.016339233612137717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.8787,
      "mean_token_accuracy": 0.10586154833436012,
      "num_tokens": 366099.0,
      "step": 210
    },
    {
      "entropy": 7.5005217552185055,
      "epoch": 0.016728262983855283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000107,
      "loss": 7.0379,
      "mean_token_accuracy": 0.09788483828306198,
      "num_tokens": 374926.0,
      "step": 215
    },
    {
      "entropy": 7.418247509002685,
      "epoch": 0.017117292355572845,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001095,
      "loss": 6.946,
      "mean_token_accuracy": 0.098393777012825,
      "num_tokens": 383846.0,
      "step": 220
    },
    {
      "entropy": 7.378245735168457,
      "epoch": 0.01750632172729041,
      "grad_norm": 1.171875,
      "learning_rate": 0.000112,
      "loss": 6.8894,
      "mean_token_accuracy": 0.1021482452750206,
      "num_tokens": 393678.0,
      "step": 225
    },
    {
      "entropy": 7.4670178413391115,
      "epoch": 0.017895351099007976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001145,
      "loss": 6.9755,
      "mean_token_accuracy": 0.09847855567932129,
      "num_tokens": 401866.0,
      "step": 230
    },
    {
      "entropy": 7.473496532440185,
      "epoch": 0.01828438047072554,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.9365,
      "mean_token_accuracy": 0.10552924722433091,
      "num_tokens": 411057.0,
      "step": 235
    },
    {
      "entropy": 7.397612524032593,
      "epoch": 0.018673409842443103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.9395,
      "mean_token_accuracy": 0.10260751843452454,
      "num_tokens": 419083.0,
      "step": 240
    },
    {
      "entropy": 7.325595569610596,
      "epoch": 0.01906243921416067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000122,
      "loss": 6.7993,
      "mean_token_accuracy": 0.10741342976689339,
      "num_tokens": 427503.0,
      "step": 245
    },
    {
      "entropy": 7.431865072250366,
      "epoch": 0.019451468585878234,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001245,
      "loss": 6.9643,
      "mean_token_accuracy": 0.09718445464968681,
      "num_tokens": 435820.0,
      "step": 250
    },
    {
      "entropy": 7.301518154144287,
      "epoch": 0.0198404979575958,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000127,
      "loss": 6.8424,
      "mean_token_accuracy": 0.10523844137787819,
      "num_tokens": 444692.0,
      "step": 255
    },
    {
      "entropy": 7.322385978698731,
      "epoch": 0.020229527329313362,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001295,
      "loss": 6.8813,
      "mean_token_accuracy": 0.09604761153459548,
      "num_tokens": 453536.0,
      "step": 260
    },
    {
      "entropy": 7.500930261611939,
      "epoch": 0.020618556701030927,
      "grad_norm": 1.265625,
      "learning_rate": 0.000132,
      "loss": 6.9135,
      "mean_token_accuracy": 0.10178809389472007,
      "num_tokens": 462264.0,
      "step": 265
    },
    {
      "entropy": 7.466741132736206,
      "epoch": 0.021007586072748493,
      "grad_norm": 1.09375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0681,
      "mean_token_accuracy": 0.09854813367128372,
      "num_tokens": 471819.0,
      "step": 270
    },
    {
      "entropy": 7.39691653251648,
      "epoch": 0.02139661544446606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.9438,
      "mean_token_accuracy": 0.0978664368391037,
      "num_tokens": 480878.0,
      "step": 275
    },
    {
      "entropy": 7.483294820785522,
      "epoch": 0.02178564481618362,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001395,
      "loss": 6.9213,
      "mean_token_accuracy": 0.1020102322101593,
      "num_tokens": 489454.0,
      "step": 280
    },
    {
      "entropy": 7.298464393615722,
      "epoch": 0.022174674187901186,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.8179,
      "mean_token_accuracy": 0.10413895472884178,
      "num_tokens": 498378.0,
      "step": 285
    },
    {
      "entropy": 7.501266813278198,
      "epoch": 0.02256370355961875,
      "grad_norm": 1.234375,
      "learning_rate": 0.0001445,
      "loss": 7.0766,
      "mean_token_accuracy": 0.09656166881322861,
      "num_tokens": 507638.0,
      "step": 290
    },
    {
      "entropy": 7.435027980804444,
      "epoch": 0.022952732931336317,
      "grad_norm": 1.265625,
      "learning_rate": 0.000147,
      "loss": 6.9631,
      "mean_token_accuracy": 0.10347767695784568,
      "num_tokens": 517047.0,
      "step": 295
    },
    {
      "entropy": 7.468677520751953,
      "epoch": 0.02334176230305388,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001495,
      "loss": 6.8928,
      "mean_token_accuracy": 0.10784424841403961,
      "num_tokens": 525287.0,
      "step": 300
    },
    {
      "entropy": 7.284840393066406,
      "epoch": 0.023730791674771445,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000152,
      "loss": 6.827,
      "mean_token_accuracy": 0.09975214079022407,
      "num_tokens": 533266.0,
      "step": 305
    },
    {
      "entropy": 7.383808851242065,
      "epoch": 0.02411982104648901,
      "grad_norm": 1.125,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9865,
      "mean_token_accuracy": 0.09793580844998359,
      "num_tokens": 542448.0,
      "step": 310
    },
    {
      "entropy": 7.3777937412261965,
      "epoch": 0.024508850418206576,
      "grad_norm": 1.40625,
      "learning_rate": 0.000157,
      "loss": 6.8559,
      "mean_token_accuracy": 0.10171961411833763,
      "num_tokens": 550872.0,
      "step": 315
    },
    {
      "entropy": 7.398427724838257,
      "epoch": 0.024897879789924138,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001595,
      "loss": 6.945,
      "mean_token_accuracy": 0.10726686492562294,
      "num_tokens": 559616.0,
      "step": 320
    },
    {
      "entropy": 7.43606333732605,
      "epoch": 0.025286909161641703,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000162,
      "loss": 6.892,
      "mean_token_accuracy": 0.1080216646194458,
      "num_tokens": 568379.0,
      "step": 325
    },
    {
      "entropy": 7.339574384689331,
      "epoch": 0.02567593853335927,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.9344,
      "mean_token_accuracy": 0.09556459337472915,
      "num_tokens": 577288.0,
      "step": 330
    },
    {
      "entropy": 7.394017219543457,
      "epoch": 0.026064967905076834,
      "grad_norm": 1.140625,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8893,
      "mean_token_accuracy": 0.09673169776797294,
      "num_tokens": 587054.0,
      "step": 335
    },
    {
      "entropy": 7.3556236743927,
      "epoch": 0.026453997276794396,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0419,
      "mean_token_accuracy": 0.10046090409159661,
      "num_tokens": 595405.0,
      "step": 340
    },
    {
      "entropy": 7.431601905822754,
      "epoch": 0.026843026648511962,
      "grad_norm": 1.1875,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8507,
      "mean_token_accuracy": 0.10375126153230667,
      "num_tokens": 604095.0,
      "step": 345
    },
    {
      "entropy": 7.273375415802002,
      "epoch": 0.027232056020229527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9004,
      "mean_token_accuracy": 0.10230252891778946,
      "num_tokens": 613518.0,
      "step": 350
    },
    {
      "entropy": 7.408315944671631,
      "epoch": 0.027621085391947093,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000177,
      "loss": 6.7702,
      "mean_token_accuracy": 0.1080463670194149,
      "num_tokens": 621212.0,
      "step": 355
    },
    {
      "entropy": 7.285730218887329,
      "epoch": 0.028010114763664655,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001795,
      "loss": 6.8426,
      "mean_token_accuracy": 0.1072110302746296,
      "num_tokens": 629326.0,
      "step": 360
    },
    {
      "entropy": 7.37860689163208,
      "epoch": 0.02839914413538222,
      "grad_norm": 1.203125,
      "learning_rate": 0.000182,
      "loss": 6.8696,
      "mean_token_accuracy": 0.10251942202448845,
      "num_tokens": 637034.0,
      "step": 365
    },
    {
      "entropy": 7.277339124679566,
      "epoch": 0.028788173507099786,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001845,
      "loss": 6.7723,
      "mean_token_accuracy": 0.10722191333770752,
      "num_tokens": 645924.0,
      "step": 370
    },
    {
      "entropy": 7.407964897155762,
      "epoch": 0.02917720287881735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000187,
      "loss": 7.0054,
      "mean_token_accuracy": 0.10154261440038681,
      "num_tokens": 655020.0,
      "step": 375
    },
    {
      "entropy": 7.300928163528442,
      "epoch": 0.029566232250534917,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001895,
      "loss": 6.8082,
      "mean_token_accuracy": 0.10665615424513816,
      "num_tokens": 663451.0,
      "step": 380
    },
    {
      "entropy": 7.367715358734131,
      "epoch": 0.02995526162225248,
      "grad_norm": 1.1875,
      "learning_rate": 0.000192,
      "loss": 6.9123,
      "mean_token_accuracy": 0.1059771090745926,
      "num_tokens": 671973.0,
      "step": 385
    },
    {
      "entropy": 7.3022441387176515,
      "epoch": 0.030344290993970045,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001945,
      "loss": 6.797,
      "mean_token_accuracy": 0.10909434482455253,
      "num_tokens": 680583.0,
      "step": 390
    },
    {
      "entropy": 7.2918174266815186,
      "epoch": 0.03073332036568761,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8136,
      "mean_token_accuracy": 0.10621452257037163,
      "num_tokens": 689000.0,
      "step": 395
    },
    {
      "entropy": 7.341504812240601,
      "epoch": 0.031122349737405176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8056,
      "mean_token_accuracy": 0.10639143735170364,
      "num_tokens": 698047.0,
      "step": 400
    },
    {
      "entropy": 7.3614107131958,
      "epoch": 0.03151137910912274,
      "grad_norm": 1.078125,
      "learning_rate": 0.000202,
      "loss": 6.9386,
      "mean_token_accuracy": 0.10114315450191498,
      "num_tokens": 707192.0,
      "step": 405
    },
    {
      "entropy": 7.2852061748504635,
      "epoch": 0.0319004084808403,
      "grad_norm": 1.421875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.735,
      "mean_token_accuracy": 0.11116274818778038,
      "num_tokens": 715895.0,
      "step": 410
    },
    {
      "entropy": 7.310044527053833,
      "epoch": 0.03228943785255787,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000207,
      "loss": 6.847,
      "mean_token_accuracy": 0.10309791564941406,
      "num_tokens": 724891.0,
      "step": 415
    },
    {
      "entropy": 7.340569972991943,
      "epoch": 0.032678467224275434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002095,
      "loss": 6.8936,
      "mean_token_accuracy": 0.10329565331339836,
      "num_tokens": 733378.0,
      "step": 420
    },
    {
      "entropy": 7.354770565032959,
      "epoch": 0.033067496595993,
      "grad_norm": 1.265625,
      "learning_rate": 0.000212,
      "loss": 6.875,
      "mean_token_accuracy": 0.10341520011425018,
      "num_tokens": 741098.0,
      "step": 425
    },
    {
      "entropy": 7.338204622268677,
      "epoch": 0.033456525967710565,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002145,
      "loss": 6.8378,
      "mean_token_accuracy": 0.10154023990035058,
      "num_tokens": 749902.0,
      "step": 430
    },
    {
      "entropy": 7.19447169303894,
      "epoch": 0.033845555339428124,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8057,
      "mean_token_accuracy": 0.10481407195329666,
      "num_tokens": 758638.0,
      "step": 435
    },
    {
      "entropy": 7.39279146194458,
      "epoch": 0.03423458471114569,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002195,
      "loss": 6.9067,
      "mean_token_accuracy": 0.10383950993418693,
      "num_tokens": 768170.0,
      "step": 440
    },
    {
      "entropy": 7.251432418823242,
      "epoch": 0.034623614082863255,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000222,
      "loss": 6.7917,
      "mean_token_accuracy": 0.10390266478061676,
      "num_tokens": 776745.0,
      "step": 445
    },
    {
      "entropy": 7.344013214111328,
      "epoch": 0.03501264345458082,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002245,
      "loss": 6.9049,
      "mean_token_accuracy": 0.10472740978002548,
      "num_tokens": 785030.0,
      "step": 450
    },
    {
      "entropy": 7.348719453811645,
      "epoch": 0.035401672826298386,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8656,
      "mean_token_accuracy": 0.10142385959625244,
      "num_tokens": 793543.0,
      "step": 455
    },
    {
      "entropy": 7.12188196182251,
      "epoch": 0.03579070219801595,
      "grad_norm": 1.25,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7096,
      "mean_token_accuracy": 0.1019783414900303,
      "num_tokens": 802401.0,
      "step": 460
    },
    {
      "entropy": 7.367281770706176,
      "epoch": 0.03617973156973352,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8284,
      "mean_token_accuracy": 0.10576049014925956,
      "num_tokens": 811625.0,
      "step": 465
    },
    {
      "entropy": 7.186238098144531,
      "epoch": 0.03656876094145108,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7775,
      "mean_token_accuracy": 0.10242061167955399,
      "num_tokens": 820742.0,
      "step": 470
    },
    {
      "entropy": 7.257758092880249,
      "epoch": 0.03695779031316864,
      "grad_norm": 1.03125,
      "learning_rate": 0.000237,
      "loss": 6.8914,
      "mean_token_accuracy": 0.10146046429872513,
      "num_tokens": 829580.0,
      "step": 475
    },
    {
      "entropy": 7.289563941955566,
      "epoch": 0.03734681968488621,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002395,
      "loss": 6.8122,
      "mean_token_accuracy": 0.10589663460850715,
      "num_tokens": 838818.0,
      "step": 480
    },
    {
      "entropy": 7.339207744598388,
      "epoch": 0.03773584905660377,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000242,
      "loss": 6.8286,
      "mean_token_accuracy": 0.09875347390770912,
      "num_tokens": 847804.0,
      "step": 485
    },
    {
      "entropy": 7.0710687160491945,
      "epoch": 0.03812487842832134,
      "grad_norm": 1.25,
      "learning_rate": 0.0002445,
      "loss": 6.5909,
      "mean_token_accuracy": 0.11272379085421562,
      "num_tokens": 856395.0,
      "step": 490
    },
    {
      "entropy": 7.308010816574097,
      "epoch": 0.0385139078000389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000247,
      "loss": 6.8648,
      "mean_token_accuracy": 0.10699901208281518,
      "num_tokens": 865219.0,
      "step": 495
    },
    {
      "entropy": 7.108584594726563,
      "epoch": 0.03890293717175647,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002495,
      "loss": 6.6471,
      "mean_token_accuracy": 0.1104583702981472,
      "num_tokens": 873868.0,
      "step": 500
    },
    {
      "entropy": 7.20792145729065,
      "epoch": 0.039291966543474034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000252,
      "loss": 6.7601,
      "mean_token_accuracy": 0.10320616662502288,
      "num_tokens": 882698.0,
      "step": 505
    },
    {
      "entropy": 7.190888500213623,
      "epoch": 0.0396809959151916,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002545,
      "loss": 6.6947,
      "mean_token_accuracy": 0.11407066881656647,
      "num_tokens": 891280.0,
      "step": 510
    },
    {
      "entropy": 7.244538879394531,
      "epoch": 0.04007002528690916,
      "grad_norm": 1.25,
      "learning_rate": 0.000257,
      "loss": 6.7425,
      "mean_token_accuracy": 0.09894983395934105,
      "num_tokens": 900044.0,
      "step": 515
    },
    {
      "entropy": 7.08066873550415,
      "epoch": 0.040459054658626724,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002595,
      "loss": 6.7217,
      "mean_token_accuracy": 0.0992046169936657,
      "num_tokens": 909297.0,
      "step": 520
    },
    {
      "entropy": 7.1543323516845705,
      "epoch": 0.04084808403034429,
      "grad_norm": 1.25,
      "learning_rate": 0.000262,
      "loss": 6.7075,
      "mean_token_accuracy": 0.10772032216191292,
      "num_tokens": 917514.0,
      "step": 525
    },
    {
      "entropy": 7.194704627990722,
      "epoch": 0.041237113402061855,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6951,
      "mean_token_accuracy": 0.1101997010409832,
      "num_tokens": 925842.0,
      "step": 530
    },
    {
      "entropy": 7.067590141296387,
      "epoch": 0.04162614277377942,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8094,
      "mean_token_accuracy": 0.09961830377578736,
      "num_tokens": 934676.0,
      "step": 535
    },
    {
      "entropy": 7.194618606567383,
      "epoch": 0.042015172145496986,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7182,
      "mean_token_accuracy": 0.10596418976783753,
      "num_tokens": 943763.0,
      "step": 540
    },
    {
      "entropy": 7.158230352401733,
      "epoch": 0.04240420151721455,
      "grad_norm": 1.25,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7183,
      "mean_token_accuracy": 0.1088571310043335,
      "num_tokens": 952721.0,
      "step": 545
    },
    {
      "entropy": 7.143984985351563,
      "epoch": 0.04279323088893212,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002745,
      "loss": 6.7143,
      "mean_token_accuracy": 0.11388995125889778,
      "num_tokens": 960899.0,
      "step": 550
    },
    {
      "entropy": 7.069172191619873,
      "epoch": 0.04318226026064968,
      "grad_norm": 1.375,
      "learning_rate": 0.000277,
      "loss": 6.7067,
      "mean_token_accuracy": 0.1085039384663105,
      "num_tokens": 969362.0,
      "step": 555
    },
    {
      "entropy": 7.178910207748413,
      "epoch": 0.04357128963236724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002795,
      "loss": 6.6775,
      "mean_token_accuracy": 0.11293213590979576,
      "num_tokens": 978028.0,
      "step": 560
    },
    {
      "entropy": 7.113482236862183,
      "epoch": 0.04396031900408481,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6594,
      "mean_token_accuracy": 0.11223076805472373,
      "num_tokens": 985709.0,
      "step": 565
    },
    {
      "entropy": 7.220449066162109,
      "epoch": 0.04434934837580237,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002845,
      "loss": 6.7666,
      "mean_token_accuracy": 0.10906944498419761,
      "num_tokens": 994769.0,
      "step": 570
    },
    {
      "entropy": 7.125522947311401,
      "epoch": 0.04473837774751994,
      "grad_norm": 1.234375,
      "learning_rate": 0.000287,
      "loss": 6.7365,
      "mean_token_accuracy": 0.11124032214283944,
      "num_tokens": 1003997.0,
      "step": 575
    },
    {
      "entropy": 7.162062454223633,
      "epoch": 0.0451274071192375,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002895,
      "loss": 6.7353,
      "mean_token_accuracy": 0.10934066548943519,
      "num_tokens": 1012931.0,
      "step": 580
    },
    {
      "entropy": 7.170513248443603,
      "epoch": 0.04551643649095507,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000292,
      "loss": 6.7534,
      "mean_token_accuracy": 0.10608652755618095,
      "num_tokens": 1021149.0,
      "step": 585
    },
    {
      "entropy": 7.184502792358399,
      "epoch": 0.045905465862672634,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002945,
      "loss": 6.7344,
      "mean_token_accuracy": 0.11254923641681672,
      "num_tokens": 1029421.0,
      "step": 590
    },
    {
      "entropy": 7.233080053329468,
      "epoch": 0.0462944952343902,
      "grad_norm": 1.453125,
      "learning_rate": 0.000297,
      "loss": 6.8544,
      "mean_token_accuracy": 0.10738511979579926,
      "num_tokens": 1037222.0,
      "step": 595
    },
    {
      "entropy": 7.11895523071289,
      "epoch": 0.04668352460610776,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002995,
      "loss": 6.7686,
      "mean_token_accuracy": 0.10719419568777085,
      "num_tokens": 1046903.0,
      "step": 600
    },
    {
      "entropy": 7.1854307651519775,
      "epoch": 0.047072553977825324,
      "grad_norm": 1.1875,
      "learning_rate": 0.000302,
      "loss": 6.8112,
      "mean_token_accuracy": 0.10291046276688576,
      "num_tokens": 1055516.0,
      "step": 605
    },
    {
      "entropy": 7.158183145523071,
      "epoch": 0.04746158334954289,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003045,
      "loss": 6.738,
      "mean_token_accuracy": 0.10869069769978523,
      "num_tokens": 1063846.0,
      "step": 610
    },
    {
      "entropy": 7.114259338378906,
      "epoch": 0.047850612721260455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000307,
      "loss": 6.7027,
      "mean_token_accuracy": 0.10939538329839707,
      "num_tokens": 1072500.0,
      "step": 615
    },
    {
      "entropy": 7.126255798339844,
      "epoch": 0.04823964209297802,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003095,
      "loss": 6.7245,
      "mean_token_accuracy": 0.10950846374034881,
      "num_tokens": 1081399.0,
      "step": 620
    },
    {
      "entropy": 7.04192967414856,
      "epoch": 0.048628671464695586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000312,
      "loss": 6.6774,
      "mean_token_accuracy": 0.11167947575449944,
      "num_tokens": 1089666.0,
      "step": 625
    },
    {
      "entropy": 7.143467807769776,
      "epoch": 0.04901770083641315,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003145,
      "loss": 6.5626,
      "mean_token_accuracy": 0.11535401195287705,
      "num_tokens": 1097496.0,
      "step": 630
    },
    {
      "entropy": 7.058351707458496,
      "epoch": 0.04940673020813072,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000317,
      "loss": 6.7028,
      "mean_token_accuracy": 0.11161455810070038,
      "num_tokens": 1105840.0,
      "step": 635
    },
    {
      "entropy": 7.135726547241211,
      "epoch": 0.049795759579848276,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003195,
      "loss": 6.7684,
      "mean_token_accuracy": 0.10728443264961243,
      "num_tokens": 1115507.0,
      "step": 640
    },
    {
      "entropy": 7.041813230514526,
      "epoch": 0.05018478895156584,
      "grad_norm": 1.078125,
      "learning_rate": 0.000322,
      "loss": 6.7167,
      "mean_token_accuracy": 0.10211084112524986,
      "num_tokens": 1124820.0,
      "step": 645
    },
    {
      "entropy": 7.087520885467529,
      "epoch": 0.05057381832328341,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5609,
      "mean_token_accuracy": 0.11316258385777474,
      "num_tokens": 1133641.0,
      "step": 650
    },
    {
      "entropy": 6.982738447189331,
      "epoch": 0.05096284769500097,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.616,
      "mean_token_accuracy": 0.11105628833174705,
      "num_tokens": 1142182.0,
      "step": 655
    },
    {
      "entropy": 7.091386175155639,
      "epoch": 0.05135187706671854,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6894,
      "mean_token_accuracy": 0.11387185752391815,
      "num_tokens": 1150737.0,
      "step": 660
    },
    {
      "entropy": 6.989086723327636,
      "epoch": 0.0517409064384361,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5525,
      "mean_token_accuracy": 0.11128119602799416,
      "num_tokens": 1158899.0,
      "step": 665
    },
    {
      "entropy": 7.124065494537353,
      "epoch": 0.05212993581015367,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6359,
      "mean_token_accuracy": 0.11705312058329582,
      "num_tokens": 1167397.0,
      "step": 670
    },
    {
      "entropy": 7.050602674484253,
      "epoch": 0.052518965181871234,
      "grad_norm": 1.1875,
      "learning_rate": 0.000337,
      "loss": 6.6545,
      "mean_token_accuracy": 0.10680999159812928,
      "num_tokens": 1177044.0,
      "step": 675
    },
    {
      "entropy": 6.930141878128052,
      "epoch": 0.05290799455358879,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003395,
      "loss": 6.524,
      "mean_token_accuracy": 0.11300303936004638,
      "num_tokens": 1185740.0,
      "step": 680
    },
    {
      "entropy": 6.976816177368164,
      "epoch": 0.05329702392530636,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000342,
      "loss": 6.6738,
      "mean_token_accuracy": 0.11296947225928307,
      "num_tokens": 1194970.0,
      "step": 685
    },
    {
      "entropy": 7.070580577850341,
      "epoch": 0.053686053297023924,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6134,
      "mean_token_accuracy": 0.11393867209553718,
      "num_tokens": 1203565.0,
      "step": 690
    },
    {
      "entropy": 7.023531770706176,
      "epoch": 0.05407508266874149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000347,
      "loss": 6.6592,
      "mean_token_accuracy": 0.10934962257742882,
      "num_tokens": 1213226.0,
      "step": 695
    },
    {
      "entropy": 6.905843925476074,
      "epoch": 0.054464112040459055,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003495,
      "loss": 6.5837,
      "mean_token_accuracy": 0.10722511857748032,
      "num_tokens": 1221710.0,
      "step": 700
    },
    {
      "entropy": 7.033835506439209,
      "epoch": 0.05485314141217662,
      "grad_norm": 1.046875,
      "learning_rate": 0.000352,
      "loss": 6.6413,
      "mean_token_accuracy": 0.10701307356357574,
      "num_tokens": 1231125.0,
      "step": 705
    },
    {
      "entropy": 7.04109468460083,
      "epoch": 0.055242170783894186,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003545,
      "loss": 6.5952,
      "mean_token_accuracy": 0.11328281164169311,
      "num_tokens": 1239521.0,
      "step": 710
    },
    {
      "entropy": 6.99030909538269,
      "epoch": 0.05563120015561175,
      "grad_norm": 1.34375,
      "learning_rate": 0.000357,
      "loss": 6.4967,
      "mean_token_accuracy": 0.11905190572142602,
      "num_tokens": 1247363.0,
      "step": 715
    },
    {
      "entropy": 6.949197673797608,
      "epoch": 0.05602022952732931,
      "grad_norm": 1.25,
      "learning_rate": 0.0003595,
      "loss": 6.5756,
      "mean_token_accuracy": 0.10785499513149262,
      "num_tokens": 1255395.0,
      "step": 720
    },
    {
      "entropy": 7.050080490112305,
      "epoch": 0.056409258899046875,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000362,
      "loss": 6.6461,
      "mean_token_accuracy": 0.10883620008826256,
      "num_tokens": 1264135.0,
      "step": 725
    },
    {
      "entropy": 7.0190126419067385,
      "epoch": 0.05679828827076444,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003645,
      "loss": 6.5703,
      "mean_token_accuracy": 0.1117468774318695,
      "num_tokens": 1272375.0,
      "step": 730
    },
    {
      "entropy": 6.988955736160278,
      "epoch": 0.057187317642482006,
      "grad_norm": 1.125,
      "learning_rate": 0.000367,
      "loss": 6.6249,
      "mean_token_accuracy": 0.10667285323143005,
      "num_tokens": 1280960.0,
      "step": 735
    },
    {
      "entropy": 7.041416883468628,
      "epoch": 0.05757634701419957,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003695,
      "loss": 6.6797,
      "mean_token_accuracy": 0.11113242357969284,
      "num_tokens": 1289273.0,
      "step": 740
    },
    {
      "entropy": 6.986820077896118,
      "epoch": 0.05796537638591714,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000372,
      "loss": 6.5673,
      "mean_token_accuracy": 0.11231251433491707,
      "num_tokens": 1297571.0,
      "step": 745
    },
    {
      "entropy": 7.002435827255249,
      "epoch": 0.0583544057576347,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003745,
      "loss": 6.6165,
      "mean_token_accuracy": 0.11419098302721978,
      "num_tokens": 1305686.0,
      "step": 750
    },
    {
      "entropy": 6.96313853263855,
      "epoch": 0.05874343512935227,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377,
      "loss": 6.6372,
      "mean_token_accuracy": 0.1087796851992607,
      "num_tokens": 1314483.0,
      "step": 755
    },
    {
      "entropy": 6.946442985534668,
      "epoch": 0.059132464501069834,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003795,
      "loss": 6.5161,
      "mean_token_accuracy": 0.10849898308515549,
      "num_tokens": 1323183.0,
      "step": 760
    },
    {
      "entropy": 6.952772951126098,
      "epoch": 0.05952149387278739,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000382,
      "loss": 6.6204,
      "mean_token_accuracy": 0.11015321984887123,
      "num_tokens": 1331964.0,
      "step": 765
    },
    {
      "entropy": 6.933269834518432,
      "epoch": 0.05991052324450496,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003845,
      "loss": 6.5707,
      "mean_token_accuracy": 0.11390582174062729,
      "num_tokens": 1340185.0,
      "step": 770
    },
    {
      "entropy": 7.053455209732055,
      "epoch": 0.060299552616222524,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6715,
      "mean_token_accuracy": 0.1104957066476345,
      "num_tokens": 1349522.0,
      "step": 775
    },
    {
      "entropy": 7.057522010803223,
      "epoch": 0.06068858198794009,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6406,
      "mean_token_accuracy": 0.10729108974337578,
      "num_tokens": 1357893.0,
      "step": 780
    },
    {
      "entropy": 6.856075191497803,
      "epoch": 0.061077611359657655,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5272,
      "mean_token_accuracy": 0.11494160592555999,
      "num_tokens": 1366685.0,
      "step": 785
    },
    {
      "entropy": 6.961356449127197,
      "epoch": 0.06146664073137522,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4661,
      "mean_token_accuracy": 0.11778462678194046,
      "num_tokens": 1375245.0,
      "step": 790
    },
    {
      "entropy": 6.9094297885894775,
      "epoch": 0.061855670103092786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.6459,
      "mean_token_accuracy": 0.1119820773601532,
      "num_tokens": 1383724.0,
      "step": 795
    },
    {
      "entropy": 7.064331769943237,
      "epoch": 0.06224469947481035,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003995,
      "loss": 6.6246,
      "mean_token_accuracy": 0.11086539849638939,
      "num_tokens": 1392455.0,
      "step": 800
    },
    {
      "entropy": 7.040333318710327,
      "epoch": 0.06263372884652792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000402,
      "loss": 6.6322,
      "mean_token_accuracy": 0.11222699806094169,
      "num_tokens": 1401262.0,
      "step": 805
    },
    {
      "entropy": 6.833669853210449,
      "epoch": 0.06302275821824548,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004045,
      "loss": 6.528,
      "mean_token_accuracy": 0.11181755363941193,
      "num_tokens": 1409991.0,
      "step": 810
    },
    {
      "entropy": 6.826368951797486,
      "epoch": 0.06341178758996305,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5143,
      "mean_token_accuracy": 0.11217057034373283,
      "num_tokens": 1418208.0,
      "step": 815
    },
    {
      "entropy": 7.009344673156738,
      "epoch": 0.0638008169616806,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004095,
      "loss": 6.5279,
      "mean_token_accuracy": 0.1156628392636776,
      "num_tokens": 1426665.0,
      "step": 820
    },
    {
      "entropy": 6.799173259735108,
      "epoch": 0.06418984633339817,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000412,
      "loss": 6.4582,
      "mean_token_accuracy": 0.11917970553040505,
      "num_tokens": 1435247.0,
      "step": 825
    },
    {
      "entropy": 6.870002889633179,
      "epoch": 0.06457887570511574,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004145,
      "loss": 6.5052,
      "mean_token_accuracy": 0.11340074241161346,
      "num_tokens": 1443678.0,
      "step": 830
    },
    {
      "entropy": 6.767337274551392,
      "epoch": 0.0649679050768333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000417,
      "loss": 6.4409,
      "mean_token_accuracy": 0.12152031734585762,
      "num_tokens": 1451567.0,
      "step": 835
    },
    {
      "entropy": 6.837034368515015,
      "epoch": 0.06535693444855087,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004195,
      "loss": 6.4578,
      "mean_token_accuracy": 0.11555799394845963,
      "num_tokens": 1460618.0,
      "step": 840
    },
    {
      "entropy": 6.84786286354065,
      "epoch": 0.06574596382026843,
      "grad_norm": 1.203125,
      "learning_rate": 0.000422,
      "loss": 6.521,
      "mean_token_accuracy": 0.1165824718773365,
      "num_tokens": 1469821.0,
      "step": 845
    },
    {
      "entropy": 6.948866891860962,
      "epoch": 0.066134993191986,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004245,
      "loss": 6.596,
      "mean_token_accuracy": 0.11146286725997925,
      "num_tokens": 1477989.0,
      "step": 850
    },
    {
      "entropy": 6.937360382080078,
      "epoch": 0.06652402256370356,
      "grad_norm": 1.203125,
      "learning_rate": 0.000427,
      "loss": 6.6151,
      "mean_token_accuracy": 0.11400183513760567,
      "num_tokens": 1486822.0,
      "step": 855
    },
    {
      "entropy": 6.887311553955078,
      "epoch": 0.06691305193542113,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004295,
      "loss": 6.4833,
      "mean_token_accuracy": 0.11117951571941376,
      "num_tokens": 1496604.0,
      "step": 860
    },
    {
      "entropy": 6.801223611831665,
      "epoch": 0.06730208130713869,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000432,
      "loss": 6.5104,
      "mean_token_accuracy": 0.11625081971287728,
      "num_tokens": 1504725.0,
      "step": 865
    },
    {
      "entropy": 6.889442777633667,
      "epoch": 0.06769111067885625,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004345,
      "loss": 6.5156,
      "mean_token_accuracy": 0.11729802563786507,
      "num_tokens": 1513222.0,
      "step": 870
    },
    {
      "entropy": 6.880262184143066,
      "epoch": 0.06808014005057382,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000437,
      "loss": 6.4837,
      "mean_token_accuracy": 0.1164153091609478,
      "num_tokens": 1522183.0,
      "step": 875
    },
    {
      "entropy": 6.819140386581421,
      "epoch": 0.06846916942229138,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004395,
      "loss": 6.5135,
      "mean_token_accuracy": 0.11310733184218406,
      "num_tokens": 1531316.0,
      "step": 880
    },
    {
      "entropy": 6.924764108657837,
      "epoch": 0.06885819879400895,
      "grad_norm": 1.265625,
      "learning_rate": 0.000442,
      "loss": 6.579,
      "mean_token_accuracy": 0.11591333150863647,
      "num_tokens": 1540053.0,
      "step": 885
    },
    {
      "entropy": 6.83437933921814,
      "epoch": 0.06924722816572651,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004445,
      "loss": 6.4385,
      "mean_token_accuracy": 0.1197456382215023,
      "num_tokens": 1549391.0,
      "step": 890
    },
    {
      "entropy": 6.9253668785095215,
      "epoch": 0.06963625753744408,
      "grad_norm": 1.25,
      "learning_rate": 0.000447,
      "loss": 6.5571,
      "mean_token_accuracy": 0.1145814374089241,
      "num_tokens": 1557364.0,
      "step": 895
    },
    {
      "entropy": 6.799445247650146,
      "epoch": 0.07002528690916164,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3947,
      "mean_token_accuracy": 0.1110591433942318,
      "num_tokens": 1567011.0,
      "step": 900
    },
    {
      "entropy": 6.749784517288208,
      "epoch": 0.0704143162808792,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5116,
      "mean_token_accuracy": 0.11603075638413429,
      "num_tokens": 1575441.0,
      "step": 905
    },
    {
      "entropy": 6.821898317337036,
      "epoch": 0.07080334565259677,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4575,
      "mean_token_accuracy": 0.11487730145454407,
      "num_tokens": 1583757.0,
      "step": 910
    },
    {
      "entropy": 6.860511064529419,
      "epoch": 0.07119237502431433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.5019,
      "mean_token_accuracy": 0.11874779388308525,
      "num_tokens": 1592406.0,
      "step": 915
    },
    {
      "entropy": 6.792093992233276,
      "epoch": 0.0715814043960319,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4787,
      "mean_token_accuracy": 0.11637907922267914,
      "num_tokens": 1601665.0,
      "step": 920
    },
    {
      "entropy": 6.8411633491516115,
      "epoch": 0.07197043376774946,
      "grad_norm": 1.46875,
      "learning_rate": 0.000462,
      "loss": 6.502,
      "mean_token_accuracy": 0.11683260723948478,
      "num_tokens": 1609788.0,
      "step": 925
    },
    {
      "entropy": 6.7901835441589355,
      "epoch": 0.07235946313946703,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004645,
      "loss": 6.5025,
      "mean_token_accuracy": 0.11982274875044822,
      "num_tokens": 1618030.0,
      "step": 930
    },
    {
      "entropy": 6.767443609237671,
      "epoch": 0.07274849251118459,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000467,
      "loss": 6.5022,
      "mean_token_accuracy": 0.11648002341389656,
      "num_tokens": 1626901.0,
      "step": 935
    },
    {
      "entropy": 6.773926687240601,
      "epoch": 0.07313752188290217,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004695,
      "loss": 6.3908,
      "mean_token_accuracy": 0.1169219508767128,
      "num_tokens": 1635372.0,
      "step": 940
    },
    {
      "entropy": 6.745931243896484,
      "epoch": 0.07352655125461972,
      "grad_norm": 1.15625,
      "learning_rate": 0.000472,
      "loss": 6.4347,
      "mean_token_accuracy": 0.1165583461523056,
      "num_tokens": 1643791.0,
      "step": 945
    },
    {
      "entropy": 6.671934700012207,
      "epoch": 0.07391558062633728,
      "grad_norm": 1.25,
      "learning_rate": 0.0004745,
      "loss": 6.4064,
      "mean_token_accuracy": 0.12068933174014092,
      "num_tokens": 1652161.0,
      "step": 950
    },
    {
      "entropy": 6.832337713241577,
      "epoch": 0.07430460999805485,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000477,
      "loss": 6.4351,
      "mean_token_accuracy": 0.11309069246053696,
      "num_tokens": 1661077.0,
      "step": 955
    },
    {
      "entropy": 6.688000345230103,
      "epoch": 0.07469363936977241,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004795,
      "loss": 6.4195,
      "mean_token_accuracy": 0.11566439867019654,
      "num_tokens": 1669010.0,
      "step": 960
    },
    {
      "entropy": 6.826863241195679,
      "epoch": 0.07508266874148999,
      "grad_norm": 1.15625,
      "learning_rate": 0.000482,
      "loss": 6.5798,
      "mean_token_accuracy": 0.11118761152029037,
      "num_tokens": 1678002.0,
      "step": 965
    },
    {
      "entropy": 6.840054130554199,
      "epoch": 0.07547169811320754,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004845,
      "loss": 6.532,
      "mean_token_accuracy": 0.11202086359262467,
      "num_tokens": 1687403.0,
      "step": 970
    },
    {
      "entropy": 6.791687536239624,
      "epoch": 0.07586072748492512,
      "grad_norm": 1.203125,
      "learning_rate": 0.000487,
      "loss": 6.4514,
      "mean_token_accuracy": 0.11927441358566285,
      "num_tokens": 1695512.0,
      "step": 975
    },
    {
      "entropy": 6.763000774383545,
      "epoch": 0.07624975685664268,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004895,
      "loss": 6.4496,
      "mean_token_accuracy": 0.1148236095905304,
      "num_tokens": 1705116.0,
      "step": 980
    },
    {
      "entropy": 6.659470129013061,
      "epoch": 0.07663878622836025,
      "grad_norm": 1.265625,
      "learning_rate": 0.000492,
      "loss": 6.3587,
      "mean_token_accuracy": 0.11876563057303428,
      "num_tokens": 1713715.0,
      "step": 985
    },
    {
      "entropy": 6.745743465423584,
      "epoch": 0.0770278156000778,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004945,
      "loss": 6.4189,
      "mean_token_accuracy": 0.11852910742163658,
      "num_tokens": 1721881.0,
      "step": 990
    },
    {
      "entropy": 6.7183277130126955,
      "epoch": 0.07741684497179536,
      "grad_norm": 1.1875,
      "learning_rate": 0.000497,
      "loss": 6.4432,
      "mean_token_accuracy": 0.1176742009818554,
      "num_tokens": 1730716.0,
      "step": 995
    },
    {
      "entropy": 6.888888120651245,
      "epoch": 0.07780587434351294,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004995,
      "loss": 6.5998,
      "mean_token_accuracy": 0.11513158455491065,
      "num_tokens": 1740257.0,
      "step": 1000
    },
    {
      "entropy": 6.787261962890625,
      "epoch": 0.0781949037152305,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999989075146,
      "loss": 6.5037,
      "mean_token_accuracy": 0.1148828312754631,
      "num_tokens": 1749071.0,
      "step": 1005
    },
    {
      "entropy": 6.743206596374511,
      "epoch": 0.07858393308694807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999944692927,
      "loss": 6.5174,
      "mean_token_accuracy": 0.10950450450181962,
      "num_tokens": 1757952.0,
      "step": 1010
    },
    {
      "entropy": 6.725841903686524,
      "epoch": 0.07897296245866563,
      "grad_norm": 1.125,
      "learning_rate": 0.000499999986617054,
      "loss": 6.4453,
      "mean_token_accuracy": 0.11559172496199607,
      "num_tokens": 1766027.0,
      "step": 1015
    },
    {
      "entropy": 6.709922885894775,
      "epoch": 0.0793619918303832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999753507986,
      "loss": 6.4473,
      "mean_token_accuracy": 0.11136003509163857,
      "num_tokens": 1775039.0,
      "step": 1020
    },
    {
      "entropy": 6.766154432296753,
      "epoch": 0.07975102120210076,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999606705267,
      "loss": 6.4508,
      "mean_token_accuracy": 0.12302565276622772,
      "num_tokens": 1783971.0,
      "step": 1025
    },
    {
      "entropy": 6.6978371143341064,
      "epoch": 0.08014005057381832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999999425762385,
      "loss": 6.3723,
      "mean_token_accuracy": 0.11690968126058579,
      "num_tokens": 1792743.0,
      "step": 1030
    },
    {
      "entropy": 6.746917724609375,
      "epoch": 0.08052907994553589,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999210679344,
      "loss": 6.5249,
      "mean_token_accuracy": 0.10790274739265442,
      "num_tokens": 1801435.0,
      "step": 1035
    },
    {
      "entropy": 6.829019927978516,
      "epoch": 0.08091810931725345,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999998961456145,
      "loss": 6.4185,
      "mean_token_accuracy": 0.11439956873655319,
      "num_tokens": 1810147.0,
      "step": 1040
    },
    {
      "entropy": 6.664861822128296,
      "epoch": 0.08130713868897102,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004999998678092794,
      "loss": 6.3459,
      "mean_token_accuracy": 0.12141675725579262,
      "num_tokens": 1818672.0,
      "step": 1045
    },
    {
      "entropy": 6.779337644577026,
      "epoch": 0.08169616806068858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999998360589293,
      "loss": 6.5038,
      "mean_token_accuracy": 0.11229550465941429,
      "num_tokens": 1827590.0,
      "step": 1050
    },
    {
      "entropy": 6.68260350227356,
      "epoch": 0.08208519743240615,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499999800894565,
      "loss": 6.4138,
      "mean_token_accuracy": 0.11804961189627647,
      "num_tokens": 1836467.0,
      "step": 1055
    },
    {
      "entropy": 6.659322786331177,
      "epoch": 0.08247422680412371,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999997623161866,
      "loss": 6.3545,
      "mean_token_accuracy": 0.11814910471439362,
      "num_tokens": 1845537.0,
      "step": 1060
    },
    {
      "entropy": 6.668975210189819,
      "epoch": 0.08286325617584128,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999997203237951,
      "loss": 6.3591,
      "mean_token_accuracy": 0.12657058462500573,
      "num_tokens": 1854433.0,
      "step": 1065
    },
    {
      "entropy": 6.699471855163575,
      "epoch": 0.08325228554755884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499999674917391,
      "loss": 6.4629,
      "mean_token_accuracy": 0.11414268612861633,
      "num_tokens": 1863708.0,
      "step": 1070
    },
    {
      "entropy": 6.6790022373199465,
      "epoch": 0.0836413149192764,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999996260969748,
      "loss": 6.3401,
      "mean_token_accuracy": 0.12237241193652153,
      "num_tokens": 1871854.0,
      "step": 1075
    },
    {
      "entropy": 6.618637943267823,
      "epoch": 0.08403034429099397,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999995738625475,
      "loss": 6.2953,
      "mean_token_accuracy": 0.12108489871025085,
      "num_tokens": 1880471.0,
      "step": 1080
    },
    {
      "entropy": 6.6455847263336185,
      "epoch": 0.08441937366271153,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999995182141099,
      "loss": 6.3643,
      "mean_token_accuracy": 0.12223557531833648,
      "num_tokens": 1889882.0,
      "step": 1085
    },
    {
      "entropy": 6.545784616470337,
      "epoch": 0.0848084030344291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999994591516626,
      "loss": 6.255,
      "mean_token_accuracy": 0.12821755334734916,
      "num_tokens": 1898701.0,
      "step": 1090
    },
    {
      "entropy": 6.611828470230103,
      "epoch": 0.08519743240614666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999993966752066,
      "loss": 6.3715,
      "mean_token_accuracy": 0.11571631282567978,
      "num_tokens": 1908165.0,
      "step": 1095
    },
    {
      "entropy": 6.640139150619507,
      "epoch": 0.08558646177786423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499999330784743,
      "loss": 6.4811,
      "mean_token_accuracy": 0.1190013438463211,
      "num_tokens": 1916295.0,
      "step": 1100
    },
    {
      "entropy": 6.72911229133606,
      "epoch": 0.08597549114958179,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999992614802725,
      "loss": 6.3657,
      "mean_token_accuracy": 0.12518487498164177,
      "num_tokens": 1924643.0,
      "step": 1105
    },
    {
      "entropy": 6.63273024559021,
      "epoch": 0.08636452052129936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999991887617966,
      "loss": 6.3742,
      "mean_token_accuracy": 0.11607154682278634,
      "num_tokens": 1933822.0,
      "step": 1110
    },
    {
      "entropy": 6.567139911651611,
      "epoch": 0.08675354989301692,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999991126293158,
      "loss": 6.3671,
      "mean_token_accuracy": 0.119285686314106,
      "num_tokens": 1943278.0,
      "step": 1115
    },
    {
      "entropy": 6.764870882034302,
      "epoch": 0.08714257926473448,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999990330828318,
      "loss": 6.4038,
      "mean_token_accuracy": 0.11785744428634644,
      "num_tokens": 1951883.0,
      "step": 1120
    },
    {
      "entropy": 6.537186765670777,
      "epoch": 0.08753160863645205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999989501223456,
      "loss": 6.2868,
      "mean_token_accuracy": 0.12786889150738717,
      "num_tokens": 1959567.0,
      "step": 1125
    },
    {
      "entropy": 6.700858688354492,
      "epoch": 0.08792063800816961,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999988637478584,
      "loss": 6.3722,
      "mean_token_accuracy": 0.12128480300307273,
      "num_tokens": 1968190.0,
      "step": 1130
    },
    {
      "entropy": 6.507136249542237,
      "epoch": 0.08830966737988719,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999987739593716,
      "loss": 6.2919,
      "mean_token_accuracy": 0.1241019494831562,
      "num_tokens": 1977519.0,
      "step": 1135
    },
    {
      "entropy": 6.6155976295471195,
      "epoch": 0.08869869675160474,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999986807568865,
      "loss": 6.2728,
      "mean_token_accuracy": 0.12157176062464714,
      "num_tokens": 1985796.0,
      "step": 1140
    },
    {
      "entropy": 6.703161907196045,
      "epoch": 0.08908772612332232,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999985841404045,
      "loss": 6.4198,
      "mean_token_accuracy": 0.11606316342949867,
      "num_tokens": 1993874.0,
      "step": 1145
    },
    {
      "entropy": 6.698982381820679,
      "epoch": 0.08947675549503988,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999984841099271,
      "loss": 6.3686,
      "mean_token_accuracy": 0.12184037938714028,
      "num_tokens": 2001663.0,
      "step": 1150
    },
    {
      "entropy": 6.559295749664306,
      "epoch": 0.08986578486675743,
      "grad_norm": 1.203125,
      "learning_rate": 0.000499998380665456,
      "loss": 6.3539,
      "mean_token_accuracy": 0.1229154959321022,
      "num_tokens": 2010179.0,
      "step": 1155
    },
    {
      "entropy": 6.653844833374023,
      "epoch": 0.090254814238475,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999982738069924,
      "loss": 6.2589,
      "mean_token_accuracy": 0.12252619490027428,
      "num_tokens": 2020013.0,
      "step": 1160
    },
    {
      "entropy": 6.594068193435669,
      "epoch": 0.09064384361019256,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004999981635345382,
      "loss": 6.3751,
      "mean_token_accuracy": 0.12590979486703874,
      "num_tokens": 2028396.0,
      "step": 1165
    },
    {
      "entropy": 6.726529788970947,
      "epoch": 0.09103287298191014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999980498480949,
      "loss": 6.4733,
      "mean_token_accuracy": 0.11403553262352943,
      "num_tokens": 2036947.0,
      "step": 1170
    },
    {
      "entropy": 6.585730981826782,
      "epoch": 0.0914219023536277,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999979327476644,
      "loss": 6.3254,
      "mean_token_accuracy": 0.11759492009878159,
      "num_tokens": 2045162.0,
      "step": 1175
    },
    {
      "entropy": 6.6485522270202635,
      "epoch": 0.09181093172534527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999978122332485,
      "loss": 6.4095,
      "mean_token_accuracy": 0.11743757203221321,
      "num_tokens": 2054198.0,
      "step": 1180
    },
    {
      "entropy": 6.583405160903931,
      "epoch": 0.09219996109706283,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999976883048487,
      "loss": 6.3297,
      "mean_token_accuracy": 0.11797131523489952,
      "num_tokens": 2062558.0,
      "step": 1185
    },
    {
      "entropy": 6.551700019836426,
      "epoch": 0.0925889904687804,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999975609624673,
      "loss": 6.2095,
      "mean_token_accuracy": 0.11727948859333992,
      "num_tokens": 2070816.0,
      "step": 1190
    },
    {
      "entropy": 6.552899169921875,
      "epoch": 0.09297801984049796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999974302061059,
      "loss": 6.2712,
      "mean_token_accuracy": 0.12385502606630325,
      "num_tokens": 2079058.0,
      "step": 1195
    },
    {
      "entropy": 6.530488681793213,
      "epoch": 0.09336704921221552,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999972960357666,
      "loss": 6.2343,
      "mean_token_accuracy": 0.12390355318784714,
      "num_tokens": 2086679.0,
      "step": 1200
    },
    {
      "entropy": 6.622032976150512,
      "epoch": 0.09375607858393309,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004999971584514516,
      "loss": 6.3462,
      "mean_token_accuracy": 0.12203274443745613,
      "num_tokens": 2095788.0,
      "step": 1205
    },
    {
      "entropy": 6.5792169094085695,
      "epoch": 0.09414510795565065,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999970174531627,
      "loss": 6.3628,
      "mean_token_accuracy": 0.11934652253985405,
      "num_tokens": 2105435.0,
      "step": 1210
    },
    {
      "entropy": 6.570254611968994,
      "epoch": 0.09453413732736822,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999968730409022,
      "loss": 6.3156,
      "mean_token_accuracy": 0.1212031789124012,
      "num_tokens": 2114318.0,
      "step": 1215
    },
    {
      "entropy": 6.521538877487183,
      "epoch": 0.09492316669908578,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999967252146723,
      "loss": 6.245,
      "mean_token_accuracy": 0.13023234903812408,
      "num_tokens": 2122707.0,
      "step": 1220
    },
    {
      "entropy": 6.497583961486816,
      "epoch": 0.09531219607080335,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999965739744752,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1240180067718029,
      "num_tokens": 2131176.0,
      "step": 1225
    },
    {
      "entropy": 6.681749296188355,
      "epoch": 0.09570122544252091,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999964193203132,
      "loss": 6.3785,
      "mean_token_accuracy": 0.12624216973781585,
      "num_tokens": 2139872.0,
      "step": 1230
    },
    {
      "entropy": 6.508302974700928,
      "epoch": 0.09609025481423847,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999962612521888,
      "loss": 6.2386,
      "mean_token_accuracy": 0.1306472271680832,
      "num_tokens": 2148668.0,
      "step": 1235
    },
    {
      "entropy": 6.534418439865112,
      "epoch": 0.09647928418595604,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999960997701041,
      "loss": 6.24,
      "mean_token_accuracy": 0.12787432298064233,
      "num_tokens": 2156628.0,
      "step": 1240
    },
    {
      "entropy": 6.563996267318726,
      "epoch": 0.0968683135576736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999959348740618,
      "loss": 6.3335,
      "mean_token_accuracy": 0.11790938824415206,
      "num_tokens": 2165620.0,
      "step": 1245
    },
    {
      "entropy": 6.626491832733154,
      "epoch": 0.09725734292939117,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999957665640641,
      "loss": 6.3967,
      "mean_token_accuracy": 0.11869805678725243,
      "num_tokens": 2174493.0,
      "step": 1250
    },
    {
      "entropy": 6.55765929222107,
      "epoch": 0.09764637230110873,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499995594840114,
      "loss": 6.2909,
      "mean_token_accuracy": 0.11683699488639832,
      "num_tokens": 2183258.0,
      "step": 1255
    },
    {
      "entropy": 6.548435878753662,
      "epoch": 0.0980354016728263,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999954197022138,
      "loss": 6.3454,
      "mean_token_accuracy": 0.11895403861999512,
      "num_tokens": 2192263.0,
      "step": 1260
    },
    {
      "entropy": 6.651948308944702,
      "epoch": 0.09842443104454386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999952411503661,
      "loss": 6.3889,
      "mean_token_accuracy": 0.1177945502102375,
      "num_tokens": 2201797.0,
      "step": 1265
    },
    {
      "entropy": 6.354124736785889,
      "epoch": 0.09881346041626143,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999950591845738,
      "loss": 6.1299,
      "mean_token_accuracy": 0.1311613693833351,
      "num_tokens": 2209454.0,
      "step": 1270
    },
    {
      "entropy": 6.649595832824707,
      "epoch": 0.09920248978797899,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999948738048397,
      "loss": 6.3022,
      "mean_token_accuracy": 0.12611282840371132,
      "num_tokens": 2217761.0,
      "step": 1275
    },
    {
      "entropy": 6.47292742729187,
      "epoch": 0.09959151915969655,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999946850111663,
      "loss": 6.2906,
      "mean_token_accuracy": 0.12207216396927834,
      "num_tokens": 2225793.0,
      "step": 1280
    },
    {
      "entropy": 6.592188453674316,
      "epoch": 0.09998054853141412,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999944928035569,
      "loss": 6.1805,
      "mean_token_accuracy": 0.1319039449095726,
      "num_tokens": 2234620.0,
      "step": 1285
    },
    {
      "entropy": 6.4829436302185055,
      "epoch": 0.10036957790313168,
      "grad_norm": 1.125,
      "learning_rate": 0.000499994297182014,
      "loss": 6.2643,
      "mean_token_accuracy": 0.12326281145215034,
      "num_tokens": 2242829.0,
      "step": 1290
    },
    {
      "entropy": 6.57083477973938,
      "epoch": 0.10075860727484925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999940981465408,
      "loss": 6.3014,
      "mean_token_accuracy": 0.11634257808327675,
      "num_tokens": 2251356.0,
      "step": 1295
    },
    {
      "entropy": 6.523905658721924,
      "epoch": 0.10114763664656681,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999938956971404,
      "loss": 6.3684,
      "mean_token_accuracy": 0.11852610781788826,
      "num_tokens": 2259327.0,
      "step": 1300
    },
    {
      "entropy": 6.59006552696228,
      "epoch": 0.10153666601828439,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999936898338156,
      "loss": 6.3871,
      "mean_token_accuracy": 0.11884624063968659,
      "num_tokens": 2267948.0,
      "step": 1305
    },
    {
      "entropy": 6.446884679794311,
      "epoch": 0.10192569539000194,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999934805565698,
      "loss": 6.194,
      "mean_token_accuracy": 0.1225943498313427,
      "num_tokens": 2275828.0,
      "step": 1310
    },
    {
      "entropy": 6.5004541873931885,
      "epoch": 0.10231472476171952,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499993267865406,
      "loss": 6.1373,
      "mean_token_accuracy": 0.1289021112024784,
      "num_tokens": 2284246.0,
      "step": 1315
    },
    {
      "entropy": 6.385093450546265,
      "epoch": 0.10270375413343708,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999930517603275,
      "loss": 6.1261,
      "mean_token_accuracy": 0.13380536437034607,
      "num_tokens": 2291837.0,
      "step": 1320
    },
    {
      "entropy": 6.546138572692871,
      "epoch": 0.10309278350515463,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999928322413374,
      "loss": 6.206,
      "mean_token_accuracy": 0.12851327955722808,
      "num_tokens": 2299853.0,
      "step": 1325
    },
    {
      "entropy": 6.574787330627442,
      "epoch": 0.1034818128768722,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999926093084394,
      "loss": 6.3075,
      "mean_token_accuracy": 0.1206931158900261,
      "num_tokens": 2308416.0,
      "step": 1330
    },
    {
      "entropy": 6.507684230804443,
      "epoch": 0.10387084224858976,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999923829616365,
      "loss": 6.3388,
      "mean_token_accuracy": 0.12228097319602967,
      "num_tokens": 2318104.0,
      "step": 1335
    },
    {
      "entropy": 6.501170110702515,
      "epoch": 0.10425987162030734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999921532009323,
      "loss": 6.2612,
      "mean_token_accuracy": 0.12383703589439392,
      "num_tokens": 2326807.0,
      "step": 1340
    },
    {
      "entropy": 6.528704977035522,
      "epoch": 0.1046489009920249,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999919200263304,
      "loss": 6.259,
      "mean_token_accuracy": 0.12394311875104905,
      "num_tokens": 2335921.0,
      "step": 1345
    },
    {
      "entropy": 6.505633020401001,
      "epoch": 0.10503793036374247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999916834378342,
      "loss": 6.2739,
      "mean_token_accuracy": 0.12304315939545632,
      "num_tokens": 2344914.0,
      "step": 1350
    },
    {
      "entropy": 6.447750616073608,
      "epoch": 0.10542695973546003,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999914434354472,
      "loss": 6.2044,
      "mean_token_accuracy": 0.125525876134634,
      "num_tokens": 2353211.0,
      "step": 1355
    },
    {
      "entropy": 6.4893646240234375,
      "epoch": 0.10581598910717759,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999912000191733,
      "loss": 6.3487,
      "mean_token_accuracy": 0.12199961394071579,
      "num_tokens": 2361717.0,
      "step": 1360
    },
    {
      "entropy": 6.525090074539184,
      "epoch": 0.10620501847889516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990953189016,
      "loss": 6.2712,
      "mean_token_accuracy": 0.13059643879532815,
      "num_tokens": 2370114.0,
      "step": 1365
    },
    {
      "entropy": 6.630150699615479,
      "epoch": 0.10659404785061272,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999907029449791,
      "loss": 6.263,
      "mean_token_accuracy": 0.12237593159079552,
      "num_tokens": 2378820.0,
      "step": 1370
    },
    {
      "entropy": 6.354615974426269,
      "epoch": 0.10698307722233029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999904492870663,
      "loss": 6.1715,
      "mean_token_accuracy": 0.12704389467835425,
      "num_tokens": 2387119.0,
      "step": 1375
    },
    {
      "entropy": 6.53058009147644,
      "epoch": 0.10737210659404785,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999901922152816,
      "loss": 6.2259,
      "mean_token_accuracy": 0.12630244940519333,
      "num_tokens": 2396246.0,
      "step": 1380
    },
    {
      "entropy": 6.406842041015625,
      "epoch": 0.10776113596576542,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999899317296288,
      "loss": 6.2855,
      "mean_token_accuracy": 0.12351492792367935,
      "num_tokens": 2405428.0,
      "step": 1385
    },
    {
      "entropy": 6.490512895584106,
      "epoch": 0.10815016533748298,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999896678301121,
      "loss": 6.2421,
      "mean_token_accuracy": 0.1267989382147789,
      "num_tokens": 2414669.0,
      "step": 1390
    },
    {
      "entropy": 6.563437604904175,
      "epoch": 0.10853919470920055,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999894005167352,
      "loss": 6.3,
      "mean_token_accuracy": 0.12090388685464859,
      "num_tokens": 2423193.0,
      "step": 1395
    },
    {
      "entropy": 6.500105094909668,
      "epoch": 0.10892822408091811,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499989129789502,
      "loss": 6.2664,
      "mean_token_accuracy": 0.12048013284802436,
      "num_tokens": 2431150.0,
      "step": 1400
    },
    {
      "entropy": 6.445392656326294,
      "epoch": 0.10931725345263567,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999888556484172,
      "loss": 6.2446,
      "mean_token_accuracy": 0.12127585709095001,
      "num_tokens": 2440337.0,
      "step": 1405
    },
    {
      "entropy": 6.441078138351441,
      "epoch": 0.10970628282435324,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999885780934845,
      "loss": 6.1736,
      "mean_token_accuracy": 0.1296399176120758,
      "num_tokens": 2449246.0,
      "step": 1410
    },
    {
      "entropy": 6.434943819046021,
      "epoch": 0.1100953121960708,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999882971247081,
      "loss": 6.1752,
      "mean_token_accuracy": 0.13090624362230302,
      "num_tokens": 2457982.0,
      "step": 1415
    },
    {
      "entropy": 6.44070029258728,
      "epoch": 0.11048434156778837,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999880127420926,
      "loss": 6.1661,
      "mean_token_accuracy": 0.12823255807161332,
      "num_tokens": 2466697.0,
      "step": 1420
    },
    {
      "entropy": 6.419079399108886,
      "epoch": 0.11087337093950593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999877249456421,
      "loss": 6.1397,
      "mean_token_accuracy": 0.13144427835941314,
      "num_tokens": 2475457.0,
      "step": 1425
    },
    {
      "entropy": 6.336713647842407,
      "epoch": 0.1112624003112235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999874337353609,
      "loss": 6.1393,
      "mean_token_accuracy": 0.12938348650932313,
      "num_tokens": 2484753.0,
      "step": 1430
    },
    {
      "entropy": 6.465570974349975,
      "epoch": 0.11165142968294106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999871391112535,
      "loss": 6.2556,
      "mean_token_accuracy": 0.124443519115448,
      "num_tokens": 2492921.0,
      "step": 1435
    },
    {
      "entropy": 6.46670618057251,
      "epoch": 0.11204045905465862,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999868410733244,
      "loss": 6.224,
      "mean_token_accuracy": 0.12451541647315026,
      "num_tokens": 2501371.0,
      "step": 1440
    },
    {
      "entropy": 6.471292209625244,
      "epoch": 0.11242948842637619,
      "grad_norm": 1.234375,
      "learning_rate": 0.000499986539621578,
      "loss": 6.1023,
      "mean_token_accuracy": 0.12834953069686889,
      "num_tokens": 2509521.0,
      "step": 1445
    },
    {
      "entropy": 6.416775035858154,
      "epoch": 0.11281851779809375,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999862347560191,
      "loss": 6.2143,
      "mean_token_accuracy": 0.12635174617171288,
      "num_tokens": 2518316.0,
      "step": 1450
    },
    {
      "entropy": 6.514144420623779,
      "epoch": 0.11320754716981132,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999859264766521,
      "loss": 6.309,
      "mean_token_accuracy": 0.1220340982079506,
      "num_tokens": 2527549.0,
      "step": 1455
    },
    {
      "entropy": 6.501974773406983,
      "epoch": 0.11359657654152888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999856147834817,
      "loss": 6.33,
      "mean_token_accuracy": 0.12365049570798874,
      "num_tokens": 2536341.0,
      "step": 1460
    },
    {
      "entropy": 6.390389823913575,
      "epoch": 0.11398560591324645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999852996765129,
      "loss": 6.2144,
      "mean_token_accuracy": 0.13640521019697188,
      "num_tokens": 2544869.0,
      "step": 1465
    },
    {
      "entropy": 6.442426967620849,
      "epoch": 0.11437463528496401,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999849811557502,
      "loss": 6.1124,
      "mean_token_accuracy": 0.1380823515355587,
      "num_tokens": 2553432.0,
      "step": 1470
    },
    {
      "entropy": 6.435215473175049,
      "epoch": 0.11476366465668159,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999846592211984,
      "loss": 6.245,
      "mean_token_accuracy": 0.12433946654200553,
      "num_tokens": 2561996.0,
      "step": 1475
    },
    {
      "entropy": 6.435883617401123,
      "epoch": 0.11515269402839914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999843338728626,
      "loss": 6.1732,
      "mean_token_accuracy": 0.13035295605659486,
      "num_tokens": 2571315.0,
      "step": 1480
    },
    {
      "entropy": 6.433837747573852,
      "epoch": 0.1155417234001167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999840051107476,
      "loss": 6.2935,
      "mean_token_accuracy": 0.12482402846217155,
      "num_tokens": 2579589.0,
      "step": 1485
    },
    {
      "entropy": 6.433605575561524,
      "epoch": 0.11593075277183428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999836729348584,
      "loss": 6.0607,
      "mean_token_accuracy": 0.1283522941172123,
      "num_tokens": 2588528.0,
      "step": 1490
    },
    {
      "entropy": 6.3824395656585695,
      "epoch": 0.11631978214355183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999833373452,
      "loss": 6.1744,
      "mean_token_accuracy": 0.12705983147025107,
      "num_tokens": 2596820.0,
      "step": 1495
    },
    {
      "entropy": 6.30913896560669,
      "epoch": 0.1167088115152694,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999829983417778,
      "loss": 6.1168,
      "mean_token_accuracy": 0.12735813558101655,
      "num_tokens": 2605502.0,
      "step": 1500
    },
    {
      "entropy": 6.378449249267578,
      "epoch": 0.11709784088698696,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999826559245965,
      "loss": 6.1193,
      "mean_token_accuracy": 0.13392857909202577,
      "num_tokens": 2613974.0,
      "step": 1505
    },
    {
      "entropy": 6.370792722702026,
      "epoch": 0.11748687025870454,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999823100936615,
      "loss": 6.1965,
      "mean_token_accuracy": 0.12880947291851044,
      "num_tokens": 2622581.0,
      "step": 1510
    },
    {
      "entropy": 6.402956151962281,
      "epoch": 0.1178758996304221,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999819608489781,
      "loss": 6.0841,
      "mean_token_accuracy": 0.1251236990094185,
      "num_tokens": 2631868.0,
      "step": 1515
    },
    {
      "entropy": 6.339415788650513,
      "epoch": 0.11826492900213967,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999816081905515,
      "loss": 6.1093,
      "mean_token_accuracy": 0.1355234570801258,
      "num_tokens": 2640501.0,
      "step": 1520
    },
    {
      "entropy": 6.433193206787109,
      "epoch": 0.11865395837385723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999812521183872,
      "loss": 6.1035,
      "mean_token_accuracy": 0.13361863121390344,
      "num_tokens": 2648916.0,
      "step": 1525
    },
    {
      "entropy": 6.406140279769898,
      "epoch": 0.11904298774557479,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999808926324903,
      "loss": 6.1896,
      "mean_token_accuracy": 0.1334061346948147,
      "num_tokens": 2657725.0,
      "step": 1530
    },
    {
      "entropy": 6.353977298736572,
      "epoch": 0.11943201711729236,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999805297328665,
      "loss": 6.1751,
      "mean_token_accuracy": 0.1281532824039459,
      "num_tokens": 2665983.0,
      "step": 1535
    },
    {
      "entropy": 6.392253303527832,
      "epoch": 0.11982104648900992,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999801634195214,
      "loss": 6.1957,
      "mean_token_accuracy": 0.13333539366722108,
      "num_tokens": 2674546.0,
      "step": 1540
    },
    {
      "entropy": 6.389540004730224,
      "epoch": 0.12021007586072749,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999797936924603,
      "loss": 6.1091,
      "mean_token_accuracy": 0.13411504104733468,
      "num_tokens": 2684088.0,
      "step": 1545
    },
    {
      "entropy": 6.37681622505188,
      "epoch": 0.12059910523244505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499979420551689,
      "loss": 6.1664,
      "mean_token_accuracy": 0.12699728310108185,
      "num_tokens": 2693289.0,
      "step": 1550
    },
    {
      "entropy": 6.318667316436768,
      "epoch": 0.12098813460416262,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499979043997213,
      "loss": 6.095,
      "mean_token_accuracy": 0.13104904145002366,
      "num_tokens": 2701744.0,
      "step": 1555
    },
    {
      "entropy": 6.411156368255615,
      "epoch": 0.12137716397588018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999786640290381,
      "loss": 6.1499,
      "mean_token_accuracy": 0.12439182624220849,
      "num_tokens": 2710584.0,
      "step": 1560
    },
    {
      "entropy": 6.411997509002686,
      "epoch": 0.12176619334759774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999782806471701,
      "loss": 6.1491,
      "mean_token_accuracy": 0.13031079471111298,
      "num_tokens": 2719478.0,
      "step": 1565
    },
    {
      "entropy": 6.465567684173584,
      "epoch": 0.12215522271931531,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999778938516147,
      "loss": 6.2286,
      "mean_token_accuracy": 0.12305473163723946,
      "num_tokens": 2729105.0,
      "step": 1570
    },
    {
      "entropy": 6.401231145858764,
      "epoch": 0.12254425209103287,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499977503642378,
      "loss": 6.1908,
      "mean_token_accuracy": 0.1272389329969883,
      "num_tokens": 2738300.0,
      "step": 1575
    },
    {
      "entropy": 6.384681034088135,
      "epoch": 0.12293328146275044,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999771100194656,
      "loss": 6.1335,
      "mean_token_accuracy": 0.12494338452816009,
      "num_tokens": 2747236.0,
      "step": 1580
    },
    {
      "entropy": 6.289424848556519,
      "epoch": 0.123322310834468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999767129828837,
      "loss": 6.0956,
      "mean_token_accuracy": 0.13419784530997275,
      "num_tokens": 2755919.0,
      "step": 1585
    },
    {
      "entropy": 6.305175161361694,
      "epoch": 0.12371134020618557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999763125326383,
      "loss": 6.0768,
      "mean_token_accuracy": 0.1307978369295597,
      "num_tokens": 2764469.0,
      "step": 1590
    },
    {
      "entropy": 6.355486297607422,
      "epoch": 0.12410036957790313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999759086687355,
      "loss": 6.081,
      "mean_token_accuracy": 0.1366819642484188,
      "num_tokens": 2772663.0,
      "step": 1595
    },
    {
      "entropy": 6.419379472732544,
      "epoch": 0.1244893989496207,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999755013911813,
      "loss": 6.1161,
      "mean_token_accuracy": 0.12876046895980836,
      "num_tokens": 2780719.0,
      "step": 1600
    },
    {
      "entropy": 6.330186557769776,
      "epoch": 0.12487842832133826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499975090699982,
      "loss": 6.1714,
      "mean_token_accuracy": 0.128857558965683,
      "num_tokens": 2789186.0,
      "step": 1605
    },
    {
      "entropy": 6.337528562545776,
      "epoch": 0.12526745769305583,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999746765951438,
      "loss": 6.0856,
      "mean_token_accuracy": 0.13346567153930664,
      "num_tokens": 2797993.0,
      "step": 1610
    },
    {
      "entropy": 6.312220811843872,
      "epoch": 0.1256564870647734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999742590766729,
      "loss": 6.0464,
      "mean_token_accuracy": 0.13435178622603416,
      "num_tokens": 2806375.0,
      "step": 1615
    },
    {
      "entropy": 6.440674829483032,
      "epoch": 0.12604551643649095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999738381445757,
      "loss": 6.1942,
      "mean_token_accuracy": 0.13562749326229095,
      "num_tokens": 2814772.0,
      "step": 1620
    },
    {
      "entropy": 6.202065896987915,
      "epoch": 0.1264345458082085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999734137988586,
      "loss": 6.0582,
      "mean_token_accuracy": 0.13596921488642694,
      "num_tokens": 2824231.0,
      "step": 1625
    },
    {
      "entropy": 6.376500511169434,
      "epoch": 0.1268235751799261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499972986039528,
      "loss": 6.1245,
      "mean_token_accuracy": 0.12782634794712067,
      "num_tokens": 2833470.0,
      "step": 1630
    },
    {
      "entropy": 6.353403902053833,
      "epoch": 0.12721260455164365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999725548665904,
      "loss": 6.1327,
      "mean_token_accuracy": 0.1281338281929493,
      "num_tokens": 2843010.0,
      "step": 1635
    },
    {
      "entropy": 6.346253395080566,
      "epoch": 0.1276016339233612,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999721202800524,
      "loss": 6.0231,
      "mean_token_accuracy": 0.13678104057908058,
      "num_tokens": 2851545.0,
      "step": 1640
    },
    {
      "entropy": 6.269683694839477,
      "epoch": 0.12799066329507877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999716822799206,
      "loss": 6.1796,
      "mean_token_accuracy": 0.11961811035871506,
      "num_tokens": 2861230.0,
      "step": 1645
    },
    {
      "entropy": 6.395336675643921,
      "epoch": 0.12837969266679633,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999712408662016,
      "loss": 6.1356,
      "mean_token_accuracy": 0.1289409004151821,
      "num_tokens": 2869436.0,
      "step": 1650
    },
    {
      "entropy": 6.306952619552613,
      "epoch": 0.12876872203851392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999707960389021,
      "loss": 5.9996,
      "mean_token_accuracy": 0.13151076734066008,
      "num_tokens": 2877688.0,
      "step": 1655
    },
    {
      "entropy": 6.365733051300049,
      "epoch": 0.12915775141023147,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999703477980288,
      "loss": 6.1389,
      "mean_token_accuracy": 0.13166750520467757,
      "num_tokens": 2887043.0,
      "step": 1660
    },
    {
      "entropy": 6.262747430801392,
      "epoch": 0.12954678078194903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999698961435885,
      "loss": 6.0721,
      "mean_token_accuracy": 0.13516117706894876,
      "num_tokens": 2895732.0,
      "step": 1665
    },
    {
      "entropy": 6.256302070617676,
      "epoch": 0.1299358101536666,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999694410755882,
      "loss": 6.039,
      "mean_token_accuracy": 0.13851652070879936,
      "num_tokens": 2904162.0,
      "step": 1670
    },
    {
      "entropy": 6.356344509124756,
      "epoch": 0.13032483952538418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999689825940347,
      "loss": 6.1758,
      "mean_token_accuracy": 0.13067347109317778,
      "num_tokens": 2912783.0,
      "step": 1675
    },
    {
      "entropy": 6.319488096237182,
      "epoch": 0.13071386889710174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999685206989349,
      "loss": 6.0558,
      "mean_token_accuracy": 0.13287809044122695,
      "num_tokens": 2921027.0,
      "step": 1680
    },
    {
      "entropy": 6.344622087478638,
      "epoch": 0.1311028982688193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999680553902959,
      "loss": 6.093,
      "mean_token_accuracy": 0.13244052454829217,
      "num_tokens": 2930398.0,
      "step": 1685
    },
    {
      "entropy": 6.316963481903076,
      "epoch": 0.13149192764053685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999675866681247,
      "loss": 6.1126,
      "mean_token_accuracy": 0.1291515350341797,
      "num_tokens": 2938703.0,
      "step": 1690
    },
    {
      "entropy": 6.272860383987426,
      "epoch": 0.1318809570122544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999671145324286,
      "loss": 6.0394,
      "mean_token_accuracy": 0.13433808013796805,
      "num_tokens": 2947007.0,
      "step": 1695
    },
    {
      "entropy": 6.319776296615601,
      "epoch": 0.132269986383972,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999666389832144,
      "loss": 6.1923,
      "mean_token_accuracy": 0.13449379578232765,
      "num_tokens": 2955485.0,
      "step": 1700
    },
    {
      "entropy": 6.353970575332641,
      "epoch": 0.13265901575568956,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999661600204897,
      "loss": 6.0481,
      "mean_token_accuracy": 0.1327715963125229,
      "num_tokens": 2964416.0,
      "step": 1705
    },
    {
      "entropy": 6.239326429367066,
      "epoch": 0.13304804512740712,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999656776442615,
      "loss": 6.0398,
      "mean_token_accuracy": 0.1316245622932911,
      "num_tokens": 2972737.0,
      "step": 1710
    },
    {
      "entropy": 6.29252290725708,
      "epoch": 0.13343707449912467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999651918545372,
      "loss": 6.084,
      "mean_token_accuracy": 0.13347245752811432,
      "num_tokens": 2981166.0,
      "step": 1715
    },
    {
      "entropy": 6.313875722885132,
      "epoch": 0.13382610387084226,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999647026513243,
      "loss": 6.124,
      "mean_token_accuracy": 0.13230146765708922,
      "num_tokens": 2989689.0,
      "step": 1720
    },
    {
      "entropy": 6.354332733154297,
      "epoch": 0.13421513324255982,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00049996421003463,
      "loss": 6.0353,
      "mean_token_accuracy": 0.1331894539296627,
      "num_tokens": 2998793.0,
      "step": 1725
    },
    {
      "entropy": 6.235461759567261,
      "epoch": 0.13460416261427738,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999637140044619,
      "loss": 6.1931,
      "mean_token_accuracy": 0.12982380986213685,
      "num_tokens": 3007320.0,
      "step": 1730
    },
    {
      "entropy": 6.328571414947509,
      "epoch": 0.13499319198599494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999632145608275,
      "loss": 6.0911,
      "mean_token_accuracy": 0.13207585886120796,
      "num_tokens": 3015024.0,
      "step": 1735
    },
    {
      "entropy": 6.351219606399536,
      "epoch": 0.1353822213577125,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999627117037343,
      "loss": 6.1831,
      "mean_token_accuracy": 0.1294490687549114,
      "num_tokens": 3023058.0,
      "step": 1740
    },
    {
      "entropy": 6.363010501861572,
      "epoch": 0.13577125072943008,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049996220543319,
      "loss": 6.0924,
      "mean_token_accuracy": 0.13458985537290574,
      "num_tokens": 3031308.0,
      "step": 1745
    },
    {
      "entropy": 6.337850189208984,
      "epoch": 0.13616028010114764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999616957492025,
      "loss": 6.1513,
      "mean_token_accuracy": 0.1289216734468937,
      "num_tokens": 3039598.0,
      "step": 1750
    },
    {
      "entropy": 6.253518009185791,
      "epoch": 0.1365493094728652,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999611826517793,
      "loss": 5.9714,
      "mean_token_accuracy": 0.1398603707551956,
      "num_tokens": 3047598.0,
      "step": 1755
    },
    {
      "entropy": 6.273720216751099,
      "epoch": 0.13693833884458276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999606661409281,
      "loss": 6.1199,
      "mean_token_accuracy": 0.1361253760755062,
      "num_tokens": 3055702.0,
      "step": 1760
    },
    {
      "entropy": 6.330643892288208,
      "epoch": 0.13732736821630034,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999601462166569,
      "loss": 5.9841,
      "mean_token_accuracy": 0.14291464164853096,
      "num_tokens": 3063908.0,
      "step": 1765
    },
    {
      "entropy": 6.251221227645874,
      "epoch": 0.1377163975880179,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999596228789736,
      "loss": 6.0754,
      "mean_token_accuracy": 0.1333239495754242,
      "num_tokens": 3072077.0,
      "step": 1770
    },
    {
      "entropy": 6.303908920288086,
      "epoch": 0.13810542695973546,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999590961278859,
      "loss": 6.0791,
      "mean_token_accuracy": 0.13312057927250862,
      "num_tokens": 3080498.0,
      "step": 1775
    },
    {
      "entropy": 6.31710524559021,
      "epoch": 0.13849445633145302,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999585659634021,
      "loss": 6.0814,
      "mean_token_accuracy": 0.1339930735528469,
      "num_tokens": 3089055.0,
      "step": 1780
    },
    {
      "entropy": 6.248744344711303,
      "epoch": 0.13888348570317058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999580323855301,
      "loss": 6.0718,
      "mean_token_accuracy": 0.1378532275557518,
      "num_tokens": 3096944.0,
      "step": 1785
    },
    {
      "entropy": 6.228180599212647,
      "epoch": 0.13927251507488816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999574953942781,
      "loss": 5.965,
      "mean_token_accuracy": 0.13897001892328262,
      "num_tokens": 3105596.0,
      "step": 1790
    },
    {
      "entropy": 6.191677761077881,
      "epoch": 0.13966154444660572,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999569549896541,
      "loss": 5.958,
      "mean_token_accuracy": 0.14424095302820206,
      "num_tokens": 3114349.0,
      "step": 1795
    },
    {
      "entropy": 6.258058547973633,
      "epoch": 0.14005057381832328,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999564111716665,
      "loss": 6.0799,
      "mean_token_accuracy": 0.12551943436264992,
      "num_tokens": 3122490.0,
      "step": 1800
    },
    {
      "entropy": 6.355373287200928,
      "epoch": 0.14043960319004084,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999558639403232,
      "loss": 6.1203,
      "mean_token_accuracy": 0.13192636370658875,
      "num_tokens": 3130988.0,
      "step": 1805
    },
    {
      "entropy": 6.246408748626709,
      "epoch": 0.1408286325617584,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999553132956328,
      "loss": 5.9784,
      "mean_token_accuracy": 0.13481251522898674,
      "num_tokens": 3139488.0,
      "step": 1810
    },
    {
      "entropy": 6.221766948699951,
      "epoch": 0.14121766193347599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999547592376035,
      "loss": 6.0288,
      "mean_token_accuracy": 0.13629044219851494,
      "num_tokens": 3147955.0,
      "step": 1815
    },
    {
      "entropy": 6.249310064315796,
      "epoch": 0.14160669130519354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999542017662438,
      "loss": 6.0759,
      "mean_token_accuracy": 0.1276159830391407,
      "num_tokens": 3157608.0,
      "step": 1820
    },
    {
      "entropy": 6.328677940368652,
      "epoch": 0.1419957206769111,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999536408815623,
      "loss": 6.0303,
      "mean_token_accuracy": 0.130377709120512,
      "num_tokens": 3166124.0,
      "step": 1825
    },
    {
      "entropy": 6.159740209579468,
      "epoch": 0.14238475004862866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999530765835672,
      "loss": 6.0358,
      "mean_token_accuracy": 0.13254115357995033,
      "num_tokens": 3175069.0,
      "step": 1830
    },
    {
      "entropy": 6.292894077301026,
      "epoch": 0.14277377942034625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999525088722673,
      "loss": 6.0664,
      "mean_token_accuracy": 0.13529409915208818,
      "num_tokens": 3184259.0,
      "step": 1835
    },
    {
      "entropy": 6.275870656967163,
      "epoch": 0.1431628087920638,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999519377476711,
      "loss": 5.9672,
      "mean_token_accuracy": 0.1381848506629467,
      "num_tokens": 3193015.0,
      "step": 1840
    },
    {
      "entropy": 6.267138957977295,
      "epoch": 0.14355183816378136,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999513632097872,
      "loss": 5.9147,
      "mean_token_accuracy": 0.14502743408083915,
      "num_tokens": 3201624.0,
      "step": 1845
    },
    {
      "entropy": 6.208207750320435,
      "epoch": 0.14394086753549892,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999507852586245,
      "loss": 6.0294,
      "mean_token_accuracy": 0.13966942057013512,
      "num_tokens": 3209474.0,
      "step": 1850
    },
    {
      "entropy": 6.183934068679809,
      "epoch": 0.14432989690721648,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999502038941917,
      "loss": 6.0023,
      "mean_token_accuracy": 0.13654958382248877,
      "num_tokens": 3218550.0,
      "step": 1855
    },
    {
      "entropy": 6.118973255157471,
      "epoch": 0.14471892627893407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999496191164975,
      "loss": 5.9653,
      "mean_token_accuracy": 0.13626407235860824,
      "num_tokens": 3227108.0,
      "step": 1860
    },
    {
      "entropy": 6.250623989105224,
      "epoch": 0.14510795565065163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499949030925551,
      "loss": 6.0032,
      "mean_token_accuracy": 0.13892803117632865,
      "num_tokens": 3235785.0,
      "step": 1865
    },
    {
      "entropy": 6.293147802352905,
      "epoch": 0.14549698502236919,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999484393213609,
      "loss": 6.0377,
      "mean_token_accuracy": 0.1293921411037445,
      "num_tokens": 3244180.0,
      "step": 1870
    },
    {
      "entropy": 6.158266830444336,
      "epoch": 0.14588601439408674,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999478443039363,
      "loss": 5.9157,
      "mean_token_accuracy": 0.14264118522405625,
      "num_tokens": 3253098.0,
      "step": 1875
    },
    {
      "entropy": 6.26281852722168,
      "epoch": 0.14627504376580433,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999472458732862,
      "loss": 5.962,
      "mean_token_accuracy": 0.14164620488882065,
      "num_tokens": 3261726.0,
      "step": 1880
    },
    {
      "entropy": 6.122699928283692,
      "epoch": 0.1466640731375219,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999466440294198,
      "loss": 5.9218,
      "mean_token_accuracy": 0.13797034472227096,
      "num_tokens": 3270362.0,
      "step": 1885
    },
    {
      "entropy": 6.218784666061401,
      "epoch": 0.14705310250923945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499946038772346,
      "loss": 5.9119,
      "mean_token_accuracy": 0.1409492813050747,
      "num_tokens": 3278332.0,
      "step": 1890
    },
    {
      "entropy": 6.106905126571656,
      "epoch": 0.147442131880957,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999454301020741,
      "loss": 5.8984,
      "mean_token_accuracy": 0.14537974447011948,
      "num_tokens": 3286800.0,
      "step": 1895
    },
    {
      "entropy": 6.261422252655029,
      "epoch": 0.14783116125267456,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999448180186133,
      "loss": 6.1243,
      "mean_token_accuracy": 0.12735687866806983,
      "num_tokens": 3296186.0,
      "step": 1900
    },
    {
      "entropy": 6.170867204666138,
      "epoch": 0.14822019062439215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499944202521973,
      "loss": 5.9874,
      "mean_token_accuracy": 0.14373966082930564,
      "num_tokens": 3303973.0,
      "step": 1905
    },
    {
      "entropy": 6.154075765609742,
      "epoch": 0.1486092199961097,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999435836121624,
      "loss": 5.9417,
      "mean_token_accuracy": 0.13721608817577363,
      "num_tokens": 3312550.0,
      "step": 1910
    },
    {
      "entropy": 6.231463432312012,
      "epoch": 0.14899824936782727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499942961289191,
      "loss": 6.0136,
      "mean_token_accuracy": 0.13601985946297646,
      "num_tokens": 3321342.0,
      "step": 1915
    },
    {
      "entropy": 6.136085319519043,
      "epoch": 0.14938727873954483,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999423355530683,
      "loss": 5.8833,
      "mean_token_accuracy": 0.14982874765992166,
      "num_tokens": 3329293.0,
      "step": 1920
    },
    {
      "entropy": 6.077541828155518,
      "epoch": 0.1497763081112624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999417064038035,
      "loss": 5.947,
      "mean_token_accuracy": 0.135797318816185,
      "num_tokens": 3338278.0,
      "step": 1925
    },
    {
      "entropy": 6.248018217086792,
      "epoch": 0.15016533748297997,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999410738414065,
      "loss": 6.0446,
      "mean_token_accuracy": 0.1355997994542122,
      "num_tokens": 3346757.0,
      "step": 1930
    },
    {
      "entropy": 6.1532470226287845,
      "epoch": 0.15055436685469753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999404378658866,
      "loss": 6.0451,
      "mean_token_accuracy": 0.1347198210656643,
      "num_tokens": 3355903.0,
      "step": 1935
    },
    {
      "entropy": 6.208436155319214,
      "epoch": 0.1509433962264151,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999397984772536,
      "loss": 6.0141,
      "mean_token_accuracy": 0.12824626341462136,
      "num_tokens": 3365440.0,
      "step": 1940
    },
    {
      "entropy": 6.216444635391236,
      "epoch": 0.15133242559813265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999391556755172,
      "loss": 5.9763,
      "mean_token_accuracy": 0.1436944492161274,
      "num_tokens": 3373672.0,
      "step": 1945
    },
    {
      "entropy": 6.172102642059326,
      "epoch": 0.15172145496985023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999385094606872,
      "loss": 5.9552,
      "mean_token_accuracy": 0.13290092200040818,
      "num_tokens": 3382733.0,
      "step": 1950
    },
    {
      "entropy": 6.100259685516358,
      "epoch": 0.1521104843415678,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999378598327733,
      "loss": 5.9076,
      "mean_token_accuracy": 0.1429830349981785,
      "num_tokens": 3390842.0,
      "step": 1955
    },
    {
      "entropy": 6.200795269012451,
      "epoch": 0.15249951371328535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999372067917854,
      "loss": 6.061,
      "mean_token_accuracy": 0.13028706163167952,
      "num_tokens": 3399761.0,
      "step": 1960
    },
    {
      "entropy": 6.219650220870972,
      "epoch": 0.1528885430850029,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999365503377335,
      "loss": 5.9837,
      "mean_token_accuracy": 0.13638316243886947,
      "num_tokens": 3408986.0,
      "step": 1965
    },
    {
      "entropy": 6.199184322357178,
      "epoch": 0.1532775724567205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999358904706273,
      "loss": 5.9356,
      "mean_token_accuracy": 0.13719403147697448,
      "num_tokens": 3418051.0,
      "step": 1970
    },
    {
      "entropy": 6.225965929031372,
      "epoch": 0.15366660182843805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999352271904771,
      "loss": 6.0436,
      "mean_token_accuracy": 0.13756765872240068,
      "num_tokens": 3427185.0,
      "step": 1975
    },
    {
      "entropy": 6.179631471633911,
      "epoch": 0.1540556312001556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999345604972928,
      "loss": 5.9495,
      "mean_token_accuracy": 0.13750605583190917,
      "num_tokens": 3435367.0,
      "step": 1980
    },
    {
      "entropy": 6.154869127273559,
      "epoch": 0.15444466057187317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999338903910846,
      "loss": 5.9934,
      "mean_token_accuracy": 0.1389468789100647,
      "num_tokens": 3443451.0,
      "step": 1985
    },
    {
      "entropy": 6.200529718399048,
      "epoch": 0.15483368994359073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999332168718626,
      "loss": 5.9893,
      "mean_token_accuracy": 0.1442691780626774,
      "num_tokens": 3451038.0,
      "step": 1990
    },
    {
      "entropy": 6.408123874664307,
      "epoch": 0.15522271931530832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999325399396371,
      "loss": 6.1241,
      "mean_token_accuracy": 0.1256192870438099,
      "num_tokens": 3459693.0,
      "step": 1995
    },
    {
      "entropy": 6.269746255874634,
      "epoch": 0.15561174868702587,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999318595944183,
      "loss": 6.0106,
      "mean_token_accuracy": 0.13469778150320053,
      "num_tokens": 3468347.0,
      "step": 2000
    },
    {
      "entropy": 6.145987224578858,
      "epoch": 0.15600077805874343,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999311758362166,
      "loss": 5.986,
      "mean_token_accuracy": 0.13514223098754882,
      "num_tokens": 3476030.0,
      "step": 2005
    },
    {
      "entropy": 6.12372317314148,
      "epoch": 0.156389807430461,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999304886650422,
      "loss": 5.9609,
      "mean_token_accuracy": 0.13282474353909493,
      "num_tokens": 3484840.0,
      "step": 2010
    },
    {
      "entropy": 6.290483140945435,
      "epoch": 0.15677883680217858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999297980809058,
      "loss": 6.0489,
      "mean_token_accuracy": 0.13376422077417374,
      "num_tokens": 3492662.0,
      "step": 2015
    },
    {
      "entropy": 6.2352056980133055,
      "epoch": 0.15716786617389614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999291040838176,
      "loss": 5.9688,
      "mean_token_accuracy": 0.13237249329686165,
      "num_tokens": 3501582.0,
      "step": 2020
    },
    {
      "entropy": 6.179763507843018,
      "epoch": 0.1575568955456137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999284066737884,
      "loss": 5.9639,
      "mean_token_accuracy": 0.14239997863769532,
      "num_tokens": 3510141.0,
      "step": 2025
    },
    {
      "entropy": 6.209929418563843,
      "epoch": 0.15794592491733125,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999277058508285,
      "loss": 6.0049,
      "mean_token_accuracy": 0.14213063940405846,
      "num_tokens": 3518530.0,
      "step": 2030
    },
    {
      "entropy": 6.1083790302276615,
      "epoch": 0.1583349542890488,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999270016149486,
      "loss": 5.9775,
      "mean_token_accuracy": 0.1358012966811657,
      "num_tokens": 3526815.0,
      "step": 2035
    },
    {
      "entropy": 6.179563379287719,
      "epoch": 0.1587239836607664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999262939661596,
      "loss": 6.0282,
      "mean_token_accuracy": 0.13272086828947066,
      "num_tokens": 3535840.0,
      "step": 2040
    },
    {
      "entropy": 6.240152835845947,
      "epoch": 0.15911301303248396,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999255829044721,
      "loss": 6.0579,
      "mean_token_accuracy": 0.13129353299736976,
      "num_tokens": 3544876.0,
      "step": 2045
    },
    {
      "entropy": 6.140031433105468,
      "epoch": 0.15950204240420152,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999248684298968,
      "loss": 5.7949,
      "mean_token_accuracy": 0.14725889563560485,
      "num_tokens": 3553018.0,
      "step": 2050
    },
    {
      "entropy": 6.0304547309875485,
      "epoch": 0.15989107177591907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999241505424447,
      "loss": 5.8511,
      "mean_token_accuracy": 0.1435144364833832,
      "num_tokens": 3560909.0,
      "step": 2055
    },
    {
      "entropy": 6.116934061050415,
      "epoch": 0.16028010114763663,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999234292421265,
      "loss": 5.9999,
      "mean_token_accuracy": 0.1397472143173218,
      "num_tokens": 3569235.0,
      "step": 2060
    },
    {
      "entropy": 6.105283117294311,
      "epoch": 0.16066913051935422,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999227045289535,
      "loss": 5.8792,
      "mean_token_accuracy": 0.14063235372304916,
      "num_tokens": 3577646.0,
      "step": 2065
    },
    {
      "entropy": 6.068844556808472,
      "epoch": 0.16105815989107178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999219764029363,
      "loss": 5.9745,
      "mean_token_accuracy": 0.1359837457537651,
      "num_tokens": 3586499.0,
      "step": 2070
    },
    {
      "entropy": 6.221855163574219,
      "epoch": 0.16144718926278934,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999212448640861,
      "loss": 5.9812,
      "mean_token_accuracy": 0.14166299402713775,
      "num_tokens": 3595391.0,
      "step": 2075
    },
    {
      "entropy": 6.117988729476929,
      "epoch": 0.1618362186345069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999205099124141,
      "loss": 5.903,
      "mean_token_accuracy": 0.14056897163391113,
      "num_tokens": 3603774.0,
      "step": 2080
    },
    {
      "entropy": 6.214944982528687,
      "epoch": 0.16222524800622448,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999197715479313,
      "loss": 5.9841,
      "mean_token_accuracy": 0.13269646242260932,
      "num_tokens": 3611980.0,
      "step": 2085
    },
    {
      "entropy": 6.1437524318695065,
      "epoch": 0.16261427737794204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999190297706489,
      "loss": 5.902,
      "mean_token_accuracy": 0.1393118217587471,
      "num_tokens": 3620902.0,
      "step": 2090
    },
    {
      "entropy": 6.299076747894287,
      "epoch": 0.1630033067496596,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999182845805783,
      "loss": 6.1871,
      "mean_token_accuracy": 0.12861791327595712,
      "num_tokens": 3628924.0,
      "step": 2095
    },
    {
      "entropy": 6.2256190299987795,
      "epoch": 0.16339233612137716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999175359777306,
      "loss": 6.0231,
      "mean_token_accuracy": 0.13702933117747307,
      "num_tokens": 3637538.0,
      "step": 2100
    },
    {
      "entropy": 6.148886060714721,
      "epoch": 0.16378136549309472,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999167839621174,
      "loss": 6.0424,
      "mean_token_accuracy": 0.130807376652956,
      "num_tokens": 3646309.0,
      "step": 2105
    },
    {
      "entropy": 6.18635630607605,
      "epoch": 0.1641703948648123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999160285337501,
      "loss": 5.8807,
      "mean_token_accuracy": 0.14085181877017022,
      "num_tokens": 3655204.0,
      "step": 2110
    },
    {
      "entropy": 6.139962005615234,
      "epoch": 0.16455942423652986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999152696926399,
      "loss": 6.031,
      "mean_token_accuracy": 0.1334392912685871,
      "num_tokens": 3663939.0,
      "step": 2115
    },
    {
      "entropy": 6.174469661712647,
      "epoch": 0.16494845360824742,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999145074387985,
      "loss": 5.9681,
      "mean_token_accuracy": 0.13877132534980774,
      "num_tokens": 3673167.0,
      "step": 2120
    },
    {
      "entropy": 6.170883893966675,
      "epoch": 0.16533748297996498,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999137417722374,
      "loss": 5.9079,
      "mean_token_accuracy": 0.1417030766606331,
      "num_tokens": 3682234.0,
      "step": 2125
    },
    {
      "entropy": 6.110505771636963,
      "epoch": 0.16572651235168256,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999129726929684,
      "loss": 5.904,
      "mean_token_accuracy": 0.14176712557673454,
      "num_tokens": 3690835.0,
      "step": 2130
    },
    {
      "entropy": 6.160515642166137,
      "epoch": 0.16611554172340012,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999122002010029,
      "loss": 5.9654,
      "mean_token_accuracy": 0.13802013844251632,
      "num_tokens": 3699348.0,
      "step": 2135
    },
    {
      "entropy": 6.086480808258057,
      "epoch": 0.16650457109511768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999114242963527,
      "loss": 5.8711,
      "mean_token_accuracy": 0.1402495212852955,
      "num_tokens": 3708205.0,
      "step": 2140
    },
    {
      "entropy": 6.106543445587159,
      "epoch": 0.16689360046683524,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999106449790298,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1402938574552536,
      "num_tokens": 3717016.0,
      "step": 2145
    },
    {
      "entropy": 6.246592807769775,
      "epoch": 0.1672826298385528,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999098622490458,
      "loss": 5.9544,
      "mean_token_accuracy": 0.13701125010848045,
      "num_tokens": 3725509.0,
      "step": 2150
    },
    {
      "entropy": 6.154575967788697,
      "epoch": 0.16767165921027039,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999090761064125,
      "loss": 6.0286,
      "mean_token_accuracy": 0.1339248612523079,
      "num_tokens": 3734674.0,
      "step": 2155
    },
    {
      "entropy": 6.060240173339844,
      "epoch": 0.16806068858198794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499908286551142,
      "loss": 5.8456,
      "mean_token_accuracy": 0.1407431736588478,
      "num_tokens": 3742930.0,
      "step": 2160
    },
    {
      "entropy": 6.154189491271973,
      "epoch": 0.1684497179537055,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999074935832463,
      "loss": 5.9369,
      "mean_token_accuracy": 0.1409689135849476,
      "num_tokens": 3751314.0,
      "step": 2165
    },
    {
      "entropy": 6.216179037094117,
      "epoch": 0.16883874732542306,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999066972027373,
      "loss": 6.0535,
      "mean_token_accuracy": 0.13319066762924195,
      "num_tokens": 3760229.0,
      "step": 2170
    },
    {
      "entropy": 6.285803413391113,
      "epoch": 0.16922777669714065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999058974096271,
      "loss": 6.0545,
      "mean_token_accuracy": 0.1335298717021942,
      "num_tokens": 3769037.0,
      "step": 2175
    },
    {
      "entropy": 6.052758312225341,
      "epoch": 0.1696168060688582,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499905094203928,
      "loss": 5.9238,
      "mean_token_accuracy": 0.13934388458728791,
      "num_tokens": 3777686.0,
      "step": 2180
    },
    {
      "entropy": 6.218620347976684,
      "epoch": 0.17000583544057576,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499904287585652,
      "loss": 6.0044,
      "mean_token_accuracy": 0.13825156837701796,
      "num_tokens": 3786412.0,
      "step": 2185
    },
    {
      "entropy": 6.192234373092651,
      "epoch": 0.17039486481229332,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999034775548113,
      "loss": 6.0351,
      "mean_token_accuracy": 0.13853697702288628,
      "num_tokens": 3794981.0,
      "step": 2190
    },
    {
      "entropy": 6.155090808868408,
      "epoch": 0.17078389418401088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999026641114185,
      "loss": 5.9303,
      "mean_token_accuracy": 0.14125299602746963,
      "num_tokens": 3803069.0,
      "step": 2195
    },
    {
      "entropy": 6.151627492904663,
      "epoch": 0.17117292355572847,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999018472554857,
      "loss": 6.0112,
      "mean_token_accuracy": 0.1394211769104004,
      "num_tokens": 3811986.0,
      "step": 2200
    },
    {
      "entropy": 6.219188404083252,
      "epoch": 0.17156195292744603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999010269870253,
      "loss": 5.9302,
      "mean_token_accuracy": 0.13892962485551835,
      "num_tokens": 3820626.0,
      "step": 2205
    },
    {
      "entropy": 6.105683422088623,
      "epoch": 0.17195098229916359,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999002033060498,
      "loss": 5.9254,
      "mean_token_accuracy": 0.1396380692720413,
      "num_tokens": 3828608.0,
      "step": 2210
    },
    {
      "entropy": 6.041509485244751,
      "epoch": 0.17234001167088114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998993762125716,
      "loss": 5.8795,
      "mean_token_accuracy": 0.1437774755060673,
      "num_tokens": 3837059.0,
      "step": 2215
    },
    {
      "entropy": 6.148551082611084,
      "epoch": 0.17272904104259873,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998985457066034,
      "loss": 5.8953,
      "mean_token_accuracy": 0.13735355138778688,
      "num_tokens": 3845137.0,
      "step": 2220
    },
    {
      "entropy": 6.128031158447266,
      "epoch": 0.1731180704143163,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998977117881576,
      "loss": 5.8861,
      "mean_token_accuracy": 0.14574987590312957,
      "num_tokens": 3854012.0,
      "step": 2225
    },
    {
      "entropy": 6.015646696090698,
      "epoch": 0.17350709978603385,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998968744572472,
      "loss": 5.8295,
      "mean_token_accuracy": 0.1409018814563751,
      "num_tokens": 3862308.0,
      "step": 2230
    },
    {
      "entropy": 6.067210721969604,
      "epoch": 0.1738961291577514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998960337138845,
      "loss": 5.7717,
      "mean_token_accuracy": 0.13879043981432915,
      "num_tokens": 3870975.0,
      "step": 2235
    },
    {
      "entropy": 6.087196779251099,
      "epoch": 0.17428515852946896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998951895580826,
      "loss": 5.9544,
      "mean_token_accuracy": 0.14272485896945,
      "num_tokens": 3879621.0,
      "step": 2240
    },
    {
      "entropy": 6.093369865417481,
      "epoch": 0.17467418790118655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998943419898542,
      "loss": 5.9166,
      "mean_token_accuracy": 0.13817212730646133,
      "num_tokens": 3889214.0,
      "step": 2245
    },
    {
      "entropy": 6.146738815307617,
      "epoch": 0.1750632172729041,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998934910092119,
      "loss": 5.9504,
      "mean_token_accuracy": 0.14284390732645988,
      "num_tokens": 3898250.0,
      "step": 2250
    },
    {
      "entropy": 6.129539108276367,
      "epoch": 0.17545224664462167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998926366161689,
      "loss": 5.9798,
      "mean_token_accuracy": 0.13206198140978814,
      "num_tokens": 3906971.0,
      "step": 2255
    },
    {
      "entropy": 6.1653541088104244,
      "epoch": 0.17584127601633923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998917788107383,
      "loss": 6.0139,
      "mean_token_accuracy": 0.13972242698073387,
      "num_tokens": 3916663.0,
      "step": 2260
    },
    {
      "entropy": 6.103351974487305,
      "epoch": 0.17623030538805678,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998909175929326,
      "loss": 5.8942,
      "mean_token_accuracy": 0.14154435321688652,
      "num_tokens": 3926848.0,
      "step": 2265
    },
    {
      "entropy": 6.034053564071655,
      "epoch": 0.17661933475977437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998900529627653,
      "loss": 5.8613,
      "mean_token_accuracy": 0.1404354229569435,
      "num_tokens": 3935729.0,
      "step": 2270
    },
    {
      "entropy": 6.141492509841919,
      "epoch": 0.17700836413149193,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998891849202495,
      "loss": 5.8766,
      "mean_token_accuracy": 0.1424740418791771,
      "num_tokens": 3943408.0,
      "step": 2275
    },
    {
      "entropy": 6.052723979949951,
      "epoch": 0.1773973935032095,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499888313465398,
      "loss": 5.8509,
      "mean_token_accuracy": 0.14049414843320845,
      "num_tokens": 3952473.0,
      "step": 2280
    },
    {
      "entropy": 6.178114128112793,
      "epoch": 0.17778642287492705,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998874385982247,
      "loss": 6.0018,
      "mean_token_accuracy": 0.13325022533535957,
      "num_tokens": 3960878.0,
      "step": 2285
    },
    {
      "entropy": 6.208947801589966,
      "epoch": 0.17817545224664463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998865603187421,
      "loss": 5.945,
      "mean_token_accuracy": 0.13925340324640273,
      "num_tokens": 3969282.0,
      "step": 2290
    },
    {
      "entropy": 6.132673740386963,
      "epoch": 0.1785644816183622,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499885678626964,
      "loss": 5.9267,
      "mean_token_accuracy": 0.14256497994065284,
      "num_tokens": 3978516.0,
      "step": 2295
    },
    {
      "entropy": 6.066269874572754,
      "epoch": 0.17895351099007975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998847935229038,
      "loss": 5.7855,
      "mean_token_accuracy": 0.15060577318072318,
      "num_tokens": 3986580.0,
      "step": 2300
    },
    {
      "entropy": 5.910192060470581,
      "epoch": 0.1793425403617973,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998839050065746,
      "loss": 5.8052,
      "mean_token_accuracy": 0.14743324518203735,
      "num_tokens": 3995144.0,
      "step": 2305
    },
    {
      "entropy": 6.101610565185547,
      "epoch": 0.17973156973351487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998830130779902,
      "loss": 5.8075,
      "mean_token_accuracy": 0.13889141082763673,
      "num_tokens": 4003887.0,
      "step": 2310
    },
    {
      "entropy": 6.0803686618804935,
      "epoch": 0.18012059910523245,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499882117737164,
      "loss": 5.9209,
      "mean_token_accuracy": 0.14422064125537873,
      "num_tokens": 4012607.0,
      "step": 2315
    },
    {
      "entropy": 6.053954172134399,
      "epoch": 0.18050962847695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998812189841096,
      "loss": 6.0034,
      "mean_token_accuracy": 0.13622427582740784,
      "num_tokens": 4021950.0,
      "step": 2320
    },
    {
      "entropy": 6.1665068626403805,
      "epoch": 0.18089865784866757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998803168188407,
      "loss": 5.9668,
      "mean_token_accuracy": 0.1415689304471016,
      "num_tokens": 4031076.0,
      "step": 2325
    },
    {
      "entropy": 5.990097284317017,
      "epoch": 0.18128768722038513,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998794112413708,
      "loss": 5.8054,
      "mean_token_accuracy": 0.14801988750696182,
      "num_tokens": 4040414.0,
      "step": 2330
    },
    {
      "entropy": 6.18867506980896,
      "epoch": 0.18167671659210272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998785022517137,
      "loss": 5.8924,
      "mean_token_accuracy": 0.14523087441921234,
      "num_tokens": 4048931.0,
      "step": 2335
    },
    {
      "entropy": 5.915593385696411,
      "epoch": 0.18206574596382027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998775898498835,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1424514263868332,
      "num_tokens": 4057688.0,
      "step": 2340
    },
    {
      "entropy": 6.047796154022217,
      "epoch": 0.18245477533553783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998766740358936,
      "loss": 5.9117,
      "mean_token_accuracy": 0.14413589462637902,
      "num_tokens": 4066365.0,
      "step": 2345
    },
    {
      "entropy": 6.0719398021697994,
      "epoch": 0.1828438047072554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998757548097582,
      "loss": 5.7968,
      "mean_token_accuracy": 0.14763821363449098,
      "num_tokens": 4075136.0,
      "step": 2350
    },
    {
      "entropy": 6.112574434280395,
      "epoch": 0.18323283407897295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998748321714911,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1457724839448929,
      "num_tokens": 4083337.0,
      "step": 2355
    },
    {
      "entropy": 6.00722303390503,
      "epoch": 0.18362186345069054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998739061211065,
      "loss": 5.8555,
      "mean_token_accuracy": 0.14413827881217003,
      "num_tokens": 4091201.0,
      "step": 2360
    },
    {
      "entropy": 6.140283298492432,
      "epoch": 0.1840108928224081,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998729766586181,
      "loss": 5.9072,
      "mean_token_accuracy": 0.14183114990592002,
      "num_tokens": 4099480.0,
      "step": 2365
    },
    {
      "entropy": 5.984076595306396,
      "epoch": 0.18439992219412565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998720437840403,
      "loss": 5.8158,
      "mean_token_accuracy": 0.1437339261174202,
      "num_tokens": 4108151.0,
      "step": 2370
    },
    {
      "entropy": 5.961683034896851,
      "epoch": 0.1847889515658432,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998711074973871,
      "loss": 5.8251,
      "mean_token_accuracy": 0.1457428365945816,
      "num_tokens": 4117380.0,
      "step": 2375
    },
    {
      "entropy": 6.036372423171997,
      "epoch": 0.1851779809375608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998701677986728,
      "loss": 5.8689,
      "mean_token_accuracy": 0.14735556095838548,
      "num_tokens": 4126757.0,
      "step": 2380
    },
    {
      "entropy": 6.031168508529663,
      "epoch": 0.18556701030927836,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998692246879116,
      "loss": 5.7891,
      "mean_token_accuracy": 0.146223184466362,
      "num_tokens": 4135326.0,
      "step": 2385
    },
    {
      "entropy": 6.053634834289551,
      "epoch": 0.18595603968099592,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998682781651178,
      "loss": 5.7988,
      "mean_token_accuracy": 0.15181397199630736,
      "num_tokens": 4143624.0,
      "step": 2390
    },
    {
      "entropy": 5.992155838012695,
      "epoch": 0.18634506905271347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998673282303059,
      "loss": 5.7659,
      "mean_token_accuracy": 0.14670938700437547,
      "num_tokens": 4152193.0,
      "step": 2395
    },
    {
      "entropy": 6.03981647491455,
      "epoch": 0.18673409842443103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998663748834901,
      "loss": 5.8835,
      "mean_token_accuracy": 0.141094571352005,
      "num_tokens": 4160614.0,
      "step": 2400
    },
    {
      "entropy": 6.138073205947876,
      "epoch": 0.18712312779614862,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998654181246851,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1435561805963516,
      "num_tokens": 4168715.0,
      "step": 2405
    },
    {
      "entropy": 6.1046959400177006,
      "epoch": 0.18751215716786618,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998644579539052,
      "loss": 5.956,
      "mean_token_accuracy": 0.1427804246544838,
      "num_tokens": 4176676.0,
      "step": 2410
    },
    {
      "entropy": 6.041570472717285,
      "epoch": 0.18790118653958374,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499863494371165,
      "loss": 5.9196,
      "mean_token_accuracy": 0.13990830332040788,
      "num_tokens": 4185613.0,
      "step": 2415
    },
    {
      "entropy": 6.056922721862793,
      "epoch": 0.1882902159113013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998625273764793,
      "loss": 5.9221,
      "mean_token_accuracy": 0.1435866177082062,
      "num_tokens": 4193985.0,
      "step": 2420
    },
    {
      "entropy": 6.057398271560669,
      "epoch": 0.18867924528301888,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998615569698626,
      "loss": 5.8347,
      "mean_token_accuracy": 0.1472941130399704,
      "num_tokens": 4202713.0,
      "step": 2425
    },
    {
      "entropy": 6.15912766456604,
      "epoch": 0.18906827465473644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998605831513296,
      "loss": 5.9598,
      "mean_token_accuracy": 0.13513548597693442,
      "num_tokens": 4210858.0,
      "step": 2430
    },
    {
      "entropy": 6.027268743515014,
      "epoch": 0.189457304026454,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998596059208953,
      "loss": 5.876,
      "mean_token_accuracy": 0.14263733327388764,
      "num_tokens": 4219570.0,
      "step": 2435
    },
    {
      "entropy": 6.034002113342285,
      "epoch": 0.18984633339817156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998586252785743,
      "loss": 5.7873,
      "mean_token_accuracy": 0.14603268951177598,
      "num_tokens": 4228026.0,
      "step": 2440
    },
    {
      "entropy": 6.043663167953492,
      "epoch": 0.19023536276988912,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998576412243816,
      "loss": 5.8084,
      "mean_token_accuracy": 0.14466246217489243,
      "num_tokens": 4235993.0,
      "step": 2445
    },
    {
      "entropy": 6.009197854995728,
      "epoch": 0.1906243921416067,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998566537583321,
      "loss": 5.8151,
      "mean_token_accuracy": 0.147155824303627,
      "num_tokens": 4244225.0,
      "step": 2450
    },
    {
      "entropy": 5.965597009658813,
      "epoch": 0.19101342151332426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998556628804408,
      "loss": 5.7885,
      "mean_token_accuracy": 0.1485578991472721,
      "num_tokens": 4252989.0,
      "step": 2455
    },
    {
      "entropy": 6.069110679626465,
      "epoch": 0.19140245088504182,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998546685907225,
      "loss": 5.8298,
      "mean_token_accuracy": 0.14657569378614427,
      "num_tokens": 4261297.0,
      "step": 2460
    },
    {
      "entropy": 6.0928350448608395,
      "epoch": 0.19179148025675938,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998536708891927,
      "loss": 5.9644,
      "mean_token_accuracy": 0.13758677765727043,
      "num_tokens": 4270299.0,
      "step": 2465
    },
    {
      "entropy": 6.128661823272705,
      "epoch": 0.19218050962847694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998526697758663,
      "loss": 5.9728,
      "mean_token_accuracy": 0.13565355464816092,
      "num_tokens": 4279662.0,
      "step": 2470
    },
    {
      "entropy": 6.051914119720459,
      "epoch": 0.19256953900019452,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998516652507585,
      "loss": 5.7902,
      "mean_token_accuracy": 0.1471126437187195,
      "num_tokens": 4287876.0,
      "step": 2475
    },
    {
      "entropy": 6.014107513427734,
      "epoch": 0.19295856837191208,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998506573138846,
      "loss": 5.9218,
      "mean_token_accuracy": 0.14491380155086517,
      "num_tokens": 4296415.0,
      "step": 2480
    },
    {
      "entropy": 5.982778406143188,
      "epoch": 0.19334759774362964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998496459652598,
      "loss": 5.7484,
      "mean_token_accuracy": 0.14843220636248589,
      "num_tokens": 4304434.0,
      "step": 2485
    },
    {
      "entropy": 6.051380062103272,
      "epoch": 0.1937366271153472,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998486312048995,
      "loss": 5.9265,
      "mean_token_accuracy": 0.1381976492702961,
      "num_tokens": 4313266.0,
      "step": 2490
    },
    {
      "entropy": 6.0365697860717775,
      "epoch": 0.19412565648706478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998476130328193,
      "loss": 5.8013,
      "mean_token_accuracy": 0.14801377803087234,
      "num_tokens": 4322121.0,
      "step": 2495
    },
    {
      "entropy": 5.980331611633301,
      "epoch": 0.19451468585878234,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998465914490343,
      "loss": 5.8202,
      "mean_token_accuracy": 0.14618535935878754,
      "num_tokens": 4330596.0,
      "step": 2500
    },
    {
      "entropy": 6.062447881698608,
      "epoch": 0.1949037152304999,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049984556645356,
      "loss": 5.9022,
      "mean_token_accuracy": 0.1466861292719841,
      "num_tokens": 4337779.0,
      "step": 2505
    },
    {
      "entropy": 6.0627391815185545,
      "epoch": 0.19529274460221746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998445380464124,
      "loss": 5.7454,
      "mean_token_accuracy": 0.14693358018994332,
      "num_tokens": 4346623.0,
      "step": 2510
    },
    {
      "entropy": 5.929016637802124,
      "epoch": 0.19568177397393502,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998435062276066,
      "loss": 5.8526,
      "mean_token_accuracy": 0.14088083803653717,
      "num_tokens": 4355435.0,
      "step": 2515
    },
    {
      "entropy": 6.1320250034332275,
      "epoch": 0.1960708033456526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998424709971586,
      "loss": 5.902,
      "mean_token_accuracy": 0.14303931146860122,
      "num_tokens": 4363636.0,
      "step": 2520
    },
    {
      "entropy": 5.970055866241455,
      "epoch": 0.19645983271737016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998414323550839,
      "loss": 5.7269,
      "mean_token_accuracy": 0.15075851678848268,
      "num_tokens": 4372337.0,
      "step": 2525
    },
    {
      "entropy": 5.979619741439819,
      "epoch": 0.19684886208908772,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004998403903013984,
      "loss": 5.7561,
      "mean_token_accuracy": 0.15269921571016312,
      "num_tokens": 4380528.0,
      "step": 2530
    },
    {
      "entropy": 5.998526477813721,
      "epoch": 0.19723789146080528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998393448361179,
      "loss": 5.8501,
      "mean_token_accuracy": 0.13916658014059066,
      "num_tokens": 4389464.0,
      "step": 2535
    },
    {
      "entropy": 6.082704496383667,
      "epoch": 0.19762692083252287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499838295959258,
      "loss": 5.9085,
      "mean_token_accuracy": 0.14364177733659744,
      "num_tokens": 4397879.0,
      "step": 2540
    },
    {
      "entropy": 6.062457323074341,
      "epoch": 0.19801595020424043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499837243670835,
      "loss": 5.932,
      "mean_token_accuracy": 0.14348311796784402,
      "num_tokens": 4406145.0,
      "step": 2545
    },
    {
      "entropy": 6.0163922786712645,
      "epoch": 0.19840497957595798,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998361879708646,
      "loss": 5.8382,
      "mean_token_accuracy": 0.14055323377251625,
      "num_tokens": 4415480.0,
      "step": 2550
    },
    {
      "entropy": 6.038650321960449,
      "epoch": 0.19879400894767554,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499835128859363,
      "loss": 5.8457,
      "mean_token_accuracy": 0.14373011589050294,
      "num_tokens": 4423355.0,
      "step": 2555
    },
    {
      "entropy": 5.971053647994995,
      "epoch": 0.1991830383193931,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998340663363461,
      "loss": 5.7288,
      "mean_token_accuracy": 0.15200028717517852,
      "num_tokens": 4432059.0,
      "step": 2560
    },
    {
      "entropy": 5.9389901638031,
      "epoch": 0.1995720676911107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998330004018301,
      "loss": 5.8266,
      "mean_token_accuracy": 0.14117742702364922,
      "num_tokens": 4441146.0,
      "step": 2565
    },
    {
      "entropy": 5.973790550231934,
      "epoch": 0.19996109706282825,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998319310558312,
      "loss": 5.6701,
      "mean_token_accuracy": 0.15412732362747192,
      "num_tokens": 4448553.0,
      "step": 2570
    },
    {
      "entropy": 5.836313152313233,
      "epoch": 0.2003501264345458,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004998308582983657,
      "loss": 5.7375,
      "mean_token_accuracy": 0.15181366503238677,
      "num_tokens": 4456930.0,
      "step": 2575
    },
    {
      "entropy": 6.073364543914795,
      "epoch": 0.20073915580626336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998297821294497,
      "loss": 5.9386,
      "mean_token_accuracy": 0.14404268860816954,
      "num_tokens": 4464932.0,
      "step": 2580
    },
    {
      "entropy": 5.992941284179688,
      "epoch": 0.20112818517798095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998287025490995,
      "loss": 5.6672,
      "mean_token_accuracy": 0.15902086794376374,
      "num_tokens": 4473695.0,
      "step": 2585
    },
    {
      "entropy": 5.925492858886718,
      "epoch": 0.2015172145496985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998276195573317,
      "loss": 5.8921,
      "mean_token_accuracy": 0.14221425950527192,
      "num_tokens": 4483375.0,
      "step": 2590
    },
    {
      "entropy": 6.11337890625,
      "epoch": 0.20190624392141607,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998265331541627,
      "loss": 5.9193,
      "mean_token_accuracy": 0.13467642813920974,
      "num_tokens": 4491695.0,
      "step": 2595
    },
    {
      "entropy": 6.003162908554077,
      "epoch": 0.20229527329313363,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998254433396088,
      "loss": 5.8721,
      "mean_token_accuracy": 0.1440841868519783,
      "num_tokens": 4500361.0,
      "step": 2600
    },
    {
      "entropy": 5.987391805648803,
      "epoch": 0.20268430266485118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998243501136867,
      "loss": 5.8079,
      "mean_token_accuracy": 0.14264530092477798,
      "num_tokens": 4509368.0,
      "step": 2605
    },
    {
      "entropy": 6.039334154129028,
      "epoch": 0.20307333203656877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998232534764129,
      "loss": 5.7889,
      "mean_token_accuracy": 0.14281965047121048,
      "num_tokens": 4518146.0,
      "step": 2610
    },
    {
      "entropy": 6.01565580368042,
      "epoch": 0.20346236140828633,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998221534278041,
      "loss": 5.8405,
      "mean_token_accuracy": 0.14281932562589644,
      "num_tokens": 4526128.0,
      "step": 2615
    },
    {
      "entropy": 6.036164903640747,
      "epoch": 0.2038513907800039,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998210499678769,
      "loss": 5.8847,
      "mean_token_accuracy": 0.14803217351436615,
      "num_tokens": 4534724.0,
      "step": 2620
    },
    {
      "entropy": 5.968258047103882,
      "epoch": 0.20424042015172145,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998199430966481,
      "loss": 5.7427,
      "mean_token_accuracy": 0.14365783631801604,
      "num_tokens": 4543725.0,
      "step": 2625
    },
    {
      "entropy": 6.012339067459107,
      "epoch": 0.20462944952343903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998188328141346,
      "loss": 5.8252,
      "mean_token_accuracy": 0.14723241925239564,
      "num_tokens": 4551758.0,
      "step": 2630
    },
    {
      "entropy": 5.894060182571411,
      "epoch": 0.2050184788951566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998177191203531,
      "loss": 5.7079,
      "mean_token_accuracy": 0.15085572972893715,
      "num_tokens": 4559990.0,
      "step": 2635
    },
    {
      "entropy": 5.93325366973877,
      "epoch": 0.20540750826687415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998166020153204,
      "loss": 5.7558,
      "mean_token_accuracy": 0.1438131585717201,
      "num_tokens": 4568651.0,
      "step": 2640
    },
    {
      "entropy": 5.900708532333374,
      "epoch": 0.2057965376385917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998154814990538,
      "loss": 5.6758,
      "mean_token_accuracy": 0.15049925595521926,
      "num_tokens": 4577981.0,
      "step": 2645
    },
    {
      "entropy": 5.955266237258911,
      "epoch": 0.20618556701030927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049981435757157,
      "loss": 5.8312,
      "mean_token_accuracy": 0.147544726729393,
      "num_tokens": 4586480.0,
      "step": 2650
    },
    {
      "entropy": 5.980217790603637,
      "epoch": 0.20657459638202685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998132302328862,
      "loss": 5.7291,
      "mean_token_accuracy": 0.1525867000222206,
      "num_tokens": 4595554.0,
      "step": 2655
    },
    {
      "entropy": 5.923847293853759,
      "epoch": 0.2069636257537444,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998120994830195,
      "loss": 5.7566,
      "mean_token_accuracy": 0.15113697126507758,
      "num_tokens": 4604091.0,
      "step": 2660
    },
    {
      "entropy": 6.023667430877685,
      "epoch": 0.20735265512546197,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499810965321987,
      "loss": 5.8615,
      "mean_token_accuracy": 0.14118051454424857,
      "num_tokens": 4612603.0,
      "step": 2665
    },
    {
      "entropy": 6.032278680801392,
      "epoch": 0.20774168449717953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998098277498059,
      "loss": 5.7592,
      "mean_token_accuracy": 0.1470929801464081,
      "num_tokens": 4621041.0,
      "step": 2670
    },
    {
      "entropy": 5.90060658454895,
      "epoch": 0.2081307138688971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998086867664934,
      "loss": 5.7983,
      "mean_token_accuracy": 0.15032607913017274,
      "num_tokens": 4629849.0,
      "step": 2675
    },
    {
      "entropy": 6.054356479644776,
      "epoch": 0.20851974324061467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499807542372067,
      "loss": 5.8687,
      "mean_token_accuracy": 0.14306920617818833,
      "num_tokens": 4639426.0,
      "step": 2680
    },
    {
      "entropy": 5.9633255958557125,
      "epoch": 0.20890877261233223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998063945665439,
      "loss": 5.7888,
      "mean_token_accuracy": 0.14868860244750975,
      "num_tokens": 4647257.0,
      "step": 2685
    },
    {
      "entropy": 5.958910369873047,
      "epoch": 0.2092978019840498,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998052433499416,
      "loss": 5.7688,
      "mean_token_accuracy": 0.14971182197332383,
      "num_tokens": 4655773.0,
      "step": 2690
    },
    {
      "entropy": 5.995905542373658,
      "epoch": 0.20968683135576735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998040887222776,
      "loss": 5.8141,
      "mean_token_accuracy": 0.14208999648690224,
      "num_tokens": 4664292.0,
      "step": 2695
    },
    {
      "entropy": 6.006470203399658,
      "epoch": 0.21007586072748494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998029306835693,
      "loss": 5.8357,
      "mean_token_accuracy": 0.1457445926964283,
      "num_tokens": 4672774.0,
      "step": 2700
    },
    {
      "entropy": 6.1288737773895265,
      "epoch": 0.2104648900992025,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998017692338344,
      "loss": 5.954,
      "mean_token_accuracy": 0.14017487764358522,
      "num_tokens": 4681855.0,
      "step": 2705
    },
    {
      "entropy": 5.958335494995117,
      "epoch": 0.21085391947092005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998006043730905,
      "loss": 5.7565,
      "mean_token_accuracy": 0.14842477068305016,
      "num_tokens": 4689771.0,
      "step": 2710
    },
    {
      "entropy": 6.003065395355224,
      "epoch": 0.2112429488426376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997994361013551,
      "loss": 5.8986,
      "mean_token_accuracy": 0.1353430561721325,
      "num_tokens": 4698808.0,
      "step": 2715
    },
    {
      "entropy": 5.980752849578858,
      "epoch": 0.21163197821435517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997982644186461,
      "loss": 5.7782,
      "mean_token_accuracy": 0.14798913151025772,
      "num_tokens": 4707627.0,
      "step": 2720
    },
    {
      "entropy": 5.949896383285522,
      "epoch": 0.21202100758607276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997970893249813,
      "loss": 5.7576,
      "mean_token_accuracy": 0.14555923044681549,
      "num_tokens": 4716091.0,
      "step": 2725
    },
    {
      "entropy": 5.944676637649536,
      "epoch": 0.21241003695779032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997959108203785,
      "loss": 5.788,
      "mean_token_accuracy": 0.1488296665251255,
      "num_tokens": 4724042.0,
      "step": 2730
    },
    {
      "entropy": 5.887519454956054,
      "epoch": 0.21279906632950787,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997947289048554,
      "loss": 5.7058,
      "mean_token_accuracy": 0.15533653944730758,
      "num_tokens": 4731745.0,
      "step": 2735
    },
    {
      "entropy": 5.945332145690918,
      "epoch": 0.21318809570122543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997935435784302,
      "loss": 5.872,
      "mean_token_accuracy": 0.146851022541523,
      "num_tokens": 4740386.0,
      "step": 2740
    },
    {
      "entropy": 6.041448068618775,
      "epoch": 0.21357712507294302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997923548411208,
      "loss": 5.8834,
      "mean_token_accuracy": 0.1391231372952461,
      "num_tokens": 4749809.0,
      "step": 2745
    },
    {
      "entropy": 6.063974905014038,
      "epoch": 0.21396615444466058,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499791162692945,
      "loss": 5.9768,
      "mean_token_accuracy": 0.1409013733267784,
      "num_tokens": 4759507.0,
      "step": 2750
    },
    {
      "entropy": 5.960179090499878,
      "epoch": 0.21435518381637814,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997899671339214,
      "loss": 5.7479,
      "mean_token_accuracy": 0.1449030451476574,
      "num_tokens": 4768793.0,
      "step": 2755
    },
    {
      "entropy": 6.029723405838013,
      "epoch": 0.2147442131880957,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997887681640675,
      "loss": 5.9008,
      "mean_token_accuracy": 0.14465020522475242,
      "num_tokens": 4777344.0,
      "step": 2760
    },
    {
      "entropy": 6.030321025848389,
      "epoch": 0.21513324255981325,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997875657834021,
      "loss": 5.8127,
      "mean_token_accuracy": 0.1441117987036705,
      "num_tokens": 4785934.0,
      "step": 2765
    },
    {
      "entropy": 5.947022724151611,
      "epoch": 0.21552227193153084,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997863599919432,
      "loss": 5.7731,
      "mean_token_accuracy": 0.1485394760966301,
      "num_tokens": 4794889.0,
      "step": 2770
    },
    {
      "entropy": 5.944930124282837,
      "epoch": 0.2159113013032484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997851507897089,
      "loss": 5.8881,
      "mean_token_accuracy": 0.14024131819605828,
      "num_tokens": 4803597.0,
      "step": 2775
    },
    {
      "entropy": 5.933191776275635,
      "epoch": 0.21630033067496596,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997839381767178,
      "loss": 5.6613,
      "mean_token_accuracy": 0.15584157705307006,
      "num_tokens": 4812706.0,
      "step": 2780
    },
    {
      "entropy": 5.90777177810669,
      "epoch": 0.21668936004668352,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997827221529882,
      "loss": 5.7295,
      "mean_token_accuracy": 0.1480869472026825,
      "num_tokens": 4821790.0,
      "step": 2785
    },
    {
      "entropy": 5.968650197982788,
      "epoch": 0.2170783894184011,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997815027185386,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1485279232263565,
      "num_tokens": 4830123.0,
      "step": 2790
    },
    {
      "entropy": 5.94366946220398,
      "epoch": 0.21746741879011866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997802798733874,
      "loss": 5.7194,
      "mean_token_accuracy": 0.1537564590573311,
      "num_tokens": 4838066.0,
      "step": 2795
    },
    {
      "entropy": 5.9353564262390135,
      "epoch": 0.21785644816183622,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997790536175534,
      "loss": 5.7944,
      "mean_token_accuracy": 0.14718551486730574,
      "num_tokens": 4846877.0,
      "step": 2800
    },
    {
      "entropy": 5.991869354248047,
      "epoch": 0.21824547753355378,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997778239510548,
      "loss": 5.8426,
      "mean_token_accuracy": 0.14051205813884735,
      "num_tokens": 4856960.0,
      "step": 2805
    },
    {
      "entropy": 6.004691219329834,
      "epoch": 0.21863450690527134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997765908739106,
      "loss": 5.8542,
      "mean_token_accuracy": 0.14763535112142562,
      "num_tokens": 4865440.0,
      "step": 2810
    },
    {
      "entropy": 5.893306064605713,
      "epoch": 0.21902353627698892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997753543861395,
      "loss": 5.7417,
      "mean_token_accuracy": 0.1489579737186432,
      "num_tokens": 4874187.0,
      "step": 2815
    },
    {
      "entropy": 6.102828645706177,
      "epoch": 0.21941256564870648,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049977411448776,
      "loss": 5.8831,
      "mean_token_accuracy": 0.14227399230003357,
      "num_tokens": 4883151.0,
      "step": 2820
    },
    {
      "entropy": 5.983131647109985,
      "epoch": 0.21980159502042404,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997728711787912,
      "loss": 5.7977,
      "mean_token_accuracy": 0.14272094666957855,
      "num_tokens": 4890968.0,
      "step": 2825
    },
    {
      "entropy": 5.879379606246948,
      "epoch": 0.2201906243921416,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997716244592518,
      "loss": 5.6825,
      "mean_token_accuracy": 0.15600915551185607,
      "num_tokens": 4899389.0,
      "step": 2830
    },
    {
      "entropy": 5.905595302581787,
      "epoch": 0.22057965376385918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997703743291607,
      "loss": 5.7509,
      "mean_token_accuracy": 0.15158967077732086,
      "num_tokens": 4907622.0,
      "step": 2835
    },
    {
      "entropy": 5.943753623962403,
      "epoch": 0.22096868313557674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499769120788537,
      "loss": 5.7491,
      "mean_token_accuracy": 0.1479824811220169,
      "num_tokens": 4916151.0,
      "step": 2840
    },
    {
      "entropy": 5.892854356765747,
      "epoch": 0.2213577125072943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997678638373995,
      "loss": 5.6692,
      "mean_token_accuracy": 0.15641618967056276,
      "num_tokens": 4923938.0,
      "step": 2845
    },
    {
      "entropy": 5.87450475692749,
      "epoch": 0.22174674187901186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997666034757676,
      "loss": 5.7613,
      "mean_token_accuracy": 0.16034080535173417,
      "num_tokens": 4932027.0,
      "step": 2850
    },
    {
      "entropy": 5.947768974304199,
      "epoch": 0.22213577125072942,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997653397036603,
      "loss": 5.7766,
      "mean_token_accuracy": 0.1560262620449066,
      "num_tokens": 4940229.0,
      "step": 2855
    },
    {
      "entropy": 5.907529211044311,
      "epoch": 0.222524800622447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997640725210965,
      "loss": 5.8405,
      "mean_token_accuracy": 0.14572847858071328,
      "num_tokens": 4948662.0,
      "step": 2860
    },
    {
      "entropy": 5.958647584915161,
      "epoch": 0.22291382999416456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997628019280958,
      "loss": 5.7854,
      "mean_token_accuracy": 0.1521115556359291,
      "num_tokens": 4956847.0,
      "step": 2865
    },
    {
      "entropy": 5.970917129516602,
      "epoch": 0.22330285936588212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997615279246773,
      "loss": 5.7205,
      "mean_token_accuracy": 0.14754560068249703,
      "num_tokens": 4965139.0,
      "step": 2870
    },
    {
      "entropy": 6.0411519527435305,
      "epoch": 0.22369188873759968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997602505108603,
      "loss": 5.8952,
      "mean_token_accuracy": 0.14166620448231698,
      "num_tokens": 4973671.0,
      "step": 2875
    },
    {
      "entropy": 5.933946704864502,
      "epoch": 0.22408091810931724,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997589696866644,
      "loss": 5.7652,
      "mean_token_accuracy": 0.14397089332342147,
      "num_tokens": 4981970.0,
      "step": 2880
    },
    {
      "entropy": 5.935157251358032,
      "epoch": 0.22446994748103483,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997576854521088,
      "loss": 5.7921,
      "mean_token_accuracy": 0.14724297374486922,
      "num_tokens": 4990591.0,
      "step": 2885
    },
    {
      "entropy": 5.9578015327453615,
      "epoch": 0.22485897685275238,
      "grad_norm": 1.125,
      "learning_rate": 0.000499756397807213,
      "loss": 5.7954,
      "mean_token_accuracy": 0.14315731078386307,
      "num_tokens": 4999623.0,
      "step": 2890
    },
    {
      "entropy": 6.011578512191773,
      "epoch": 0.22524800622446994,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997551067519966,
      "loss": 5.8545,
      "mean_token_accuracy": 0.1422098271548748,
      "num_tokens": 5008387.0,
      "step": 2895
    },
    {
      "entropy": 5.949732828140259,
      "epoch": 0.2256370355961875,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997538122864792,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1461174599826336,
      "num_tokens": 5016788.0,
      "step": 2900
    },
    {
      "entropy": 6.023969125747681,
      "epoch": 0.2260260649679051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997525144106804,
      "loss": 5.8104,
      "mean_token_accuracy": 0.14314667582511903,
      "num_tokens": 5025508.0,
      "step": 2905
    },
    {
      "entropy": 5.927273511886597,
      "epoch": 0.22641509433962265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049975121312462,
      "loss": 5.7821,
      "mean_token_accuracy": 0.14801414608955382,
      "num_tokens": 5035081.0,
      "step": 2910
    },
    {
      "entropy": 5.879549169540406,
      "epoch": 0.2268041237113402,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997499084283177,
      "loss": 5.7173,
      "mean_token_accuracy": 0.15523723959922792,
      "num_tokens": 5043655.0,
      "step": 2915
    },
    {
      "entropy": 5.91579909324646,
      "epoch": 0.22719315308305776,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997486003217932,
      "loss": 5.74,
      "mean_token_accuracy": 0.15187966972589492,
      "num_tokens": 5052080.0,
      "step": 2920
    },
    {
      "entropy": 5.888910627365112,
      "epoch": 0.22758218245477532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997472888050664,
      "loss": 5.7076,
      "mean_token_accuracy": 0.15852134227752684,
      "num_tokens": 5061325.0,
      "step": 2925
    },
    {
      "entropy": 5.961522054672241,
      "epoch": 0.2279712118264929,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004997459738781573,
      "loss": 5.8208,
      "mean_token_accuracy": 0.15095880776643752,
      "num_tokens": 5069461.0,
      "step": 2930
    },
    {
      "entropy": 5.988435316085815,
      "epoch": 0.22836024119821047,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997446555410857,
      "loss": 5.7807,
      "mean_token_accuracy": 0.14658573567867278,
      "num_tokens": 5077747.0,
      "step": 2935
    },
    {
      "entropy": 5.988274765014649,
      "epoch": 0.22874927056992803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997433337938716,
      "loss": 5.8365,
      "mean_token_accuracy": 0.15124830231070518,
      "num_tokens": 5086300.0,
      "step": 2940
    },
    {
      "entropy": 5.844106388092041,
      "epoch": 0.22913829994164558,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997420086365351,
      "loss": 5.6151,
      "mean_token_accuracy": 0.15467342734336853,
      "num_tokens": 5094762.0,
      "step": 2945
    },
    {
      "entropy": 5.89040789604187,
      "epoch": 0.22952732931336317,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004997406800690966,
      "loss": 5.7742,
      "mean_token_accuracy": 0.14963481277227403,
      "num_tokens": 5103364.0,
      "step": 2950
    },
    {
      "entropy": 6.023597526550293,
      "epoch": 0.22991635868508073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997393480915758,
      "loss": 5.8799,
      "mean_token_accuracy": 0.14407353550195695,
      "num_tokens": 5111730.0,
      "step": 2955
    },
    {
      "entropy": 5.970983409881592,
      "epoch": 0.2303053880567983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997380127039931,
      "loss": 5.6786,
      "mean_token_accuracy": 0.1486763596534729,
      "num_tokens": 5119878.0,
      "step": 2960
    },
    {
      "entropy": 5.9219104766845705,
      "epoch": 0.23069441742851585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997366739063688,
      "loss": 5.7901,
      "mean_token_accuracy": 0.1522405095398426,
      "num_tokens": 5129146.0,
      "step": 2965
    },
    {
      "entropy": 5.892798137664795,
      "epoch": 0.2310834468002334,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997353316987231,
      "loss": 5.8065,
      "mean_token_accuracy": 0.14348602443933486,
      "num_tokens": 5137989.0,
      "step": 2970
    },
    {
      "entropy": 6.039399528503418,
      "epoch": 0.231472476171951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997339860810765,
      "loss": 5.8268,
      "mean_token_accuracy": 0.14684822633862496,
      "num_tokens": 5146722.0,
      "step": 2975
    },
    {
      "entropy": 6.007096481323242,
      "epoch": 0.23186150554366855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997326370534495,
      "loss": 5.8128,
      "mean_token_accuracy": 0.14832943975925444,
      "num_tokens": 5154902.0,
      "step": 2980
    },
    {
      "entropy": 5.963329792022705,
      "epoch": 0.2322505349153861,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997312846158622,
      "loss": 5.827,
      "mean_token_accuracy": 0.14517123103141785,
      "num_tokens": 5163594.0,
      "step": 2985
    },
    {
      "entropy": 5.939759683609009,
      "epoch": 0.23263956428710367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997299287683355,
      "loss": 5.7065,
      "mean_token_accuracy": 0.1531409054994583,
      "num_tokens": 5172139.0,
      "step": 2990
    },
    {
      "entropy": 5.937909650802612,
      "epoch": 0.23302859365882125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997285695108898,
      "loss": 5.8315,
      "mean_token_accuracy": 0.15084856152534484,
      "num_tokens": 5180077.0,
      "step": 2995
    },
    {
      "entropy": 5.849994754791259,
      "epoch": 0.2334176230305388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997272068435458,
      "loss": 5.6851,
      "mean_token_accuracy": 0.14938049018383026,
      "num_tokens": 5188500.0,
      "step": 3000
    },
    {
      "epoch": 0.2334176230305388,
      "eval_entropy": 5.659352120260876,
      "eval_loss": 5.808426856994629,
      "eval_mean_token_accuracy": 0.15537977798754946,
      "eval_num_tokens": 5188500.0,
      "eval_runtime": 21.8125,
      "eval_samples_per_second": 1905.236,
      "eval_steps_per_second": 238.166,
      "step": 3000
    },
    {
      "entropy": 5.93622145652771,
      "epoch": 0.23380665240225637,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499725840766324,
      "loss": 5.7657,
      "mean_token_accuracy": 0.14401972368359567,
      "num_tokens": 5197214.0,
      "step": 3005
    },
    {
      "entropy": 6.025457191467285,
      "epoch": 0.23419568177397393,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997244712792453,
      "loss": 5.8495,
      "mean_token_accuracy": 0.14099710583686828,
      "num_tokens": 5206348.0,
      "step": 3010
    },
    {
      "entropy": 5.808966779708863,
      "epoch": 0.2345847111456915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997230983823305,
      "loss": 5.6655,
      "mean_token_accuracy": 0.14906769394874572,
      "num_tokens": 5215425.0,
      "step": 3015
    },
    {
      "entropy": 6.039634037017822,
      "epoch": 0.23497374051740907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997217220756003,
      "loss": 5.891,
      "mean_token_accuracy": 0.13933431655168532,
      "num_tokens": 5224899.0,
      "step": 3020
    },
    {
      "entropy": 5.964243030548095,
      "epoch": 0.23536276988912663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997203423590757,
      "loss": 5.8292,
      "mean_token_accuracy": 0.14440490528941155,
      "num_tokens": 5233829.0,
      "step": 3025
    },
    {
      "entropy": 5.998610305786133,
      "epoch": 0.2357517992608442,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997189592327775,
      "loss": 5.9395,
      "mean_token_accuracy": 0.1378275215625763,
      "num_tokens": 5243144.0,
      "step": 3030
    },
    {
      "entropy": 5.9781769752502445,
      "epoch": 0.23614082863256175,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997175726967268,
      "loss": 5.7548,
      "mean_token_accuracy": 0.1460421398282051,
      "num_tokens": 5251097.0,
      "step": 3035
    },
    {
      "entropy": 5.962513780593872,
      "epoch": 0.23652985800427934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997161827509447,
      "loss": 5.8404,
      "mean_token_accuracy": 0.15355018228292466,
      "num_tokens": 5260050.0,
      "step": 3040
    },
    {
      "entropy": 5.938783931732178,
      "epoch": 0.2369188873759969,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997147893954521,
      "loss": 5.645,
      "mean_token_accuracy": 0.15660878717899324,
      "num_tokens": 5268363.0,
      "step": 3045
    },
    {
      "entropy": 5.887162923812866,
      "epoch": 0.23730791674771445,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997133926302702,
      "loss": 5.8357,
      "mean_token_accuracy": 0.1478561833500862,
      "num_tokens": 5276821.0,
      "step": 3050
    },
    {
      "entropy": 5.9527363777160645,
      "epoch": 0.237696946119432,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997119924554204,
      "loss": 5.7226,
      "mean_token_accuracy": 0.14752280563116074,
      "num_tokens": 5285419.0,
      "step": 3055
    },
    {
      "entropy": 5.866193866729736,
      "epoch": 0.23808597549114957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997105888709236,
      "loss": 5.6553,
      "mean_token_accuracy": 0.1510017544031143,
      "num_tokens": 5294004.0,
      "step": 3060
    },
    {
      "entropy": 5.8557802677154545,
      "epoch": 0.23847500486286716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997091818768015,
      "loss": 5.8301,
      "mean_token_accuracy": 0.14936043918132783,
      "num_tokens": 5302699.0,
      "step": 3065
    },
    {
      "entropy": 5.966806888580322,
      "epoch": 0.23886403423458472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499707771473075,
      "loss": 5.7575,
      "mean_token_accuracy": 0.15072435289621353,
      "num_tokens": 5311115.0,
      "step": 3070
    },
    {
      "entropy": 5.901711511611938,
      "epoch": 0.23925306360630227,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997063576597659,
      "loss": 5.7648,
      "mean_token_accuracy": 0.15388039052486419,
      "num_tokens": 5319795.0,
      "step": 3075
    },
    {
      "entropy": 5.978531169891357,
      "epoch": 0.23964209297801983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997049404368954,
      "loss": 5.8031,
      "mean_token_accuracy": 0.15001416355371475,
      "num_tokens": 5328273.0,
      "step": 3080
    },
    {
      "entropy": 5.9556351661682125,
      "epoch": 0.2400311223497374,
      "grad_norm": 1.125,
      "learning_rate": 0.000499703519804485,
      "loss": 5.7934,
      "mean_token_accuracy": 0.1462474137544632,
      "num_tokens": 5336394.0,
      "step": 3085
    },
    {
      "entropy": 5.87838306427002,
      "epoch": 0.24042015172145498,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997020957625564,
      "loss": 5.7355,
      "mean_token_accuracy": 0.15258182883262633,
      "num_tokens": 5344480.0,
      "step": 3090
    },
    {
      "entropy": 5.896993970870971,
      "epoch": 0.24080918109317254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997006683111312,
      "loss": 5.797,
      "mean_token_accuracy": 0.141014863550663,
      "num_tokens": 5353445.0,
      "step": 3095
    },
    {
      "entropy": 5.892958879470825,
      "epoch": 0.2411982104648901,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499699237450231,
      "loss": 5.6596,
      "mean_token_accuracy": 0.16075871959328653,
      "num_tokens": 5362356.0,
      "step": 3100
    },
    {
      "entropy": 5.916681575775146,
      "epoch": 0.24158723983660765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996978031798774,
      "loss": 5.7964,
      "mean_token_accuracy": 0.14848088175058366,
      "num_tokens": 5371467.0,
      "step": 3105
    },
    {
      "entropy": 5.812461185455322,
      "epoch": 0.24197626920832524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996963655000924,
      "loss": 5.6558,
      "mean_token_accuracy": 0.15510347113013268,
      "num_tokens": 5380256.0,
      "step": 3110
    },
    {
      "entropy": 5.971241188049317,
      "epoch": 0.2423652985800428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996949244108977,
      "loss": 5.7135,
      "mean_token_accuracy": 0.15098942071199417,
      "num_tokens": 5388884.0,
      "step": 3115
    },
    {
      "entropy": 5.897061204910278,
      "epoch": 0.24275432795176036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499693479912315,
      "loss": 5.6853,
      "mean_token_accuracy": 0.15718052834272384,
      "num_tokens": 5397980.0,
      "step": 3120
    },
    {
      "entropy": 5.859615325927734,
      "epoch": 0.24314335732347792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996920320043666,
      "loss": 5.7922,
      "mean_token_accuracy": 0.14827542155981063,
      "num_tokens": 5406879.0,
      "step": 3125
    },
    {
      "entropy": 5.936743116378784,
      "epoch": 0.24353238669519547,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996905806870742,
      "loss": 5.7959,
      "mean_token_accuracy": 0.14135870039463044,
      "num_tokens": 5416161.0,
      "step": 3130
    },
    {
      "entropy": 5.887562036514282,
      "epoch": 0.24392141606691306,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996891259604598,
      "loss": 5.6741,
      "mean_token_accuracy": 0.15898128598928452,
      "num_tokens": 5424323.0,
      "step": 3135
    },
    {
      "entropy": 5.883194923400879,
      "epoch": 0.24431044543863062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996876678245455,
      "loss": 5.8032,
      "mean_token_accuracy": 0.14068420529365538,
      "num_tokens": 5433750.0,
      "step": 3140
    },
    {
      "entropy": 5.986171770095825,
      "epoch": 0.24469947481034818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996862062793536,
      "loss": 5.821,
      "mean_token_accuracy": 0.14756249487400055,
      "num_tokens": 5442113.0,
      "step": 3145
    },
    {
      "entropy": 5.898403453826904,
      "epoch": 0.24508850418206574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996847413249061,
      "loss": 5.6867,
      "mean_token_accuracy": 0.1511594235897064,
      "num_tokens": 5450403.0,
      "step": 3150
    },
    {
      "entropy": 5.896078729629517,
      "epoch": 0.24547753355378332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996832729612252,
      "loss": 5.7644,
      "mean_token_accuracy": 0.14809978604316712,
      "num_tokens": 5459018.0,
      "step": 3155
    },
    {
      "entropy": 5.9197851657867435,
      "epoch": 0.24586656292550088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996818011883333,
      "loss": 5.7983,
      "mean_token_accuracy": 0.14402209520339965,
      "num_tokens": 5468314.0,
      "step": 3160
    },
    {
      "entropy": 5.9504999160766605,
      "epoch": 0.24625559229721844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996803260062527,
      "loss": 5.7793,
      "mean_token_accuracy": 0.1457969695329666,
      "num_tokens": 5476950.0,
      "step": 3165
    },
    {
      "entropy": 5.940018224716186,
      "epoch": 0.246644621668936,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996788474150057,
      "loss": 5.7464,
      "mean_token_accuracy": 0.1534033253788948,
      "num_tokens": 5485201.0,
      "step": 3170
    },
    {
      "entropy": 6.006494426727295,
      "epoch": 0.24703365104065356,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996773654146149,
      "loss": 5.8628,
      "mean_token_accuracy": 0.14921573102474212,
      "num_tokens": 5493096.0,
      "step": 3175
    },
    {
      "entropy": 5.871295690536499,
      "epoch": 0.24742268041237114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996758800051026,
      "loss": 5.6369,
      "mean_token_accuracy": 0.15377923771739005,
      "num_tokens": 5502413.0,
      "step": 3180
    },
    {
      "entropy": 5.906892967224121,
      "epoch": 0.2478117097840887,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004996743911864916,
      "loss": 5.781,
      "mean_token_accuracy": 0.15396690219640732,
      "num_tokens": 5511670.0,
      "step": 3185
    },
    {
      "entropy": 5.927646780014038,
      "epoch": 0.24820073915580626,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996728989588041,
      "loss": 5.6606,
      "mean_token_accuracy": 0.15735298097133638,
      "num_tokens": 5519344.0,
      "step": 3190
    },
    {
      "entropy": 5.830929899215699,
      "epoch": 0.24858976852752382,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996714033220632,
      "loss": 5.6786,
      "mean_token_accuracy": 0.14968922585248948,
      "num_tokens": 5528413.0,
      "step": 3195
    },
    {
      "entropy": 5.8730100154876705,
      "epoch": 0.2489787978992414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996699042762911,
      "loss": 5.7738,
      "mean_token_accuracy": 0.14690972194075586,
      "num_tokens": 5537201.0,
      "step": 3200
    },
    {
      "entropy": 5.923853874206543,
      "epoch": 0.24936782727095896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996684018215111,
      "loss": 5.7804,
      "mean_token_accuracy": 0.14298839941620828,
      "num_tokens": 5545975.0,
      "step": 3205
    },
    {
      "entropy": 5.8805577754974365,
      "epoch": 0.24975685664267652,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996668959577453,
      "loss": 5.8066,
      "mean_token_accuracy": 0.1442289024591446,
      "num_tokens": 5555237.0,
      "step": 3210
    },
    {
      "entropy": 5.989409494400024,
      "epoch": 0.2501458860143941,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996653866850173,
      "loss": 5.6911,
      "mean_token_accuracy": 0.15224729627370834,
      "num_tokens": 5564048.0,
      "step": 3215
    },
    {
      "entropy": 5.877629709243775,
      "epoch": 0.25053491538611167,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996638740033495,
      "loss": 5.7427,
      "mean_token_accuracy": 0.154438579082489,
      "num_tokens": 5573615.0,
      "step": 3220
    },
    {
      "entropy": 5.876510715484619,
      "epoch": 0.2509239447578292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996623579127651,
      "loss": 5.6247,
      "mean_token_accuracy": 0.15758016407489778,
      "num_tokens": 5582211.0,
      "step": 3225
    },
    {
      "entropy": 5.906939220428467,
      "epoch": 0.2513129741295468,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996608384132868,
      "loss": 5.8569,
      "mean_token_accuracy": 0.14443435817956923,
      "num_tokens": 5592542.0,
      "step": 3230
    },
    {
      "entropy": 5.988466691970825,
      "epoch": 0.25170200350126437,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499659315504938,
      "loss": 5.8257,
      "mean_token_accuracy": 0.15278731137514115,
      "num_tokens": 5601279.0,
      "step": 3235
    },
    {
      "entropy": 5.927760648727417,
      "epoch": 0.2520910328729819,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004996577891877417,
      "loss": 5.7578,
      "mean_token_accuracy": 0.1475792944431305,
      "num_tokens": 5610266.0,
      "step": 3240
    },
    {
      "entropy": 5.8933803081512455,
      "epoch": 0.2524800622446995,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996562594617209,
      "loss": 5.8036,
      "mean_token_accuracy": 0.14620354622602463,
      "num_tokens": 5618799.0,
      "step": 3245
    },
    {
      "entropy": 6.05321683883667,
      "epoch": 0.252869091616417,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996547263268991,
      "loss": 5.7447,
      "mean_token_accuracy": 0.15469980388879775,
      "num_tokens": 5627861.0,
      "step": 3250
    },
    {
      "entropy": 5.931217432022095,
      "epoch": 0.2532581209881346,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004996531897832993,
      "loss": 5.7375,
      "mean_token_accuracy": 0.14861182644963264,
      "num_tokens": 5636229.0,
      "step": 3255
    },
    {
      "entropy": 5.938875341415406,
      "epoch": 0.2536471503598522,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499651649830945,
      "loss": 5.9558,
      "mean_token_accuracy": 0.14124528542160988,
      "num_tokens": 5645828.0,
      "step": 3260
    },
    {
      "entropy": 5.8967883586883545,
      "epoch": 0.2540361797315697,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996501064698593,
      "loss": 5.5671,
      "mean_token_accuracy": 0.15032654106616974,
      "num_tokens": 5654750.0,
      "step": 3265
    },
    {
      "entropy": 5.8799279689788815,
      "epoch": 0.2544252091032873,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499648559700066,
      "loss": 5.7448,
      "mean_token_accuracy": 0.14809657484292985,
      "num_tokens": 5664017.0,
      "step": 3270
    },
    {
      "entropy": 5.874881553649902,
      "epoch": 0.25481423847500484,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996470095215884,
      "loss": 5.7406,
      "mean_token_accuracy": 0.14475095346570016,
      "num_tokens": 5672220.0,
      "step": 3275
    },
    {
      "entropy": 5.830359411239624,
      "epoch": 0.2552032678467224,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996454559344498,
      "loss": 5.6864,
      "mean_token_accuracy": 0.1500364437699318,
      "num_tokens": 5680921.0,
      "step": 3280
    },
    {
      "entropy": 5.955457019805908,
      "epoch": 0.25559229721844,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996438989386741,
      "loss": 5.6874,
      "mean_token_accuracy": 0.1482978232204914,
      "num_tokens": 5689226.0,
      "step": 3285
    },
    {
      "entropy": 5.929180049896241,
      "epoch": 0.25598132659015754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996423385342847,
      "loss": 5.8345,
      "mean_token_accuracy": 0.143748040497303,
      "num_tokens": 5698413.0,
      "step": 3290
    },
    {
      "entropy": 5.851541137695312,
      "epoch": 0.25637035596187513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996407747213055,
      "loss": 5.718,
      "mean_token_accuracy": 0.14506666213274003,
      "num_tokens": 5707724.0,
      "step": 3295
    },
    {
      "entropy": 5.939691019058228,
      "epoch": 0.25675938533359266,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049963920749976,
      "loss": 5.7271,
      "mean_token_accuracy": 0.14950018227100373,
      "num_tokens": 5716566.0,
      "step": 3300
    },
    {
      "entropy": 5.84193868637085,
      "epoch": 0.25714841470531025,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996376368696721,
      "loss": 5.6666,
      "mean_token_accuracy": 0.1517062321305275,
      "num_tokens": 5725463.0,
      "step": 3305
    },
    {
      "entropy": 5.895420980453491,
      "epoch": 0.25753744407702783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996360628310656,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1558502957224846,
      "num_tokens": 5734143.0,
      "step": 3310
    },
    {
      "entropy": 5.89042387008667,
      "epoch": 0.25792647344874536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996344853839644,
      "loss": 5.7478,
      "mean_token_accuracy": 0.15216899216175078,
      "num_tokens": 5742981.0,
      "step": 3315
    },
    {
      "entropy": 5.7684619426727295,
      "epoch": 0.25831550282046295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996329045283924,
      "loss": 5.625,
      "mean_token_accuracy": 0.1628153458237648,
      "num_tokens": 5751294.0,
      "step": 3320
    },
    {
      "entropy": 5.869237899780273,
      "epoch": 0.25870453219218054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996313202643737,
      "loss": 5.7406,
      "mean_token_accuracy": 0.1526793658733368,
      "num_tokens": 5760958.0,
      "step": 3325
    },
    {
      "entropy": 5.935721826553345,
      "epoch": 0.25909356156389807,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996297325919321,
      "loss": 5.7595,
      "mean_token_accuracy": 0.1518444985151291,
      "num_tokens": 5769077.0,
      "step": 3330
    },
    {
      "entropy": 5.96410346031189,
      "epoch": 0.25948259093561565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996281415110919,
      "loss": 5.7934,
      "mean_token_accuracy": 0.14791057929396628,
      "num_tokens": 5778093.0,
      "step": 3335
    },
    {
      "entropy": 5.904092693328858,
      "epoch": 0.2598716203073332,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996265470218773,
      "loss": 5.7365,
      "mean_token_accuracy": 0.15056012198328972,
      "num_tokens": 5786996.0,
      "step": 3340
    },
    {
      "entropy": 5.893521738052368,
      "epoch": 0.26026064967905077,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996249491243122,
      "loss": 5.7462,
      "mean_token_accuracy": 0.1535051003098488,
      "num_tokens": 5794831.0,
      "step": 3345
    },
    {
      "entropy": 5.789423561096191,
      "epoch": 0.26064967905076836,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996233478184211,
      "loss": 5.5761,
      "mean_token_accuracy": 0.15884069055318834,
      "num_tokens": 5802286.0,
      "step": 3350
    },
    {
      "entropy": 5.788263034820557,
      "epoch": 0.2610387084224859,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996217431042282,
      "loss": 5.7195,
      "mean_token_accuracy": 0.15141943246126174,
      "num_tokens": 5810985.0,
      "step": 3355
    },
    {
      "entropy": 5.955795001983643,
      "epoch": 0.2614277377942035,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499620134981758,
      "loss": 5.7707,
      "mean_token_accuracy": 0.15107223987579346,
      "num_tokens": 5820037.0,
      "step": 3360
    },
    {
      "entropy": 5.795082426071167,
      "epoch": 0.261816767165921,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996185234510346,
      "loss": 5.6242,
      "mean_token_accuracy": 0.16446445733308793,
      "num_tokens": 5828232.0,
      "step": 3365
    },
    {
      "entropy": 5.909881448745727,
      "epoch": 0.2622057965376386,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996169085120828,
      "loss": 5.8446,
      "mean_token_accuracy": 0.15064912438392639,
      "num_tokens": 5838184.0,
      "step": 3370
    },
    {
      "entropy": 5.951297903060913,
      "epoch": 0.2625948259093562,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996152901649268,
      "loss": 5.6708,
      "mean_token_accuracy": 0.15528088957071304,
      "num_tokens": 5846705.0,
      "step": 3375
    },
    {
      "entropy": 5.891704511642456,
      "epoch": 0.2629838552810737,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996136684095913,
      "loss": 5.7557,
      "mean_token_accuracy": 0.14960285723209382,
      "num_tokens": 5855959.0,
      "step": 3380
    },
    {
      "entropy": 5.8450446128845215,
      "epoch": 0.2633728846527913,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996120432461009,
      "loss": 5.7099,
      "mean_token_accuracy": 0.15203075855970383,
      "num_tokens": 5865283.0,
      "step": 3385
    },
    {
      "entropy": 5.913759851455689,
      "epoch": 0.2637619140245088,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996104146744804,
      "loss": 5.7391,
      "mean_token_accuracy": 0.1565643608570099,
      "num_tokens": 5873833.0,
      "step": 3390
    },
    {
      "entropy": 5.948887634277344,
      "epoch": 0.2641509433962264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996087826947541,
      "loss": 5.6684,
      "mean_token_accuracy": 0.16087514907121658,
      "num_tokens": 5882027.0,
      "step": 3395
    },
    {
      "entropy": 5.710987138748169,
      "epoch": 0.264539972767944,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996071473069471,
      "loss": 5.548,
      "mean_token_accuracy": 0.16150110214948654,
      "num_tokens": 5890650.0,
      "step": 3400
    },
    {
      "entropy": 5.778266859054566,
      "epoch": 0.26492900213966153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996055085110842,
      "loss": 5.6004,
      "mean_token_accuracy": 0.16101501286029815,
      "num_tokens": 5898595.0,
      "step": 3405
    },
    {
      "entropy": 5.916549825668335,
      "epoch": 0.2653180315113791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996038663071902,
      "loss": 5.7157,
      "mean_token_accuracy": 0.1523623764514923,
      "num_tokens": 5907234.0,
      "step": 3410
    },
    {
      "entropy": 5.877885484695435,
      "epoch": 0.26570706088309665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996022206952901,
      "loss": 5.6562,
      "mean_token_accuracy": 0.15936153680086135,
      "num_tokens": 5915690.0,
      "step": 3415
    },
    {
      "entropy": 5.798541498184204,
      "epoch": 0.26609609025481423,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996005716754086,
      "loss": 5.6053,
      "mean_token_accuracy": 0.16079822927713394,
      "num_tokens": 5924481.0,
      "step": 3420
    },
    {
      "entropy": 5.808663415908813,
      "epoch": 0.2664851196265318,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499598919247571,
      "loss": 5.6951,
      "mean_token_accuracy": 0.1579988867044449,
      "num_tokens": 5933059.0,
      "step": 3425
    },
    {
      "entropy": 5.887850189208985,
      "epoch": 0.26687414899824935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995972634118023,
      "loss": 5.8016,
      "mean_token_accuracy": 0.14508845061063766,
      "num_tokens": 5941767.0,
      "step": 3430
    },
    {
      "entropy": 5.844564151763916,
      "epoch": 0.26726317836996694,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995956041681277,
      "loss": 5.691,
      "mean_token_accuracy": 0.15415164232254028,
      "num_tokens": 5950202.0,
      "step": 3435
    },
    {
      "entropy": 5.848789262771606,
      "epoch": 0.2676522077416845,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995939415165722,
      "loss": 5.7359,
      "mean_token_accuracy": 0.14812117516994477,
      "num_tokens": 5959242.0,
      "step": 3440
    },
    {
      "entropy": 5.846918964385987,
      "epoch": 0.26804123711340205,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995922754571611,
      "loss": 5.6836,
      "mean_token_accuracy": 0.15515549182891847,
      "num_tokens": 5968404.0,
      "step": 3445
    },
    {
      "entropy": 5.860494327545166,
      "epoch": 0.26843026648511964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995906059899197,
      "loss": 5.8005,
      "mean_token_accuracy": 0.15509908199310302,
      "num_tokens": 5976711.0,
      "step": 3450
    },
    {
      "entropy": 5.894256687164306,
      "epoch": 0.26881929585683717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995889331148733,
      "loss": 5.6551,
      "mean_token_accuracy": 0.14889176189899445,
      "num_tokens": 5985049.0,
      "step": 3455
    },
    {
      "entropy": 5.8000787734985355,
      "epoch": 0.26920832522855476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995872568320474,
      "loss": 5.6823,
      "mean_token_accuracy": 0.15802373141050338,
      "num_tokens": 5993904.0,
      "step": 3460
    },
    {
      "entropy": 5.915686511993409,
      "epoch": 0.26959735460027234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995855771414673,
      "loss": 5.7404,
      "mean_token_accuracy": 0.15314059257507323,
      "num_tokens": 6001832.0,
      "step": 3465
    },
    {
      "entropy": 5.825476264953613,
      "epoch": 0.2699863839719899,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995838940431585,
      "loss": 5.6599,
      "mean_token_accuracy": 0.1504707530140877,
      "num_tokens": 6010609.0,
      "step": 3470
    },
    {
      "entropy": 5.851345634460449,
      "epoch": 0.27037541334370746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995822075371466,
      "loss": 5.7193,
      "mean_token_accuracy": 0.15108724534511567,
      "num_tokens": 6019299.0,
      "step": 3475
    },
    {
      "entropy": 5.943111848831177,
      "epoch": 0.270764442715425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995805176234571,
      "loss": 5.7676,
      "mean_token_accuracy": 0.14575971737504007,
      "num_tokens": 6028743.0,
      "step": 3480
    },
    {
      "entropy": 5.879656744003296,
      "epoch": 0.2711534720871426,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995788243021158,
      "loss": 5.6713,
      "mean_token_accuracy": 0.14724702388048172,
      "num_tokens": 6038064.0,
      "step": 3485
    },
    {
      "entropy": 5.793353033065796,
      "epoch": 0.27154250145886016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995771275731483,
      "loss": 5.6612,
      "mean_token_accuracy": 0.1524479627609253,
      "num_tokens": 6047031.0,
      "step": 3490
    },
    {
      "entropy": 5.889426136016846,
      "epoch": 0.2719315308305777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995754274365802,
      "loss": 5.7498,
      "mean_token_accuracy": 0.15119700282812118,
      "num_tokens": 6056147.0,
      "step": 3495
    },
    {
      "entropy": 5.890491199493408,
      "epoch": 0.2723205602022953,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995737238924376,
      "loss": 5.6689,
      "mean_token_accuracy": 0.15356390327215194,
      "num_tokens": 6064641.0,
      "step": 3500
    },
    {
      "entropy": 5.796261215209961,
      "epoch": 0.2727095895740128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995720169407461,
      "loss": 5.5947,
      "mean_token_accuracy": 0.15325731337070464,
      "num_tokens": 6073664.0,
      "step": 3505
    },
    {
      "entropy": 5.811582803726196,
      "epoch": 0.2730986189457304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995703065815317,
      "loss": 5.5752,
      "mean_token_accuracy": 0.16186247318983077,
      "num_tokens": 6081480.0,
      "step": 3510
    },
    {
      "entropy": 5.8565483570098875,
      "epoch": 0.273487648317448,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995685928148203,
      "loss": 5.7021,
      "mean_token_accuracy": 0.14859129190444947,
      "num_tokens": 6090408.0,
      "step": 3515
    },
    {
      "entropy": 5.866174983978271,
      "epoch": 0.2738766776891655,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499566875640638,
      "loss": 5.699,
      "mean_token_accuracy": 0.15143164694309236,
      "num_tokens": 6099076.0,
      "step": 3520
    },
    {
      "entropy": 5.874391031265259,
      "epoch": 0.2742657070608831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995651550590108,
      "loss": 5.7466,
      "mean_token_accuracy": 0.14636368602514266,
      "num_tokens": 6108346.0,
      "step": 3525
    },
    {
      "entropy": 5.837771892547607,
      "epoch": 0.2746547364326007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995634310699647,
      "loss": 5.7037,
      "mean_token_accuracy": 0.15670324489474297,
      "num_tokens": 6116974.0,
      "step": 3530
    },
    {
      "entropy": 5.913712024688721,
      "epoch": 0.2750437658043182,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995617036735261,
      "loss": 5.7323,
      "mean_token_accuracy": 0.1500618875026703,
      "num_tokens": 6125510.0,
      "step": 3535
    },
    {
      "entropy": 5.865264272689819,
      "epoch": 0.2754327951760358,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995599728697211,
      "loss": 5.7712,
      "mean_token_accuracy": 0.14675354063510895,
      "num_tokens": 6134959.0,
      "step": 3540
    },
    {
      "entropy": 5.930070304870606,
      "epoch": 0.27582182454775334,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995582386585759,
      "loss": 5.7049,
      "mean_token_accuracy": 0.145357196778059,
      "num_tokens": 6144485.0,
      "step": 3545
    },
    {
      "entropy": 5.86058931350708,
      "epoch": 0.2762108539194709,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995565010401167,
      "loss": 5.6487,
      "mean_token_accuracy": 0.15821592509746552,
      "num_tokens": 6153991.0,
      "step": 3550
    },
    {
      "entropy": 5.739012575149536,
      "epoch": 0.2765998832911885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995547600143702,
      "loss": 5.6425,
      "mean_token_accuracy": 0.1544785276055336,
      "num_tokens": 6162649.0,
      "step": 3555
    },
    {
      "entropy": 5.813714075088501,
      "epoch": 0.27698891266290604,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995530155813625,
      "loss": 5.5463,
      "mean_token_accuracy": 0.15806780904531478,
      "num_tokens": 6171525.0,
      "step": 3560
    },
    {
      "entropy": 5.872299385070801,
      "epoch": 0.2773779420346236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995512677411203,
      "loss": 5.7726,
      "mean_token_accuracy": 0.15298861116170884,
      "num_tokens": 6180403.0,
      "step": 3565
    },
    {
      "entropy": 5.834949064254761,
      "epoch": 0.27776697140634116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995495164936698,
      "loss": 5.6142,
      "mean_token_accuracy": 0.15946034938097,
      "num_tokens": 6188361.0,
      "step": 3570
    },
    {
      "entropy": 5.807054090499878,
      "epoch": 0.27815600077805874,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995477618390381,
      "loss": 5.6716,
      "mean_token_accuracy": 0.14874202236533166,
      "num_tokens": 6197418.0,
      "step": 3575
    },
    {
      "entropy": 5.810306215286255,
      "epoch": 0.27854503014977633,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995460037772513,
      "loss": 5.6813,
      "mean_token_accuracy": 0.15601591318845748,
      "num_tokens": 6206040.0,
      "step": 3580
    },
    {
      "entropy": 5.79304084777832,
      "epoch": 0.27893405952149386,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995442423083365,
      "loss": 5.5968,
      "mean_token_accuracy": 0.15546937584877013,
      "num_tokens": 6214433.0,
      "step": 3585
    },
    {
      "entropy": 5.769221782684326,
      "epoch": 0.27932308889321145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995424774323201,
      "loss": 5.5765,
      "mean_token_accuracy": 0.15947029739618301,
      "num_tokens": 6223375.0,
      "step": 3590
    },
    {
      "entropy": 5.744332695007325,
      "epoch": 0.279712118264929,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499540709149229,
      "loss": 5.6898,
      "mean_token_accuracy": 0.15379903316497803,
      "num_tokens": 6232597.0,
      "step": 3595
    },
    {
      "entropy": 5.785550165176391,
      "epoch": 0.28010114763664656,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995389374590901,
      "loss": 5.5429,
      "mean_token_accuracy": 0.16784969419240953,
      "num_tokens": 6240634.0,
      "step": 3600
    },
    {
      "entropy": 5.780906867980957,
      "epoch": 0.28049017700836415,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995371623619301,
      "loss": 5.5015,
      "mean_token_accuracy": 0.15682514160871505,
      "num_tokens": 6249098.0,
      "step": 3605
    },
    {
      "entropy": 5.739667129516602,
      "epoch": 0.2808792063800817,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499535383857776,
      "loss": 5.5472,
      "mean_token_accuracy": 0.16006743013858796,
      "num_tokens": 6257572.0,
      "step": 3610
    },
    {
      "entropy": 5.859043931961059,
      "epoch": 0.28126823575179927,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499533601946655,
      "loss": 5.8165,
      "mean_token_accuracy": 0.14751632660627365,
      "num_tokens": 6266492.0,
      "step": 3615
    },
    {
      "entropy": 5.865448904037476,
      "epoch": 0.2816572651235168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995318166285938,
      "loss": 5.6829,
      "mean_token_accuracy": 0.155305153131485,
      "num_tokens": 6275149.0,
      "step": 3620
    },
    {
      "entropy": 5.874626731872558,
      "epoch": 0.2820462944952344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995300279036199,
      "loss": 5.6956,
      "mean_token_accuracy": 0.15312935113906861,
      "num_tokens": 6283208.0,
      "step": 3625
    },
    {
      "entropy": 5.855358982086182,
      "epoch": 0.28243532386695197,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049952823577176,
      "loss": 5.6347,
      "mean_token_accuracy": 0.15589332580566406,
      "num_tokens": 6292084.0,
      "step": 3630
    },
    {
      "entropy": 5.784860134124756,
      "epoch": 0.2828243532386695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995264402330416,
      "loss": 5.6336,
      "mean_token_accuracy": 0.1528566986322403,
      "num_tokens": 6301855.0,
      "step": 3635
    },
    {
      "entropy": 5.845756483078003,
      "epoch": 0.2832133826103871,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995246412874919,
      "loss": 5.6379,
      "mean_token_accuracy": 0.15769062787294388,
      "num_tokens": 6310166.0,
      "step": 3640
    },
    {
      "entropy": 5.783097553253174,
      "epoch": 0.2836024119821047,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499522838935138,
      "loss": 5.6228,
      "mean_token_accuracy": 0.16248167604207991,
      "num_tokens": 6318928.0,
      "step": 3645
    },
    {
      "entropy": 5.826745843887329,
      "epoch": 0.2839914413538222,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995210331760074,
      "loss": 5.6509,
      "mean_token_accuracy": 0.1659294068813324,
      "num_tokens": 6327790.0,
      "step": 3650
    },
    {
      "entropy": 5.85683012008667,
      "epoch": 0.2843804707255398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995192240101275,
      "loss": 5.7141,
      "mean_token_accuracy": 0.1522120863199234,
      "num_tokens": 6335517.0,
      "step": 3655
    },
    {
      "entropy": 5.789180564880371,
      "epoch": 0.2847695000972573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995174114375258,
      "loss": 5.7049,
      "mean_token_accuracy": 0.15419832319021226,
      "num_tokens": 6344218.0,
      "step": 3660
    },
    {
      "entropy": 5.769765949249267,
      "epoch": 0.2851585294689749,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995155954582297,
      "loss": 5.5136,
      "mean_token_accuracy": 0.16411114037036895,
      "num_tokens": 6353125.0,
      "step": 3665
    },
    {
      "entropy": 5.8027407169342045,
      "epoch": 0.2855475588406925,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995137760722668,
      "loss": 5.6528,
      "mean_token_accuracy": 0.1551785096526146,
      "num_tokens": 6362113.0,
      "step": 3670
    },
    {
      "entropy": 5.844509744644165,
      "epoch": 0.28593658821241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995119532796646,
      "loss": 5.7323,
      "mean_token_accuracy": 0.15433150082826613,
      "num_tokens": 6370916.0,
      "step": 3675
    },
    {
      "entropy": 5.866036510467529,
      "epoch": 0.2863256175841276,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499510127080451,
      "loss": 5.5915,
      "mean_token_accuracy": 0.1586878553032875,
      "num_tokens": 6379256.0,
      "step": 3680
    },
    {
      "entropy": 5.825314474105835,
      "epoch": 0.28671464695584514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995082974746535,
      "loss": 5.723,
      "mean_token_accuracy": 0.15312020778656005,
      "num_tokens": 6388098.0,
      "step": 3685
    },
    {
      "entropy": 5.7031796932220455,
      "epoch": 0.28710367632756273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995064644622999,
      "loss": 5.5249,
      "mean_token_accuracy": 0.16168365329504014,
      "num_tokens": 6396292.0,
      "step": 3690
    },
    {
      "entropy": 5.792140960693359,
      "epoch": 0.2874927056992803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995046280434181,
      "loss": 5.5502,
      "mean_token_accuracy": 0.16206852197647095,
      "num_tokens": 6404968.0,
      "step": 3695
    },
    {
      "entropy": 5.691653776168823,
      "epoch": 0.28788173507099785,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995027882180358,
      "loss": 5.6329,
      "mean_token_accuracy": 0.15900261849164962,
      "num_tokens": 6412839.0,
      "step": 3700
    },
    {
      "entropy": 5.784019470214844,
      "epoch": 0.28827076444271543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995009449861812,
      "loss": 5.6493,
      "mean_token_accuracy": 0.14958156421780586,
      "num_tokens": 6421943.0,
      "step": 3705
    },
    {
      "entropy": 5.855062103271484,
      "epoch": 0.28865979381443296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499499098347882,
      "loss": 5.7285,
      "mean_token_accuracy": 0.14923914447426795,
      "num_tokens": 6430731.0,
      "step": 3710
    },
    {
      "entropy": 5.851906061172485,
      "epoch": 0.28904882318615055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994972483031662,
      "loss": 5.6556,
      "mean_token_accuracy": 0.1556018888950348,
      "num_tokens": 6439365.0,
      "step": 3715
    },
    {
      "entropy": 5.8089118003845215,
      "epoch": 0.28943785255786814,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499495394852062,
      "loss": 5.6021,
      "mean_token_accuracy": 0.16112566590309144,
      "num_tokens": 6448085.0,
      "step": 3720
    },
    {
      "entropy": 5.684608078002929,
      "epoch": 0.28982688192958567,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994935379945977,
      "loss": 5.6652,
      "mean_token_accuracy": 0.14997439682483674,
      "num_tokens": 6457630.0,
      "step": 3725
    },
    {
      "entropy": 5.838469409942627,
      "epoch": 0.29021591130130325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004994916777308012,
      "loss": 5.5251,
      "mean_token_accuracy": 0.1629138931632042,
      "num_tokens": 6465781.0,
      "step": 3730
    },
    {
      "entropy": 5.844767808914185,
      "epoch": 0.29060494067302084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994898140607007,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1529698833823204,
      "num_tokens": 6473723.0,
      "step": 3735
    },
    {
      "entropy": 5.711420583724975,
      "epoch": 0.29099397004473837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994879469843247,
      "loss": 5.6375,
      "mean_token_accuracy": 0.1572704315185547,
      "num_tokens": 6482535.0,
      "step": 3740
    },
    {
      "entropy": 5.8504077911376955,
      "epoch": 0.29138299941645596,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994860765017013,
      "loss": 5.6747,
      "mean_token_accuracy": 0.1574747622013092,
      "num_tokens": 6491297.0,
      "step": 3745
    },
    {
      "entropy": 5.773846578598023,
      "epoch": 0.2917720287881735,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499484202612859,
      "loss": 5.6295,
      "mean_token_accuracy": 0.15734697729349137,
      "num_tokens": 6499809.0,
      "step": 3750
    },
    {
      "entropy": 5.7556627750396725,
      "epoch": 0.2921610581598911,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994823253178264,
      "loss": 5.6075,
      "mean_token_accuracy": 0.16184348315000535,
      "num_tokens": 6507965.0,
      "step": 3755
    },
    {
      "entropy": 5.852641582489014,
      "epoch": 0.29255008753160866,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004994804446166317,
      "loss": 5.6552,
      "mean_token_accuracy": 0.15120026841759682,
      "num_tokens": 6517106.0,
      "step": 3760
    },
    {
      "entropy": 5.845433378219605,
      "epoch": 0.2929391169033262,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994785605093035,
      "loss": 5.7337,
      "mean_token_accuracy": 0.15373013466596602,
      "num_tokens": 6526663.0,
      "step": 3765
    },
    {
      "entropy": 5.927438497543335,
      "epoch": 0.2933281462750438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994766729958705,
      "loss": 5.7739,
      "mean_token_accuracy": 0.1520956665277481,
      "num_tokens": 6535732.0,
      "step": 3770
    },
    {
      "entropy": 5.808690547943115,
      "epoch": 0.2937171756467613,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994747820763612,
      "loss": 5.6301,
      "mean_token_accuracy": 0.15487213283777237,
      "num_tokens": 6544938.0,
      "step": 3775
    },
    {
      "entropy": 5.802787160873413,
      "epoch": 0.2941062050184789,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994728877508046,
      "loss": 5.6336,
      "mean_token_accuracy": 0.1528394863009453,
      "num_tokens": 6554507.0,
      "step": 3780
    },
    {
      "entropy": 5.839257526397705,
      "epoch": 0.2944952343901965,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994709900192289,
      "loss": 5.6781,
      "mean_token_accuracy": 0.15294660478830338,
      "num_tokens": 6563285.0,
      "step": 3785
    },
    {
      "entropy": 5.752003622055054,
      "epoch": 0.294884263761914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994690888816635,
      "loss": 5.56,
      "mean_token_accuracy": 0.15720225274562835,
      "num_tokens": 6572223.0,
      "step": 3790
    },
    {
      "entropy": 5.823800420761108,
      "epoch": 0.2952732931336316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994671843381367,
      "loss": 5.7438,
      "mean_token_accuracy": 0.14936941713094712,
      "num_tokens": 6580686.0,
      "step": 3795
    },
    {
      "entropy": 5.855613756179809,
      "epoch": 0.29566232250534913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994652763886778,
      "loss": 5.6053,
      "mean_token_accuracy": 0.1560147821903229,
      "num_tokens": 6589056.0,
      "step": 3800
    },
    {
      "entropy": 5.8202354431152346,
      "epoch": 0.2960513518770667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994633650333155,
      "loss": 5.6695,
      "mean_token_accuracy": 0.15734007209539413,
      "num_tokens": 6598221.0,
      "step": 3805
    },
    {
      "entropy": 5.879884719848633,
      "epoch": 0.2964403812487843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994614502720792,
      "loss": 5.72,
      "mean_token_accuracy": 0.15723774582147598,
      "num_tokens": 6606657.0,
      "step": 3810
    },
    {
      "entropy": 5.883637285232544,
      "epoch": 0.29682941062050183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994595321049974,
      "loss": 5.8057,
      "mean_token_accuracy": 0.14732917174696922,
      "num_tokens": 6616203.0,
      "step": 3815
    },
    {
      "entropy": 5.904540967941284,
      "epoch": 0.2972184399922194,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994576105320994,
      "loss": 5.6617,
      "mean_token_accuracy": 0.15442878901958465,
      "num_tokens": 6624982.0,
      "step": 3820
    },
    {
      "entropy": 5.843383693695069,
      "epoch": 0.29760746936393695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994556855534146,
      "loss": 5.6234,
      "mean_token_accuracy": 0.16022102981805803,
      "num_tokens": 6632548.0,
      "step": 3825
    },
    {
      "entropy": 5.7950318336486815,
      "epoch": 0.29799649873565454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994537571689719,
      "loss": 5.7647,
      "mean_token_accuracy": 0.14879212528467178,
      "num_tokens": 6641943.0,
      "step": 3830
    },
    {
      "entropy": 5.838181400299073,
      "epoch": 0.2983855281073721,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994518253788007,
      "loss": 5.6,
      "mean_token_accuracy": 0.16140060871839523,
      "num_tokens": 6650371.0,
      "step": 3835
    },
    {
      "entropy": 5.819985818862915,
      "epoch": 0.29877455747908965,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994498901829304,
      "loss": 5.6843,
      "mean_token_accuracy": 0.15791809260845185,
      "num_tokens": 6659324.0,
      "step": 3840
    },
    {
      "entropy": 5.810818195343018,
      "epoch": 0.29916358685080724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994479515813903,
      "loss": 5.6492,
      "mean_token_accuracy": 0.1515626162290573,
      "num_tokens": 6668237.0,
      "step": 3845
    },
    {
      "entropy": 5.808945417404175,
      "epoch": 0.2995526162225248,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994460095742096,
      "loss": 5.5772,
      "mean_token_accuracy": 0.16233526021242142,
      "num_tokens": 6677213.0,
      "step": 3850
    },
    {
      "entropy": 5.7874250411987305,
      "epoch": 0.29994164559424236,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499444064161418,
      "loss": 5.587,
      "mean_token_accuracy": 0.15759442001581192,
      "num_tokens": 6686850.0,
      "step": 3855
    },
    {
      "entropy": 5.77463321685791,
      "epoch": 0.30033067496595994,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499442115343045,
      "loss": 5.5608,
      "mean_token_accuracy": 0.16466309428215026,
      "num_tokens": 6695697.0,
      "step": 3860
    },
    {
      "entropy": 5.764589834213257,
      "epoch": 0.3007197043376775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994401631191202,
      "loss": 5.6185,
      "mean_token_accuracy": 0.16456811875104904,
      "num_tokens": 6704525.0,
      "step": 3865
    },
    {
      "entropy": 5.749872732162475,
      "epoch": 0.30110873370939506,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994382074896731,
      "loss": 5.555,
      "mean_token_accuracy": 0.1628976732492447,
      "num_tokens": 6713562.0,
      "step": 3870
    },
    {
      "entropy": 5.841018533706665,
      "epoch": 0.30149776308111265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994362484547335,
      "loss": 5.6633,
      "mean_token_accuracy": 0.15924778133630751,
      "num_tokens": 6721817.0,
      "step": 3875
    },
    {
      "entropy": 5.8318479537963865,
      "epoch": 0.3018867924528302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994342860143309,
      "loss": 5.5799,
      "mean_token_accuracy": 0.16334821432828903,
      "num_tokens": 6729757.0,
      "step": 3880
    },
    {
      "entropy": 5.756866598129273,
      "epoch": 0.30227582182454776,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994323201684953,
      "loss": 5.6585,
      "mean_token_accuracy": 0.1475125730037689,
      "num_tokens": 6738354.0,
      "step": 3885
    },
    {
      "entropy": 5.819000101089477,
      "epoch": 0.3026648511962653,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994303509172566,
      "loss": 5.7076,
      "mean_token_accuracy": 0.15129155218601226,
      "num_tokens": 6746505.0,
      "step": 3890
    },
    {
      "entropy": 5.904794931411743,
      "epoch": 0.3030538805679829,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994283782606444,
      "loss": 5.6111,
      "mean_token_accuracy": 0.15827509462833406,
      "num_tokens": 6754547.0,
      "step": 3895
    },
    {
      "entropy": 5.7554058074951175,
      "epoch": 0.30344290993970047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994264021986888,
      "loss": 5.5293,
      "mean_token_accuracy": 0.16018292903900147,
      "num_tokens": 6763213.0,
      "step": 3900
    },
    {
      "entropy": 5.691382646560669,
      "epoch": 0.303831939311418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994244227314197,
      "loss": 5.5787,
      "mean_token_accuracy": 0.15872986018657684,
      "num_tokens": 6771060.0,
      "step": 3905
    },
    {
      "entropy": 5.808207893371582,
      "epoch": 0.3042209686831356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994224398588672,
      "loss": 5.7216,
      "mean_token_accuracy": 0.15398526936769485,
      "num_tokens": 6779819.0,
      "step": 3910
    },
    {
      "entropy": 5.835114002227783,
      "epoch": 0.3046099980548531,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994204535810615,
      "loss": 5.618,
      "mean_token_accuracy": 0.15699677169322968,
      "num_tokens": 6788660.0,
      "step": 3915
    },
    {
      "entropy": 5.723716831207275,
      "epoch": 0.3049990274265707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994184638980326,
      "loss": 5.5708,
      "mean_token_accuracy": 0.16314460784196855,
      "num_tokens": 6797298.0,
      "step": 3920
    },
    {
      "entropy": 5.743786954879761,
      "epoch": 0.3053880567982883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994164708098107,
      "loss": 5.6223,
      "mean_token_accuracy": 0.15216543674468994,
      "num_tokens": 6806148.0,
      "step": 3925
    },
    {
      "entropy": 5.764825057983399,
      "epoch": 0.3057770861700058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499414474316426,
      "loss": 5.6592,
      "mean_token_accuracy": 0.157614828646183,
      "num_tokens": 6815494.0,
      "step": 3930
    },
    {
      "entropy": 5.904352045059204,
      "epoch": 0.3061661155417234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994124744179088,
      "loss": 5.7051,
      "mean_token_accuracy": 0.15148080736398697,
      "num_tokens": 6824852.0,
      "step": 3935
    },
    {
      "entropy": 5.787218236923218,
      "epoch": 0.306555144913441,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994104711142896,
      "loss": 5.5069,
      "mean_token_accuracy": 0.16816734969615937,
      "num_tokens": 6834200.0,
      "step": 3940
    },
    {
      "entropy": 5.774364185333252,
      "epoch": 0.3069441742851585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994084644055986,
      "loss": 5.6567,
      "mean_token_accuracy": 0.15507219433784486,
      "num_tokens": 6842869.0,
      "step": 3945
    },
    {
      "entropy": 5.814112663269043,
      "epoch": 0.3073332036568761,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994064542918664,
      "loss": 5.6178,
      "mean_token_accuracy": 0.1599549800157547,
      "num_tokens": 6851668.0,
      "step": 3950
    },
    {
      "entropy": 5.831446838378906,
      "epoch": 0.30772223302859364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994044407731235,
      "loss": 5.6875,
      "mean_token_accuracy": 0.16353656947612763,
      "num_tokens": 6860710.0,
      "step": 3955
    },
    {
      "entropy": 5.75144510269165,
      "epoch": 0.3081112624003112,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994024238494002,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1537766396999359,
      "num_tokens": 6869608.0,
      "step": 3960
    },
    {
      "entropy": 5.827413606643677,
      "epoch": 0.3085002917720288,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994004035207274,
      "loss": 5.639,
      "mean_token_accuracy": 0.1554042413830757,
      "num_tokens": 6878108.0,
      "step": 3965
    },
    {
      "entropy": 5.779709196090698,
      "epoch": 0.30888932114374634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993983797871356,
      "loss": 5.5448,
      "mean_token_accuracy": 0.15877590030431749,
      "num_tokens": 6886369.0,
      "step": 3970
    },
    {
      "entropy": 5.7670244693756105,
      "epoch": 0.30927835051546393,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993963526486555,
      "loss": 5.607,
      "mean_token_accuracy": 0.15788756608963012,
      "num_tokens": 6895410.0,
      "step": 3975
    },
    {
      "entropy": 5.797284984588623,
      "epoch": 0.30966737988718146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499394322105318,
      "loss": 5.5952,
      "mean_token_accuracy": 0.15986649543046952,
      "num_tokens": 6904380.0,
      "step": 3980
    },
    {
      "entropy": 5.797864580154419,
      "epoch": 0.31005640925889905,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993922881571538,
      "loss": 5.6206,
      "mean_token_accuracy": 0.15683696866035463,
      "num_tokens": 6913012.0,
      "step": 3985
    },
    {
      "entropy": 5.7702795505523685,
      "epoch": 0.31044543863061663,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993902508041939,
      "loss": 5.6387,
      "mean_token_accuracy": 0.15548964440822602,
      "num_tokens": 6921696.0,
      "step": 3990
    },
    {
      "entropy": 5.791179227828979,
      "epoch": 0.31083446800233416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499388210046469,
      "loss": 5.579,
      "mean_token_accuracy": 0.15531976968050004,
      "num_tokens": 6930017.0,
      "step": 3995
    },
    {
      "entropy": 5.709838199615478,
      "epoch": 0.31122349737405175,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993861658840102,
      "loss": 5.5464,
      "mean_token_accuracy": 0.1598093330860138,
      "num_tokens": 6938410.0,
      "step": 4000
    },
    {
      "entropy": 5.782442045211792,
      "epoch": 0.3116125267457693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993841183168484,
      "loss": 5.6157,
      "mean_token_accuracy": 0.15706539303064346,
      "num_tokens": 6946227.0,
      "step": 4005
    },
    {
      "entropy": 5.797301340103149,
      "epoch": 0.31200155611748687,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993820673450148,
      "loss": 5.6096,
      "mean_token_accuracy": 0.15807232558727263,
      "num_tokens": 6954747.0,
      "step": 4010
    },
    {
      "entropy": 5.788407516479492,
      "epoch": 0.31239058548920445,
      "grad_norm": 1.125,
      "learning_rate": 0.0004993800129685404,
      "loss": 5.6804,
      "mean_token_accuracy": 0.15833715796470643,
      "num_tokens": 6962990.0,
      "step": 4015
    },
    {
      "entropy": 5.777428436279297,
      "epoch": 0.312779614860922,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993779551874565,
      "loss": 5.638,
      "mean_token_accuracy": 0.1595762073993683,
      "num_tokens": 6972351.0,
      "step": 4020
    },
    {
      "entropy": 5.789421892166137,
      "epoch": 0.31316864423263957,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993758940017943,
      "loss": 5.5714,
      "mean_token_accuracy": 0.16144972741603852,
      "num_tokens": 6980387.0,
      "step": 4025
    },
    {
      "entropy": 5.658299160003662,
      "epoch": 0.31355767360435716,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499373829411585,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1596735969185829,
      "num_tokens": 6989388.0,
      "step": 4030
    },
    {
      "entropy": 5.87131175994873,
      "epoch": 0.3139467029760747,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049937176141686,
      "loss": 5.5914,
      "mean_token_accuracy": 0.16546436548233032,
      "num_tokens": 6997714.0,
      "step": 4035
    },
    {
      "entropy": 5.819481420516968,
      "epoch": 0.3143357323477923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993696900176506,
      "loss": 5.5748,
      "mean_token_accuracy": 0.16345698684453963,
      "num_tokens": 7006453.0,
      "step": 4040
    },
    {
      "entropy": 5.764643812179566,
      "epoch": 0.3147247617195098,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993676152139883,
      "loss": 5.6514,
      "mean_token_accuracy": 0.15526093393564225,
      "num_tokens": 7015398.0,
      "step": 4045
    },
    {
      "entropy": 5.767642402648926,
      "epoch": 0.3151137910912274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993655370059046,
      "loss": 5.5952,
      "mean_token_accuracy": 0.1578880578279495,
      "num_tokens": 7023796.0,
      "step": 4050
    },
    {
      "entropy": 5.703687334060669,
      "epoch": 0.315502820462945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993634553934309,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1635144382715225,
      "num_tokens": 7032453.0,
      "step": 4055
    },
    {
      "entropy": 5.7386377334594725,
      "epoch": 0.3158918498346625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993613703765989,
      "loss": 5.6841,
      "mean_token_accuracy": 0.15243002027273178,
      "num_tokens": 7041034.0,
      "step": 4060
    },
    {
      "entropy": 5.8850563049316404,
      "epoch": 0.3162808792063801,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993592819554402,
      "loss": 5.7752,
      "mean_token_accuracy": 0.1502375230193138,
      "num_tokens": 7050289.0,
      "step": 4065
    },
    {
      "entropy": 5.781525039672852,
      "epoch": 0.3166699085780976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993571901299866,
      "loss": 5.6123,
      "mean_token_accuracy": 0.16497506499290465,
      "num_tokens": 7059638.0,
      "step": 4070
    },
    {
      "entropy": 5.751022434234619,
      "epoch": 0.3170589379498152,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993550949002697,
      "loss": 5.622,
      "mean_token_accuracy": 0.15959509760141372,
      "num_tokens": 7067763.0,
      "step": 4075
    },
    {
      "entropy": 5.7526812076568605,
      "epoch": 0.3174479673215328,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993529962663213,
      "loss": 5.5854,
      "mean_token_accuracy": 0.16497846245765685,
      "num_tokens": 7076126.0,
      "step": 4080
    },
    {
      "entropy": 5.754148960113525,
      "epoch": 0.31783699669325033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993508942281732,
      "loss": 5.5271,
      "mean_token_accuracy": 0.16019581109285355,
      "num_tokens": 7084403.0,
      "step": 4085
    },
    {
      "entropy": 5.6758214950561525,
      "epoch": 0.3182260260649679,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993487887858575,
      "loss": 5.4989,
      "mean_token_accuracy": 0.1651519536972046,
      "num_tokens": 7093450.0,
      "step": 4090
    },
    {
      "entropy": 5.695490217208862,
      "epoch": 0.31861505543668545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499346679939406,
      "loss": 5.4938,
      "mean_token_accuracy": 0.1639673426747322,
      "num_tokens": 7101722.0,
      "step": 4095
    },
    {
      "entropy": 5.70269193649292,
      "epoch": 0.31900408480840303,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993445676888507,
      "loss": 5.5973,
      "mean_token_accuracy": 0.15606338530778885,
      "num_tokens": 7111285.0,
      "step": 4100
    },
    {
      "entropy": 5.855121707916259,
      "epoch": 0.3193931141801206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993424520342236,
      "loss": 5.6664,
      "mean_token_accuracy": 0.1561015322804451,
      "num_tokens": 7120281.0,
      "step": 4105
    },
    {
      "entropy": 5.717962598800659,
      "epoch": 0.31978214355183815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993403329755569,
      "loss": 5.5438,
      "mean_token_accuracy": 0.16015197783708573,
      "num_tokens": 7128584.0,
      "step": 4110
    },
    {
      "entropy": 5.808613395690918,
      "epoch": 0.32017117292355574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993382105128828,
      "loss": 5.6513,
      "mean_token_accuracy": 0.15905333682894707,
      "num_tokens": 7137570.0,
      "step": 4115
    },
    {
      "entropy": 5.773133039474487,
      "epoch": 0.32056020229527327,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993360846462333,
      "loss": 5.6899,
      "mean_token_accuracy": 0.15812519043684006,
      "num_tokens": 7146965.0,
      "step": 4120
    },
    {
      "entropy": 5.787432909011841,
      "epoch": 0.32094923166699085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993339553756408,
      "loss": 5.6272,
      "mean_token_accuracy": 0.15931697189807892,
      "num_tokens": 7155657.0,
      "step": 4125
    },
    {
      "entropy": 5.8046619415283205,
      "epoch": 0.32133826103870844,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993318227011378,
      "loss": 5.5615,
      "mean_token_accuracy": 0.1596067264676094,
      "num_tokens": 7164369.0,
      "step": 4130
    },
    {
      "entropy": 5.7777173042297365,
      "epoch": 0.32172729041042597,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993296866227562,
      "loss": 5.5399,
      "mean_token_accuracy": 0.1697876438498497,
      "num_tokens": 7172161.0,
      "step": 4135
    },
    {
      "entropy": 5.706020784378052,
      "epoch": 0.32211631978214356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993275471405288,
      "loss": 5.5862,
      "mean_token_accuracy": 0.15171700865030288,
      "num_tokens": 7180606.0,
      "step": 4140
    },
    {
      "entropy": 5.728547096252441,
      "epoch": 0.32250534915386114,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993254042544879,
      "loss": 5.6262,
      "mean_token_accuracy": 0.1590207114815712,
      "num_tokens": 7188912.0,
      "step": 4145
    },
    {
      "entropy": 5.788024091720581,
      "epoch": 0.3228943785255787,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499323257964666,
      "loss": 5.6215,
      "mean_token_accuracy": 0.15666755810379981,
      "num_tokens": 7197477.0,
      "step": 4150
    },
    {
      "entropy": 5.801720523834229,
      "epoch": 0.32328340789729626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993211082710957,
      "loss": 5.6026,
      "mean_token_accuracy": 0.16264097541570663,
      "num_tokens": 7206571.0,
      "step": 4155
    },
    {
      "entropy": 5.751149368286133,
      "epoch": 0.3236724372690138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993189551738097,
      "loss": 5.4462,
      "mean_token_accuracy": 0.15989296287298202,
      "num_tokens": 7215319.0,
      "step": 4160
    },
    {
      "entropy": 5.73461594581604,
      "epoch": 0.3240614666407314,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993167986728405,
      "loss": 5.6286,
      "mean_token_accuracy": 0.1550238996744156,
      "num_tokens": 7223389.0,
      "step": 4165
    },
    {
      "entropy": 5.771875619888306,
      "epoch": 0.32445049601244896,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499314638768221,
      "loss": 5.6018,
      "mean_token_accuracy": 0.15770421475172042,
      "num_tokens": 7231853.0,
      "step": 4170
    },
    {
      "entropy": 5.761664533615113,
      "epoch": 0.3248395253841665,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993124754599839,
      "loss": 5.5293,
      "mean_token_accuracy": 0.16403157413005828,
      "num_tokens": 7241081.0,
      "step": 4175
    },
    {
      "entropy": 5.71790828704834,
      "epoch": 0.3252285547558841,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993103087481619,
      "loss": 5.5487,
      "mean_token_accuracy": 0.1609074741601944,
      "num_tokens": 7249742.0,
      "step": 4180
    },
    {
      "entropy": 5.8293393611907955,
      "epoch": 0.3256175841276016,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993081386327881,
      "loss": 5.6852,
      "mean_token_accuracy": 0.15708462446928023,
      "num_tokens": 7258285.0,
      "step": 4185
    },
    {
      "entropy": 5.750807237625122,
      "epoch": 0.3260066134993192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993059651138953,
      "loss": 5.5018,
      "mean_token_accuracy": 0.16263256967067719,
      "num_tokens": 7266843.0,
      "step": 4190
    },
    {
      "entropy": 5.682654571533203,
      "epoch": 0.3263956428710368,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993037881915165,
      "loss": 5.5913,
      "mean_token_accuracy": 0.16370514035224915,
      "num_tokens": 7276041.0,
      "step": 4195
    },
    {
      "entropy": 5.738144969940185,
      "epoch": 0.3267846722427543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993016078656847,
      "loss": 5.6502,
      "mean_token_accuracy": 0.15509369671344758,
      "num_tokens": 7285144.0,
      "step": 4200
    },
    {
      "entropy": 5.65037989616394,
      "epoch": 0.3271737016144719,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499299424136433,
      "loss": 5.4993,
      "mean_token_accuracy": 0.16100571006536485,
      "num_tokens": 7293976.0,
      "step": 4205
    },
    {
      "entropy": 5.765503311157227,
      "epoch": 0.32756273098618943,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992972370037946,
      "loss": 5.5695,
      "mean_token_accuracy": 0.15914326161146164,
      "num_tokens": 7302365.0,
      "step": 4210
    },
    {
      "entropy": 5.791704416275024,
      "epoch": 0.327951760357907,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992950464678026,
      "loss": 5.6453,
      "mean_token_accuracy": 0.15589653253555297,
      "num_tokens": 7310909.0,
      "step": 4215
    },
    {
      "entropy": 5.762666463851929,
      "epoch": 0.3283407897296246,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992928525284903,
      "loss": 5.6041,
      "mean_token_accuracy": 0.15364096760749818,
      "num_tokens": 7319391.0,
      "step": 4220
    },
    {
      "entropy": 5.86359977722168,
      "epoch": 0.32872981910134214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499290655185891,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16046358346939088,
      "num_tokens": 7328076.0,
      "step": 4225
    },
    {
      "entropy": 5.7980348587036135,
      "epoch": 0.3291188484730597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499288454440038,
      "loss": 5.6062,
      "mean_token_accuracy": 0.15722377076745034,
      "num_tokens": 7336833.0,
      "step": 4230
    },
    {
      "entropy": 5.711167240142823,
      "epoch": 0.3295078778447773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992862502909645,
      "loss": 5.5223,
      "mean_token_accuracy": 0.1646954208612442,
      "num_tokens": 7344862.0,
      "step": 4235
    },
    {
      "entropy": 5.775886392593383,
      "epoch": 0.32989690721649484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992840427387043,
      "loss": 5.5541,
      "mean_token_accuracy": 0.16591128408908845,
      "num_tokens": 7354503.0,
      "step": 4240
    },
    {
      "entropy": 5.760515594482422,
      "epoch": 0.3302859365882124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992818317832908,
      "loss": 5.5568,
      "mean_token_accuracy": 0.16480007916688919,
      "num_tokens": 7363000.0,
      "step": 4245
    },
    {
      "entropy": 5.701372575759888,
      "epoch": 0.33067496595992996,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992796174247574,
      "loss": 5.6312,
      "mean_token_accuracy": 0.15584687665104865,
      "num_tokens": 7372619.0,
      "step": 4250
    },
    {
      "entropy": 5.803398752212525,
      "epoch": 0.33106399533164754,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992773996631378,
      "loss": 5.5015,
      "mean_token_accuracy": 0.17060409784317015,
      "num_tokens": 7380531.0,
      "step": 4255
    },
    {
      "entropy": 5.745937347412109,
      "epoch": 0.33145302470336513,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992751784984656,
      "loss": 5.6194,
      "mean_token_accuracy": 0.16034240871667862,
      "num_tokens": 7389153.0,
      "step": 4260
    },
    {
      "entropy": 5.725397109985352,
      "epoch": 0.33184205407508266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992729539307745,
      "loss": 5.5189,
      "mean_token_accuracy": 0.16305640786886216,
      "num_tokens": 7397880.0,
      "step": 4265
    },
    {
      "entropy": 5.78955659866333,
      "epoch": 0.33223108344680025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992707259600984,
      "loss": 5.5854,
      "mean_token_accuracy": 0.15720919668674468,
      "num_tokens": 7406180.0,
      "step": 4270
    },
    {
      "entropy": 5.7014524936676025,
      "epoch": 0.3326201128185178,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992684945864709,
      "loss": 5.593,
      "mean_token_accuracy": 0.16094412356615068,
      "num_tokens": 7415069.0,
      "step": 4275
    },
    {
      "entropy": 5.744061803817749,
      "epoch": 0.33300914219023536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992662598099261,
      "loss": 5.6493,
      "mean_token_accuracy": 0.15555099695920943,
      "num_tokens": 7423921.0,
      "step": 4280
    },
    {
      "entropy": 5.696127796173096,
      "epoch": 0.33339817156195295,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992640216304975,
      "loss": 5.3854,
      "mean_token_accuracy": 0.17118807137012482,
      "num_tokens": 7432655.0,
      "step": 4285
    },
    {
      "entropy": 5.776129102706909,
      "epoch": 0.3337872009336705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992617800482196,
      "loss": 5.578,
      "mean_token_accuracy": 0.16193160861730577,
      "num_tokens": 7441858.0,
      "step": 4290
    },
    {
      "entropy": 5.6602355480194095,
      "epoch": 0.33417623030538807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499259535063126,
      "loss": 5.5398,
      "mean_token_accuracy": 0.16309095025062562,
      "num_tokens": 7450030.0,
      "step": 4295
    },
    {
      "entropy": 5.785944986343384,
      "epoch": 0.3345652596771056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499257286675251,
      "loss": 5.6258,
      "mean_token_accuracy": 0.15881484746932983,
      "num_tokens": 7458256.0,
      "step": 4300
    },
    {
      "entropy": 5.786884784698486,
      "epoch": 0.3349542890488232,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992550348846285,
      "loss": 5.5915,
      "mean_token_accuracy": 0.15856461822986603,
      "num_tokens": 7466710.0,
      "step": 4305
    },
    {
      "entropy": 5.710636949539184,
      "epoch": 0.33534331842054077,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992527796912928,
      "loss": 5.5344,
      "mean_token_accuracy": 0.16328217089176178,
      "num_tokens": 7475445.0,
      "step": 4310
    },
    {
      "entropy": 5.79075436592102,
      "epoch": 0.3357323477922583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992505210952782,
      "loss": 5.6155,
      "mean_token_accuracy": 0.15852408558130265,
      "num_tokens": 7484326.0,
      "step": 4315
    },
    {
      "entropy": 5.715756559371949,
      "epoch": 0.3361213771639759,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992482590966188,
      "loss": 5.5138,
      "mean_token_accuracy": 0.15974442958831786,
      "num_tokens": 7493028.0,
      "step": 4320
    },
    {
      "entropy": 5.810898685455323,
      "epoch": 0.3365104065356934,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992459936953489,
      "loss": 5.691,
      "mean_token_accuracy": 0.1529371842741966,
      "num_tokens": 7502284.0,
      "step": 4325
    },
    {
      "entropy": 5.789745283126831,
      "epoch": 0.336899435907411,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992437248915032,
      "loss": 5.572,
      "mean_token_accuracy": 0.15930989682674407,
      "num_tokens": 7510728.0,
      "step": 4330
    },
    {
      "entropy": 5.705639839172363,
      "epoch": 0.3372884652791286,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992414526851157,
      "loss": 5.5308,
      "mean_token_accuracy": 0.1664871394634247,
      "num_tokens": 7519578.0,
      "step": 4335
    },
    {
      "entropy": 5.7081053256988525,
      "epoch": 0.3376774946508461,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992391770762212,
      "loss": 5.6538,
      "mean_token_accuracy": 0.15889098793268203,
      "num_tokens": 7528853.0,
      "step": 4340
    },
    {
      "entropy": 5.90105094909668,
      "epoch": 0.3380665240225637,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499236898064854,
      "loss": 5.6981,
      "mean_token_accuracy": 0.156148099899292,
      "num_tokens": 7537000.0,
      "step": 4345
    },
    {
      "entropy": 5.771493673324585,
      "epoch": 0.3384555533942813,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992346156510488,
      "loss": 5.6498,
      "mean_token_accuracy": 0.15839225202798843,
      "num_tokens": 7545439.0,
      "step": 4350
    },
    {
      "entropy": 5.763132333755493,
      "epoch": 0.3388445827659988,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992323298348403,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1569911375641823,
      "num_tokens": 7553808.0,
      "step": 4355
    },
    {
      "entropy": 5.793811559677124,
      "epoch": 0.3392336121377164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499230040616263,
      "loss": 5.5838,
      "mean_token_accuracy": 0.16720182895660402,
      "num_tokens": 7562124.0,
      "step": 4360
    },
    {
      "entropy": 5.779132938385009,
      "epoch": 0.33962264150943394,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992277479953518,
      "loss": 5.626,
      "mean_token_accuracy": 0.1591033309698105,
      "num_tokens": 7570692.0,
      "step": 4365
    },
    {
      "entropy": 5.69827675819397,
      "epoch": 0.34001167088115153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992254519721414,
      "loss": 5.5254,
      "mean_token_accuracy": 0.1599597990512848,
      "num_tokens": 7579380.0,
      "step": 4370
    },
    {
      "entropy": 5.712707996368408,
      "epoch": 0.3404007002528691,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992231525466666,
      "loss": 5.5748,
      "mean_token_accuracy": 0.15763231217861176,
      "num_tokens": 7587581.0,
      "step": 4375
    },
    {
      "entropy": 5.803431415557862,
      "epoch": 0.34078972962458665,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992208497189624,
      "loss": 5.6207,
      "mean_token_accuracy": 0.15856635421514512,
      "num_tokens": 7596373.0,
      "step": 4380
    },
    {
      "entropy": 5.769613647460938,
      "epoch": 0.34117875899630423,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992185434890637,
      "loss": 5.63,
      "mean_token_accuracy": 0.15952463150024415,
      "num_tokens": 7605698.0,
      "step": 4385
    },
    {
      "entropy": 5.831510257720947,
      "epoch": 0.34156778836802176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992162338570055,
      "loss": 5.6701,
      "mean_token_accuracy": 0.15391338169574736,
      "num_tokens": 7614299.0,
      "step": 4390
    },
    {
      "entropy": 5.711894607543945,
      "epoch": 0.34195681773973935,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992139208228227,
      "loss": 5.5742,
      "mean_token_accuracy": 0.1665854424238205,
      "num_tokens": 7622244.0,
      "step": 4395
    },
    {
      "entropy": 5.799953508377075,
      "epoch": 0.34234584711145694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992116043865506,
      "loss": 5.6835,
      "mean_token_accuracy": 0.14930743724107742,
      "num_tokens": 7631421.0,
      "step": 4400
    },
    {
      "entropy": 5.780383348464966,
      "epoch": 0.34273487648317447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992092845482244,
      "loss": 5.617,
      "mean_token_accuracy": 0.1649213418364525,
      "num_tokens": 7640483.0,
      "step": 4405
    },
    {
      "entropy": 5.734336900711059,
      "epoch": 0.34312390585489205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992069613078791,
      "loss": 5.5434,
      "mean_token_accuracy": 0.16003944873809814,
      "num_tokens": 7648928.0,
      "step": 4410
    },
    {
      "entropy": 5.748956537246704,
      "epoch": 0.3435129352266096,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992046346655499,
      "loss": 5.556,
      "mean_token_accuracy": 0.16345997750759125,
      "num_tokens": 7658158.0,
      "step": 4415
    },
    {
      "entropy": 5.79419469833374,
      "epoch": 0.34390196459832717,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992023046212723,
      "loss": 5.6394,
      "mean_token_accuracy": 0.15967277586460113,
      "num_tokens": 7666657.0,
      "step": 4420
    },
    {
      "entropy": 5.756088972091675,
      "epoch": 0.34429099397004476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991999711750816,
      "loss": 5.5864,
      "mean_token_accuracy": 0.1604827433824539,
      "num_tokens": 7675182.0,
      "step": 4425
    },
    {
      "entropy": 5.713095378875733,
      "epoch": 0.3446800233417623,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991976343270133,
      "loss": 5.537,
      "mean_token_accuracy": 0.16038379073143005,
      "num_tokens": 7684132.0,
      "step": 4430
    },
    {
      "entropy": 5.707273960113525,
      "epoch": 0.3450690527134799,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991952940771026,
      "loss": 5.4828,
      "mean_token_accuracy": 0.16817560493946077,
      "num_tokens": 7692574.0,
      "step": 4435
    },
    {
      "entropy": 5.682709455490112,
      "epoch": 0.34545808208519746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991929504253852,
      "loss": 5.5448,
      "mean_token_accuracy": 0.1668045476078987,
      "num_tokens": 7700969.0,
      "step": 4440
    },
    {
      "entropy": 5.786290264129638,
      "epoch": 0.345847111456915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991906033718966,
      "loss": 5.6863,
      "mean_token_accuracy": 0.15646209716796874,
      "num_tokens": 7710688.0,
      "step": 4445
    },
    {
      "entropy": 5.8290328025817875,
      "epoch": 0.3462361408286326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991882529166724,
      "loss": 5.6461,
      "mean_token_accuracy": 0.1635972410440445,
      "num_tokens": 7720071.0,
      "step": 4450
    },
    {
      "entropy": 5.673279428482056,
      "epoch": 0.3466251702003501,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991858990597483,
      "loss": 5.4371,
      "mean_token_accuracy": 0.16662000864744186,
      "num_tokens": 7728418.0,
      "step": 4455
    },
    {
      "entropy": 5.733144617080688,
      "epoch": 0.3470141995720677,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991835418011601,
      "loss": 5.67,
      "mean_token_accuracy": 0.1602802112698555,
      "num_tokens": 7736849.0,
      "step": 4460
    },
    {
      "entropy": 5.787078762054444,
      "epoch": 0.3474032289437853,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991811811409434,
      "loss": 5.5981,
      "mean_token_accuracy": 0.1622970446944237,
      "num_tokens": 7745730.0,
      "step": 4465
    },
    {
      "entropy": 5.771909475326538,
      "epoch": 0.3477922583155028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991788170791341,
      "loss": 5.4965,
      "mean_token_accuracy": 0.16542282104492187,
      "num_tokens": 7755040.0,
      "step": 4470
    },
    {
      "entropy": 5.734687471389771,
      "epoch": 0.3481812876872204,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991764496157682,
      "loss": 5.5207,
      "mean_token_accuracy": 0.16281888782978057,
      "num_tokens": 7763765.0,
      "step": 4475
    },
    {
      "entropy": 5.622528505325318,
      "epoch": 0.34857031705893793,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991740787508814,
      "loss": 5.5206,
      "mean_token_accuracy": 0.16116178408265114,
      "num_tokens": 7772261.0,
      "step": 4480
    },
    {
      "entropy": 5.827115106582641,
      "epoch": 0.3489593464306555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991717044845097,
      "loss": 5.6628,
      "mean_token_accuracy": 0.15989381670951844,
      "num_tokens": 7781651.0,
      "step": 4485
    },
    {
      "entropy": 5.714611434936524,
      "epoch": 0.3493483758023731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991693268166892,
      "loss": 5.518,
      "mean_token_accuracy": 0.15781033784151077,
      "num_tokens": 7789926.0,
      "step": 4490
    },
    {
      "entropy": 5.715775346755981,
      "epoch": 0.34973740517409063,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991669457474559,
      "loss": 5.5567,
      "mean_token_accuracy": 0.15968117415904998,
      "num_tokens": 7798435.0,
      "step": 4495
    },
    {
      "entropy": 5.662638235092163,
      "epoch": 0.3501264345458082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991645612768461,
      "loss": 5.4951,
      "mean_token_accuracy": 0.15639443546533585,
      "num_tokens": 7806674.0,
      "step": 4500
    },
    {
      "entropy": 5.754309892654419,
      "epoch": 0.35051546391752575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991621734048958,
      "loss": 5.5444,
      "mean_token_accuracy": 0.16342764496803283,
      "num_tokens": 7814774.0,
      "step": 4505
    },
    {
      "entropy": 5.704548788070679,
      "epoch": 0.35090449328924334,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991597821316415,
      "loss": 5.5986,
      "mean_token_accuracy": 0.16580508202314376,
      "num_tokens": 7822879.0,
      "step": 4510
    },
    {
      "entropy": 5.696450519561767,
      "epoch": 0.3512935226609609,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991573874571192,
      "loss": 5.5416,
      "mean_token_accuracy": 0.15570010393857955,
      "num_tokens": 7831418.0,
      "step": 4515
    },
    {
      "entropy": 5.7250312805175785,
      "epoch": 0.35168255203267845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991549893813653,
      "loss": 5.5079,
      "mean_token_accuracy": 0.15775506496429442,
      "num_tokens": 7839513.0,
      "step": 4520
    },
    {
      "entropy": 5.691388750076294,
      "epoch": 0.35207158140439604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991525879044162,
      "loss": 5.5324,
      "mean_token_accuracy": 0.16139761209487916,
      "num_tokens": 7847653.0,
      "step": 4525
    },
    {
      "entropy": 5.782090950012207,
      "epoch": 0.35246061077611357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991501830263084,
      "loss": 5.6185,
      "mean_token_accuracy": 0.1575747847557068,
      "num_tokens": 7857093.0,
      "step": 4530
    },
    {
      "entropy": 5.737625694274902,
      "epoch": 0.35284964014783116,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991477747470783,
      "loss": 5.582,
      "mean_token_accuracy": 0.1566581904888153,
      "num_tokens": 7866129.0,
      "step": 4535
    },
    {
      "entropy": 5.70094780921936,
      "epoch": 0.35323866951954874,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991453630667625,
      "loss": 5.6615,
      "mean_token_accuracy": 0.15456809997558593,
      "num_tokens": 7874613.0,
      "step": 4540
    },
    {
      "entropy": 5.6414001941680905,
      "epoch": 0.3536276988912663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991429479853976,
      "loss": 5.4164,
      "mean_token_accuracy": 0.16817961633205414,
      "num_tokens": 7882292.0,
      "step": 4545
    },
    {
      "entropy": 5.737434244155883,
      "epoch": 0.35401672826298386,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991405295030202,
      "loss": 5.4426,
      "mean_token_accuracy": 0.17088190019130706,
      "num_tokens": 7890810.0,
      "step": 4550
    },
    {
      "entropy": 5.730095767974854,
      "epoch": 0.35440575763470145,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991381076196671,
      "loss": 5.5862,
      "mean_token_accuracy": 0.15629213452339172,
      "num_tokens": 7899964.0,
      "step": 4555
    },
    {
      "entropy": 5.672501754760742,
      "epoch": 0.354794787006419,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991356823353748,
      "loss": 5.4754,
      "mean_token_accuracy": 0.16402625739574433,
      "num_tokens": 7908600.0,
      "step": 4560
    },
    {
      "entropy": 5.677447080612183,
      "epoch": 0.35518381637813656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991332536501804,
      "loss": 5.5912,
      "mean_token_accuracy": 0.15798643827438355,
      "num_tokens": 7916893.0,
      "step": 4565
    },
    {
      "entropy": 5.731996774673462,
      "epoch": 0.3555728457498541,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991308215641205,
      "loss": 5.4517,
      "mean_token_accuracy": 0.1716650128364563,
      "num_tokens": 7925230.0,
      "step": 4570
    },
    {
      "entropy": 5.719581079483032,
      "epoch": 0.3559618751215717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991283860772323,
      "loss": 5.5052,
      "mean_token_accuracy": 0.16825538873672485,
      "num_tokens": 7933487.0,
      "step": 4575
    },
    {
      "entropy": 5.671692514419556,
      "epoch": 0.35635090449328927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991259471895525,
      "loss": 5.5453,
      "mean_token_accuracy": 0.16380455940961838,
      "num_tokens": 7941957.0,
      "step": 4580
    },
    {
      "entropy": 5.7017723560333256,
      "epoch": 0.3567399338650068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991235049011182,
      "loss": 5.4838,
      "mean_token_accuracy": 0.16157088428735733,
      "num_tokens": 7950912.0,
      "step": 4585
    },
    {
      "entropy": 5.68826847076416,
      "epoch": 0.3571289632367244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991210592119663,
      "loss": 5.5453,
      "mean_token_accuracy": 0.16106254905462264,
      "num_tokens": 7959724.0,
      "step": 4590
    },
    {
      "entropy": 5.739056968688965,
      "epoch": 0.3575179926084419,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991186101221342,
      "loss": 5.5182,
      "mean_token_accuracy": 0.1671526774764061,
      "num_tokens": 7968683.0,
      "step": 4595
    },
    {
      "entropy": 5.694250869750976,
      "epoch": 0.3579070219801595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499116157631659,
      "loss": 5.4915,
      "mean_token_accuracy": 0.1625325247645378,
      "num_tokens": 7977004.0,
      "step": 4600
    },
    {
      "entropy": 5.774178075790405,
      "epoch": 0.3582960513518771,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991137017405777,
      "loss": 5.6947,
      "mean_token_accuracy": 0.1508305624127388,
      "num_tokens": 7985537.0,
      "step": 4605
    },
    {
      "entropy": 5.810094594955444,
      "epoch": 0.3586850807235946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991112424489277,
      "loss": 5.5914,
      "mean_token_accuracy": 0.16004215627908708,
      "num_tokens": 7994298.0,
      "step": 4610
    },
    {
      "entropy": 5.675488615036011,
      "epoch": 0.3590741100953122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991087797567462,
      "loss": 5.5485,
      "mean_token_accuracy": 0.16430403739213945,
      "num_tokens": 8003145.0,
      "step": 4615
    },
    {
      "entropy": 5.653676176071167,
      "epoch": 0.35946313946702974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991063136640709,
      "loss": 5.5177,
      "mean_token_accuracy": 0.16057982444763183,
      "num_tokens": 8011839.0,
      "step": 4620
    },
    {
      "entropy": 5.8171083450317385,
      "epoch": 0.3598521688387473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991038441709388,
      "loss": 5.5952,
      "mean_token_accuracy": 0.15525781512260436,
      "num_tokens": 8020248.0,
      "step": 4625
    },
    {
      "entropy": 5.770076322555542,
      "epoch": 0.3602411982104649,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991013712773877,
      "loss": 5.6226,
      "mean_token_accuracy": 0.1595271795988083,
      "num_tokens": 8029360.0,
      "step": 4630
    },
    {
      "entropy": 5.830290365219116,
      "epoch": 0.36063022758218244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990988949834549,
      "loss": 5.6756,
      "mean_token_accuracy": 0.15414261519908906,
      "num_tokens": 8037407.0,
      "step": 4635
    },
    {
      "entropy": 5.773358583450317,
      "epoch": 0.3610192569539,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499096415289178,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1608835220336914,
      "num_tokens": 8045935.0,
      "step": 4640
    },
    {
      "entropy": 5.71015191078186,
      "epoch": 0.3614082863256176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990939321945948,
      "loss": 5.6419,
      "mean_token_accuracy": 0.14878360480070113,
      "num_tokens": 8054640.0,
      "step": 4645
    },
    {
      "entropy": 5.802717781066894,
      "epoch": 0.36179731569733514,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990914456997427,
      "loss": 5.6215,
      "mean_token_accuracy": 0.15971666425466538,
      "num_tokens": 8063845.0,
      "step": 4650
    },
    {
      "entropy": 5.790019845962524,
      "epoch": 0.36218634506905273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990889558046598,
      "loss": 5.5653,
      "mean_token_accuracy": 0.15723626762628556,
      "num_tokens": 8072389.0,
      "step": 4655
    },
    {
      "entropy": 5.651761722564697,
      "epoch": 0.36257537444077026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990864625093836,
      "loss": 5.4354,
      "mean_token_accuracy": 0.1628458485007286,
      "num_tokens": 8080683.0,
      "step": 4660
    },
    {
      "entropy": 5.626598167419433,
      "epoch": 0.36296440381248785,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990839658139519,
      "loss": 5.469,
      "mean_token_accuracy": 0.1658550351858139,
      "num_tokens": 8088945.0,
      "step": 4665
    },
    {
      "entropy": 5.662598133087158,
      "epoch": 0.36335343318420543,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990814657184028,
      "loss": 5.528,
      "mean_token_accuracy": 0.16415927559137344,
      "num_tokens": 8097447.0,
      "step": 4670
    },
    {
      "entropy": 5.654489183425904,
      "epoch": 0.36374246255592296,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990789622227741,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18078008890151978,
      "num_tokens": 8105570.0,
      "step": 4675
    },
    {
      "entropy": 5.64568943977356,
      "epoch": 0.36413149192764055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990764553271038,
      "loss": 5.5065,
      "mean_token_accuracy": 0.16293679773807526,
      "num_tokens": 8114842.0,
      "step": 4680
    },
    {
      "entropy": 5.6961933135986325,
      "epoch": 0.3645205212993581,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990739450314299,
      "loss": 5.4941,
      "mean_token_accuracy": 0.16237769424915313,
      "num_tokens": 8122960.0,
      "step": 4685
    },
    {
      "entropy": 5.800394201278687,
      "epoch": 0.36490955067107567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990714313357906,
      "loss": 5.6298,
      "mean_token_accuracy": 0.16036661416292192,
      "num_tokens": 8131850.0,
      "step": 4690
    },
    {
      "entropy": 5.730320596694947,
      "epoch": 0.36529858004279325,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499068914240224,
      "loss": 5.5025,
      "mean_token_accuracy": 0.16529248505830765,
      "num_tokens": 8139818.0,
      "step": 4695
    },
    {
      "entropy": 5.6371876239776615,
      "epoch": 0.3656876094145108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990663937447682,
      "loss": 5.5769,
      "mean_token_accuracy": 0.16349330991506578,
      "num_tokens": 8147995.0,
      "step": 4700
    },
    {
      "entropy": 5.758992433547974,
      "epoch": 0.36607663878622837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990638698494615,
      "loss": 5.5741,
      "mean_token_accuracy": 0.16156561821699142,
      "num_tokens": 8156879.0,
      "step": 4705
    },
    {
      "entropy": 5.740928316116333,
      "epoch": 0.3664656681579459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990613425543423,
      "loss": 5.5433,
      "mean_token_accuracy": 0.150849449634552,
      "num_tokens": 8164860.0,
      "step": 4710
    },
    {
      "entropy": 5.624187231063843,
      "epoch": 0.3668546975296635,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990588118594487,
      "loss": 5.3721,
      "mean_token_accuracy": 0.17311475276947022,
      "num_tokens": 8173319.0,
      "step": 4715
    },
    {
      "entropy": 5.68842396736145,
      "epoch": 0.3672437269013811,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990562777648194,
      "loss": 5.5738,
      "mean_token_accuracy": 0.16338748335838318,
      "num_tokens": 8182385.0,
      "step": 4720
    },
    {
      "entropy": 5.749617624282837,
      "epoch": 0.3676327562730986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990537402704928,
      "loss": 5.494,
      "mean_token_accuracy": 0.16597676873207093,
      "num_tokens": 8190394.0,
      "step": 4725
    },
    {
      "entropy": 5.677096652984619,
      "epoch": 0.3680217856448162,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990511993765071,
      "loss": 5.5141,
      "mean_token_accuracy": 0.156700736284256,
      "num_tokens": 8198532.0,
      "step": 4730
    },
    {
      "entropy": 5.582213068008423,
      "epoch": 0.3684108150165337,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990486550829011,
      "loss": 5.4755,
      "mean_token_accuracy": 0.16821785122156144,
      "num_tokens": 8207436.0,
      "step": 4735
    },
    {
      "entropy": 5.749109268188477,
      "epoch": 0.3687998443882513,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990461073897134,
      "loss": 5.5093,
      "mean_token_accuracy": 0.16389241069555283,
      "num_tokens": 8216316.0,
      "step": 4740
    },
    {
      "entropy": 5.682870817184448,
      "epoch": 0.3691888737599689,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990435562969827,
      "loss": 5.4634,
      "mean_token_accuracy": 0.16937032341957092,
      "num_tokens": 8224918.0,
      "step": 4745
    },
    {
      "entropy": 5.652633953094482,
      "epoch": 0.3695779031316864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990410018047475,
      "loss": 5.4855,
      "mean_token_accuracy": 0.16341211497783661,
      "num_tokens": 8233506.0,
      "step": 4750
    },
    {
      "entropy": 5.686648225784301,
      "epoch": 0.369966932503404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990384439130467,
      "loss": 5.5371,
      "mean_token_accuracy": 0.16136658936738968,
      "num_tokens": 8243122.0,
      "step": 4755
    },
    {
      "entropy": 5.695843458175659,
      "epoch": 0.3703559618751216,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990358826219192,
      "loss": 5.4954,
      "mean_token_accuracy": 0.16794812679290771,
      "num_tokens": 8251639.0,
      "step": 4760
    },
    {
      "entropy": 5.694685459136963,
      "epoch": 0.37074499124683913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990333179314038,
      "loss": 5.5441,
      "mean_token_accuracy": 0.1565728172659874,
      "num_tokens": 8260487.0,
      "step": 4765
    },
    {
      "entropy": 5.731024360656738,
      "epoch": 0.3711340206185567,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990307498415393,
      "loss": 5.5353,
      "mean_token_accuracy": 0.16720337867736818,
      "num_tokens": 8270111.0,
      "step": 4770
    },
    {
      "entropy": 5.732585525512695,
      "epoch": 0.37152304999027425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990281783523648,
      "loss": 5.6699,
      "mean_token_accuracy": 0.15745235085487366,
      "num_tokens": 8278171.0,
      "step": 4775
    },
    {
      "entropy": 5.728489589691162,
      "epoch": 0.37191207936199183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990256034639192,
      "loss": 5.5444,
      "mean_token_accuracy": 0.16361308097839355,
      "num_tokens": 8287278.0,
      "step": 4780
    },
    {
      "entropy": 5.801375198364258,
      "epoch": 0.3723011087337094,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990230251762418,
      "loss": 5.6581,
      "mean_token_accuracy": 0.15506795942783355,
      "num_tokens": 8296874.0,
      "step": 4785
    },
    {
      "entropy": 5.677485656738281,
      "epoch": 0.37269013810542695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990204434893713,
      "loss": 5.5292,
      "mean_token_accuracy": 0.16247798800468444,
      "num_tokens": 8305375.0,
      "step": 4790
    },
    {
      "entropy": 5.805545330047607,
      "epoch": 0.37307916747714454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990178584033474,
      "loss": 5.5798,
      "mean_token_accuracy": 0.1593550831079483,
      "num_tokens": 8313960.0,
      "step": 4795
    },
    {
      "entropy": 5.732056093215943,
      "epoch": 0.37346819684886207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990152699182089,
      "loss": 5.5091,
      "mean_token_accuracy": 0.1695062816143036,
      "num_tokens": 8321838.0,
      "step": 4800
    },
    {
      "entropy": 5.647253465652466,
      "epoch": 0.37385722622057965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990126780339953,
      "loss": 5.549,
      "mean_token_accuracy": 0.15523339807987213,
      "num_tokens": 8331325.0,
      "step": 4805
    },
    {
      "entropy": 5.7562426090240475,
      "epoch": 0.37424625559229724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990100827507459,
      "loss": 5.5963,
      "mean_token_accuracy": 0.1604474112391472,
      "num_tokens": 8340105.0,
      "step": 4810
    },
    {
      "entropy": 5.675045204162598,
      "epoch": 0.37463528496401477,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990074840684999,
      "loss": 5.4176,
      "mean_token_accuracy": 0.17652603089809418,
      "num_tokens": 8348256.0,
      "step": 4815
    },
    {
      "entropy": 5.65481014251709,
      "epoch": 0.37502431433573236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990048819872969,
      "loss": 5.6006,
      "mean_token_accuracy": 0.15629589855670928,
      "num_tokens": 8357456.0,
      "step": 4820
    },
    {
      "entropy": 5.74207820892334,
      "epoch": 0.3754133437074499,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990022765071765,
      "loss": 5.4483,
      "mean_token_accuracy": 0.16962354630231857,
      "num_tokens": 8366939.0,
      "step": 4825
    },
    {
      "entropy": 5.838831901550293,
      "epoch": 0.3758023730791675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498999667628178,
      "loss": 5.6137,
      "mean_token_accuracy": 0.15240543484687805,
      "num_tokens": 8376062.0,
      "step": 4830
    },
    {
      "entropy": 5.771680784225464,
      "epoch": 0.37619140245088506,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989970553503411,
      "loss": 5.5276,
      "mean_token_accuracy": 0.16518781930208207,
      "num_tokens": 8383803.0,
      "step": 4835
    },
    {
      "entropy": 5.6745326042175295,
      "epoch": 0.3765804318226026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989944396737054,
      "loss": 5.6889,
      "mean_token_accuracy": 0.15957241952419282,
      "num_tokens": 8392844.0,
      "step": 4840
    },
    {
      "entropy": 5.768671894073487,
      "epoch": 0.3769694611943202,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989918205983106,
      "loss": 5.5881,
      "mean_token_accuracy": 0.1660487473011017,
      "num_tokens": 8401937.0,
      "step": 4845
    },
    {
      "entropy": 5.746508407592773,
      "epoch": 0.37735849056603776,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989891981241964,
      "loss": 5.5918,
      "mean_token_accuracy": 0.15559911131858825,
      "num_tokens": 8410383.0,
      "step": 4850
    },
    {
      "entropy": 5.664513111114502,
      "epoch": 0.3777475199377553,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989865722514027,
      "loss": 5.547,
      "mean_token_accuracy": 0.1587306007742882,
      "num_tokens": 8419053.0,
      "step": 4855
    },
    {
      "entropy": 5.702281188964844,
      "epoch": 0.3781365493094729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989839429799692,
      "loss": 5.5249,
      "mean_token_accuracy": 0.16742938309907912,
      "num_tokens": 8427311.0,
      "step": 4860
    },
    {
      "entropy": 5.715058994293213,
      "epoch": 0.3785255786811904,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498981310309936,
      "loss": 5.5741,
      "mean_token_accuracy": 0.15908973962068557,
      "num_tokens": 8436050.0,
      "step": 4865
    },
    {
      "entropy": 5.820055341720581,
      "epoch": 0.378914608052908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989786742413428,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1646547108888626,
      "num_tokens": 8444579.0,
      "step": 4870
    },
    {
      "entropy": 5.689808559417725,
      "epoch": 0.3793036374246256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989760347742298,
      "loss": 5.5276,
      "mean_token_accuracy": 0.1638904482126236,
      "num_tokens": 8454013.0,
      "step": 4875
    },
    {
      "entropy": 5.724506807327271,
      "epoch": 0.3796926667963431,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989733919086369,
      "loss": 5.6429,
      "mean_token_accuracy": 0.15735770165920257,
      "num_tokens": 8462564.0,
      "step": 4880
    },
    {
      "entropy": 5.7334497451782225,
      "epoch": 0.3800816961680607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989707456446043,
      "loss": 5.4841,
      "mean_token_accuracy": 0.16679838299751282,
      "num_tokens": 8471035.0,
      "step": 4885
    },
    {
      "entropy": 5.680740404129028,
      "epoch": 0.38047072553977823,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989680959821721,
      "loss": 5.4693,
      "mean_token_accuracy": 0.16390866339206694,
      "num_tokens": 8479475.0,
      "step": 4890
    },
    {
      "entropy": 5.623473501205444,
      "epoch": 0.3808597549114958,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989654429213805,
      "loss": 5.5591,
      "mean_token_accuracy": 0.1630413204431534,
      "num_tokens": 8488163.0,
      "step": 4895
    },
    {
      "entropy": 5.821042728424072,
      "epoch": 0.3812487842832134,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989627864622699,
      "loss": 5.5995,
      "mean_token_accuracy": 0.15810228884220123,
      "num_tokens": 8497735.0,
      "step": 4900
    },
    {
      "entropy": 5.743423938751221,
      "epoch": 0.38163781365493094,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989601266048804,
      "loss": 5.5459,
      "mean_token_accuracy": 0.16006243228912354,
      "num_tokens": 8506986.0,
      "step": 4905
    },
    {
      "entropy": 5.6776947498321535,
      "epoch": 0.3820268430266485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989574633492525,
      "loss": 5.5709,
      "mean_token_accuracy": 0.16254400610923767,
      "num_tokens": 8515168.0,
      "step": 4910
    },
    {
      "entropy": 5.690753602981568,
      "epoch": 0.38241587239836605,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989547966954266,
      "loss": 5.5509,
      "mean_token_accuracy": 0.16618261039257048,
      "num_tokens": 8523980.0,
      "step": 4915
    },
    {
      "entropy": 5.658663177490235,
      "epoch": 0.38280490177008364,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498952126643443,
      "loss": 5.5032,
      "mean_token_accuracy": 0.16129754930734636,
      "num_tokens": 8532938.0,
      "step": 4920
    },
    {
      "entropy": 5.67104868888855,
      "epoch": 0.3831939311418012,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989494531933424,
      "loss": 5.4671,
      "mean_token_accuracy": 0.16780146807432175,
      "num_tokens": 8541951.0,
      "step": 4925
    },
    {
      "entropy": 5.74264087677002,
      "epoch": 0.38358296051351876,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989467763451652,
      "loss": 5.6436,
      "mean_token_accuracy": 0.15836040526628495,
      "num_tokens": 8550981.0,
      "step": 4930
    },
    {
      "entropy": 5.670666360855103,
      "epoch": 0.38397198988523634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989440960989523,
      "loss": 5.5569,
      "mean_token_accuracy": 0.15649997144937516,
      "num_tokens": 8559711.0,
      "step": 4935
    },
    {
      "entropy": 5.681796836853027,
      "epoch": 0.3843610192569539,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498941412454744,
      "loss": 5.4853,
      "mean_token_accuracy": 0.15973234325647354,
      "num_tokens": 8567797.0,
      "step": 4940
    },
    {
      "entropy": 5.7254406929016115,
      "epoch": 0.38475004862867146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989387254125813,
      "loss": 5.5791,
      "mean_token_accuracy": 0.15818444043397903,
      "num_tokens": 8576719.0,
      "step": 4945
    },
    {
      "entropy": 5.737820959091186,
      "epoch": 0.38513907800038905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989360349725049,
      "loss": 5.5682,
      "mean_token_accuracy": 0.1667224258184433,
      "num_tokens": 8585871.0,
      "step": 4950
    },
    {
      "entropy": 5.708989763259888,
      "epoch": 0.3855281073721066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989333411345555,
      "loss": 5.4759,
      "mean_token_accuracy": 0.16250301152467728,
      "num_tokens": 8594018.0,
      "step": 4955
    },
    {
      "entropy": 5.680680894851685,
      "epoch": 0.38591713674382416,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989306438987742,
      "loss": 5.5976,
      "mean_token_accuracy": 0.15720874518156053,
      "num_tokens": 8602788.0,
      "step": 4960
    },
    {
      "entropy": 5.760657072067261,
      "epoch": 0.38630616611554175,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989279432652018,
      "loss": 5.5509,
      "mean_token_accuracy": 0.16338287740945817,
      "num_tokens": 8611837.0,
      "step": 4965
    },
    {
      "entropy": 5.704029893875122,
      "epoch": 0.3866951954872593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989252392338791,
      "loss": 5.5267,
      "mean_token_accuracy": 0.16973906457424165,
      "num_tokens": 8620114.0,
      "step": 4970
    },
    {
      "entropy": 5.8096014022827145,
      "epoch": 0.38708422485897687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989225318048475,
      "loss": 5.6419,
      "mean_token_accuracy": 0.15283987894654275,
      "num_tokens": 8628777.0,
      "step": 4975
    },
    {
      "entropy": 5.7505690574646,
      "epoch": 0.3874732542306944,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989198209781478,
      "loss": 5.5492,
      "mean_token_accuracy": 0.15922651141881944,
      "num_tokens": 8637088.0,
      "step": 4980
    },
    {
      "entropy": 5.732375717163086,
      "epoch": 0.387862283602412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989171067538211,
      "loss": 5.5598,
      "mean_token_accuracy": 0.16299061924219133,
      "num_tokens": 8645261.0,
      "step": 4985
    },
    {
      "entropy": 5.73756160736084,
      "epoch": 0.38825131297412957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498914389131909,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16376165598630904,
      "num_tokens": 8653492.0,
      "step": 4990
    },
    {
      "entropy": 5.660993385314941,
      "epoch": 0.3886403423458471,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989116681124523,
      "loss": 5.5167,
      "mean_token_accuracy": 0.16651235818862914,
      "num_tokens": 8661909.0,
      "step": 4995
    },
    {
      "entropy": 5.676572227478028,
      "epoch": 0.3890293717175647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989089436954924,
      "loss": 5.4822,
      "mean_token_accuracy": 0.16759965121746062,
      "num_tokens": 8670332.0,
      "step": 5000
    },
    {
      "entropy": 5.762911462783814,
      "epoch": 0.3894184010892822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989062158810706,
      "loss": 5.5234,
      "mean_token_accuracy": 0.1680043324828148,
      "num_tokens": 8678537.0,
      "step": 5005
    },
    {
      "entropy": 5.692707681655884,
      "epoch": 0.3898074304609998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989034846692284,
      "loss": 5.4863,
      "mean_token_accuracy": 0.1656753972172737,
      "num_tokens": 8686839.0,
      "step": 5010
    },
    {
      "entropy": 5.6931657791137695,
      "epoch": 0.3901964598327174,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989007500600073,
      "loss": 5.5426,
      "mean_token_accuracy": 0.16253531575202942,
      "num_tokens": 8694407.0,
      "step": 5015
    },
    {
      "entropy": 5.662943887710571,
      "epoch": 0.3905854892044349,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988980120534486,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1695681780576706,
      "num_tokens": 8703484.0,
      "step": 5020
    },
    {
      "entropy": 5.807813596725464,
      "epoch": 0.3909745185761525,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498895270649594,
      "loss": 5.6039,
      "mean_token_accuracy": 0.1552526980638504,
      "num_tokens": 8712634.0,
      "step": 5025
    },
    {
      "entropy": 5.683317804336548,
      "epoch": 0.39136354794787004,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498892525848485,
      "loss": 5.487,
      "mean_token_accuracy": 0.1599026545882225,
      "num_tokens": 8721229.0,
      "step": 5030
    },
    {
      "entropy": 5.661917400360108,
      "epoch": 0.3917525773195876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988897776501633,
      "loss": 5.541,
      "mean_token_accuracy": 0.16514813005924225,
      "num_tokens": 8730010.0,
      "step": 5035
    },
    {
      "entropy": 5.674532079696656,
      "epoch": 0.3921416066913052,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988870260546705,
      "loss": 5.5342,
      "mean_token_accuracy": 0.1578855812549591,
      "num_tokens": 8739046.0,
      "step": 5040
    },
    {
      "entropy": 5.742646265029907,
      "epoch": 0.39253063606302274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988842710620486,
      "loss": 5.6198,
      "mean_token_accuracy": 0.16325782537460326,
      "num_tokens": 8747650.0,
      "step": 5045
    },
    {
      "entropy": 5.667177534103393,
      "epoch": 0.39291966543474033,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988815126723391,
      "loss": 5.5398,
      "mean_token_accuracy": 0.16407661586999894,
      "num_tokens": 8756319.0,
      "step": 5050
    },
    {
      "entropy": 5.6189886093139645,
      "epoch": 0.3933086948064579,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988787508855841,
      "loss": 5.554,
      "mean_token_accuracy": 0.16317882537841796,
      "num_tokens": 8765448.0,
      "step": 5055
    },
    {
      "entropy": 5.711760187149048,
      "epoch": 0.39369772417817545,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988759857018253,
      "loss": 5.5024,
      "mean_token_accuracy": 0.15837739109992982,
      "num_tokens": 8774146.0,
      "step": 5060
    },
    {
      "entropy": 5.642172765731812,
      "epoch": 0.39408675354989303,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988732171211048,
      "loss": 5.3808,
      "mean_token_accuracy": 0.16725590229034423,
      "num_tokens": 8782311.0,
      "step": 5065
    },
    {
      "entropy": 5.620117568969727,
      "epoch": 0.39447578292161056,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988704451434644,
      "loss": 5.4233,
      "mean_token_accuracy": 0.167247074842453,
      "num_tokens": 8790196.0,
      "step": 5070
    },
    {
      "entropy": 5.670018005371094,
      "epoch": 0.39486481229332815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988676697689465,
      "loss": 5.5098,
      "mean_token_accuracy": 0.1597319647669792,
      "num_tokens": 8798836.0,
      "step": 5075
    },
    {
      "entropy": 5.671954202651977,
      "epoch": 0.39525384166504574,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498864890997593,
      "loss": 5.5441,
      "mean_token_accuracy": 0.16164022088050842,
      "num_tokens": 8806942.0,
      "step": 5080
    },
    {
      "entropy": 5.761148405075073,
      "epoch": 0.39564287103676327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988621088294461,
      "loss": 5.7312,
      "mean_token_accuracy": 0.15544241815805435,
      "num_tokens": 8816354.0,
      "step": 5085
    },
    {
      "entropy": 5.7759716510772705,
      "epoch": 0.39603190040848085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498859323264548,
      "loss": 5.4963,
      "mean_token_accuracy": 0.1662037640810013,
      "num_tokens": 8824802.0,
      "step": 5090
    },
    {
      "entropy": 5.677498435974121,
      "epoch": 0.3964209297801984,
      "grad_norm": 1.0,
      "learning_rate": 0.000498856534302941,
      "loss": 5.4777,
      "mean_token_accuracy": 0.1677809789776802,
      "num_tokens": 8833348.0,
      "step": 5095
    },
    {
      "entropy": 5.632452917098999,
      "epoch": 0.39680995915191597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988537419446673,
      "loss": 5.4872,
      "mean_token_accuracy": 0.16559952795505523,
      "num_tokens": 8841131.0,
      "step": 5100
    },
    {
      "entropy": 5.647958040237427,
      "epoch": 0.39719898852363356,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988509461897694,
      "loss": 5.4972,
      "mean_token_accuracy": 0.16812338382005693,
      "num_tokens": 8849768.0,
      "step": 5105
    },
    {
      "entropy": 5.73589940071106,
      "epoch": 0.3975880178953511,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988481470382897,
      "loss": 5.466,
      "mean_token_accuracy": 0.17053375840187074,
      "num_tokens": 8858170.0,
      "step": 5110
    },
    {
      "entropy": 5.652495384216309,
      "epoch": 0.3979770472670687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988453444902708,
      "loss": 5.4898,
      "mean_token_accuracy": 0.16427329331636428,
      "num_tokens": 8866720.0,
      "step": 5115
    },
    {
      "entropy": 5.744022846221924,
      "epoch": 0.3983660766387862,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988425385457549,
      "loss": 5.5027,
      "mean_token_accuracy": 0.1700969159603119,
      "num_tokens": 8875141.0,
      "step": 5120
    },
    {
      "entropy": 5.721825456619262,
      "epoch": 0.3987551060105038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988397292047848,
      "loss": 5.5215,
      "mean_token_accuracy": 0.17009811848402023,
      "num_tokens": 8884512.0,
      "step": 5125
    },
    {
      "entropy": 5.744118022918701,
      "epoch": 0.3991441353822214,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988369164674032,
      "loss": 5.5924,
      "mean_token_accuracy": 0.1634611889719963,
      "num_tokens": 8893487.0,
      "step": 5130
    },
    {
      "entropy": 5.661117696762085,
      "epoch": 0.3995331647539389,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988341003336526,
      "loss": 5.5359,
      "mean_token_accuracy": 0.1611662432551384,
      "num_tokens": 8902909.0,
      "step": 5135
    },
    {
      "entropy": 5.72418098449707,
      "epoch": 0.3999221941256565,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988312808035757,
      "loss": 5.4443,
      "mean_token_accuracy": 0.1671905130147934,
      "num_tokens": 8911311.0,
      "step": 5140
    },
    {
      "entropy": 5.624884366989136,
      "epoch": 0.400311223497374,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988284578772155,
      "loss": 5.4728,
      "mean_token_accuracy": 0.1723196566104889,
      "num_tokens": 8920222.0,
      "step": 5145
    },
    {
      "entropy": 5.642570638656617,
      "epoch": 0.4007002528690916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988256315546146,
      "loss": 5.5248,
      "mean_token_accuracy": 0.16120615154504775,
      "num_tokens": 8928763.0,
      "step": 5150
    },
    {
      "entropy": 5.795141744613647,
      "epoch": 0.4010892822408092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988228018358161,
      "loss": 5.5824,
      "mean_token_accuracy": 0.16073987632989883,
      "num_tokens": 8937832.0,
      "step": 5155
    },
    {
      "entropy": 5.649924993515015,
      "epoch": 0.40147831161252673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988199687208628,
      "loss": 5.4415,
      "mean_token_accuracy": 0.1690294101834297,
      "num_tokens": 8945662.0,
      "step": 5160
    },
    {
      "entropy": 5.631535243988037,
      "epoch": 0.4018673409842443,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988171322097977,
      "loss": 5.5209,
      "mean_token_accuracy": 0.16340990066528321,
      "num_tokens": 8954570.0,
      "step": 5165
    },
    {
      "entropy": 5.723238515853882,
      "epoch": 0.4022563703559619,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988142923026638,
      "loss": 5.5174,
      "mean_token_accuracy": 0.16258451640605925,
      "num_tokens": 8962827.0,
      "step": 5170
    },
    {
      "entropy": 5.70035228729248,
      "epoch": 0.40264539972767943,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988114489995044,
      "loss": 5.5363,
      "mean_token_accuracy": 0.16639768928289414,
      "num_tokens": 8972119.0,
      "step": 5175
    },
    {
      "entropy": 5.6737593650817875,
      "epoch": 0.403034429099397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988086023003624,
      "loss": 5.4556,
      "mean_token_accuracy": 0.16658223271369935,
      "num_tokens": 8980498.0,
      "step": 5180
    },
    {
      "entropy": 5.702155590057373,
      "epoch": 0.40342345847111455,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988057522052811,
      "loss": 5.6045,
      "mean_token_accuracy": 0.16114473044872285,
      "num_tokens": 8989819.0,
      "step": 5185
    },
    {
      "entropy": 5.76625337600708,
      "epoch": 0.40381248784283214,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988028987143037,
      "loss": 5.6453,
      "mean_token_accuracy": 0.15599432587623596,
      "num_tokens": 8998884.0,
      "step": 5190
    },
    {
      "entropy": 5.686150741577149,
      "epoch": 0.4042015172145497,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988000418274736,
      "loss": 5.5014,
      "mean_token_accuracy": 0.1671816885471344,
      "num_tokens": 9007806.0,
      "step": 5195
    },
    {
      "entropy": 5.734596157073975,
      "epoch": 0.40459054658626725,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987971815448341,
      "loss": 5.6647,
      "mean_token_accuracy": 0.15147211849689485,
      "num_tokens": 9016541.0,
      "step": 5200
    },
    {
      "entropy": 5.6630021095275875,
      "epoch": 0.40497957595798484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987943178664285,
      "loss": 5.429,
      "mean_token_accuracy": 0.16656740009784698,
      "num_tokens": 9024503.0,
      "step": 5205
    },
    {
      "entropy": 5.718490028381348,
      "epoch": 0.40536860532970237,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987914507923004,
      "loss": 5.5085,
      "mean_token_accuracy": 0.16654182076454163,
      "num_tokens": 9033182.0,
      "step": 5210
    },
    {
      "entropy": 5.655713081359863,
      "epoch": 0.40575763470141996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987885803224931,
      "loss": 5.5342,
      "mean_token_accuracy": 0.1643572121858597,
      "num_tokens": 9042281.0,
      "step": 5215
    },
    {
      "entropy": 5.745838260650634,
      "epoch": 0.40614666407313754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987857064570505,
      "loss": 5.447,
      "mean_token_accuracy": 0.1690574824810028,
      "num_tokens": 9050570.0,
      "step": 5220
    },
    {
      "entropy": 5.760504627227784,
      "epoch": 0.4065356934448551,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987828291960158,
      "loss": 5.6251,
      "mean_token_accuracy": 0.1569422021508217,
      "num_tokens": 9059782.0,
      "step": 5225
    },
    {
      "entropy": 5.624711227416992,
      "epoch": 0.40692472281657266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498779948539433,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1710534617304802,
      "num_tokens": 9067484.0,
      "step": 5230
    },
    {
      "entropy": 5.667460870742798,
      "epoch": 0.4073137521882902,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987770644873455,
      "loss": 5.5824,
      "mean_token_accuracy": 0.15947526395320893,
      "num_tokens": 9075740.0,
      "step": 5235
    },
    {
      "entropy": 5.768729782104492,
      "epoch": 0.4077027815600078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987741770397974,
      "loss": 5.6286,
      "mean_token_accuracy": 0.15769142508506775,
      "num_tokens": 9084137.0,
      "step": 5240
    },
    {
      "entropy": 5.665251207351685,
      "epoch": 0.40809181093172536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987712861968323,
      "loss": 5.4091,
      "mean_token_accuracy": 0.17634232938289643,
      "num_tokens": 9093245.0,
      "step": 5245
    },
    {
      "entropy": 5.578675079345703,
      "epoch": 0.4084808403034429,
      "grad_norm": 1.0,
      "learning_rate": 0.000498768391958494,
      "loss": 5.4075,
      "mean_token_accuracy": 0.16961877197027206,
      "num_tokens": 9102837.0,
      "step": 5250
    },
    {
      "entropy": 5.6759144306182865,
      "epoch": 0.4088698696751605,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987654943248266,
      "loss": 5.4679,
      "mean_token_accuracy": 0.16916224360466003,
      "num_tokens": 9111087.0,
      "step": 5255
    },
    {
      "entropy": 5.666718244552612,
      "epoch": 0.40925889904687807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987625932958739,
      "loss": 5.4955,
      "mean_token_accuracy": 0.17213426232337953,
      "num_tokens": 9120506.0,
      "step": 5260
    },
    {
      "entropy": 5.754262399673462,
      "epoch": 0.4096479284185956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987596888716801,
      "loss": 5.6924,
      "mean_token_accuracy": 0.15665965899825096,
      "num_tokens": 9129637.0,
      "step": 5265
    },
    {
      "entropy": 5.753048419952393,
      "epoch": 0.4100369577903132,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987567810522892,
      "loss": 5.4725,
      "mean_token_accuracy": 0.16619387418031692,
      "num_tokens": 9139338.0,
      "step": 5270
    },
    {
      "entropy": 5.6472631931304935,
      "epoch": 0.4104259871620307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987538698377451,
      "loss": 5.5164,
      "mean_token_accuracy": 0.16488752961158754,
      "num_tokens": 9148034.0,
      "step": 5275
    },
    {
      "entropy": 5.662676572799683,
      "epoch": 0.4108150165337483,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987509552280924,
      "loss": 5.4633,
      "mean_token_accuracy": 0.16608307063579558,
      "num_tokens": 9156744.0,
      "step": 5280
    },
    {
      "entropy": 5.606441593170166,
      "epoch": 0.4112040459054659,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498748037223375,
      "loss": 5.4493,
      "mean_token_accuracy": 0.1717175617814064,
      "num_tokens": 9165174.0,
      "step": 5285
    },
    {
      "entropy": 5.640111970901489,
      "epoch": 0.4115930752771834,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987451158236372,
      "loss": 5.451,
      "mean_token_accuracy": 0.17056777477264404,
      "num_tokens": 9173945.0,
      "step": 5290
    },
    {
      "entropy": 5.71455397605896,
      "epoch": 0.411982104648901,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987421910289234,
      "loss": 5.4518,
      "mean_token_accuracy": 0.16707825809717178,
      "num_tokens": 9182728.0,
      "step": 5295
    },
    {
      "entropy": 5.591234254837036,
      "epoch": 0.41237113402061853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987392628392781,
      "loss": 5.4896,
      "mean_token_accuracy": 0.1636035606265068,
      "num_tokens": 9191494.0,
      "step": 5300
    },
    {
      "entropy": 5.655320119857788,
      "epoch": 0.4127601633923361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987363312547456,
      "loss": 5.5143,
      "mean_token_accuracy": 0.16841454058885574,
      "num_tokens": 9199190.0,
      "step": 5305
    },
    {
      "entropy": 5.651890897750855,
      "epoch": 0.4131491927640537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987333962753703,
      "loss": 5.4463,
      "mean_token_accuracy": 0.16756607592105865,
      "num_tokens": 9208066.0,
      "step": 5310
    },
    {
      "entropy": 5.643588495254517,
      "epoch": 0.41353822213577124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987304579011967,
      "loss": 5.4525,
      "mean_token_accuracy": 0.1683342546224594,
      "num_tokens": 9217103.0,
      "step": 5315
    },
    {
      "entropy": 5.648463869094849,
      "epoch": 0.4139272515074888,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987275161322697,
      "loss": 5.4995,
      "mean_token_accuracy": 0.16448337435722352,
      "num_tokens": 9225786.0,
      "step": 5320
    },
    {
      "entropy": 5.739774513244629,
      "epoch": 0.41431628087920636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987245709686337,
      "loss": 5.5035,
      "mean_token_accuracy": 0.171517077088356,
      "num_tokens": 9234182.0,
      "step": 5325
    },
    {
      "entropy": 5.693743753433227,
      "epoch": 0.41470531025092394,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987216224103334,
      "loss": 5.4811,
      "mean_token_accuracy": 0.1606791913509369,
      "num_tokens": 9243844.0,
      "step": 5330
    },
    {
      "entropy": 5.649951553344726,
      "epoch": 0.41509433962264153,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987186704574136,
      "loss": 5.4631,
      "mean_token_accuracy": 0.17001460939645768,
      "num_tokens": 9252901.0,
      "step": 5335
    },
    {
      "entropy": 5.588115978240967,
      "epoch": 0.41548336899435906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987157151099189,
      "loss": 5.4564,
      "mean_token_accuracy": 0.16492853313684464,
      "num_tokens": 9260947.0,
      "step": 5340
    },
    {
      "entropy": 5.7257404804229735,
      "epoch": 0.41587239836607665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987127563678945,
      "loss": 5.5146,
      "mean_token_accuracy": 0.1641767904162407,
      "num_tokens": 9269249.0,
      "step": 5345
    },
    {
      "entropy": 5.650511598587036,
      "epoch": 0.4162614277377942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987097942313852,
      "loss": 5.5138,
      "mean_token_accuracy": 0.16471074074506759,
      "num_tokens": 9277668.0,
      "step": 5350
    },
    {
      "entropy": 5.667751789093018,
      "epoch": 0.41665045710951176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987068287004355,
      "loss": 5.5103,
      "mean_token_accuracy": 0.15986161753535272,
      "num_tokens": 9286000.0,
      "step": 5355
    },
    {
      "entropy": 5.684542274475097,
      "epoch": 0.41703948648122935,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987038597750909,
      "loss": 5.5558,
      "mean_token_accuracy": 0.15900709629058837,
      "num_tokens": 9295090.0,
      "step": 5360
    },
    {
      "entropy": 5.656639814376831,
      "epoch": 0.4174285158529469,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987008874553964,
      "loss": 5.4706,
      "mean_token_accuracy": 0.16817761659622193,
      "num_tokens": 9302996.0,
      "step": 5365
    },
    {
      "entropy": 5.613350296020508,
      "epoch": 0.41781754522466447,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986979117413969,
      "loss": 5.3777,
      "mean_token_accuracy": 0.17125458717346193,
      "num_tokens": 9311636.0,
      "step": 5370
    },
    {
      "entropy": 5.645034265518189,
      "epoch": 0.41820657459638205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986949326331376,
      "loss": 5.4855,
      "mean_token_accuracy": 0.16353526413440705,
      "num_tokens": 9319884.0,
      "step": 5375
    },
    {
      "entropy": 5.59897813796997,
      "epoch": 0.4185956039680996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986919501306638,
      "loss": 5.4166,
      "mean_token_accuracy": 0.16931099370121955,
      "num_tokens": 9328273.0,
      "step": 5380
    },
    {
      "entropy": 5.65986647605896,
      "epoch": 0.41898463333981717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986889642340207,
      "loss": 5.5473,
      "mean_token_accuracy": 0.1573350176215172,
      "num_tokens": 9337696.0,
      "step": 5385
    },
    {
      "entropy": 5.6967016696929935,
      "epoch": 0.4193736627115347,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986859749432536,
      "loss": 5.5193,
      "mean_token_accuracy": 0.16329908072948457,
      "num_tokens": 9346596.0,
      "step": 5390
    },
    {
      "entropy": 5.662880325317383,
      "epoch": 0.4197626920832523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986829822584078,
      "loss": 5.4472,
      "mean_token_accuracy": 0.16471709311008453,
      "num_tokens": 9355480.0,
      "step": 5395
    },
    {
      "entropy": 5.650768995285034,
      "epoch": 0.4201517214549699,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986799861795289,
      "loss": 5.5032,
      "mean_token_accuracy": 0.166390161216259,
      "num_tokens": 9364422.0,
      "step": 5400
    },
    {
      "entropy": 5.634923124313355,
      "epoch": 0.4205407508266874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986769867066622,
      "loss": 5.3819,
      "mean_token_accuracy": 0.16537950336933135,
      "num_tokens": 9372311.0,
      "step": 5405
    },
    {
      "entropy": 5.655995321273804,
      "epoch": 0.420929780198405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986739838398532,
      "loss": 5.5105,
      "mean_token_accuracy": 0.16196997761726378,
      "num_tokens": 9380936.0,
      "step": 5410
    },
    {
      "entropy": 5.6495442390441895,
      "epoch": 0.4213188095701225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986709775791475,
      "loss": 5.4565,
      "mean_token_accuracy": 0.16568768620491028,
      "num_tokens": 9390008.0,
      "step": 5415
    },
    {
      "entropy": 5.68368182182312,
      "epoch": 0.4217078389418401,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986679679245907,
      "loss": 5.539,
      "mean_token_accuracy": 0.16439356058835983,
      "num_tokens": 9398771.0,
      "step": 5420
    },
    {
      "entropy": 5.65579195022583,
      "epoch": 0.4220968683135577,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986649548762286,
      "loss": 5.5314,
      "mean_token_accuracy": 0.16600719690322877,
      "num_tokens": 9407105.0,
      "step": 5425
    },
    {
      "entropy": 5.6679205894470215,
      "epoch": 0.4224858976852752,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986619384341066,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1649916112422943,
      "num_tokens": 9415737.0,
      "step": 5430
    },
    {
      "entropy": 5.6381096839904785,
      "epoch": 0.4228749270569928,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986589185982708,
      "loss": 5.4651,
      "mean_token_accuracy": 0.16958317011594773,
      "num_tokens": 9424661.0,
      "step": 5435
    },
    {
      "entropy": 5.731217908859253,
      "epoch": 0.42326395642871034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986558953687671,
      "loss": 5.535,
      "mean_token_accuracy": 0.1611144796013832,
      "num_tokens": 9433971.0,
      "step": 5440
    },
    {
      "entropy": 5.628467655181884,
      "epoch": 0.42365298580042793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986528687456409,
      "loss": 5.4786,
      "mean_token_accuracy": 0.16527963280677796,
      "num_tokens": 9442905.0,
      "step": 5445
    },
    {
      "entropy": 5.631464242935181,
      "epoch": 0.4240420151721455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986498387289384,
      "loss": 5.5061,
      "mean_token_accuracy": 0.1702937290072441,
      "num_tokens": 9451258.0,
      "step": 5450
    },
    {
      "entropy": 5.669587421417236,
      "epoch": 0.42443104454386305,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986468053187058,
      "loss": 5.4643,
      "mean_token_accuracy": 0.16737881749868394,
      "num_tokens": 9460218.0,
      "step": 5455
    },
    {
      "entropy": 5.6664165496826175,
      "epoch": 0.42482007391558063,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986437685149887,
      "loss": 5.5074,
      "mean_token_accuracy": 0.1663854479789734,
      "num_tokens": 9468756.0,
      "step": 5460
    },
    {
      "entropy": 5.695584869384765,
      "epoch": 0.4252091032872982,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986407283178334,
      "loss": 5.528,
      "mean_token_accuracy": 0.16710674911737441,
      "num_tokens": 9477622.0,
      "step": 5465
    },
    {
      "entropy": 5.6477516174316404,
      "epoch": 0.42559813265901575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986376847272861,
      "loss": 5.4023,
      "mean_token_accuracy": 0.17134957611560822,
      "num_tokens": 9486332.0,
      "step": 5470
    },
    {
      "entropy": 5.6369672298431395,
      "epoch": 0.42598716203073334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986346377433929,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1770221471786499,
      "num_tokens": 9494703.0,
      "step": 5475
    },
    {
      "entropy": 5.6113409996032715,
      "epoch": 0.42637619140245087,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986315873661999,
      "loss": 5.3682,
      "mean_token_accuracy": 0.17947474718093873,
      "num_tokens": 9503008.0,
      "step": 5480
    },
    {
      "entropy": 5.647426748275757,
      "epoch": 0.42676522077416845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986285335957536,
      "loss": 5.475,
      "mean_token_accuracy": 0.16509030908346176,
      "num_tokens": 9511752.0,
      "step": 5485
    },
    {
      "entropy": 5.595451688766479,
      "epoch": 0.42715425014588604,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004986254764321002,
      "loss": 5.3896,
      "mean_token_accuracy": 0.16756030768156052,
      "num_tokens": 9520490.0,
      "step": 5490
    },
    {
      "entropy": 5.626486682891846,
      "epoch": 0.42754327951760357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986224158752861,
      "loss": 5.4421,
      "mean_token_accuracy": 0.16567257046699524,
      "num_tokens": 9528474.0,
      "step": 5495
    },
    {
      "entropy": 5.688561153411865,
      "epoch": 0.42793230888932116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986193519253578,
      "loss": 5.5345,
      "mean_token_accuracy": 0.16336488127708435,
      "num_tokens": 9537747.0,
      "step": 5500
    },
    {
      "entropy": 5.632611274719238,
      "epoch": 0.4283213382610387,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986162845823618,
      "loss": 5.4534,
      "mean_token_accuracy": 0.16207062304019929,
      "num_tokens": 9546414.0,
      "step": 5505
    },
    {
      "entropy": 5.664350414276123,
      "epoch": 0.4287103676327563,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986132138463446,
      "loss": 5.4661,
      "mean_token_accuracy": 0.1694653421640396,
      "num_tokens": 9555166.0,
      "step": 5510
    },
    {
      "entropy": 5.688541412353516,
      "epoch": 0.42909939700447386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986101397173528,
      "loss": 5.519,
      "mean_token_accuracy": 0.16417355090379715,
      "num_tokens": 9563217.0,
      "step": 5515
    },
    {
      "entropy": 5.645942831039429,
      "epoch": 0.4294884263761914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498607062195433,
      "loss": 5.5671,
      "mean_token_accuracy": 0.15941138565540314,
      "num_tokens": 9572837.0,
      "step": 5520
    },
    {
      "entropy": 5.718004369735718,
      "epoch": 0.429877455747909,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498603981280632,
      "loss": 5.462,
      "mean_token_accuracy": 0.16318256556987762,
      "num_tokens": 9581297.0,
      "step": 5525
    },
    {
      "entropy": 5.6322996616363525,
      "epoch": 0.4302664851196265,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986008969729964,
      "loss": 5.5443,
      "mean_token_accuracy": 0.1617195874452591,
      "num_tokens": 9590431.0,
      "step": 5530
    },
    {
      "entropy": 5.723177671432495,
      "epoch": 0.4306555144913441,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985978092725731,
      "loss": 5.5917,
      "mean_token_accuracy": 0.16330402940511704,
      "num_tokens": 9598719.0,
      "step": 5535
    },
    {
      "entropy": 5.7271167755126955,
      "epoch": 0.4310445438630617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498594718179409,
      "loss": 5.5861,
      "mean_token_accuracy": 0.15970134884119033,
      "num_tokens": 9606935.0,
      "step": 5540
    },
    {
      "entropy": 5.710335636138916,
      "epoch": 0.4314335732347792,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985916236935508,
      "loss": 5.5288,
      "mean_token_accuracy": 0.17187016010284423,
      "num_tokens": 9615104.0,
      "step": 5545
    },
    {
      "entropy": 5.803662633895874,
      "epoch": 0.4318226026064968,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985885258150458,
      "loss": 5.5578,
      "mean_token_accuracy": 0.15836212933063507,
      "num_tokens": 9623332.0,
      "step": 5550
    },
    {
      "entropy": 5.581446504592895,
      "epoch": 0.43221163197821433,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985854245439408,
      "loss": 5.3979,
      "mean_token_accuracy": 0.16912583708763124,
      "num_tokens": 9631874.0,
      "step": 5555
    },
    {
      "entropy": 5.609959077835083,
      "epoch": 0.4326006613499319,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985823198802827,
      "loss": 5.4509,
      "mean_token_accuracy": 0.1646541029214859,
      "num_tokens": 9641150.0,
      "step": 5560
    },
    {
      "entropy": 5.571775960922241,
      "epoch": 0.4329896907216495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985792118241188,
      "loss": 5.2773,
      "mean_token_accuracy": 0.1755918726325035,
      "num_tokens": 9649116.0,
      "step": 5565
    },
    {
      "entropy": 5.6169108867645265,
      "epoch": 0.43337872009336703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985761003754961,
      "loss": 5.5178,
      "mean_token_accuracy": 0.166967211663723,
      "num_tokens": 9657579.0,
      "step": 5570
    },
    {
      "entropy": 5.635841417312622,
      "epoch": 0.4337677494650846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985729855344622,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1659495487809181,
      "num_tokens": 9666242.0,
      "step": 5575
    },
    {
      "entropy": 5.669478797912598,
      "epoch": 0.4341567788368022,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498569867301064,
      "loss": 5.4663,
      "mean_token_accuracy": 0.16676506102085115,
      "num_tokens": 9675196.0,
      "step": 5580
    },
    {
      "entropy": 5.643025684356689,
      "epoch": 0.43454580820851973,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985667456753489,
      "loss": 5.4287,
      "mean_token_accuracy": 0.16679195463657379,
      "num_tokens": 9683681.0,
      "step": 5585
    },
    {
      "entropy": 5.592366361618042,
      "epoch": 0.4349348375802373,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985636206573642,
      "loss": 5.4964,
      "mean_token_accuracy": 0.1720845341682434,
      "num_tokens": 9692445.0,
      "step": 5590
    },
    {
      "entropy": 5.6447841167449955,
      "epoch": 0.43532386695195485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985604922471575,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1695060133934021,
      "num_tokens": 9701256.0,
      "step": 5595
    },
    {
      "entropy": 5.647923994064331,
      "epoch": 0.43571289632367244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985573604447761,
      "loss": 5.4284,
      "mean_token_accuracy": 0.16670031696558,
      "num_tokens": 9710351.0,
      "step": 5600
    },
    {
      "entropy": 5.681913232803344,
      "epoch": 0.43610192569539,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985542252502676,
      "loss": 5.5098,
      "mean_token_accuracy": 0.15940970182418823,
      "num_tokens": 9718795.0,
      "step": 5605
    },
    {
      "entropy": 5.679136896133423,
      "epoch": 0.43649095506710756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985510866636796,
      "loss": 5.4323,
      "mean_token_accuracy": 0.16713555306196212,
      "num_tokens": 9727136.0,
      "step": 5610
    },
    {
      "entropy": 5.578325414657593,
      "epoch": 0.43687998443882514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985479446850596,
      "loss": 5.4182,
      "mean_token_accuracy": 0.17000139653682708,
      "num_tokens": 9735712.0,
      "step": 5615
    },
    {
      "entropy": 5.6629225730896,
      "epoch": 0.4372690138105427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985447993144554,
      "loss": 5.4542,
      "mean_token_accuracy": 0.16782625317573546,
      "num_tokens": 9744027.0,
      "step": 5620
    },
    {
      "entropy": 5.750279188156128,
      "epoch": 0.43765804318226026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985416505519147,
      "loss": 5.59,
      "mean_token_accuracy": 0.1612232133746147,
      "num_tokens": 9753083.0,
      "step": 5625
    },
    {
      "entropy": 5.72616057395935,
      "epoch": 0.43804707255397785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985384983974853,
      "loss": 5.5179,
      "mean_token_accuracy": 0.16758409440517424,
      "num_tokens": 9762098.0,
      "step": 5630
    },
    {
      "entropy": 5.692498731613159,
      "epoch": 0.4384361019256954,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985353428512148,
      "loss": 5.4615,
      "mean_token_accuracy": 0.16436221897602082,
      "num_tokens": 9770956.0,
      "step": 5635
    },
    {
      "entropy": 5.574732780456543,
      "epoch": 0.43882513129741296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985321839131514,
      "loss": 5.4127,
      "mean_token_accuracy": 0.1726594388484955,
      "num_tokens": 9779427.0,
      "step": 5640
    },
    {
      "entropy": 5.621200513839722,
      "epoch": 0.4392141606691305,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985290215833428,
      "loss": 5.4777,
      "mean_token_accuracy": 0.17316209226846696,
      "num_tokens": 9787655.0,
      "step": 5645
    },
    {
      "entropy": 5.668504667282105,
      "epoch": 0.4396031900408481,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985258558618372,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1715027719736099,
      "num_tokens": 9796850.0,
      "step": 5650
    },
    {
      "entropy": 5.651020812988281,
      "epoch": 0.43999221941256567,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985226867486825,
      "loss": 5.5034,
      "mean_token_accuracy": 0.17284105122089385,
      "num_tokens": 9805317.0,
      "step": 5655
    },
    {
      "entropy": 5.650268125534057,
      "epoch": 0.4403812487842832,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985195142439267,
      "loss": 5.4416,
      "mean_token_accuracy": 0.17098015695810317,
      "num_tokens": 9814249.0,
      "step": 5660
    },
    {
      "entropy": 5.67411699295044,
      "epoch": 0.4407702781560008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985163383476181,
      "loss": 5.539,
      "mean_token_accuracy": 0.16412654519081116,
      "num_tokens": 9823122.0,
      "step": 5665
    },
    {
      "entropy": 5.595461463928222,
      "epoch": 0.44115930752771837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985131590598048,
      "loss": 5.4086,
      "mean_token_accuracy": 0.17494799494743346,
      "num_tokens": 9831470.0,
      "step": 5670
    },
    {
      "entropy": 5.623825263977051,
      "epoch": 0.4415483368994359,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985099763805352,
      "loss": 5.5128,
      "mean_token_accuracy": 0.16640195101499558,
      "num_tokens": 9840096.0,
      "step": 5675
    },
    {
      "entropy": 5.687500905990601,
      "epoch": 0.4419373662711535,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985067903098574,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17098743170499803,
      "num_tokens": 9848595.0,
      "step": 5680
    },
    {
      "entropy": 5.637724256515503,
      "epoch": 0.442326395642871,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985036008478197,
      "loss": 5.4073,
      "mean_token_accuracy": 0.16370895206928254,
      "num_tokens": 9857353.0,
      "step": 5685
    },
    {
      "entropy": 5.63226637840271,
      "epoch": 0.4427154250145886,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985004079944707,
      "loss": 5.46,
      "mean_token_accuracy": 0.16471842527389527,
      "num_tokens": 9865828.0,
      "step": 5690
    },
    {
      "entropy": 5.593899774551391,
      "epoch": 0.4431044543863062,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984972117498587,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16496108174324037,
      "num_tokens": 9874540.0,
      "step": 5695
    },
    {
      "entropy": 5.677505302429199,
      "epoch": 0.4434934837580237,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984940121140323,
      "loss": 5.4622,
      "mean_token_accuracy": 0.16470122337341309,
      "num_tokens": 9883046.0,
      "step": 5700
    },
    {
      "entropy": 5.6196901321411135,
      "epoch": 0.4438825131297413,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049849080908704,
      "loss": 5.4166,
      "mean_token_accuracy": 0.16411941945552827,
      "num_tokens": 9892282.0,
      "step": 5705
    },
    {
      "entropy": 5.635061359405517,
      "epoch": 0.44427154250145884,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984876026689302,
      "loss": 5.3502,
      "mean_token_accuracy": 0.17350826859474183,
      "num_tokens": 9900653.0,
      "step": 5710
    },
    {
      "entropy": 5.568204784393311,
      "epoch": 0.4446605718731764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498484392859752,
      "loss": 5.3852,
      "mean_token_accuracy": 0.17214204668998717,
      "num_tokens": 9909323.0,
      "step": 5715
    },
    {
      "entropy": 5.582053184509277,
      "epoch": 0.445049601244894,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984811796595538,
      "loss": 5.5187,
      "mean_token_accuracy": 0.16376986354589462,
      "num_tokens": 9918155.0,
      "step": 5720
    },
    {
      "entropy": 5.747524881362915,
      "epoch": 0.44543863061661154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984779630683843,
      "loss": 5.4763,
      "mean_token_accuracy": 0.16589940339326859,
      "num_tokens": 9925977.0,
      "step": 5725
    },
    {
      "entropy": 5.637293529510498,
      "epoch": 0.44582765998832913,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984747430862924,
      "loss": 5.4612,
      "mean_token_accuracy": 0.16873055398464204,
      "num_tokens": 9934606.0,
      "step": 5730
    },
    {
      "entropy": 5.580166482925415,
      "epoch": 0.44621668936004666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984715197133271,
      "loss": 5.4339,
      "mean_token_accuracy": 0.16639334559440613,
      "num_tokens": 9943638.0,
      "step": 5735
    },
    {
      "entropy": 5.641740894317627,
      "epoch": 0.44660571873176425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984682929495371,
      "loss": 5.4527,
      "mean_token_accuracy": 0.16659653931856155,
      "num_tokens": 9952686.0,
      "step": 5740
    },
    {
      "entropy": 5.693140602111816,
      "epoch": 0.44699474810348183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984650627949715,
      "loss": 5.6415,
      "mean_token_accuracy": 0.15615570396184922,
      "num_tokens": 9961947.0,
      "step": 5745
    },
    {
      "entropy": 5.670505428314209,
      "epoch": 0.44738377747519936,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984618292496792,
      "loss": 5.4647,
      "mean_token_accuracy": 0.16673058420419692,
      "num_tokens": 9971218.0,
      "step": 5750
    },
    {
      "entropy": 5.664007711410522,
      "epoch": 0.44777280684691695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984585923137093,
      "loss": 5.4916,
      "mean_token_accuracy": 0.1662154719233513,
      "num_tokens": 9980741.0,
      "step": 5755
    },
    {
      "entropy": 5.694485855102539,
      "epoch": 0.4481618362186345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498455351987111,
      "loss": 5.4705,
      "mean_token_accuracy": 0.16652709394693374,
      "num_tokens": 9989483.0,
      "step": 5760
    },
    {
      "entropy": 5.693090534210205,
      "epoch": 0.44855086559035207,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984521082699333,
      "loss": 5.4978,
      "mean_token_accuracy": 0.16687163412570954,
      "num_tokens": 9998491.0,
      "step": 5765
    },
    {
      "entropy": 5.543629884719849,
      "epoch": 0.44893989496206965,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984488611622256,
      "loss": 5.3806,
      "mean_token_accuracy": 0.17117684185504914,
      "num_tokens": 10006276.0,
      "step": 5770
    },
    {
      "entropy": 5.5802089214324955,
      "epoch": 0.4493289243337872,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498445610664037,
      "loss": 5.4696,
      "mean_token_accuracy": 0.16785627007484435,
      "num_tokens": 10015088.0,
      "step": 5775
    },
    {
      "entropy": 5.663906669616699,
      "epoch": 0.44971795370550477,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498442356775417,
      "loss": 5.4266,
      "mean_token_accuracy": 0.16554765850305558,
      "num_tokens": 10023305.0,
      "step": 5780
    },
    {
      "entropy": 5.575909042358399,
      "epoch": 0.45010698307722236,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984390994964148,
      "loss": 5.3805,
      "mean_token_accuracy": 0.16881942003965378,
      "num_tokens": 10031540.0,
      "step": 5785
    },
    {
      "entropy": 5.6279661655426025,
      "epoch": 0.4504960124489399,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049843583882708,
      "loss": 5.4733,
      "mean_token_accuracy": 0.16908168643712998,
      "num_tokens": 10040025.0,
      "step": 5790
    },
    {
      "entropy": 5.6401628971099855,
      "epoch": 0.4508850418206575,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498432574767462,
      "loss": 5.48,
      "mean_token_accuracy": 0.16680154353380203,
      "num_tokens": 10048904.0,
      "step": 5795
    },
    {
      "entropy": 5.644533729553222,
      "epoch": 0.451274071192375,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984293073176103,
      "loss": 5.5133,
      "mean_token_accuracy": 0.16642902791500092,
      "num_tokens": 10057608.0,
      "step": 5800
    },
    {
      "entropy": 5.713375568389893,
      "epoch": 0.4516631005640926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984260364775744,
      "loss": 5.4701,
      "mean_token_accuracy": 0.17314359098672866,
      "num_tokens": 10066587.0,
      "step": 5805
    },
    {
      "entropy": 5.608742380142212,
      "epoch": 0.4520521299358102,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498422762247404,
      "loss": 5.4281,
      "mean_token_accuracy": 0.1667902648448944,
      "num_tokens": 10075651.0,
      "step": 5810
    },
    {
      "entropy": 5.65361647605896,
      "epoch": 0.4524411593075277,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984194846271489,
      "loss": 5.5379,
      "mean_token_accuracy": 0.16456001549959182,
      "num_tokens": 10084158.0,
      "step": 5815
    },
    {
      "entropy": 5.714414310455322,
      "epoch": 0.4528301886792453,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984162036168586,
      "loss": 5.4715,
      "mean_token_accuracy": 0.16263072341680526,
      "num_tokens": 10093730.0,
      "step": 5820
    },
    {
      "entropy": 5.636698055267334,
      "epoch": 0.4532192180509628,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984129192165831,
      "loss": 5.3956,
      "mean_token_accuracy": 0.16606880724430084,
      "num_tokens": 10102457.0,
      "step": 5825
    },
    {
      "entropy": 5.657480669021607,
      "epoch": 0.4536082474226804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984096314263722,
      "loss": 5.5361,
      "mean_token_accuracy": 0.1659406006336212,
      "num_tokens": 10111188.0,
      "step": 5830
    },
    {
      "entropy": 5.691649389266968,
      "epoch": 0.453997276794398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984063402462757,
      "loss": 5.4321,
      "mean_token_accuracy": 0.1693912699818611,
      "num_tokens": 10120268.0,
      "step": 5835
    },
    {
      "entropy": 5.646623325347901,
      "epoch": 0.45438630616611553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984030456763435,
      "loss": 5.5018,
      "mean_token_accuracy": 0.159344282746315,
      "num_tokens": 10129262.0,
      "step": 5840
    },
    {
      "entropy": 5.604523992538452,
      "epoch": 0.4547753355378331,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983997477166257,
      "loss": 5.4425,
      "mean_token_accuracy": 0.17042097300291062,
      "num_tokens": 10137417.0,
      "step": 5845
    },
    {
      "entropy": 5.60966420173645,
      "epoch": 0.45516436490955064,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983964463671723,
      "loss": 5.3859,
      "mean_token_accuracy": 0.169807331264019,
      "num_tokens": 10145804.0,
      "step": 5850
    },
    {
      "entropy": 5.6830277919769285,
      "epoch": 0.45555339428126823,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983931416280334,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1644147291779518,
      "num_tokens": 10154421.0,
      "step": 5855
    },
    {
      "entropy": 5.750033950805664,
      "epoch": 0.4559424236529858,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983898334992591,
      "loss": 5.5348,
      "mean_token_accuracy": 0.16695573478937148,
      "num_tokens": 10163631.0,
      "step": 5860
    },
    {
      "entropy": 5.64238224029541,
      "epoch": 0.45633145302470335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983865219808998,
      "loss": 5.4525,
      "mean_token_accuracy": 0.1701275646686554,
      "num_tokens": 10172926.0,
      "step": 5865
    },
    {
      "entropy": 5.679831504821777,
      "epoch": 0.45672048239642093,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004983832070730055,
      "loss": 5.5529,
      "mean_token_accuracy": 0.15790846645832063,
      "num_tokens": 10181671.0,
      "step": 5870
    },
    {
      "entropy": 5.6128734111785885,
      "epoch": 0.4571095117681385,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983798887756265,
      "loss": 5.3875,
      "mean_token_accuracy": 0.17633408457040786,
      "num_tokens": 10189626.0,
      "step": 5875
    },
    {
      "entropy": 5.681553602218628,
      "epoch": 0.45749854113985605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983765670888133,
      "loss": 5.5232,
      "mean_token_accuracy": 0.16761491298675538,
      "num_tokens": 10198212.0,
      "step": 5880
    },
    {
      "entropy": 5.586812591552734,
      "epoch": 0.45788757051157364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983732420126163,
      "loss": 5.4185,
      "mean_token_accuracy": 0.16471427530050278,
      "num_tokens": 10207342.0,
      "step": 5885
    },
    {
      "entropy": 5.704931020736694,
      "epoch": 0.45827659988329117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983699135470858,
      "loss": 5.5231,
      "mean_token_accuracy": 0.15820106565952302,
      "num_tokens": 10216413.0,
      "step": 5890
    },
    {
      "entropy": 5.657903099060059,
      "epoch": 0.45866562925500876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983665816922723,
      "loss": 5.4524,
      "mean_token_accuracy": 0.16937243938446045,
      "num_tokens": 10225889.0,
      "step": 5895
    },
    {
      "entropy": 5.589326333999634,
      "epoch": 0.45905465862672634,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983632464482267,
      "loss": 5.4063,
      "mean_token_accuracy": 0.17112260311841965,
      "num_tokens": 10234518.0,
      "step": 5900
    },
    {
      "entropy": 5.695744514465332,
      "epoch": 0.4594436879984439,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983599078149991,
      "loss": 5.5391,
      "mean_token_accuracy": 0.16143682301044465,
      "num_tokens": 10243074.0,
      "step": 5905
    },
    {
      "entropy": 5.771560573577881,
      "epoch": 0.45983271737016146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983565657926405,
      "loss": 5.4887,
      "mean_token_accuracy": 0.16415552347898482,
      "num_tokens": 10253473.0,
      "step": 5910
    },
    {
      "entropy": 5.69731650352478,
      "epoch": 0.460221746741879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983532203812017,
      "loss": 5.5747,
      "mean_token_accuracy": 0.1648534968495369,
      "num_tokens": 10261866.0,
      "step": 5915
    },
    {
      "entropy": 5.649386692047119,
      "epoch": 0.4606107761135966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983498715807331,
      "loss": 5.4187,
      "mean_token_accuracy": 0.1710118681192398,
      "num_tokens": 10270350.0,
      "step": 5920
    },
    {
      "entropy": 5.647009754180909,
      "epoch": 0.46099980548531416,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983465193912857,
      "loss": 5.5711,
      "mean_token_accuracy": 0.15795120000839233,
      "num_tokens": 10279556.0,
      "step": 5925
    },
    {
      "entropy": 5.586729669570923,
      "epoch": 0.4613888348570317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983431638129104,
      "loss": 5.3409,
      "mean_token_accuracy": 0.1722436711192131,
      "num_tokens": 10287489.0,
      "step": 5930
    },
    {
      "entropy": 5.604912328720093,
      "epoch": 0.4617778642287493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983398048456581,
      "loss": 5.3242,
      "mean_token_accuracy": 0.16806061565876007,
      "num_tokens": 10296280.0,
      "step": 5935
    },
    {
      "entropy": 5.619849300384521,
      "epoch": 0.4621668936004668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983364424895796,
      "loss": 5.4869,
      "mean_token_accuracy": 0.1619819775223732,
      "num_tokens": 10304091.0,
      "step": 5940
    },
    {
      "entropy": 5.668075275421143,
      "epoch": 0.4625559229721844,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983330767447262,
      "loss": 5.5026,
      "mean_token_accuracy": 0.15887704938650132,
      "num_tokens": 10313058.0,
      "step": 5945
    },
    {
      "entropy": 5.622173261642456,
      "epoch": 0.462944952343902,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983297076111489,
      "loss": 5.4749,
      "mean_token_accuracy": 0.1640453204512596,
      "num_tokens": 10321642.0,
      "step": 5950
    },
    {
      "entropy": 5.577755784988403,
      "epoch": 0.4633339817156195,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983263350888987,
      "loss": 5.2922,
      "mean_token_accuracy": 0.17254432886838914,
      "num_tokens": 10329829.0,
      "step": 5955
    },
    {
      "entropy": 5.640766525268555,
      "epoch": 0.4637230110873371,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983229591780268,
      "loss": 5.5284,
      "mean_token_accuracy": 0.16286519914865494,
      "num_tokens": 10338182.0,
      "step": 5960
    },
    {
      "entropy": 5.673437929153442,
      "epoch": 0.46411204045905463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983195798785844,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1703182965517044,
      "num_tokens": 10346377.0,
      "step": 5965
    },
    {
      "entropy": 5.668019008636475,
      "epoch": 0.4645010698307722,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983161971906228,
      "loss": 5.4089,
      "mean_token_accuracy": 0.17171140760183334,
      "num_tokens": 10354937.0,
      "step": 5970
    },
    {
      "entropy": 5.543028974533081,
      "epoch": 0.4648900992024898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983128111141935,
      "loss": 5.3114,
      "mean_token_accuracy": 0.17455710768699645,
      "num_tokens": 10363706.0,
      "step": 5975
    },
    {
      "entropy": 5.574260568618774,
      "epoch": 0.46527912857420733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983094216493475,
      "loss": 5.4645,
      "mean_token_accuracy": 0.16650742590427398,
      "num_tokens": 10372292.0,
      "step": 5980
    },
    {
      "entropy": 5.671193408966064,
      "epoch": 0.4656681579459249,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983060287961367,
      "loss": 5.4386,
      "mean_token_accuracy": 0.17165132462978364,
      "num_tokens": 10381260.0,
      "step": 5985
    },
    {
      "entropy": 5.648577165603638,
      "epoch": 0.4660571873176425,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983026325546123,
      "loss": 5.4539,
      "mean_token_accuracy": 0.16947742700576782,
      "num_tokens": 10390646.0,
      "step": 5990
    },
    {
      "entropy": 5.625260925292968,
      "epoch": 0.46644621668936004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982992329248257,
      "loss": 5.5166,
      "mean_token_accuracy": 0.16828270852565766,
      "num_tokens": 10399056.0,
      "step": 5995
    },
    {
      "entropy": 5.6372960090637205,
      "epoch": 0.4668352460610776,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982958299068289,
      "loss": 5.5048,
      "mean_token_accuracy": 0.16254437565803528,
      "num_tokens": 10408746.0,
      "step": 6000
    },
    {
      "epoch": 0.4668352460610776,
      "eval_entropy": 5.468773723818914,
      "eval_loss": 5.478248119354248,
      "eval_mean_token_accuracy": 0.17529629864178678,
      "eval_num_tokens": 10408746.0,
      "eval_runtime": 21.786,
      "eval_samples_per_second": 1907.559,
      "eval_steps_per_second": 238.456,
      "step": 6000
    },
    {
      "entropy": 5.667443513870239,
      "epoch": 0.46722427543279516,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498292423500673,
      "loss": 5.4276,
      "mean_token_accuracy": 0.16848916113376616,
      "num_tokens": 10416937.0,
      "step": 6005
    },
    {
      "entropy": 5.6209752559661865,
      "epoch": 0.46761330480451274,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982890137064102,
      "loss": 5.4023,
      "mean_token_accuracy": 0.17630034387111665,
      "num_tokens": 10426343.0,
      "step": 6010
    },
    {
      "entropy": 5.6079370975494385,
      "epoch": 0.46800233417623033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498285600524092,
      "loss": 5.4317,
      "mean_token_accuracy": 0.1709451973438263,
      "num_tokens": 10435014.0,
      "step": 6015
    },
    {
      "entropy": 5.557825231552124,
      "epoch": 0.46839136354794786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982821839537701,
      "loss": 5.447,
      "mean_token_accuracy": 0.16832986921072007,
      "num_tokens": 10443522.0,
      "step": 6020
    },
    {
      "entropy": 5.6745398998260494,
      "epoch": 0.46878039291966545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982787639954966,
      "loss": 5.4939,
      "mean_token_accuracy": 0.1648608475923538,
      "num_tokens": 10452202.0,
      "step": 6025
    },
    {
      "entropy": 5.631177759170532,
      "epoch": 0.469169422291383,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982753406493232,
      "loss": 5.4735,
      "mean_token_accuracy": 0.16007108837366105,
      "num_tokens": 10462293.0,
      "step": 6030
    },
    {
      "entropy": 5.586612892150879,
      "epoch": 0.46955845166310056,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982719139153018,
      "loss": 5.4051,
      "mean_token_accuracy": 0.1679594933986664,
      "num_tokens": 10470322.0,
      "step": 6035
    },
    {
      "entropy": 5.635045433044434,
      "epoch": 0.46994748103481815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982684837934845,
      "loss": 5.3935,
      "mean_token_accuracy": 0.17465801239013673,
      "num_tokens": 10479581.0,
      "step": 6040
    },
    {
      "entropy": 5.655844640731812,
      "epoch": 0.4703365104065357,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982650502839234,
      "loss": 5.4006,
      "mean_token_accuracy": 0.1719515100121498,
      "num_tokens": 10488207.0,
      "step": 6045
    },
    {
      "entropy": 5.579202795028687,
      "epoch": 0.47072553977825327,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982616133866705,
      "loss": 5.4117,
      "mean_token_accuracy": 0.16743631064891815,
      "num_tokens": 10497360.0,
      "step": 6050
    },
    {
      "entropy": 5.578354835510254,
      "epoch": 0.4711145691499708,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982581731017779,
      "loss": 5.3959,
      "mean_token_accuracy": 0.16868714839220048,
      "num_tokens": 10506131.0,
      "step": 6055
    },
    {
      "entropy": 5.615327072143555,
      "epoch": 0.4715035985216884,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982547294292979,
      "loss": 5.4404,
      "mean_token_accuracy": 0.16701407134532928,
      "num_tokens": 10514955.0,
      "step": 6060
    },
    {
      "entropy": 5.593543672561646,
      "epoch": 0.47189262789340597,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982512823692828,
      "loss": 5.4119,
      "mean_token_accuracy": 0.16133114844560623,
      "num_tokens": 10523743.0,
      "step": 6065
    },
    {
      "entropy": 5.605992555618286,
      "epoch": 0.4722816572651235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982478319217848,
      "loss": 5.3815,
      "mean_token_accuracy": 0.1775616154074669,
      "num_tokens": 10533149.0,
      "step": 6070
    },
    {
      "entropy": 5.565766048431397,
      "epoch": 0.4726706866368411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982443780868562,
      "loss": 5.417,
      "mean_token_accuracy": 0.17523194700479508,
      "num_tokens": 10541291.0,
      "step": 6075
    },
    {
      "entropy": 5.651870632171631,
      "epoch": 0.4730597160085587,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982409208645496,
      "loss": 5.4601,
      "mean_token_accuracy": 0.16842615902423858,
      "num_tokens": 10550005.0,
      "step": 6080
    },
    {
      "entropy": 5.692563199996949,
      "epoch": 0.4734487453802762,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982374602549173,
      "loss": 5.4792,
      "mean_token_accuracy": 0.16938978135585786,
      "num_tokens": 10558957.0,
      "step": 6085
    },
    {
      "entropy": 5.64795618057251,
      "epoch": 0.4738377747519938,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982339962580119,
      "loss": 5.3678,
      "mean_token_accuracy": 0.16741629391908647,
      "num_tokens": 10567293.0,
      "step": 6090
    },
    {
      "entropy": 5.580725526809692,
      "epoch": 0.4742268041237113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982305288738859,
      "loss": 5.3384,
      "mean_token_accuracy": 0.17617610692977906,
      "num_tokens": 10575546.0,
      "step": 6095
    },
    {
      "entropy": 5.528723192214966,
      "epoch": 0.4746158334954289,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982270581025919,
      "loss": 5.465,
      "mean_token_accuracy": 0.17209935039281846,
      "num_tokens": 10584779.0,
      "step": 6100
    },
    {
      "entropy": 5.587629985809326,
      "epoch": 0.4750048628671465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982235839441826,
      "loss": 5.3955,
      "mean_token_accuracy": 0.17494287937879563,
      "num_tokens": 10592852.0,
      "step": 6105
    },
    {
      "entropy": 5.655500841140747,
      "epoch": 0.475393892238864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982201063987108,
      "loss": 5.4999,
      "mean_token_accuracy": 0.1644456207752228,
      "num_tokens": 10601875.0,
      "step": 6110
    },
    {
      "entropy": 5.611720180511474,
      "epoch": 0.4757829216105816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982166254662292,
      "loss": 5.4688,
      "mean_token_accuracy": 0.17126654237508773,
      "num_tokens": 10610438.0,
      "step": 6115
    },
    {
      "entropy": 5.627661180496216,
      "epoch": 0.47617195098229914,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982131411467904,
      "loss": 5.4636,
      "mean_token_accuracy": 0.16586989760398865,
      "num_tokens": 10619479.0,
      "step": 6120
    },
    {
      "entropy": 5.612498712539673,
      "epoch": 0.4765609803540167,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982096534404476,
      "loss": 5.3893,
      "mean_token_accuracy": 0.17240849733352662,
      "num_tokens": 10628519.0,
      "step": 6125
    },
    {
      "entropy": 5.640328454971313,
      "epoch": 0.4769500097257343,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982061623472535,
      "loss": 5.3526,
      "mean_token_accuracy": 0.17426417917013168,
      "num_tokens": 10636522.0,
      "step": 6130
    },
    {
      "entropy": 5.569852495193482,
      "epoch": 0.47733903909745184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982026678672612,
      "loss": 5.3648,
      "mean_token_accuracy": 0.16899030655622482,
      "num_tokens": 10645728.0,
      "step": 6135
    },
    {
      "entropy": 5.596399021148682,
      "epoch": 0.47772806846916943,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981991700005238,
      "loss": 5.4789,
      "mean_token_accuracy": 0.1673264682292938,
      "num_tokens": 10654868.0,
      "step": 6140
    },
    {
      "entropy": 5.548848628997803,
      "epoch": 0.47811709784088696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498195668747094,
      "loss": 5.3646,
      "mean_token_accuracy": 0.17227984219789505,
      "num_tokens": 10662958.0,
      "step": 6145
    },
    {
      "entropy": 5.6338269233703615,
      "epoch": 0.47850612721260455,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981921641070254,
      "loss": 5.3796,
      "mean_token_accuracy": 0.17123519778251647,
      "num_tokens": 10671333.0,
      "step": 6150
    },
    {
      "entropy": 5.627652263641357,
      "epoch": 0.47889515658432213,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981886560803708,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1699555292725563,
      "num_tokens": 10680523.0,
      "step": 6155
    },
    {
      "entropy": 5.619808626174927,
      "epoch": 0.47928418595603967,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981851446671838,
      "loss": 5.4687,
      "mean_token_accuracy": 0.16397960782051085,
      "num_tokens": 10689656.0,
      "step": 6160
    },
    {
      "entropy": 5.666473770141602,
      "epoch": 0.47967321532775725,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004981816298675173,
      "loss": 5.5385,
      "mean_token_accuracy": 0.16096996814012526,
      "num_tokens": 10698423.0,
      "step": 6165
    },
    {
      "entropy": 5.625974845886231,
      "epoch": 0.4800622446994748,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981781116814248,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1723208859562874,
      "num_tokens": 10706760.0,
      "step": 6170
    },
    {
      "entropy": 5.707794189453125,
      "epoch": 0.48045127407119237,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981745901089596,
      "loss": 5.5088,
      "mean_token_accuracy": 0.1589950516819954,
      "num_tokens": 10716221.0,
      "step": 6175
    },
    {
      "entropy": 5.5495470523834225,
      "epoch": 0.48084030344290996,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981710651501753,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18165120333433152,
      "num_tokens": 10724603.0,
      "step": 6180
    },
    {
      "entropy": 5.645468711853027,
      "epoch": 0.4812293328146275,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981675368051254,
      "loss": 5.5351,
      "mean_token_accuracy": 0.15899534821510314,
      "num_tokens": 10733577.0,
      "step": 6185
    },
    {
      "entropy": 5.712690258026123,
      "epoch": 0.4816183621863451,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981640050738633,
      "loss": 5.4466,
      "mean_token_accuracy": 0.17270689308643342,
      "num_tokens": 10743199.0,
      "step": 6190
    },
    {
      "entropy": 5.616707849502563,
      "epoch": 0.48200739155806266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981604699564426,
      "loss": 5.3312,
      "mean_token_accuracy": 0.17357029765844345,
      "num_tokens": 10751734.0,
      "step": 6195
    },
    {
      "entropy": 5.6283769607543945,
      "epoch": 0.4823964209297802,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004981569314529169,
      "loss": 5.4868,
      "mean_token_accuracy": 0.16709917336702346,
      "num_tokens": 10760194.0,
      "step": 6200
    },
    {
      "entropy": 5.650402545928955,
      "epoch": 0.4827854503014978,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049815338956334,
      "loss": 5.4036,
      "mean_token_accuracy": 0.17521733343601226,
      "num_tokens": 10768944.0,
      "step": 6205
    },
    {
      "entropy": 5.69504828453064,
      "epoch": 0.4831744796732153,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981498442877656,
      "loss": 5.5049,
      "mean_token_accuracy": 0.1730547532439232,
      "num_tokens": 10777685.0,
      "step": 6210
    },
    {
      "entropy": 5.642111921310425,
      "epoch": 0.4835635090449329,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981462956262474,
      "loss": 5.4765,
      "mean_token_accuracy": 0.16693192422389985,
      "num_tokens": 10787264.0,
      "step": 6215
    },
    {
      "entropy": 5.602992391586303,
      "epoch": 0.4839525384166505,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981427435788396,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1655664786696434,
      "num_tokens": 10796555.0,
      "step": 6220
    },
    {
      "entropy": 5.525999927520752,
      "epoch": 0.484341567788368,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981391881455957,
      "loss": 5.3515,
      "mean_token_accuracy": 0.17589380443096161,
      "num_tokens": 10805379.0,
      "step": 6225
    },
    {
      "entropy": 5.565594863891602,
      "epoch": 0.4847305971600856,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981356293265696,
      "loss": 5.4786,
      "mean_token_accuracy": 0.16520614624023439,
      "num_tokens": 10815406.0,
      "step": 6230
    },
    {
      "entropy": 5.658706951141357,
      "epoch": 0.4851196265318031,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981320671218157,
      "loss": 5.5369,
      "mean_token_accuracy": 0.16370739042758942,
      "num_tokens": 10824443.0,
      "step": 6235
    },
    {
      "entropy": 5.686445903778076,
      "epoch": 0.4855086559035207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981285015313877,
      "loss": 5.366,
      "mean_token_accuracy": 0.1736448109149933,
      "num_tokens": 10833497.0,
      "step": 6240
    },
    {
      "entropy": 5.557299995422364,
      "epoch": 0.4858976852752383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981249325553399,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1717936083674431,
      "num_tokens": 10841864.0,
      "step": 6245
    },
    {
      "entropy": 5.662262201309204,
      "epoch": 0.48628671464695583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981213601937264,
      "loss": 5.4643,
      "mean_token_accuracy": 0.17284016013145448,
      "num_tokens": 10850887.0,
      "step": 6250
    },
    {
      "entropy": 5.590152788162231,
      "epoch": 0.4866757440186734,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981177844466013,
      "loss": 5.3453,
      "mean_token_accuracy": 0.174080491065979,
      "num_tokens": 10859629.0,
      "step": 6255
    },
    {
      "entropy": 5.652895593643189,
      "epoch": 0.48706477339039095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981142053140192,
      "loss": 5.5285,
      "mean_token_accuracy": 0.16324191689491271,
      "num_tokens": 10868457.0,
      "step": 6260
    },
    {
      "entropy": 5.6257092475891115,
      "epoch": 0.48745380276210853,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981106227960339,
      "loss": 5.4483,
      "mean_token_accuracy": 0.16676026433706284,
      "num_tokens": 10878007.0,
      "step": 6265
    },
    {
      "entropy": 5.703521251678467,
      "epoch": 0.4878428321338261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981070368927001,
      "loss": 5.4914,
      "mean_token_accuracy": 0.16467054486274718,
      "num_tokens": 10887689.0,
      "step": 6270
    },
    {
      "entropy": 5.585311794281006,
      "epoch": 0.48823186150554365,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498103447604072,
      "loss": 5.2933,
      "mean_token_accuracy": 0.17395345568656922,
      "num_tokens": 10896214.0,
      "step": 6275
    },
    {
      "entropy": 5.55946979522705,
      "epoch": 0.48862089087726124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980998549302044,
      "loss": 5.4149,
      "mean_token_accuracy": 0.16991112530231475,
      "num_tokens": 10904946.0,
      "step": 6280
    },
    {
      "entropy": 5.693157911300659,
      "epoch": 0.4890099202489788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980962588711516,
      "loss": 5.4979,
      "mean_token_accuracy": 0.1608987919986248,
      "num_tokens": 10913382.0,
      "step": 6285
    },
    {
      "entropy": 5.672442865371704,
      "epoch": 0.48939894962069636,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498092659426968,
      "loss": 5.4338,
      "mean_token_accuracy": 0.17338544130325317,
      "num_tokens": 10922055.0,
      "step": 6290
    },
    {
      "entropy": 5.615342998504639,
      "epoch": 0.48978797899241394,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980890565977085,
      "loss": 5.4267,
      "mean_token_accuracy": 0.16947162002325059,
      "num_tokens": 10930282.0,
      "step": 6295
    },
    {
      "entropy": 5.565211153030395,
      "epoch": 0.4901770083641315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980854503834276,
      "loss": 5.4367,
      "mean_token_accuracy": 0.17593979835510254,
      "num_tokens": 10938432.0,
      "step": 6300
    },
    {
      "entropy": 5.670564031600952,
      "epoch": 0.49056603773584906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980818407841802,
      "loss": 5.4375,
      "mean_token_accuracy": 0.17329999804496765,
      "num_tokens": 10946354.0,
      "step": 6305
    },
    {
      "entropy": 5.725423860549927,
      "epoch": 0.49095506710756665,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980782278000207,
      "loss": 5.4683,
      "mean_token_accuracy": 0.16744253635406495,
      "num_tokens": 10954885.0,
      "step": 6310
    },
    {
      "entropy": 5.590633678436279,
      "epoch": 0.4913440964792842,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980746114310043,
      "loss": 5.3743,
      "mean_token_accuracy": 0.1690499573945999,
      "num_tokens": 10963173.0,
      "step": 6315
    },
    {
      "entropy": 5.634382104873657,
      "epoch": 0.49173312585100176,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980709916771858,
      "loss": 5.5258,
      "mean_token_accuracy": 0.15521946102380751,
      "num_tokens": 10972264.0,
      "step": 6320
    },
    {
      "entropy": 5.617056846618652,
      "epoch": 0.4921221552227193,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980673685386198,
      "loss": 5.3702,
      "mean_token_accuracy": 0.1731475278735161,
      "num_tokens": 10980645.0,
      "step": 6325
    },
    {
      "entropy": 5.541935157775879,
      "epoch": 0.4925111845944369,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980637420153618,
      "loss": 5.33,
      "mean_token_accuracy": 0.17462508678436278,
      "num_tokens": 10989121.0,
      "step": 6330
    },
    {
      "entropy": 5.581294679641724,
      "epoch": 0.49290021396615447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980601121074664,
      "loss": 5.3333,
      "mean_token_accuracy": 0.18154052793979644,
      "num_tokens": 10998249.0,
      "step": 6335
    },
    {
      "entropy": 5.693126916885376,
      "epoch": 0.493289243337872,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980564788149889,
      "loss": 5.448,
      "mean_token_accuracy": 0.16501189395785332,
      "num_tokens": 11007470.0,
      "step": 6340
    },
    {
      "entropy": 5.657134103775024,
      "epoch": 0.4936782727095896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980528421379844,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1677241399884224,
      "num_tokens": 11015912.0,
      "step": 6345
    },
    {
      "entropy": 5.585910654067993,
      "epoch": 0.4940673020813071,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498049202076508,
      "loss": 5.418,
      "mean_token_accuracy": 0.16667619943618775,
      "num_tokens": 11024629.0,
      "step": 6350
    },
    {
      "entropy": 5.614324712753296,
      "epoch": 0.4944563314530247,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980455586306149,
      "loss": 5.3509,
      "mean_token_accuracy": 0.1716369092464447,
      "num_tokens": 11033953.0,
      "step": 6355
    },
    {
      "entropy": 5.545697212219238,
      "epoch": 0.4948453608247423,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980419118003605,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1693047046661377,
      "num_tokens": 11042891.0,
      "step": 6360
    },
    {
      "entropy": 5.601671361923218,
      "epoch": 0.4952343901964598,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980382615858001,
      "loss": 5.5162,
      "mean_token_accuracy": 0.16135913282632827,
      "num_tokens": 11051262.0,
      "step": 6365
    },
    {
      "entropy": 5.606281852722168,
      "epoch": 0.4956234195681774,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980346079869892,
      "loss": 5.3472,
      "mean_token_accuracy": 0.1761609584093094,
      "num_tokens": 11059357.0,
      "step": 6370
    },
    {
      "entropy": 5.603213357925415,
      "epoch": 0.49601244893989493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498030951003983,
      "loss": 5.3587,
      "mean_token_accuracy": 0.1736203297972679,
      "num_tokens": 11068076.0,
      "step": 6375
    },
    {
      "entropy": 5.53122091293335,
      "epoch": 0.4964014783116125,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004980272906368371,
      "loss": 5.2448,
      "mean_token_accuracy": 0.18355217725038528,
      "num_tokens": 11076804.0,
      "step": 6380
    },
    {
      "entropy": 5.650952482223511,
      "epoch": 0.4967905076833301,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980236268856071,
      "loss": 5.4407,
      "mean_token_accuracy": 0.17113036364316941,
      "num_tokens": 11085794.0,
      "step": 6385
    },
    {
      "entropy": 5.527957010269165,
      "epoch": 0.49717953705504764,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980199597503487,
      "loss": 5.3427,
      "mean_token_accuracy": 0.16856258362531662,
      "num_tokens": 11094989.0,
      "step": 6390
    },
    {
      "entropy": 5.545766878128052,
      "epoch": 0.4975685664267652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980162892311171,
      "loss": 5.3043,
      "mean_token_accuracy": 0.18141588121652602,
      "num_tokens": 11103046.0,
      "step": 6395
    },
    {
      "entropy": 5.5342535972595215,
      "epoch": 0.4979575957984828,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980126153279684,
      "loss": 5.3777,
      "mean_token_accuracy": 0.17280755192041397,
      "num_tokens": 11111533.0,
      "step": 6400
    },
    {
      "entropy": 5.606961250305176,
      "epoch": 0.49834662517020034,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004980089380409583,
      "loss": 5.5668,
      "mean_token_accuracy": 0.15929663628339769,
      "num_tokens": 11120534.0,
      "step": 6405
    },
    {
      "entropy": 5.7170703411102295,
      "epoch": 0.4987356545419179,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980052573701424,
      "loss": 5.4978,
      "mean_token_accuracy": 0.17221135199069976,
      "num_tokens": 11128354.0,
      "step": 6410
    },
    {
      "entropy": 5.644919776916504,
      "epoch": 0.49912468391363546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980015733155767,
      "loss": 5.412,
      "mean_token_accuracy": 0.16755639761686325,
      "num_tokens": 11136201.0,
      "step": 6415
    },
    {
      "entropy": 5.673887538909912,
      "epoch": 0.49951371328535304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979978858773171,
      "loss": 5.3676,
      "mean_token_accuracy": 0.1745676428079605,
      "num_tokens": 11145131.0,
      "step": 6420
    },
    {
      "entropy": 5.552023649215698,
      "epoch": 0.49990274265707063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979941950554195,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1781318262219429,
      "num_tokens": 11154514.0,
      "step": 6425
    },
    {
      "entropy": 5.660540771484375,
      "epoch": 0.5002917720287882,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979905008499399,
      "loss": 5.3985,
      "mean_token_accuracy": 0.17282738238573075,
      "num_tokens": 11163160.0,
      "step": 6430
    },
    {
      "entropy": 5.576961517333984,
      "epoch": 0.5006808014005057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979868032609344,
      "loss": 5.3869,
      "mean_token_accuracy": 0.17956127524375914,
      "num_tokens": 11171573.0,
      "step": 6435
    },
    {
      "entropy": 5.56858458518982,
      "epoch": 0.5010698307722233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979831022884591,
      "loss": 5.3739,
      "mean_token_accuracy": 0.17165141701698303,
      "num_tokens": 11180028.0,
      "step": 6440
    },
    {
      "entropy": 5.55673394203186,
      "epoch": 0.5014588601439409,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979793979325701,
      "loss": 5.2887,
      "mean_token_accuracy": 0.17160007655620574,
      "num_tokens": 11188403.0,
      "step": 6445
    },
    {
      "entropy": 5.49115891456604,
      "epoch": 0.5018478895156584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979756901933236,
      "loss": 5.2994,
      "mean_token_accuracy": 0.18273988515138626,
      "num_tokens": 11196391.0,
      "step": 6450
    },
    {
      "entropy": 5.588707399368286,
      "epoch": 0.502236918887376,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497971979070776,
      "loss": 5.4289,
      "mean_token_accuracy": 0.1710171401500702,
      "num_tokens": 11204596.0,
      "step": 6455
    },
    {
      "entropy": 5.60586142539978,
      "epoch": 0.5026259482590936,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979682645649834,
      "loss": 5.3772,
      "mean_token_accuracy": 0.17194198518991471,
      "num_tokens": 11212629.0,
      "step": 6460
    },
    {
      "entropy": 5.6631279468536375,
      "epoch": 0.5030149776308112,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979645466760025,
      "loss": 5.4461,
      "mean_token_accuracy": 0.1727239266037941,
      "num_tokens": 11221836.0,
      "step": 6465
    },
    {
      "entropy": 5.644039964675903,
      "epoch": 0.5034040070025287,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979608254038892,
      "loss": 5.5072,
      "mean_token_accuracy": 0.161077281832695,
      "num_tokens": 11230600.0,
      "step": 6470
    },
    {
      "entropy": 5.740381383895874,
      "epoch": 0.5037930363742462,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979571007487003,
      "loss": 5.4891,
      "mean_token_accuracy": 0.16057279258966445,
      "num_tokens": 11239664.0,
      "step": 6475
    },
    {
      "entropy": 5.589925527572632,
      "epoch": 0.5041820657459638,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979533727104924,
      "loss": 5.4073,
      "mean_token_accuracy": 0.1660124585032463,
      "num_tokens": 11249111.0,
      "step": 6480
    },
    {
      "entropy": 5.558521556854248,
      "epoch": 0.5045710951176814,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497949641289322,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1722047045826912,
      "num_tokens": 11257599.0,
      "step": 6485
    },
    {
      "entropy": 5.5940399169921875,
      "epoch": 0.504960124489399,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979459064852456,
      "loss": 5.433,
      "mean_token_accuracy": 0.18156720101833343,
      "num_tokens": 11265586.0,
      "step": 6490
    },
    {
      "entropy": 5.708671045303345,
      "epoch": 0.5053491538611166,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979421682983197,
      "loss": 5.5548,
      "mean_token_accuracy": 0.16661259680986404,
      "num_tokens": 11274736.0,
      "step": 6495
    },
    {
      "entropy": 5.630670738220215,
      "epoch": 0.505738183232834,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004979384267286015,
      "loss": 5.3417,
      "mean_token_accuracy": 0.18198635578155517,
      "num_tokens": 11282596.0,
      "step": 6500
    },
    {
      "entropy": 5.623176193237304,
      "epoch": 0.5061272126045516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979346817761475,
      "loss": 5.4219,
      "mean_token_accuracy": 0.16544625163078308,
      "num_tokens": 11290974.0,
      "step": 6505
    },
    {
      "entropy": 5.506893968582153,
      "epoch": 0.5065162419762692,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979309334410144,
      "loss": 5.3031,
      "mean_token_accuracy": 0.17951635420322418,
      "num_tokens": 11299196.0,
      "step": 6510
    },
    {
      "entropy": 5.595285224914551,
      "epoch": 0.5069052713479868,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979271817232594,
      "loss": 5.4252,
      "mean_token_accuracy": 0.16717837750911713,
      "num_tokens": 11308214.0,
      "step": 6515
    },
    {
      "entropy": 5.613783502578736,
      "epoch": 0.5072943007197044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979234266229391,
      "loss": 5.4134,
      "mean_token_accuracy": 0.17198467999696732,
      "num_tokens": 11316566.0,
      "step": 6520
    },
    {
      "entropy": 5.5896409034729,
      "epoch": 0.5076833300914219,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979196681401106,
      "loss": 5.3964,
      "mean_token_accuracy": 0.17299696058034897,
      "num_tokens": 11325079.0,
      "step": 6525
    },
    {
      "entropy": 5.591872882843018,
      "epoch": 0.5080723594631394,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497915906274831,
      "loss": 5.415,
      "mean_token_accuracy": 0.16943266242742538,
      "num_tokens": 11333829.0,
      "step": 6530
    },
    {
      "entropy": 5.550478506088257,
      "epoch": 0.508461388834857,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979121410271574,
      "loss": 5.3552,
      "mean_token_accuracy": 0.1776174560189247,
      "num_tokens": 11343078.0,
      "step": 6535
    },
    {
      "entropy": 5.612886571884156,
      "epoch": 0.5088504182065746,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979083723971468,
      "loss": 5.4797,
      "mean_token_accuracy": 0.1714700222015381,
      "num_tokens": 11351475.0,
      "step": 6540
    },
    {
      "entropy": 5.579367780685425,
      "epoch": 0.5092394475782922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979046003848564,
      "loss": 5.3507,
      "mean_token_accuracy": 0.17437137365341188,
      "num_tokens": 11359810.0,
      "step": 6545
    },
    {
      "entropy": 5.637637901306152,
      "epoch": 0.5096284769500097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979008249903435,
      "loss": 5.4018,
      "mean_token_accuracy": 0.17298720329999923,
      "num_tokens": 11368585.0,
      "step": 6550
    },
    {
      "entropy": 5.618145275115967,
      "epoch": 0.5100175063217273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978970462136655,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17616247981786728,
      "num_tokens": 11377393.0,
      "step": 6555
    },
    {
      "entropy": 5.609848976135254,
      "epoch": 0.5104065356934449,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978932640548794,
      "loss": 5.4073,
      "mean_token_accuracy": 0.165021151304245,
      "num_tokens": 11385877.0,
      "step": 6560
    },
    {
      "entropy": 5.553051280975342,
      "epoch": 0.5107955650651624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978894785140429,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1768214926123619,
      "num_tokens": 11394457.0,
      "step": 6565
    },
    {
      "entropy": 5.530999422073364,
      "epoch": 0.51118459443688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978856895912132,
      "loss": 5.3406,
      "mean_token_accuracy": 0.17647128850221633,
      "num_tokens": 11403007.0,
      "step": 6570
    },
    {
      "entropy": 5.6438335418701175,
      "epoch": 0.5115736238085975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978818972864481,
      "loss": 5.4125,
      "mean_token_accuracy": 0.16737185269594193,
      "num_tokens": 11412237.0,
      "step": 6575
    },
    {
      "entropy": 5.6251890659332275,
      "epoch": 0.5119626531803151,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978781015998048,
      "loss": 5.3685,
      "mean_token_accuracy": 0.16872357279062272,
      "num_tokens": 11421219.0,
      "step": 6580
    },
    {
      "entropy": 5.618922996520996,
      "epoch": 0.5123516825520327,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978743025313413,
      "loss": 5.3626,
      "mean_token_accuracy": 0.16852298229932786,
      "num_tokens": 11428933.0,
      "step": 6585
    },
    {
      "entropy": 5.549186086654663,
      "epoch": 0.5127407119237503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978705000811148,
      "loss": 5.4657,
      "mean_token_accuracy": 0.16887375563383103,
      "num_tokens": 11437551.0,
      "step": 6590
    },
    {
      "entropy": 5.6655247688293455,
      "epoch": 0.5131297412954678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004978666942491832,
      "loss": 5.5484,
      "mean_token_accuracy": 0.16170087456703186,
      "num_tokens": 11446045.0,
      "step": 6595
    },
    {
      "entropy": 5.709177684783936,
      "epoch": 0.5135187706671853,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978628850356043,
      "loss": 5.4712,
      "mean_token_accuracy": 0.16923767775297166,
      "num_tokens": 11455159.0,
      "step": 6600
    },
    {
      "entropy": 5.635027027130127,
      "epoch": 0.5139078000389029,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978590724404358,
      "loss": 5.3689,
      "mean_token_accuracy": 0.17412376254796982,
      "num_tokens": 11463835.0,
      "step": 6605
    },
    {
      "entropy": 5.584535932540893,
      "epoch": 0.5142968294106205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978552564637356,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1742961212992668,
      "num_tokens": 11472439.0,
      "step": 6610
    },
    {
      "entropy": 5.613216876983643,
      "epoch": 0.5146858587823381,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978514371055616,
      "loss": 5.482,
      "mean_token_accuracy": 0.1656497597694397,
      "num_tokens": 11481626.0,
      "step": 6615
    },
    {
      "entropy": 5.561394691467285,
      "epoch": 0.5150748881540557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978476143659718,
      "loss": 5.3634,
      "mean_token_accuracy": 0.17834966033697128,
      "num_tokens": 11490571.0,
      "step": 6620
    },
    {
      "entropy": 5.6569291114807125,
      "epoch": 0.5154639175257731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978437882450241,
      "loss": 5.4625,
      "mean_token_accuracy": 0.1672450289130211,
      "num_tokens": 11499145.0,
      "step": 6625
    },
    {
      "entropy": 5.643627548217774,
      "epoch": 0.5158529468974907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978399587427766,
      "loss": 5.399,
      "mean_token_accuracy": 0.1676664710044861,
      "num_tokens": 11508133.0,
      "step": 6630
    },
    {
      "entropy": 5.459604072570801,
      "epoch": 0.5162419762692083,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978361258592874,
      "loss": 5.285,
      "mean_token_accuracy": 0.17830432951450348,
      "num_tokens": 11516790.0,
      "step": 6635
    },
    {
      "entropy": 5.564607620239258,
      "epoch": 0.5166310056409259,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978322895946147,
      "loss": 5.3745,
      "mean_token_accuracy": 0.16994131356477737,
      "num_tokens": 11525963.0,
      "step": 6640
    },
    {
      "entropy": 5.590848112106324,
      "epoch": 0.5170200350126435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978284499488167,
      "loss": 5.3554,
      "mean_token_accuracy": 0.17311257421970366,
      "num_tokens": 11534193.0,
      "step": 6645
    },
    {
      "entropy": 5.609114074707032,
      "epoch": 0.5174090643843611,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004978246069219516,
      "loss": 5.375,
      "mean_token_accuracy": 0.16920317709445953,
      "num_tokens": 11542517.0,
      "step": 6650
    },
    {
      "entropy": 5.574249744415283,
      "epoch": 0.5177980937560785,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978207605140777,
      "loss": 5.4459,
      "mean_token_accuracy": 0.16738450825214385,
      "num_tokens": 11550874.0,
      "step": 6655
    },
    {
      "entropy": 5.631093215942383,
      "epoch": 0.5181871231277961,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978169107252534,
      "loss": 5.4513,
      "mean_token_accuracy": 0.16461260616779327,
      "num_tokens": 11560562.0,
      "step": 6660
    },
    {
      "entropy": 5.592811727523804,
      "epoch": 0.5185761524995137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978130575555373,
      "loss": 5.3429,
      "mean_token_accuracy": 0.17423924505710603,
      "num_tokens": 11569887.0,
      "step": 6665
    },
    {
      "entropy": 5.5215840339660645,
      "epoch": 0.5189651818712313,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978092010049877,
      "loss": 5.2705,
      "mean_token_accuracy": 0.18187254369258882,
      "num_tokens": 11578608.0,
      "step": 6670
    },
    {
      "entropy": 5.581983661651611,
      "epoch": 0.5193542112429489,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497805341073663,
      "loss": 5.3912,
      "mean_token_accuracy": 0.17022958993911744,
      "num_tokens": 11587435.0,
      "step": 6675
    },
    {
      "entropy": 5.595023345947266,
      "epoch": 0.5197432406146664,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978014777616219,
      "loss": 5.4862,
      "mean_token_accuracy": 0.16470641642808914,
      "num_tokens": 11596094.0,
      "step": 6680
    },
    {
      "entropy": 5.631581449508667,
      "epoch": 0.520132269986384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977976110689229,
      "loss": 5.4378,
      "mean_token_accuracy": 0.16952349245548248,
      "num_tokens": 11604521.0,
      "step": 6685
    },
    {
      "entropy": 5.641193103790283,
      "epoch": 0.5205212993581015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497793740995625,
      "loss": 5.4025,
      "mean_token_accuracy": 0.17098213583230973,
      "num_tokens": 11613143.0,
      "step": 6690
    },
    {
      "entropy": 5.630144119262695,
      "epoch": 0.5209103287298191,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977898675417866,
      "loss": 5.445,
      "mean_token_accuracy": 0.17159462794661523,
      "num_tokens": 11622667.0,
      "step": 6695
    },
    {
      "entropy": 5.602881956100464,
      "epoch": 0.5212993581015367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977859907074664,
      "loss": 5.4167,
      "mean_token_accuracy": 0.1751892313361168,
      "num_tokens": 11632114.0,
      "step": 6700
    },
    {
      "entropy": 5.625202655792236,
      "epoch": 0.5216883874732542,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977821104927236,
      "loss": 5.3335,
      "mean_token_accuracy": 0.17328906655311585,
      "num_tokens": 11640731.0,
      "step": 6705
    },
    {
      "entropy": 5.5024106979370115,
      "epoch": 0.5220774168449718,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977782268976167,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1757868632674217,
      "num_tokens": 11648975.0,
      "step": 6710
    },
    {
      "entropy": 5.582853174209594,
      "epoch": 0.5224664462166894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004977743399222048,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17825236171483994,
      "num_tokens": 11657598.0,
      "step": 6715
    },
    {
      "entropy": 5.709061145782471,
      "epoch": 0.522855475588407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977704495665471,
      "loss": 5.5124,
      "mean_token_accuracy": 0.1737692803144455,
      "num_tokens": 11665504.0,
      "step": 6720
    },
    {
      "entropy": 5.629387521743775,
      "epoch": 0.5232445049601245,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004977665558307023,
      "loss": 5.4329,
      "mean_token_accuracy": 0.17330906242132188,
      "num_tokens": 11674337.0,
      "step": 6725
    },
    {
      "entropy": 5.485547828674316,
      "epoch": 0.523633534331842,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977626587147295,
      "loss": 5.2716,
      "mean_token_accuracy": 0.18216868340969086,
      "num_tokens": 11682626.0,
      "step": 6730
    },
    {
      "entropy": 5.547606945037842,
      "epoch": 0.5240225637035596,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497758758218688,
      "loss": 5.3259,
      "mean_token_accuracy": 0.1781942442059517,
      "num_tokens": 11691403.0,
      "step": 6735
    },
    {
      "entropy": 5.651517009735107,
      "epoch": 0.5244115930752772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977548543426368,
      "loss": 5.4553,
      "mean_token_accuracy": 0.166091488301754,
      "num_tokens": 11700590.0,
      "step": 6740
    },
    {
      "entropy": 5.625852966308594,
      "epoch": 0.5248006224469948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977509470866353,
      "loss": 5.3641,
      "mean_token_accuracy": 0.16962600648403167,
      "num_tokens": 11708799.0,
      "step": 6745
    },
    {
      "entropy": 5.567790889739991,
      "epoch": 0.5251896518187124,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977470364507427,
      "loss": 5.3627,
      "mean_token_accuracy": 0.17713444232940673,
      "num_tokens": 11716953.0,
      "step": 6750
    },
    {
      "entropy": 5.605804109573365,
      "epoch": 0.5255786811904298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977431224350184,
      "loss": 5.4306,
      "mean_token_accuracy": 0.1685882940888405,
      "num_tokens": 11725032.0,
      "step": 6755
    },
    {
      "entropy": 5.613562965393067,
      "epoch": 0.5259677105621474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977392050395217,
      "loss": 5.4337,
      "mean_token_accuracy": 0.16917459070682525,
      "num_tokens": 11733359.0,
      "step": 6760
    },
    {
      "entropy": 5.665448999404907,
      "epoch": 0.526356739933865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977352842643121,
      "loss": 5.4167,
      "mean_token_accuracy": 0.17600178420543672,
      "num_tokens": 11740936.0,
      "step": 6765
    },
    {
      "entropy": 5.61443829536438,
      "epoch": 0.5267457693055826,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497731360109449,
      "loss": 5.4559,
      "mean_token_accuracy": 0.16777198761701584,
      "num_tokens": 11749764.0,
      "step": 6770
    },
    {
      "entropy": 5.64350323677063,
      "epoch": 0.5271347986773002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977274325749922,
      "loss": 5.4631,
      "mean_token_accuracy": 0.17370951920747757,
      "num_tokens": 11757945.0,
      "step": 6775
    },
    {
      "entropy": 5.547140598297119,
      "epoch": 0.5275238280490177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004977235016610011,
      "loss": 5.3372,
      "mean_token_accuracy": 0.17899322956800462,
      "num_tokens": 11766333.0,
      "step": 6780
    },
    {
      "entropy": 5.56026520729065,
      "epoch": 0.5279128574207352,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977195673675353,
      "loss": 5.3937,
      "mean_token_accuracy": 0.17261524647474288,
      "num_tokens": 11775041.0,
      "step": 6785
    },
    {
      "entropy": 5.610231876373291,
      "epoch": 0.5283018867924528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977156296946545,
      "loss": 5.377,
      "mean_token_accuracy": 0.1705646276473999,
      "num_tokens": 11783958.0,
      "step": 6790
    },
    {
      "entropy": 5.579576539993286,
      "epoch": 0.5286909161641704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977116886424187,
      "loss": 5.4475,
      "mean_token_accuracy": 0.16637502908706664,
      "num_tokens": 11792174.0,
      "step": 6795
    },
    {
      "entropy": 5.5774702548980715,
      "epoch": 0.529079945535888,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977077442108873,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17724884301424026,
      "num_tokens": 11800514.0,
      "step": 6800
    },
    {
      "entropy": 5.590536308288574,
      "epoch": 0.5294689749076055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977037964001204,
      "loss": 5.2651,
      "mean_token_accuracy": 0.18220219165086746,
      "num_tokens": 11808918.0,
      "step": 6805
    },
    {
      "entropy": 5.654633092880249,
      "epoch": 0.5298580042793231,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497699845210178,
      "loss": 5.5187,
      "mean_token_accuracy": 0.1643660172820091,
      "num_tokens": 11818014.0,
      "step": 6810
    },
    {
      "entropy": 5.5917236328125,
      "epoch": 0.5302470336510406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976958906411197,
      "loss": 5.4426,
      "mean_token_accuracy": 0.16338089257478713,
      "num_tokens": 11826477.0,
      "step": 6815
    },
    {
      "entropy": 5.673809576034546,
      "epoch": 0.5306360630227582,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976919326930059,
      "loss": 5.4992,
      "mean_token_accuracy": 0.16627666354179382,
      "num_tokens": 11835513.0,
      "step": 6820
    },
    {
      "entropy": 5.695989656448364,
      "epoch": 0.5310250923944758,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976879713658964,
      "loss": 5.481,
      "mean_token_accuracy": 0.17325812727212905,
      "num_tokens": 11844495.0,
      "step": 6825
    },
    {
      "entropy": 5.628694486618042,
      "epoch": 0.5314141217661933,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976840066598514,
      "loss": 5.4452,
      "mean_token_accuracy": 0.16433378607034682,
      "num_tokens": 11853013.0,
      "step": 6830
    },
    {
      "entropy": 5.51334285736084,
      "epoch": 0.5318031511379109,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976800385749309,
      "loss": 5.2287,
      "mean_token_accuracy": 0.17677582651376725,
      "num_tokens": 11861899.0,
      "step": 6835
    },
    {
      "entropy": 5.5065830707550045,
      "epoch": 0.5321921805096285,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976760671111954,
      "loss": 5.3119,
      "mean_token_accuracy": 0.18060407787561417,
      "num_tokens": 11870836.0,
      "step": 6840
    },
    {
      "entropy": 5.568042039871216,
      "epoch": 0.532581209881346,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497672092268705,
      "loss": 5.3677,
      "mean_token_accuracy": 0.16628734916448593,
      "num_tokens": 11879363.0,
      "step": 6845
    },
    {
      "entropy": 5.6174729347229,
      "epoch": 0.5329702392530636,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049766811404752,
      "loss": 5.3655,
      "mean_token_accuracy": 0.17461650520563127,
      "num_tokens": 11887696.0,
      "step": 6850
    },
    {
      "entropy": 5.585926151275634,
      "epoch": 0.5333592686247812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976641324477008,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1717255249619484,
      "num_tokens": 11895867.0,
      "step": 6855
    },
    {
      "entropy": 5.632009267807007,
      "epoch": 0.5337482979964987,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976601474693077,
      "loss": 5.3103,
      "mean_token_accuracy": 0.17238852828741075,
      "num_tokens": 11904537.0,
      "step": 6860
    },
    {
      "entropy": 5.675828123092652,
      "epoch": 0.5341373273682163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976561591124014,
      "loss": 5.4333,
      "mean_token_accuracy": 0.17629104405641555,
      "num_tokens": 11913097.0,
      "step": 6865
    },
    {
      "entropy": 5.496969223022461,
      "epoch": 0.5345263567399339,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976521673770421,
      "loss": 5.2426,
      "mean_token_accuracy": 0.17388982623815535,
      "num_tokens": 11921852.0,
      "step": 6870
    },
    {
      "entropy": 5.594495582580566,
      "epoch": 0.5349153861116515,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976481722632907,
      "loss": 5.427,
      "mean_token_accuracy": 0.1668204829096794,
      "num_tokens": 11930836.0,
      "step": 6875
    },
    {
      "entropy": 5.647954893112183,
      "epoch": 0.535304415483369,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976441737712076,
      "loss": 5.4943,
      "mean_token_accuracy": 0.1578487128019333,
      "num_tokens": 11940069.0,
      "step": 6880
    },
    {
      "entropy": 5.67997374534607,
      "epoch": 0.5356934448550865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976401719008536,
      "loss": 5.4275,
      "mean_token_accuracy": 0.17309840321540831,
      "num_tokens": 11948525.0,
      "step": 6885
    },
    {
      "entropy": 5.600617599487305,
      "epoch": 0.5360824742268041,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004976361666522893,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17392354756593703,
      "num_tokens": 11956912.0,
      "step": 6890
    },
    {
      "entropy": 5.507595109939575,
      "epoch": 0.5364715035985217,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976321580255755,
      "loss": 5.5035,
      "mean_token_accuracy": 0.16616453230381012,
      "num_tokens": 11965782.0,
      "step": 6895
    },
    {
      "entropy": 5.623979187011718,
      "epoch": 0.5368605329702393,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976281460207731,
      "loss": 5.4132,
      "mean_token_accuracy": 0.17315759211778642,
      "num_tokens": 11974889.0,
      "step": 6900
    },
    {
      "entropy": 5.655584335327148,
      "epoch": 0.5372495623419569,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497624130637943,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1678052067756653,
      "num_tokens": 11983507.0,
      "step": 6905
    },
    {
      "entropy": 5.534413576126099,
      "epoch": 0.5376385917136743,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497620111877146,
      "loss": 5.275,
      "mean_token_accuracy": 0.1823041096329689,
      "num_tokens": 11991646.0,
      "step": 6910
    },
    {
      "entropy": 5.4777453422546385,
      "epoch": 0.5380276210853919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004976160897384431,
      "loss": 5.3216,
      "mean_token_accuracy": 0.17367151379585266,
      "num_tokens": 12000223.0,
      "step": 6915
    },
    {
      "entropy": 5.622479200363159,
      "epoch": 0.5384166504571095,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976120642218955,
      "loss": 5.4153,
      "mean_token_accuracy": 0.17153375297784806,
      "num_tokens": 12009012.0,
      "step": 6920
    },
    {
      "entropy": 5.717148637771606,
      "epoch": 0.5388056798288271,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497608035327564,
      "loss": 5.459,
      "mean_token_accuracy": 0.16472576260566713,
      "num_tokens": 12018001.0,
      "step": 6925
    },
    {
      "entropy": 5.5548868656158445,
      "epoch": 0.5391947092005447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049760400305551,
      "loss": 5.2903,
      "mean_token_accuracy": 0.18364150524139405,
      "num_tokens": 12025943.0,
      "step": 6930
    },
    {
      "entropy": 5.576215124130249,
      "epoch": 0.5395837385722622,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975999674057944,
      "loss": 5.3333,
      "mean_token_accuracy": 0.1743236228823662,
      "num_tokens": 12035073.0,
      "step": 6935
    },
    {
      "entropy": 5.650489425659179,
      "epoch": 0.5399727679439797,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975959283784787,
      "loss": 5.5105,
      "mean_token_accuracy": 0.16534960120916367,
      "num_tokens": 12043698.0,
      "step": 6940
    },
    {
      "entropy": 5.6009152889251705,
      "epoch": 0.5403617973156973,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975918859736241,
      "loss": 5.3966,
      "mean_token_accuracy": 0.17727003395557403,
      "num_tokens": 12052247.0,
      "step": 6945
    },
    {
      "entropy": 5.594502592086792,
      "epoch": 0.5407508266874149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975878401912919,
      "loss": 5.4101,
      "mean_token_accuracy": 0.1762482464313507,
      "num_tokens": 12061187.0,
      "step": 6950
    },
    {
      "entropy": 5.57668399810791,
      "epoch": 0.5411398560591325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975837910315436,
      "loss": 5.3585,
      "mean_token_accuracy": 0.170804201066494,
      "num_tokens": 12070121.0,
      "step": 6955
    },
    {
      "entropy": 5.609505748748779,
      "epoch": 0.54152888543085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975797384944405,
      "loss": 5.3086,
      "mean_token_accuracy": 0.18491584211587905,
      "num_tokens": 12078578.0,
      "step": 6960
    },
    {
      "entropy": 5.585539960861206,
      "epoch": 0.5419179148025676,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497575682580044,
      "loss": 5.466,
      "mean_token_accuracy": 0.1712225630879402,
      "num_tokens": 12087029.0,
      "step": 6965
    },
    {
      "entropy": 5.7061645030975345,
      "epoch": 0.5423069441742852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497571623288416,
      "loss": 5.435,
      "mean_token_accuracy": 0.16743697375059127,
      "num_tokens": 12096105.0,
      "step": 6970
    },
    {
      "entropy": 5.756121730804443,
      "epoch": 0.5426959735460027,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975675606196177,
      "loss": 5.506,
      "mean_token_accuracy": 0.16332751214504243,
      "num_tokens": 12105098.0,
      "step": 6975
    },
    {
      "entropy": 5.599440097808838,
      "epoch": 0.5430850029177203,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497563494573711,
      "loss": 5.3063,
      "mean_token_accuracy": 0.17674217969179154,
      "num_tokens": 12112894.0,
      "step": 6980
    },
    {
      "entropy": 5.5193421840667725,
      "epoch": 0.5434740322894378,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975594251507575,
      "loss": 5.4403,
      "mean_token_accuracy": 0.17357636839151383,
      "num_tokens": 12121457.0,
      "step": 6985
    },
    {
      "entropy": 5.639322137832641,
      "epoch": 0.5438630616611554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497555352350819,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1768542766571045,
      "num_tokens": 12129998.0,
      "step": 6990
    },
    {
      "entropy": 5.666340065002442,
      "epoch": 0.544252091032873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975512761739572,
      "loss": 5.389,
      "mean_token_accuracy": 0.17072022557258607,
      "num_tokens": 12137861.0,
      "step": 6995
    },
    {
      "entropy": 5.556141805648804,
      "epoch": 0.5446411204045906,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497547196620234,
      "loss": 5.3912,
      "mean_token_accuracy": 0.17440045475959778,
      "num_tokens": 12146294.0,
      "step": 7000
    },
    {
      "entropy": 5.559613847732544,
      "epoch": 0.5450301497763081,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975431136897114,
      "loss": 5.3929,
      "mean_token_accuracy": 0.17047546803951263,
      "num_tokens": 12155642.0,
      "step": 7005
    },
    {
      "entropy": 5.735659790039063,
      "epoch": 0.5454191791480256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975390273824512,
      "loss": 5.4939,
      "mean_token_accuracy": 0.16992603689432145,
      "num_tokens": 12164709.0,
      "step": 7010
    },
    {
      "entropy": 5.617656564712524,
      "epoch": 0.5458082085197432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975349376985155,
      "loss": 5.3662,
      "mean_token_accuracy": 0.17637594640254975,
      "num_tokens": 12173591.0,
      "step": 7015
    },
    {
      "entropy": 5.585110187530518,
      "epoch": 0.5461972378914608,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975308446379663,
      "loss": 5.4247,
      "mean_token_accuracy": 0.1666642650961876,
      "num_tokens": 12182050.0,
      "step": 7020
    },
    {
      "entropy": 5.649959850311279,
      "epoch": 0.5465862672631784,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975267482008657,
      "loss": 5.4722,
      "mean_token_accuracy": 0.16861061453819276,
      "num_tokens": 12190870.0,
      "step": 7025
    },
    {
      "entropy": 5.492293834686279,
      "epoch": 0.546975296634896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497522648387276,
      "loss": 5.2108,
      "mean_token_accuracy": 0.18491959273815156,
      "num_tokens": 12199624.0,
      "step": 7030
    },
    {
      "entropy": 5.6087899684906,
      "epoch": 0.5473643260066134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975185451972592,
      "loss": 5.4624,
      "mean_token_accuracy": 0.16690078377723694,
      "num_tokens": 12208087.0,
      "step": 7035
    },
    {
      "entropy": 5.680318450927734,
      "epoch": 0.547753355378331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975144386308776,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17517598122358322,
      "num_tokens": 12216531.0,
      "step": 7040
    },
    {
      "entropy": 5.615996551513672,
      "epoch": 0.5481423847500486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975103286881936,
      "loss": 5.3705,
      "mean_token_accuracy": 0.17713344991207122,
      "num_tokens": 12225094.0,
      "step": 7045
    },
    {
      "entropy": 5.560548114776611,
      "epoch": 0.5485314141217662,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975062153692696,
      "loss": 5.3765,
      "mean_token_accuracy": 0.17883004546165465,
      "num_tokens": 12233824.0,
      "step": 7050
    },
    {
      "entropy": 5.554152297973633,
      "epoch": 0.5489204434934838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975020986741678,
      "loss": 5.2767,
      "mean_token_accuracy": 0.17529497891664506,
      "num_tokens": 12242923.0,
      "step": 7055
    },
    {
      "entropy": 5.581981992721557,
      "epoch": 0.5493094728652014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974979786029509,
      "loss": 5.3156,
      "mean_token_accuracy": 0.17031585723161696,
      "num_tokens": 12251963.0,
      "step": 7060
    },
    {
      "entropy": 5.586628723144531,
      "epoch": 0.5496985022369189,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974938551556814,
      "loss": 5.4442,
      "mean_token_accuracy": 0.17018938660621644,
      "num_tokens": 12260347.0,
      "step": 7065
    },
    {
      "entropy": 5.551030397415161,
      "epoch": 0.5500875316086364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974897283324217,
      "loss": 5.4813,
      "mean_token_accuracy": 0.16576917320489884,
      "num_tokens": 12269156.0,
      "step": 7070
    },
    {
      "entropy": 5.632766628265381,
      "epoch": 0.550476560980354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974855981332343,
      "loss": 5.3835,
      "mean_token_accuracy": 0.17017189860343934,
      "num_tokens": 12277636.0,
      "step": 7075
    },
    {
      "entropy": 5.631530666351319,
      "epoch": 0.5508655903520716,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974814645581823,
      "loss": 5.3576,
      "mean_token_accuracy": 0.17621888518333434,
      "num_tokens": 12286027.0,
      "step": 7080
    },
    {
      "entropy": 5.633374929428101,
      "epoch": 0.5512546197237892,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974773276073282,
      "loss": 5.4807,
      "mean_token_accuracy": 0.16844650804996492,
      "num_tokens": 12295085.0,
      "step": 7085
    },
    {
      "entropy": 5.676600170135498,
      "epoch": 0.5516436490955067,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974731872807347,
      "loss": 5.4533,
      "mean_token_accuracy": 0.17263481616973878,
      "num_tokens": 12304275.0,
      "step": 7090
    },
    {
      "entropy": 5.550247097015381,
      "epoch": 0.5520326784672243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004974690435784647,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1946805715560913,
      "num_tokens": 12312841.0,
      "step": 7095
    },
    {
      "entropy": 5.572363948822021,
      "epoch": 0.5524217078389418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974648965005811,
      "loss": 5.3797,
      "mean_token_accuracy": 0.16614917069673538,
      "num_tokens": 12322133.0,
      "step": 7100
    },
    {
      "entropy": 5.581380891799927,
      "epoch": 0.5528107372106594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004974607460471466,
      "loss": 5.3535,
      "mean_token_accuracy": 0.17336255609989165,
      "num_tokens": 12330502.0,
      "step": 7105
    },
    {
      "entropy": 5.6251250267028805,
      "epoch": 0.553199766582377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974565922182246,
      "loss": 5.3795,
      "mean_token_accuracy": 0.17468350678682326,
      "num_tokens": 12338899.0,
      "step": 7110
    },
    {
      "entropy": 5.556831884384155,
      "epoch": 0.5535887959540945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974524350138777,
      "loss": 5.3728,
      "mean_token_accuracy": 0.17247378528118135,
      "num_tokens": 12348299.0,
      "step": 7115
    },
    {
      "entropy": 5.624515962600708,
      "epoch": 0.5539778253258121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974482744341693,
      "loss": 5.3927,
      "mean_token_accuracy": 0.17165566980838776,
      "num_tokens": 12356582.0,
      "step": 7120
    },
    {
      "entropy": 5.673182439804077,
      "epoch": 0.5543668546975297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974441104791624,
      "loss": 5.4972,
      "mean_token_accuracy": 0.15884086936712266,
      "num_tokens": 12366209.0,
      "step": 7125
    },
    {
      "entropy": 5.608900737762451,
      "epoch": 0.5547558840692473,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974399431489201,
      "loss": 5.4221,
      "mean_token_accuracy": 0.16073356568813324,
      "num_tokens": 12374777.0,
      "step": 7130
    },
    {
      "entropy": 5.546727514266967,
      "epoch": 0.5551449134409648,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974357724435057,
      "loss": 5.381,
      "mean_token_accuracy": 0.16626832336187364,
      "num_tokens": 12384405.0,
      "step": 7135
    },
    {
      "entropy": 5.6791276931762695,
      "epoch": 0.5555339428126823,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974315983629825,
      "loss": 5.4615,
      "mean_token_accuracy": 0.16987256854772567,
      "num_tokens": 12393464.0,
      "step": 7140
    },
    {
      "entropy": 5.695638513565063,
      "epoch": 0.5559229721843999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974274209074139,
      "loss": 5.4619,
      "mean_token_accuracy": 0.1698928251862526,
      "num_tokens": 12401917.0,
      "step": 7145
    },
    {
      "entropy": 5.474405193328858,
      "epoch": 0.5563120015561175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974232400768632,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1800903260707855,
      "num_tokens": 12410928.0,
      "step": 7150
    },
    {
      "entropy": 5.548608446121216,
      "epoch": 0.5567010309278351,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974190558713939,
      "loss": 5.3278,
      "mean_token_accuracy": 0.17504773437976837,
      "num_tokens": 12419747.0,
      "step": 7155
    },
    {
      "entropy": 5.63728699684143,
      "epoch": 0.5570900602995527,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974148682910694,
      "loss": 5.364,
      "mean_token_accuracy": 0.17363888621330262,
      "num_tokens": 12428608.0,
      "step": 7160
    },
    {
      "entropy": 5.69043550491333,
      "epoch": 0.5574790896712701,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974106773359533,
      "loss": 5.4498,
      "mean_token_accuracy": 0.16662293970584868,
      "num_tokens": 12437101.0,
      "step": 7165
    },
    {
      "entropy": 5.541791868209839,
      "epoch": 0.5578681190429877,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974064830061091,
      "loss": 5.4083,
      "mean_token_accuracy": 0.16861678659915924,
      "num_tokens": 12446813.0,
      "step": 7170
    },
    {
      "entropy": 5.622215843200683,
      "epoch": 0.5582571484147053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974022853016006,
      "loss": 5.3703,
      "mean_token_accuracy": 0.17372025400400162,
      "num_tokens": 12455931.0,
      "step": 7175
    },
    {
      "entropy": 5.6301921844482425,
      "epoch": 0.5586461777864229,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973980842224913,
      "loss": 5.3379,
      "mean_token_accuracy": 0.17587753683328627,
      "num_tokens": 12463586.0,
      "step": 7180
    },
    {
      "entropy": 5.559577131271363,
      "epoch": 0.5590352071581405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973938797688452,
      "loss": 5.3626,
      "mean_token_accuracy": 0.17821425795555115,
      "num_tokens": 12472063.0,
      "step": 7185
    },
    {
      "entropy": 5.597116994857788,
      "epoch": 0.559424236529858,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004973896719407259,
      "loss": 5.3745,
      "mean_token_accuracy": 0.17003683000802994,
      "num_tokens": 12480763.0,
      "step": 7190
    },
    {
      "entropy": 5.614082002639771,
      "epoch": 0.5598132659015755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973854607381972,
      "loss": 5.3305,
      "mean_token_accuracy": 0.1768320009112358,
      "num_tokens": 12488757.0,
      "step": 7195
    },
    {
      "entropy": 5.5355589389801025,
      "epoch": 0.5602022952732931,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004973812461613232,
      "loss": 5.3783,
      "mean_token_accuracy": 0.17371366620063783,
      "num_tokens": 12497030.0,
      "step": 7200
    },
    {
      "entropy": 5.5868391513824465,
      "epoch": 0.5605913246450107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973770282101677,
      "loss": 5.397,
      "mean_token_accuracy": 0.1695732444524765,
      "num_tokens": 12506067.0,
      "step": 7205
    },
    {
      "entropy": 5.5814941883087155,
      "epoch": 0.5609803540167283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973728068847947,
      "loss": 5.3318,
      "mean_token_accuracy": 0.1736765220761299,
      "num_tokens": 12514246.0,
      "step": 7210
    },
    {
      "entropy": 5.609518241882324,
      "epoch": 0.5613693833884458,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973685821852684,
      "loss": 5.3719,
      "mean_token_accuracy": 0.17128328680992128,
      "num_tokens": 12523523.0,
      "step": 7215
    },
    {
      "entropy": 5.615756368637085,
      "epoch": 0.5617584127601634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973643541116528,
      "loss": 5.3579,
      "mean_token_accuracy": 0.17098272442817689,
      "num_tokens": 12532142.0,
      "step": 7220
    },
    {
      "entropy": 5.577594757080078,
      "epoch": 0.562147442131881,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000497360122664012,
      "loss": 5.4065,
      "mean_token_accuracy": 0.16962229311466218,
      "num_tokens": 12540647.0,
      "step": 7225
    },
    {
      "entropy": 5.551422119140625,
      "epoch": 0.5625364715035985,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004973558878424104,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17154066264629364,
      "num_tokens": 12549010.0,
      "step": 7230
    },
    {
      "entropy": 5.5718786239624025,
      "epoch": 0.5629255008753161,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973516496469119,
      "loss": 5.3123,
      "mean_token_accuracy": 0.17378498762845992,
      "num_tokens": 12557428.0,
      "step": 7235
    },
    {
      "entropy": 5.51436676979065,
      "epoch": 0.5633145302470336,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973474080775811,
      "loss": 5.2595,
      "mean_token_accuracy": 0.17589602768421173,
      "num_tokens": 12566610.0,
      "step": 7240
    },
    {
      "entropy": 5.515880918502807,
      "epoch": 0.5637035596187512,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973431631344824,
      "loss": 5.3742,
      "mean_token_accuracy": 0.17670979499816894,
      "num_tokens": 12576016.0,
      "step": 7245
    },
    {
      "entropy": 5.48199200630188,
      "epoch": 0.5640925889904688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049733891481768,
      "loss": 5.2015,
      "mean_token_accuracy": 0.18723199367523194,
      "num_tokens": 12584033.0,
      "step": 7250
    },
    {
      "entropy": 5.607383584976196,
      "epoch": 0.5644816183621864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973346631272384,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1781315878033638,
      "num_tokens": 12592784.0,
      "step": 7255
    },
    {
      "entropy": 5.615625619888306,
      "epoch": 0.5648706477339039,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004973304080632222,
      "loss": 5.4719,
      "mean_token_accuracy": 0.1671131283044815,
      "num_tokens": 12601259.0,
      "step": 7260
    },
    {
      "entropy": 5.5404987812042235,
      "epoch": 0.5652596771056215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497326149625696,
      "loss": 5.3057,
      "mean_token_accuracy": 0.16903193295001984,
      "num_tokens": 12610264.0,
      "step": 7265
    },
    {
      "entropy": 5.611272859573364,
      "epoch": 0.565648706477339,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004973218878147244,
      "loss": 5.312,
      "mean_token_accuracy": 0.17460647821426392,
      "num_tokens": 12618817.0,
      "step": 7270
    },
    {
      "entropy": 5.547258949279785,
      "epoch": 0.5660377358490566,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004973176226303719,
      "loss": 5.3416,
      "mean_token_accuracy": 0.17069429606199266,
      "num_tokens": 12627224.0,
      "step": 7275
    },
    {
      "entropy": 5.619957113265992,
      "epoch": 0.5664267652207742,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973133540727033,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1835724949836731,
      "num_tokens": 12635020.0,
      "step": 7280
    },
    {
      "entropy": 5.6189040184021,
      "epoch": 0.5668157945924918,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973090821417835,
      "loss": 5.4567,
      "mean_token_accuracy": 0.1689133107662201,
      "num_tokens": 12643821.0,
      "step": 7285
    },
    {
      "entropy": 5.663434267044067,
      "epoch": 0.5672048239642093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004973048068376772,
      "loss": 5.4172,
      "mean_token_accuracy": 0.16706288903951644,
      "num_tokens": 12652091.0,
      "step": 7290
    },
    {
      "entropy": 5.639904451370239,
      "epoch": 0.5675938533359268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973005281604492,
      "loss": 5.4209,
      "mean_token_accuracy": 0.17488816380500793,
      "num_tokens": 12661389.0,
      "step": 7295
    },
    {
      "entropy": 5.608521747589111,
      "epoch": 0.5679828827076444,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972962461101646,
      "loss": 5.4461,
      "mean_token_accuracy": 0.17284394204616546,
      "num_tokens": 12670361.0,
      "step": 7300
    },
    {
      "entropy": 5.547716951370239,
      "epoch": 0.568371912079362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972919606868883,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1760653153061867,
      "num_tokens": 12677765.0,
      "step": 7305
    },
    {
      "entropy": 5.5433228492736815,
      "epoch": 0.5687609414510796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004972876718906852,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18404784053564072,
      "num_tokens": 12686461.0,
      "step": 7310
    },
    {
      "entropy": 5.5584510326385494,
      "epoch": 0.5691499708227972,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972833797216206,
      "loss": 5.3116,
      "mean_token_accuracy": 0.17475184798240662,
      "num_tokens": 12694663.0,
      "step": 7315
    },
    {
      "entropy": 5.576896142959595,
      "epoch": 0.5695390001945146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972790841797595,
      "loss": 5.3838,
      "mean_token_accuracy": 0.17167252004146577,
      "num_tokens": 12703659.0,
      "step": 7320
    },
    {
      "entropy": 5.546174049377441,
      "epoch": 0.5699280295662322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497274785265167,
      "loss": 5.3451,
      "mean_token_accuracy": 0.17827020734548568,
      "num_tokens": 12711402.0,
      "step": 7325
    },
    {
      "entropy": 5.593303918838501,
      "epoch": 0.5703170589379498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972704829779086,
      "loss": 5.4083,
      "mean_token_accuracy": 0.16783702671527861,
      "num_tokens": 12720500.0,
      "step": 7330
    },
    {
      "entropy": 5.664913368225098,
      "epoch": 0.5707060883096674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972661773180492,
      "loss": 5.4369,
      "mean_token_accuracy": 0.1647883877158165,
      "num_tokens": 12729325.0,
      "step": 7335
    },
    {
      "entropy": 5.6450872898101805,
      "epoch": 0.571095117681385,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972618682856545,
      "loss": 5.4592,
      "mean_token_accuracy": 0.16735735088586806,
      "num_tokens": 12738573.0,
      "step": 7340
    },
    {
      "entropy": 5.566001844406128,
      "epoch": 0.5714841470531025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972575558807896,
      "loss": 5.3145,
      "mean_token_accuracy": 0.17694940716028212,
      "num_tokens": 12746608.0,
      "step": 7345
    },
    {
      "entropy": 5.562113475799561,
      "epoch": 0.57187317642482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972532401035201,
      "loss": 5.3049,
      "mean_token_accuracy": 0.17716436237096786,
      "num_tokens": 12754724.0,
      "step": 7350
    },
    {
      "entropy": 5.545844411849975,
      "epoch": 0.5722622057965376,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972489209539114,
      "loss": 5.3816,
      "mean_token_accuracy": 0.1792597457766533,
      "num_tokens": 12762654.0,
      "step": 7355
    },
    {
      "entropy": 5.579734849929809,
      "epoch": 0.5726512351682552,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972445984320292,
      "loss": 5.3944,
      "mean_token_accuracy": 0.16492784172296523,
      "num_tokens": 12771615.0,
      "step": 7360
    },
    {
      "entropy": 5.691576433181763,
      "epoch": 0.5730402645399728,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972402725379388,
      "loss": 5.4626,
      "mean_token_accuracy": 0.17010326236486434,
      "num_tokens": 12780991.0,
      "step": 7365
    },
    {
      "entropy": 5.644831895828247,
      "epoch": 0.5734292939116903,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497235943271706,
      "loss": 5.3302,
      "mean_token_accuracy": 0.17996573746204375,
      "num_tokens": 12789402.0,
      "step": 7370
    },
    {
      "entropy": 5.5373939037322994,
      "epoch": 0.5738183232834079,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972316106333966,
      "loss": 5.3083,
      "mean_token_accuracy": 0.1795029029250145,
      "num_tokens": 12798095.0,
      "step": 7375
    },
    {
      "entropy": 5.63902177810669,
      "epoch": 0.5742073526551255,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972272746230761,
      "loss": 5.4044,
      "mean_token_accuracy": 0.17136958986520767,
      "num_tokens": 12806976.0,
      "step": 7380
    },
    {
      "entropy": 5.585088205337525,
      "epoch": 0.574596382026843,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972229352408104,
      "loss": 5.3162,
      "mean_token_accuracy": 0.18278878778219224,
      "num_tokens": 12815228.0,
      "step": 7385
    },
    {
      "entropy": 5.585175561904907,
      "epoch": 0.5749854113985606,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972185924866655,
      "loss": 5.4097,
      "mean_token_accuracy": 0.16959019750356674,
      "num_tokens": 12823674.0,
      "step": 7390
    },
    {
      "entropy": 5.6082837104797365,
      "epoch": 0.5753744407702781,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004972142463607071,
      "loss": 5.512,
      "mean_token_accuracy": 0.16803084313869476,
      "num_tokens": 12832810.0,
      "step": 7395
    },
    {
      "entropy": 5.687703227996826,
      "epoch": 0.5757634701419957,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972098968630012,
      "loss": 5.4418,
      "mean_token_accuracy": 0.16590020209550857,
      "num_tokens": 12841430.0,
      "step": 7400
    },
    {
      "entropy": 5.609046506881714,
      "epoch": 0.5761524995137133,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972055439936137,
      "loss": 5.3278,
      "mean_token_accuracy": 0.17771490216255187,
      "num_tokens": 12850539.0,
      "step": 7405
    },
    {
      "entropy": 5.6279497146606445,
      "epoch": 0.5765415288854309,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972011877526109,
      "loss": 5.447,
      "mean_token_accuracy": 0.17289159595966339,
      "num_tokens": 12859066.0,
      "step": 7410
    },
    {
      "entropy": 5.52370662689209,
      "epoch": 0.5769305582571485,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004971968281400587,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18174415230751037,
      "num_tokens": 12867603.0,
      "step": 7415
    },
    {
      "entropy": 5.455781126022339,
      "epoch": 0.5773195876288659,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971924651560233,
      "loss": 5.2013,
      "mean_token_accuracy": 0.18343908637762069,
      "num_tokens": 12875950.0,
      "step": 7420
    },
    {
      "entropy": 5.640551328659058,
      "epoch": 0.5777086170005835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004971880988005709,
      "loss": 5.4606,
      "mean_token_accuracy": 0.16838063895702363,
      "num_tokens": 12884154.0,
      "step": 7425
    },
    {
      "entropy": 5.654052639007569,
      "epoch": 0.5780976463723011,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971837290737677,
      "loss": 5.5061,
      "mean_token_accuracy": 0.16462810784578324,
      "num_tokens": 12892515.0,
      "step": 7430
    },
    {
      "entropy": 5.546605968475342,
      "epoch": 0.5784866757440187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971793559756801,
      "loss": 5.322,
      "mean_token_accuracy": 0.18057667464017868,
      "num_tokens": 12900919.0,
      "step": 7435
    },
    {
      "entropy": 5.5282165050506595,
      "epoch": 0.5788757051157363,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971749795063744,
      "loss": 5.2933,
      "mean_token_accuracy": 0.18205010741949082,
      "num_tokens": 12909338.0,
      "step": 7440
    },
    {
      "entropy": 5.652127838134765,
      "epoch": 0.5792647344874537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497170599665917,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1692366525530815,
      "num_tokens": 12918118.0,
      "step": 7445
    },
    {
      "entropy": 5.635789680480957,
      "epoch": 0.5796537638591713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004971662164543743,
      "loss": 5.4427,
      "mean_token_accuracy": 0.16868827491998672,
      "num_tokens": 12926878.0,
      "step": 7450
    },
    {
      "entropy": 5.559504508972168,
      "epoch": 0.5800427932308889,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971618298718131,
      "loss": 5.3067,
      "mean_token_accuracy": 0.17664028257131575,
      "num_tokens": 12934869.0,
      "step": 7455
    },
    {
      "entropy": 5.508236598968506,
      "epoch": 0.5804318226026065,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971574399182995,
      "loss": 5.3606,
      "mean_token_accuracy": 0.17187800258398056,
      "num_tokens": 12944281.0,
      "step": 7460
    },
    {
      "entropy": 5.611026191711426,
      "epoch": 0.5808208519743241,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971530465939005,
      "loss": 5.3873,
      "mean_token_accuracy": 0.17068445384502412,
      "num_tokens": 12953281.0,
      "step": 7465
    },
    {
      "entropy": 5.613557481765747,
      "epoch": 0.5812098813460417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971486498986825,
      "loss": 5.3671,
      "mean_token_accuracy": 0.1724362462759018,
      "num_tokens": 12962136.0,
      "step": 7470
    },
    {
      "entropy": 5.509874391555786,
      "epoch": 0.5815989107177592,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971442498327124,
      "loss": 5.3246,
      "mean_token_accuracy": 0.17531097084283828,
      "num_tokens": 12971288.0,
      "step": 7475
    },
    {
      "entropy": 5.574598217010498,
      "epoch": 0.5819879400894767,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004971398463960568,
      "loss": 5.2967,
      "mean_token_accuracy": 0.1769063279032707,
      "num_tokens": 12979836.0,
      "step": 7480
    },
    {
      "entropy": 5.564197444915772,
      "epoch": 0.5823769694611943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971354395887827,
      "loss": 5.3059,
      "mean_token_accuracy": 0.18093591332435607,
      "num_tokens": 12988233.0,
      "step": 7485
    },
    {
      "entropy": 5.561965847015381,
      "epoch": 0.5827659988329119,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971310294109569,
      "loss": 5.317,
      "mean_token_accuracy": 0.17434777021408082,
      "num_tokens": 12997077.0,
      "step": 7490
    },
    {
      "entropy": 5.549249458312988,
      "epoch": 0.5831550282046295,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004971266158626461,
      "loss": 5.3422,
      "mean_token_accuracy": 0.17838245630264282,
      "num_tokens": 13005058.0,
      "step": 7495
    },
    {
      "entropy": 5.6491687297821045,
      "epoch": 0.583544057576347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971221989439177,
      "loss": 5.4007,
      "mean_token_accuracy": 0.17415667623281478,
      "num_tokens": 13013347.0,
      "step": 7500
    },
    {
      "entropy": 5.656944322586059,
      "epoch": 0.5839330869480646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971177786548384,
      "loss": 5.3388,
      "mean_token_accuracy": 0.1761828675866127,
      "num_tokens": 13021555.0,
      "step": 7505
    },
    {
      "entropy": 5.568319988250733,
      "epoch": 0.5843221163197821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971133549954753,
      "loss": 5.414,
      "mean_token_accuracy": 0.17406437993049623,
      "num_tokens": 13029817.0,
      "step": 7510
    },
    {
      "entropy": 5.583380794525146,
      "epoch": 0.5847111456914997,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971089279658954,
      "loss": 5.3826,
      "mean_token_accuracy": 0.17224892526865004,
      "num_tokens": 13038119.0,
      "step": 7515
    },
    {
      "entropy": 5.566053342819214,
      "epoch": 0.5851001750632173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971044975661663,
      "loss": 5.31,
      "mean_token_accuracy": 0.17912320643663407,
      "num_tokens": 13047004.0,
      "step": 7520
    },
    {
      "entropy": 5.577041339874268,
      "epoch": 0.5854892044349348,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971000637963549,
      "loss": 5.2967,
      "mean_token_accuracy": 0.1811252474784851,
      "num_tokens": 13055045.0,
      "step": 7525
    },
    {
      "entropy": 5.614419794082641,
      "epoch": 0.5858782338066524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970956266565285,
      "loss": 5.3933,
      "mean_token_accuracy": 0.17078788280487062,
      "num_tokens": 13064535.0,
      "step": 7530
    },
    {
      "entropy": 5.615678119659424,
      "epoch": 0.58626726317837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970911861467544,
      "loss": 5.4565,
      "mean_token_accuracy": 0.16273304224014282,
      "num_tokens": 13073089.0,
      "step": 7535
    },
    {
      "entropy": 5.569165563583374,
      "epoch": 0.5866562925500876,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970867422671002,
      "loss": 5.3926,
      "mean_token_accuracy": 0.17114685624837875,
      "num_tokens": 13082161.0,
      "step": 7540
    },
    {
      "entropy": 5.559341335296631,
      "epoch": 0.5870453219218051,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497082295017633,
      "loss": 5.261,
      "mean_token_accuracy": 0.18079383820295333,
      "num_tokens": 13090632.0,
      "step": 7545
    },
    {
      "entropy": 5.624056482315064,
      "epoch": 0.5874343512935226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970778443984206,
      "loss": 5.3232,
      "mean_token_accuracy": 0.17447323203086854,
      "num_tokens": 13098825.0,
      "step": 7550
    },
    {
      "entropy": 5.634274244308472,
      "epoch": 0.5878233806652402,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004970733904095303,
      "loss": 5.3991,
      "mean_token_accuracy": 0.17641332894563674,
      "num_tokens": 13107059.0,
      "step": 7555
    },
    {
      "entropy": 5.6285295486450195,
      "epoch": 0.5882124100369578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970689330510298,
      "loss": 5.4063,
      "mean_token_accuracy": 0.16372135579586028,
      "num_tokens": 13116783.0,
      "step": 7560
    },
    {
      "entropy": 5.636148834228516,
      "epoch": 0.5886014394086754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970644723229868,
      "loss": 5.3586,
      "mean_token_accuracy": 0.17543469071388246,
      "num_tokens": 13124998.0,
      "step": 7565
    },
    {
      "entropy": 5.6454380512237545,
      "epoch": 0.588990468780393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970600082254687,
      "loss": 5.4192,
      "mean_token_accuracy": 0.1735449880361557,
      "num_tokens": 13134507.0,
      "step": 7570
    },
    {
      "entropy": 5.549423742294311,
      "epoch": 0.5893794981521104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004970555407585436,
      "loss": 5.3189,
      "mean_token_accuracy": 0.17249612510204315,
      "num_tokens": 13143682.0,
      "step": 7575
    },
    {
      "entropy": 5.58020977973938,
      "epoch": 0.589768527523828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497051069922279,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17372599989175797,
      "num_tokens": 13151924.0,
      "step": 7580
    },
    {
      "entropy": 5.5826180934906,
      "epoch": 0.5901575568955456,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970465957167429,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17643746584653855,
      "num_tokens": 13160853.0,
      "step": 7585
    },
    {
      "entropy": 5.621823501586914,
      "epoch": 0.5905465862672632,
      "grad_norm": 1.0,
      "learning_rate": 0.000497042118142003,
      "loss": 5.379,
      "mean_token_accuracy": 0.180886971950531,
      "num_tokens": 13168993.0,
      "step": 7590
    },
    {
      "entropy": 5.562651824951172,
      "epoch": 0.5909356156389808,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970376371981275,
      "loss": 5.3061,
      "mean_token_accuracy": 0.17483243048191072,
      "num_tokens": 13177266.0,
      "step": 7595
    },
    {
      "entropy": 5.621056079864502,
      "epoch": 0.5913246450106983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970331528851842,
      "loss": 5.3839,
      "mean_token_accuracy": 0.16920152455568313,
      "num_tokens": 13185328.0,
      "step": 7600
    },
    {
      "entropy": 5.540424346923828,
      "epoch": 0.5917136743824158,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970286652032412,
      "loss": 5.2685,
      "mean_token_accuracy": 0.18061549067497254,
      "num_tokens": 13194502.0,
      "step": 7605
    },
    {
      "entropy": 5.563891077041626,
      "epoch": 0.5921027037541334,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970241741523667,
      "loss": 5.4074,
      "mean_token_accuracy": 0.17013546228408813,
      "num_tokens": 13203745.0,
      "step": 7610
    },
    {
      "entropy": 5.606880235671997,
      "epoch": 0.592491733125851,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970196797326286,
      "loss": 5.4611,
      "mean_token_accuracy": 0.1726463630795479,
      "num_tokens": 13213115.0,
      "step": 7615
    },
    {
      "entropy": 5.696786546707154,
      "epoch": 0.5928807624975686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970151819440955,
      "loss": 5.484,
      "mean_token_accuracy": 0.167642742395401,
      "num_tokens": 13222231.0,
      "step": 7620
    },
    {
      "entropy": 5.590134191513061,
      "epoch": 0.5932697918692861,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970106807868351,
      "loss": 5.2815,
      "mean_token_accuracy": 0.17260773181915284,
      "num_tokens": 13231346.0,
      "step": 7625
    },
    {
      "entropy": 5.575092601776123,
      "epoch": 0.5936588212410037,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497006176260916,
      "loss": 5.4107,
      "mean_token_accuracy": 0.16619206964969635,
      "num_tokens": 13240436.0,
      "step": 7630
    },
    {
      "entropy": 5.629506921768188,
      "epoch": 0.5940478506127213,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970016683664068,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1771960213780403,
      "num_tokens": 13249501.0,
      "step": 7635
    },
    {
      "entropy": 5.4879992485046385,
      "epoch": 0.5944368799844388,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969971571033754,
      "loss": 5.2579,
      "mean_token_accuracy": 0.18234076499938964,
      "num_tokens": 13258322.0,
      "step": 7640
    },
    {
      "entropy": 5.556401205062866,
      "epoch": 0.5948259093561564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004969926424718906,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17190596908330918,
      "num_tokens": 13267418.0,
      "step": 7645
    },
    {
      "entropy": 5.677619123458863,
      "epoch": 0.5952149387278739,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969881244720207,
      "loss": 5.4237,
      "mean_token_accuracy": 0.16786353588104247,
      "num_tokens": 13276243.0,
      "step": 7650
    },
    {
      "entropy": 5.4979438304901125,
      "epoch": 0.5956039680995915,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969836031038343,
      "loss": 5.3041,
      "mean_token_accuracy": 0.16940934658050538,
      "num_tokens": 13285611.0,
      "step": 7655
    },
    {
      "entropy": 5.584447860717773,
      "epoch": 0.5959929974713091,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969790783674002,
      "loss": 5.3985,
      "mean_token_accuracy": 0.17014468610286712,
      "num_tokens": 13293949.0,
      "step": 7660
    },
    {
      "entropy": 5.576897478103637,
      "epoch": 0.5963820268430267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969745502627868,
      "loss": 5.3419,
      "mean_token_accuracy": 0.17507205307483673,
      "num_tokens": 13302593.0,
      "step": 7665
    },
    {
      "entropy": 5.589267921447754,
      "epoch": 0.5967710562147442,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969700187900629,
      "loss": 5.3911,
      "mean_token_accuracy": 0.169544081389904,
      "num_tokens": 13310947.0,
      "step": 7670
    },
    {
      "entropy": 5.620465183258057,
      "epoch": 0.5971600855864618,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004969654839492973,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1709015414118767,
      "num_tokens": 13319233.0,
      "step": 7675
    },
    {
      "entropy": 5.654470491409302,
      "epoch": 0.5975491149581793,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969609457405588,
      "loss": 5.464,
      "mean_token_accuracy": 0.16662180572748184,
      "num_tokens": 13327979.0,
      "step": 7680
    },
    {
      "entropy": 5.604493188858032,
      "epoch": 0.5979381443298969,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004969564041639163,
      "loss": 5.4169,
      "mean_token_accuracy": 0.17162067741155623,
      "num_tokens": 13336699.0,
      "step": 7685
    },
    {
      "entropy": 5.585852384567261,
      "epoch": 0.5983271737016145,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969518592194386,
      "loss": 5.299,
      "mean_token_accuracy": 0.1750283420085907,
      "num_tokens": 13344688.0,
      "step": 7690
    },
    {
      "entropy": 5.544575977325439,
      "epoch": 0.5987162030733321,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969473109071946,
      "loss": 5.3237,
      "mean_token_accuracy": 0.17674118876457215,
      "num_tokens": 13353442.0,
      "step": 7695
    },
    {
      "entropy": 5.6312174797058105,
      "epoch": 0.5991052324450497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969427592272535,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1720339149236679,
      "num_tokens": 13362388.0,
      "step": 7700
    },
    {
      "entropy": 5.483028316497803,
      "epoch": 0.5994942618167671,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969382041796843,
      "loss": 5.196,
      "mean_token_accuracy": 0.18541455566883086,
      "num_tokens": 13370796.0,
      "step": 7705
    },
    {
      "entropy": 5.510852098464966,
      "epoch": 0.5998832911884847,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496933645764556,
      "loss": 5.3521,
      "mean_token_accuracy": 0.17108747214078904,
      "num_tokens": 13379733.0,
      "step": 7710
    },
    {
      "entropy": 5.534001636505127,
      "epoch": 0.6002723205602023,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004969290839819381,
      "loss": 5.2606,
      "mean_token_accuracy": 0.1825840100646019,
      "num_tokens": 13388278.0,
      "step": 7715
    },
    {
      "entropy": 5.6160965919494625,
      "epoch": 0.6006613499319199,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969245188318994,
      "loss": 5.4387,
      "mean_token_accuracy": 0.17236709743738174,
      "num_tokens": 13397461.0,
      "step": 7720
    },
    {
      "entropy": 5.53111891746521,
      "epoch": 0.6010503793036375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969199503145093,
      "loss": 5.307,
      "mean_token_accuracy": 0.1779583439230919,
      "num_tokens": 13406087.0,
      "step": 7725
    },
    {
      "entropy": 5.592005586624145,
      "epoch": 0.601439408675355,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969153784298374,
      "loss": 5.2829,
      "mean_token_accuracy": 0.1804654434323311,
      "num_tokens": 13414760.0,
      "step": 7730
    },
    {
      "entropy": 5.533606243133545,
      "epoch": 0.6018284380470725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004969108031779527,
      "loss": 5.3577,
      "mean_token_accuracy": 0.17098020613193513,
      "num_tokens": 13423574.0,
      "step": 7735
    },
    {
      "entropy": 5.529449605941773,
      "epoch": 0.6022174674187901,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969062245589247,
      "loss": 5.286,
      "mean_token_accuracy": 0.18063054382801055,
      "num_tokens": 13431799.0,
      "step": 7740
    },
    {
      "entropy": 5.571295166015625,
      "epoch": 0.6026064967905077,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004969016425728231,
      "loss": 5.3392,
      "mean_token_accuracy": 0.17426727712154388,
      "num_tokens": 13440436.0,
      "step": 7745
    },
    {
      "entropy": 5.569225454330445,
      "epoch": 0.6029955261622253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968970572197172,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18219717293977739,
      "num_tokens": 13449236.0,
      "step": 7750
    },
    {
      "entropy": 5.485515785217285,
      "epoch": 0.6033845555339428,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968924684996765,
      "loss": 5.254,
      "mean_token_accuracy": 0.17624879032373428,
      "num_tokens": 13457697.0,
      "step": 7755
    },
    {
      "entropy": 5.540696716308593,
      "epoch": 0.6037735849056604,
      "grad_norm": 1.046875,
      "learning_rate": 0.000496887876412771,
      "loss": 5.3577,
      "mean_token_accuracy": 0.17585546374320984,
      "num_tokens": 13465939.0,
      "step": 7760
    },
    {
      "entropy": 5.566684436798096,
      "epoch": 0.6041626142773779,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004968832809590701,
      "loss": 5.2813,
      "mean_token_accuracy": 0.18138911575078964,
      "num_tokens": 13474154.0,
      "step": 7765
    },
    {
      "entropy": 5.613201427459717,
      "epoch": 0.6045516436490955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968786821386435,
      "loss": 5.2996,
      "mean_token_accuracy": 0.17543721199035645,
      "num_tokens": 13482943.0,
      "step": 7770
    },
    {
      "entropy": 5.597583436965943,
      "epoch": 0.6049406730208131,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968740799515611,
      "loss": 5.4377,
      "mean_token_accuracy": 0.1688978724181652,
      "num_tokens": 13491945.0,
      "step": 7775
    },
    {
      "entropy": 5.537314510345459,
      "epoch": 0.6053297023925306,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968694743978926,
      "loss": 5.3675,
      "mean_token_accuracy": 0.16827480494976044,
      "num_tokens": 13501277.0,
      "step": 7780
    },
    {
      "entropy": 5.613169622421265,
      "epoch": 0.6057187317642482,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968648654777081,
      "loss": 5.393,
      "mean_token_accuracy": 0.17645200192928315,
      "num_tokens": 13509609.0,
      "step": 7785
    },
    {
      "entropy": 5.581654214859009,
      "epoch": 0.6061077611359658,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968602531910773,
      "loss": 5.391,
      "mean_token_accuracy": 0.16612172424793242,
      "num_tokens": 13518707.0,
      "step": 7790
    },
    {
      "entropy": 5.587740039825439,
      "epoch": 0.6064967905076833,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968556375380704,
      "loss": 5.3508,
      "mean_token_accuracy": 0.17670366466045379,
      "num_tokens": 13527973.0,
      "step": 7795
    },
    {
      "entropy": 5.5790623188018795,
      "epoch": 0.6068858198794009,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968510185187571,
      "loss": 5.288,
      "mean_token_accuracy": 0.18264268040657045,
      "num_tokens": 13536580.0,
      "step": 7800
    },
    {
      "entropy": 5.509527015686035,
      "epoch": 0.6072748492511184,
      "grad_norm": 1.09375,
      "learning_rate": 0.000496846396133208,
      "loss": 5.3231,
      "mean_token_accuracy": 0.17557021826505662,
      "num_tokens": 13544826.0,
      "step": 7805
    },
    {
      "entropy": 5.588966751098633,
      "epoch": 0.607663878622836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004968417703814928,
      "loss": 5.3712,
      "mean_token_accuracy": 0.18416630029678344,
      "num_tokens": 13552821.0,
      "step": 7810
    },
    {
      "entropy": 5.500114679336548,
      "epoch": 0.6080529079945536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968371412636818,
      "loss": 5.2417,
      "mean_token_accuracy": 0.17580069303512574,
      "num_tokens": 13560827.0,
      "step": 7815
    },
    {
      "entropy": 5.547198343276977,
      "epoch": 0.6084419373662712,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968325087798453,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17011705040931702,
      "num_tokens": 13569311.0,
      "step": 7820
    },
    {
      "entropy": 5.649914121627807,
      "epoch": 0.6088309667379888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968278729300536,
      "loss": 5.4098,
      "mean_token_accuracy": 0.16785431355237962,
      "num_tokens": 13579035.0,
      "step": 7825
    },
    {
      "entropy": 5.596644639968872,
      "epoch": 0.6092199961097062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968232337143769,
      "loss": 5.3608,
      "mean_token_accuracy": 0.17886339426040648,
      "num_tokens": 13587508.0,
      "step": 7830
    },
    {
      "entropy": 5.5370063304901125,
      "epoch": 0.6096090254814238,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968185911328858,
      "loss": 5.4058,
      "mean_token_accuracy": 0.17701992094516755,
      "num_tokens": 13596515.0,
      "step": 7835
    },
    {
      "entropy": 5.544048547744751,
      "epoch": 0.6099980548531414,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968139451856506,
      "loss": 5.2967,
      "mean_token_accuracy": 0.1797937497496605,
      "num_tokens": 13605162.0,
      "step": 7840
    },
    {
      "entropy": 5.561970663070679,
      "epoch": 0.610387084224859,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968092958727419,
      "loss": 5.2798,
      "mean_token_accuracy": 0.17963772565126418,
      "num_tokens": 13614398.0,
      "step": 7845
    },
    {
      "entropy": 5.54109377861023,
      "epoch": 0.6107761135965766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049680464319423,
      "loss": 5.2511,
      "mean_token_accuracy": 0.17509905099868775,
      "num_tokens": 13622347.0,
      "step": 7850
    },
    {
      "entropy": 5.503065776824951,
      "epoch": 0.6111651429682942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967999871501858,
      "loss": 5.2657,
      "mean_token_accuracy": 0.17665477693080903,
      "num_tokens": 13630773.0,
      "step": 7855
    },
    {
      "entropy": 5.551439714431763,
      "epoch": 0.6115541723400116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004967953277406798,
      "loss": 5.2635,
      "mean_token_accuracy": 0.1790945053100586,
      "num_tokens": 13639857.0,
      "step": 7860
    },
    {
      "entropy": 5.6004712104797365,
      "epoch": 0.6119432017117292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967906649657828,
      "loss": 5.3727,
      "mean_token_accuracy": 0.17795006930828094,
      "num_tokens": 13649959.0,
      "step": 7865
    },
    {
      "entropy": 5.625068473815918,
      "epoch": 0.6123322310834468,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004967859988255655,
      "loss": 5.3995,
      "mean_token_accuracy": 0.17279022485017775,
      "num_tokens": 13658072.0,
      "step": 7870
    },
    {
      "entropy": 5.529886484146118,
      "epoch": 0.6127212604551644,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967813293200985,
      "loss": 5.3382,
      "mean_token_accuracy": 0.18108959943056108,
      "num_tokens": 13666816.0,
      "step": 7875
    },
    {
      "entropy": 5.525184059143067,
      "epoch": 0.613110289826882,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967766564494529,
      "loss": 5.1971,
      "mean_token_accuracy": 0.18741032481193542,
      "num_tokens": 13675415.0,
      "step": 7880
    },
    {
      "entropy": 5.510573244094848,
      "epoch": 0.6134993191985995,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004967719802136996,
      "loss": 5.2432,
      "mean_token_accuracy": 0.18263396471738816,
      "num_tokens": 13684090.0,
      "step": 7885
    },
    {
      "entropy": 5.516405916213989,
      "epoch": 0.613888348570317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004967673006129094,
      "loss": 5.2915,
      "mean_token_accuracy": 0.18200115412473677,
      "num_tokens": 13692515.0,
      "step": 7890
    },
    {
      "entropy": 5.516071128845215,
      "epoch": 0.6142773779420346,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967626176471536,
      "loss": 5.356,
      "mean_token_accuracy": 0.1775418221950531,
      "num_tokens": 13700780.0,
      "step": 7895
    },
    {
      "entropy": 5.636789226531983,
      "epoch": 0.6146664073137522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967579313165028,
      "loss": 5.3535,
      "mean_token_accuracy": 0.1797405645251274,
      "num_tokens": 13708638.0,
      "step": 7900
    },
    {
      "entropy": 5.601937341690063,
      "epoch": 0.6150554366854698,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967532416210284,
      "loss": 5.3831,
      "mean_token_accuracy": 0.1653753012418747,
      "num_tokens": 13717867.0,
      "step": 7905
    },
    {
      "entropy": 5.563804340362549,
      "epoch": 0.6154444660571873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967485485608016,
      "loss": 5.3028,
      "mean_token_accuracy": 0.17676510214805602,
      "num_tokens": 13727280.0,
      "step": 7910
    },
    {
      "entropy": 5.595907878875733,
      "epoch": 0.6158334954289049,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967438521358934,
      "loss": 5.3475,
      "mean_token_accuracy": 0.17540148198604583,
      "num_tokens": 13735255.0,
      "step": 7915
    },
    {
      "entropy": 5.532194900512695,
      "epoch": 0.6162225248006225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967391523463754,
      "loss": 5.3775,
      "mean_token_accuracy": 0.17733073532581328,
      "num_tokens": 13743418.0,
      "step": 7920
    },
    {
      "entropy": 5.543993854522705,
      "epoch": 0.61661155417234,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967344491923185,
      "loss": 5.2762,
      "mean_token_accuracy": 0.17798438519239426,
      "num_tokens": 13752714.0,
      "step": 7925
    },
    {
      "entropy": 5.6618695735931395,
      "epoch": 0.6170005835440576,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004967297426737943,
      "loss": 5.4486,
      "mean_token_accuracy": 0.16520682275295256,
      "num_tokens": 13761350.0,
      "step": 7930
    },
    {
      "entropy": 5.5297904968261715,
      "epoch": 0.6173896129157751,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967250327908742,
      "loss": 5.2824,
      "mean_token_accuracy": 0.17875937521457672,
      "num_tokens": 13769769.0,
      "step": 7935
    },
    {
      "entropy": 5.550059127807617,
      "epoch": 0.6177786422874927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967203195436297,
      "loss": 5.311,
      "mean_token_accuracy": 0.17310796827077865,
      "num_tokens": 13778432.0,
      "step": 7940
    },
    {
      "entropy": 5.5211821556091305,
      "epoch": 0.6181676716592103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967156029321322,
      "loss": 5.3308,
      "mean_token_accuracy": 0.17745684683322907,
      "num_tokens": 13786968.0,
      "step": 7945
    },
    {
      "entropy": 5.61552062034607,
      "epoch": 0.6185567010309279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967108829564532,
      "loss": 5.4223,
      "mean_token_accuracy": 0.16451818495988846,
      "num_tokens": 13795824.0,
      "step": 7950
    },
    {
      "entropy": 5.629518461227417,
      "epoch": 0.6189457304026454,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967061596166646,
      "loss": 5.3494,
      "mean_token_accuracy": 0.17811042666435242,
      "num_tokens": 13804748.0,
      "step": 7955
    },
    {
      "entropy": 5.509480619430542,
      "epoch": 0.6193347597743629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967014329128378,
      "loss": 5.3398,
      "mean_token_accuracy": 0.17735178172588348,
      "num_tokens": 13813106.0,
      "step": 7960
    },
    {
      "entropy": 5.555459022521973,
      "epoch": 0.6197237891460805,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004966967028450448,
      "loss": 5.3667,
      "mean_token_accuracy": 0.17630673050880433,
      "num_tokens": 13821787.0,
      "step": 7965
    },
    {
      "entropy": 5.603826332092285,
      "epoch": 0.6201128185177981,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496691969413357,
      "loss": 5.3539,
      "mean_token_accuracy": 0.17787329405546187,
      "num_tokens": 13830275.0,
      "step": 7970
    },
    {
      "entropy": 5.557550859451294,
      "epoch": 0.6205018478895157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966872326178466,
      "loss": 5.2798,
      "mean_token_accuracy": 0.1828768104314804,
      "num_tokens": 13838868.0,
      "step": 7975
    },
    {
      "entropy": 5.551296949386597,
      "epoch": 0.6208908772612333,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966824924585851,
      "loss": 5.3555,
      "mean_token_accuracy": 0.17854246348142624,
      "num_tokens": 13846970.0,
      "step": 7980
    },
    {
      "entropy": 5.613220357894898,
      "epoch": 0.6212799066329507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966777489356448,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17233560085296631,
      "num_tokens": 13856059.0,
      "step": 7985
    },
    {
      "entropy": 5.587001371383667,
      "epoch": 0.6216689360046683,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966730020490975,
      "loss": 5.3059,
      "mean_token_accuracy": 0.17722841799259187,
      "num_tokens": 13865461.0,
      "step": 7990
    },
    {
      "entropy": 5.5760759830474855,
      "epoch": 0.6220579653763859,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496668251799015,
      "loss": 5.2765,
      "mean_token_accuracy": 0.17371441125869752,
      "num_tokens": 13874124.0,
      "step": 7995
    },
    {
      "entropy": 5.613064813613891,
      "epoch": 0.6224469947481035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004966634981854699,
      "loss": 5.3814,
      "mean_token_accuracy": 0.17237360179424285,
      "num_tokens": 13882952.0,
      "step": 8000
    },
    {
      "entropy": 5.585406970977783,
      "epoch": 0.6228360241198211,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966587412085339,
      "loss": 5.3293,
      "mean_token_accuracy": 0.17545945644378663,
      "num_tokens": 13890687.0,
      "step": 8005
    },
    {
      "entropy": 5.577899980545044,
      "epoch": 0.6232250534915386,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966539808682793,
      "loss": 5.4074,
      "mean_token_accuracy": 0.1689345955848694,
      "num_tokens": 13899773.0,
      "step": 8010
    },
    {
      "entropy": 5.468811941146851,
      "epoch": 0.6236140828632561,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966492171647783,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18073903769254684,
      "num_tokens": 13908313.0,
      "step": 8015
    },
    {
      "entropy": 5.554126024246216,
      "epoch": 0.6240031122349737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004966444500981032,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18334477245807648,
      "num_tokens": 13916422.0,
      "step": 8020
    },
    {
      "entropy": 5.580251407623291,
      "epoch": 0.6243921416066913,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966396796683265,
      "loss": 5.3621,
      "mean_token_accuracy": 0.17060983330011367,
      "num_tokens": 13924971.0,
      "step": 8025
    },
    {
      "entropy": 5.572302627563476,
      "epoch": 0.6247811709784089,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966349058755204,
      "loss": 5.2716,
      "mean_token_accuracy": 0.17709621489048005,
      "num_tokens": 13933016.0,
      "step": 8030
    },
    {
      "entropy": 5.673058271408081,
      "epoch": 0.6251702003501264,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966301287197573,
      "loss": 5.3638,
      "mean_token_accuracy": 0.17842250913381577,
      "num_tokens": 13941814.0,
      "step": 8035
    },
    {
      "entropy": 5.531317281723022,
      "epoch": 0.625559229721844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966253482011098,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1801487997174263,
      "num_tokens": 13950344.0,
      "step": 8040
    },
    {
      "entropy": 5.555147981643676,
      "epoch": 0.6259482590935616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966205643196503,
      "loss": 5.4287,
      "mean_token_accuracy": 0.1682429000735283,
      "num_tokens": 13959456.0,
      "step": 8045
    },
    {
      "entropy": 5.592694234848023,
      "epoch": 0.6263372884652791,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966157770754516,
      "loss": 5.3065,
      "mean_token_accuracy": 0.18245015740394593,
      "num_tokens": 13967303.0,
      "step": 8050
    },
    {
      "entropy": 5.585998249053955,
      "epoch": 0.6267263178369967,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496610986468586,
      "loss": 5.3509,
      "mean_token_accuracy": 0.16850290894508363,
      "num_tokens": 13977138.0,
      "step": 8055
    },
    {
      "entropy": 5.638797283172607,
      "epoch": 0.6271153472087143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004966061924991267,
      "loss": 5.4227,
      "mean_token_accuracy": 0.17261697947978974,
      "num_tokens": 13985062.0,
      "step": 8060
    },
    {
      "entropy": 5.668380260467529,
      "epoch": 0.6275043765804318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496601395167146,
      "loss": 5.3426,
      "mean_token_accuracy": 0.1716477170586586,
      "num_tokens": 13992923.0,
      "step": 8065
    },
    {
      "entropy": 5.543324136734009,
      "epoch": 0.6278934059521494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004965965944727168,
      "loss": 5.4132,
      "mean_token_accuracy": 0.16822911947965621,
      "num_tokens": 14002116.0,
      "step": 8070
    },
    {
      "entropy": 5.580185413360596,
      "epoch": 0.628282435323867,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965917904159121,
      "loss": 5.3931,
      "mean_token_accuracy": 0.17099771350622178,
      "num_tokens": 14011053.0,
      "step": 8075
    },
    {
      "entropy": 5.630213594436645,
      "epoch": 0.6286714646955845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965869829968046,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1766078531742096,
      "num_tokens": 14019226.0,
      "step": 8080
    },
    {
      "entropy": 5.4770303726196286,
      "epoch": 0.6290604940673021,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004965821722154674,
      "loss": 5.3391,
      "mean_token_accuracy": 0.17408648133277893,
      "num_tokens": 14027630.0,
      "step": 8085
    },
    {
      "entropy": 5.544252061843872,
      "epoch": 0.6294495234390196,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965773580719734,
      "loss": 5.2561,
      "mean_token_accuracy": 0.18546278923749923,
      "num_tokens": 14036004.0,
      "step": 8090
    },
    {
      "entropy": 5.5833745956420895,
      "epoch": 0.6298385528107372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965725405663957,
      "loss": 5.3554,
      "mean_token_accuracy": 0.17322381734848022,
      "num_tokens": 14045359.0,
      "step": 8095
    },
    {
      "entropy": 5.542785596847534,
      "epoch": 0.6302275821824548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965677196988073,
      "loss": 5.2898,
      "mean_token_accuracy": 0.1769864469766617,
      "num_tokens": 14054099.0,
      "step": 8100
    },
    {
      "entropy": 5.564396667480469,
      "epoch": 0.6306166115541724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965628954692815,
      "loss": 5.3129,
      "mean_token_accuracy": 0.18053655177354813,
      "num_tokens": 14062414.0,
      "step": 8105
    },
    {
      "entropy": 5.538123321533203,
      "epoch": 0.63100564092589,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965580678778913,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17966470420360564,
      "num_tokens": 14071016.0,
      "step": 8110
    },
    {
      "entropy": 5.558867883682251,
      "epoch": 0.6313946702976074,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965532369247102,
      "loss": 5.343,
      "mean_token_accuracy": 0.17081669420003892,
      "num_tokens": 14080622.0,
      "step": 8115
    },
    {
      "entropy": 5.537043571472168,
      "epoch": 0.631783699669325,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965484026098111,
      "loss": 5.1584,
      "mean_token_accuracy": 0.18677742630243302,
      "num_tokens": 14089500.0,
      "step": 8120
    },
    {
      "entropy": 5.58234395980835,
      "epoch": 0.6321727290410426,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004965435649332679,
      "loss": 5.4148,
      "mean_token_accuracy": 0.17902248948812485,
      "num_tokens": 14098260.0,
      "step": 8125
    },
    {
      "entropy": 5.554372978210449,
      "epoch": 0.6325617584127602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965387238951536,
      "loss": 5.2376,
      "mean_token_accuracy": 0.1855595663189888,
      "num_tokens": 14106561.0,
      "step": 8130
    },
    {
      "entropy": 5.603929567337036,
      "epoch": 0.6329507877844778,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965338794955418,
      "loss": 5.338,
      "mean_token_accuracy": 0.1756177231669426,
      "num_tokens": 14115049.0,
      "step": 8135
    },
    {
      "entropy": 5.541755533218383,
      "epoch": 0.6333398171561953,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965290317345061,
      "loss": 5.2387,
      "mean_token_accuracy": 0.17843602746725082,
      "num_tokens": 14123204.0,
      "step": 8140
    },
    {
      "entropy": 5.4808355331420895,
      "epoch": 0.6337288465279128,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965241806121198,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18215357661247253,
      "num_tokens": 14132146.0,
      "step": 8145
    },
    {
      "entropy": 5.628285121917725,
      "epoch": 0.6341178758996304,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004965193261284567,
      "loss": 5.3899,
      "mean_token_accuracy": 0.16875875890254974,
      "num_tokens": 14140839.0,
      "step": 8150
    },
    {
      "entropy": 5.610815334320068,
      "epoch": 0.634506905271348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004965144682835904,
      "loss": 5.3795,
      "mean_token_accuracy": 0.1781456559896469,
      "num_tokens": 14149551.0,
      "step": 8155
    },
    {
      "entropy": 5.515106868743897,
      "epoch": 0.6348959346430656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004965096070775946,
      "loss": 5.3471,
      "mean_token_accuracy": 0.1755740150809288,
      "num_tokens": 14158414.0,
      "step": 8160
    },
    {
      "entropy": 5.577139282226563,
      "epoch": 0.6352849640147831,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965047425105431,
      "loss": 5.3882,
      "mean_token_accuracy": 0.17380691021680833,
      "num_tokens": 14167984.0,
      "step": 8165
    },
    {
      "entropy": 5.535385370254517,
      "epoch": 0.6356739933865007,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964998745825097,
      "loss": 5.316,
      "mean_token_accuracy": 0.17888084948062896,
      "num_tokens": 14177244.0,
      "step": 8170
    },
    {
      "entropy": 5.588834619522094,
      "epoch": 0.6360630227582182,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964950032935682,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1779651641845703,
      "num_tokens": 14185754.0,
      "step": 8175
    },
    {
      "entropy": 5.528890562057495,
      "epoch": 0.6364520521299358,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964901286437927,
      "loss": 5.2899,
      "mean_token_accuracy": 0.1799963116645813,
      "num_tokens": 14194242.0,
      "step": 8180
    },
    {
      "entropy": 5.524556064605713,
      "epoch": 0.6368410815016534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964852506332568,
      "loss": 5.2153,
      "mean_token_accuracy": 0.18313002586364746,
      "num_tokens": 14202496.0,
      "step": 8185
    },
    {
      "entropy": 5.5839591979980465,
      "epoch": 0.6372301108733709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496480369262035,
      "loss": 5.448,
      "mean_token_accuracy": 0.17073566168546678,
      "num_tokens": 14210510.0,
      "step": 8190
    },
    {
      "entropy": 5.605168581008911,
      "epoch": 0.6376191402450885,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496475484530201,
      "loss": 5.4347,
      "mean_token_accuracy": 0.16247031688690186,
      "num_tokens": 14219238.0,
      "step": 8195
    },
    {
      "entropy": 5.502722358703613,
      "epoch": 0.6380081696168061,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964705964378292,
      "loss": 5.243,
      "mean_token_accuracy": 0.18735675364732743,
      "num_tokens": 14227651.0,
      "step": 8200
    },
    {
      "entropy": 5.496855878829956,
      "epoch": 0.6383971989885237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964657049849934,
      "loss": 5.2097,
      "mean_token_accuracy": 0.18215781450271606,
      "num_tokens": 14236144.0,
      "step": 8205
    },
    {
      "entropy": 5.635563611984253,
      "epoch": 0.6387862283602412,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964608101717681,
      "loss": 5.3734,
      "mean_token_accuracy": 0.16718583106994628,
      "num_tokens": 14244706.0,
      "step": 8210
    },
    {
      "entropy": 5.574279880523681,
      "epoch": 0.6391752577319587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004964559119982275,
      "loss": 5.3199,
      "mean_token_accuracy": 0.18086613714694977,
      "num_tokens": 14252873.0,
      "step": 8215
    },
    {
      "entropy": 5.576375818252563,
      "epoch": 0.6395642871036763,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496451010464446,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1750391438603401,
      "num_tokens": 14261383.0,
      "step": 8220
    },
    {
      "entropy": 5.475570201873779,
      "epoch": 0.6399533164753939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964461055704978,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18437148928642272,
      "num_tokens": 14270087.0,
      "step": 8225
    },
    {
      "entropy": 5.6076884269714355,
      "epoch": 0.6403423458471115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004964411973164574,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18397365808486937,
      "num_tokens": 14278826.0,
      "step": 8230
    },
    {
      "entropy": 5.507570075988769,
      "epoch": 0.6407313752188291,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964362857023993,
      "loss": 5.2796,
      "mean_token_accuracy": 0.17902056723833085,
      "num_tokens": 14287554.0,
      "step": 8235
    },
    {
      "entropy": 5.566535139083863,
      "epoch": 0.6411204045905465,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496431370728398,
      "loss": 5.2667,
      "mean_token_accuracy": 0.17709399312734603,
      "num_tokens": 14295839.0,
      "step": 8240
    },
    {
      "entropy": 5.555580949783325,
      "epoch": 0.6415094339622641,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004964264523945281,
      "loss": 5.3024,
      "mean_token_accuracy": 0.1806720845401287,
      "num_tokens": 14304831.0,
      "step": 8245
    },
    {
      "entropy": 5.575355482101441,
      "epoch": 0.6418984633339817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964215307008643,
      "loss": 5.2754,
      "mean_token_accuracy": 0.17711037695407866,
      "num_tokens": 14313249.0,
      "step": 8250
    },
    {
      "entropy": 5.573872232437134,
      "epoch": 0.6422874927056993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964166056474811,
      "loss": 5.3412,
      "mean_token_accuracy": 0.17384667545557023,
      "num_tokens": 14321978.0,
      "step": 8255
    },
    {
      "entropy": 5.4906227588653564,
      "epoch": 0.6426765220774169,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964116772344534,
      "loss": 5.2501,
      "mean_token_accuracy": 0.17830630242824555,
      "num_tokens": 14330379.0,
      "step": 8260
    },
    {
      "entropy": 5.5215977191925045,
      "epoch": 0.6430655514491345,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964067454618559,
      "loss": 5.2935,
      "mean_token_accuracy": 0.18217869699001313,
      "num_tokens": 14339121.0,
      "step": 8265
    },
    {
      "entropy": 5.504442310333252,
      "epoch": 0.6434545808208519,
      "grad_norm": 1.125,
      "learning_rate": 0.0004964018103297634,
      "loss": 5.247,
      "mean_token_accuracy": 0.18260000348091127,
      "num_tokens": 14347443.0,
      "step": 8270
    },
    {
      "entropy": 5.5835634708404545,
      "epoch": 0.6438436101925695,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963968718382509,
      "loss": 5.3465,
      "mean_token_accuracy": 0.17404672801494597,
      "num_tokens": 14356459.0,
      "step": 8275
    },
    {
      "entropy": 5.550328874588013,
      "epoch": 0.6442326395642871,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004963919299873931,
      "loss": 5.2589,
      "mean_token_accuracy": 0.17677343338727952,
      "num_tokens": 14365359.0,
      "step": 8280
    },
    {
      "entropy": 5.450421857833862,
      "epoch": 0.6446216689360047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004963869847772653,
      "loss": 5.196,
      "mean_token_accuracy": 0.18321047723293304,
      "num_tokens": 14373345.0,
      "step": 8285
    },
    {
      "entropy": 5.513602542877197,
      "epoch": 0.6450106983077223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963820362079423,
      "loss": 5.2799,
      "mean_token_accuracy": 0.18093516677618027,
      "num_tokens": 14381977.0,
      "step": 8290
    },
    {
      "entropy": 5.462281084060669,
      "epoch": 0.6453997276794398,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004963770842794994,
      "loss": 5.1796,
      "mean_token_accuracy": 0.1848599746823311,
      "num_tokens": 14390486.0,
      "step": 8295
    },
    {
      "entropy": 5.5589361667633055,
      "epoch": 0.6457887570511573,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963721289920115,
      "loss": 5.3657,
      "mean_token_accuracy": 0.17912714034318925,
      "num_tokens": 14399064.0,
      "step": 8300
    },
    {
      "entropy": 5.539777040481567,
      "epoch": 0.6461777864228749,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963671703455538,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1839711531996727,
      "num_tokens": 14406965.0,
      "step": 8305
    },
    {
      "entropy": 5.482750988006591,
      "epoch": 0.6465668157945925,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963622083402018,
      "loss": 5.3649,
      "mean_token_accuracy": 0.17665068954229354,
      "num_tokens": 14416293.0,
      "step": 8310
    },
    {
      "entropy": 5.606808519363403,
      "epoch": 0.6469558451663101,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963572429760305,
      "loss": 5.3405,
      "mean_token_accuracy": 0.17011832147836686,
      "num_tokens": 14424515.0,
      "step": 8315
    },
    {
      "entropy": 5.556892204284668,
      "epoch": 0.6473448745380276,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963522742531155,
      "loss": 5.1785,
      "mean_token_accuracy": 0.18246434181928634,
      "num_tokens": 14432706.0,
      "step": 8320
    },
    {
      "entropy": 5.480896806716919,
      "epoch": 0.6477339039097452,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496347302171532,
      "loss": 5.2834,
      "mean_token_accuracy": 0.1854716181755066,
      "num_tokens": 14441462.0,
      "step": 8325
    },
    {
      "entropy": 5.467887020111084,
      "epoch": 0.6481229332814628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963423267313554,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1838444337248802,
      "num_tokens": 14450218.0,
      "step": 8330
    },
    {
      "entropy": 5.616094160079956,
      "epoch": 0.6485119626531803,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963373479326614,
      "loss": 5.4552,
      "mean_token_accuracy": 0.16786930859088897,
      "num_tokens": 14459825.0,
      "step": 8335
    },
    {
      "entropy": 5.570043230056763,
      "epoch": 0.6489009920248979,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963323657755255,
      "loss": 5.2699,
      "mean_token_accuracy": 0.17940721809864044,
      "num_tokens": 14468647.0,
      "step": 8340
    },
    {
      "entropy": 5.465993928909302,
      "epoch": 0.6492900213966154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963273802600231,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1850014716386795,
      "num_tokens": 14476892.0,
      "step": 8345
    },
    {
      "entropy": 5.603579473495484,
      "epoch": 0.649679050768333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049632239138623,
      "loss": 5.429,
      "mean_token_accuracy": 0.17392351776361464,
      "num_tokens": 14485541.0,
      "step": 8350
    },
    {
      "entropy": 5.520974779129029,
      "epoch": 0.6500680801400506,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963173991542219,
      "loss": 5.2557,
      "mean_token_accuracy": 0.18615519255399704,
      "num_tokens": 14494184.0,
      "step": 8355
    },
    {
      "entropy": 5.593913173675537,
      "epoch": 0.6504571095117682,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004963124035640746,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1692001700401306,
      "num_tokens": 14502151.0,
      "step": 8360
    },
    {
      "entropy": 5.55236496925354,
      "epoch": 0.6508461388834857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963074046158637,
      "loss": 5.3677,
      "mean_token_accuracy": 0.18047490417957307,
      "num_tokens": 14511254.0,
      "step": 8365
    },
    {
      "entropy": 5.6321656703948975,
      "epoch": 0.6512351682552032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963024023096652,
      "loss": 5.4118,
      "mean_token_accuracy": 0.16682166159152984,
      "num_tokens": 14520492.0,
      "step": 8370
    },
    {
      "entropy": 5.631375026702881,
      "epoch": 0.6516241976269208,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962973966455551,
      "loss": 5.3601,
      "mean_token_accuracy": 0.16875505596399307,
      "num_tokens": 14530082.0,
      "step": 8375
    },
    {
      "entropy": 5.546644496917724,
      "epoch": 0.6520132269986384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962923876236092,
      "loss": 5.3218,
      "mean_token_accuracy": 0.17745062112808227,
      "num_tokens": 14539124.0,
      "step": 8380
    },
    {
      "entropy": 5.607315874099731,
      "epoch": 0.652402256370356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004962873752439035,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1814492389559746,
      "num_tokens": 14547780.0,
      "step": 8385
    },
    {
      "entropy": 5.554334831237793,
      "epoch": 0.6527912857420736,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004962823595065141,
      "loss": 5.2198,
      "mean_token_accuracy": 0.1786617323756218,
      "num_tokens": 14556225.0,
      "step": 8390
    },
    {
      "entropy": 5.594890832901001,
      "epoch": 0.653180315113791,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004962773404115172,
      "loss": 5.3183,
      "mean_token_accuracy": 0.1761225014925003,
      "num_tokens": 14564147.0,
      "step": 8395
    },
    {
      "entropy": 5.551420640945435,
      "epoch": 0.6535693444855086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962723179589886,
      "loss": 5.3068,
      "mean_token_accuracy": 0.17587434202432634,
      "num_tokens": 14573519.0,
      "step": 8400
    },
    {
      "entropy": 5.59135627746582,
      "epoch": 0.6539583738572262,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496267292149005,
      "loss": 5.4309,
      "mean_token_accuracy": 0.16442757546901704,
      "num_tokens": 14582806.0,
      "step": 8405
    },
    {
      "entropy": 5.543904972076416,
      "epoch": 0.6543474032289438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962622629816423,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1739957958459854,
      "num_tokens": 14591274.0,
      "step": 8410
    },
    {
      "entropy": 5.530463457107544,
      "epoch": 0.6547364326006614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496257230456977,
      "loss": 5.3608,
      "mean_token_accuracy": 0.17313478589057923,
      "num_tokens": 14599902.0,
      "step": 8415
    },
    {
      "entropy": 5.493044996261597,
      "epoch": 0.6551254619723789,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962521945750855,
      "loss": 5.2596,
      "mean_token_accuracy": 0.17716848850250244,
      "num_tokens": 14608931.0,
      "step": 8420
    },
    {
      "entropy": 5.530213308334351,
      "epoch": 0.6555144913440965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000496247155336044,
      "loss": 5.2194,
      "mean_token_accuracy": 0.18342820256948472,
      "num_tokens": 14617547.0,
      "step": 8425
    },
    {
      "entropy": 5.550770044326782,
      "epoch": 0.655903520715814,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004962421127399291,
      "loss": 5.3447,
      "mean_token_accuracy": 0.17101119011640548,
      "num_tokens": 14626523.0,
      "step": 8430
    },
    {
      "entropy": 5.586200857162476,
      "epoch": 0.6562925500875316,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962370667868172,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1851767346262932,
      "num_tokens": 14634846.0,
      "step": 8435
    },
    {
      "entropy": 5.528326034545898,
      "epoch": 0.6566815794592492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962320174767851,
      "loss": 5.294,
      "mean_token_accuracy": 0.1716813012957573,
      "num_tokens": 14643092.0,
      "step": 8440
    },
    {
      "entropy": 5.591558742523193,
      "epoch": 0.6570706088309667,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496226964809909,
      "loss": 5.3704,
      "mean_token_accuracy": 0.1788558319211006,
      "num_tokens": 14651456.0,
      "step": 8445
    },
    {
      "entropy": 5.539442729949951,
      "epoch": 0.6574596382026843,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496221908786266,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18235833197832108,
      "num_tokens": 14660488.0,
      "step": 8450
    },
    {
      "entropy": 5.618861961364746,
      "epoch": 0.6578486675744019,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962168494059327,
      "loss": 5.4666,
      "mean_token_accuracy": 0.17066026628017425,
      "num_tokens": 14670131.0,
      "step": 8455
    },
    {
      "entropy": 5.613412618637085,
      "epoch": 0.6582376969461194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962117866689857,
      "loss": 5.2624,
      "mean_token_accuracy": 0.1764333188533783,
      "num_tokens": 14678402.0,
      "step": 8460
    },
    {
      "entropy": 5.541571140289307,
      "epoch": 0.658626726317837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496206720575502,
      "loss": 5.2897,
      "mean_token_accuracy": 0.16646923422813414,
      "num_tokens": 14687589.0,
      "step": 8465
    },
    {
      "entropy": 5.584652948379516,
      "epoch": 0.6590157556895546,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962016511255584,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1722184956073761,
      "num_tokens": 14696288.0,
      "step": 8470
    },
    {
      "entropy": 5.634091234207153,
      "epoch": 0.6594047850612721,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961965783192318,
      "loss": 5.3419,
      "mean_token_accuracy": 0.17304196506738662,
      "num_tokens": 14705269.0,
      "step": 8475
    },
    {
      "entropy": 5.558816766738891,
      "epoch": 0.6597938144329897,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961915021565992,
      "loss": 5.3162,
      "mean_token_accuracy": 0.17708727568387986,
      "num_tokens": 14713669.0,
      "step": 8480
    },
    {
      "entropy": 5.500200891494751,
      "epoch": 0.6601828438047073,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961864226377377,
      "loss": 5.3647,
      "mean_token_accuracy": 0.17835226505994797,
      "num_tokens": 14722529.0,
      "step": 8485
    },
    {
      "entropy": 5.575636339187622,
      "epoch": 0.6605718731764249,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004961813397627241,
      "loss": 5.3458,
      "mean_token_accuracy": 0.18031770139932632,
      "num_tokens": 14730956.0,
      "step": 8490
    },
    {
      "entropy": 5.5568760395050045,
      "epoch": 0.6609609025481424,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961762535316358,
      "loss": 5.3101,
      "mean_token_accuracy": 0.18019221276044844,
      "num_tokens": 14739320.0,
      "step": 8495
    },
    {
      "entropy": 5.523578739166259,
      "epoch": 0.6613499319198599,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961711639445499,
      "loss": 5.3665,
      "mean_token_accuracy": 0.17696610540151597,
      "num_tokens": 14748030.0,
      "step": 8500
    },
    {
      "entropy": 5.552655601501465,
      "epoch": 0.6617389612915775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961660710015435,
      "loss": 5.2504,
      "mean_token_accuracy": 0.18410542160272597,
      "num_tokens": 14756786.0,
      "step": 8505
    },
    {
      "entropy": 5.567683458328247,
      "epoch": 0.6621279906632951,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961609747026943,
      "loss": 5.3213,
      "mean_token_accuracy": 0.1770300105214119,
      "num_tokens": 14765538.0,
      "step": 8510
    },
    {
      "entropy": 5.5553281784057615,
      "epoch": 0.6625170200350127,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961558750480791,
      "loss": 5.2499,
      "mean_token_accuracy": 0.18100422173738479,
      "num_tokens": 14773845.0,
      "step": 8515
    },
    {
      "entropy": 5.499459505081177,
      "epoch": 0.6629060494067303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961507720377756,
      "loss": 5.2031,
      "mean_token_accuracy": 0.18550216108560563,
      "num_tokens": 14782022.0,
      "step": 8520
    },
    {
      "entropy": 5.564083671569824,
      "epoch": 0.6632950787784477,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961456656718611,
      "loss": 5.343,
      "mean_token_accuracy": 0.18136597871780397,
      "num_tokens": 14790150.0,
      "step": 8525
    },
    {
      "entropy": 5.549467277526856,
      "epoch": 0.6636841081501653,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496140555950413,
      "loss": 5.379,
      "mean_token_accuracy": 0.17202800512313843,
      "num_tokens": 14799239.0,
      "step": 8530
    },
    {
      "entropy": 5.526183414459228,
      "epoch": 0.6640731375218829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961354428735091,
      "loss": 5.3029,
      "mean_token_accuracy": 0.16849559545516968,
      "num_tokens": 14808740.0,
      "step": 8535
    },
    {
      "entropy": 5.681412315368652,
      "epoch": 0.6644621668936005,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961303264412267,
      "loss": 5.4267,
      "mean_token_accuracy": 0.17152373641729354,
      "num_tokens": 14817329.0,
      "step": 8540
    },
    {
      "entropy": 5.554582118988037,
      "epoch": 0.6648511962653181,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961252066536437,
      "loss": 5.279,
      "mean_token_accuracy": 0.17454319149255754,
      "num_tokens": 14826240.0,
      "step": 8545
    },
    {
      "entropy": 5.548708868026734,
      "epoch": 0.6652402256370356,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961200835108375,
      "loss": 5.299,
      "mean_token_accuracy": 0.1745982825756073,
      "num_tokens": 14834471.0,
      "step": 8550
    },
    {
      "entropy": 5.569959831237793,
      "epoch": 0.6656292550087531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496114957012886,
      "loss": 5.304,
      "mean_token_accuracy": 0.16618645787239075,
      "num_tokens": 14842724.0,
      "step": 8555
    },
    {
      "entropy": 5.517120409011841,
      "epoch": 0.6660182843804707,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961098271598671,
      "loss": 5.218,
      "mean_token_accuracy": 0.18385373651981354,
      "num_tokens": 14850725.0,
      "step": 8560
    },
    {
      "entropy": 5.458403539657593,
      "epoch": 0.6664073137521883,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961046939518585,
      "loss": 5.19,
      "mean_token_accuracy": 0.18690327554941177,
      "num_tokens": 14859412.0,
      "step": 8565
    },
    {
      "entropy": 5.5778275489807125,
      "epoch": 0.6667963431239059,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004960995573889379,
      "loss": 5.2877,
      "mean_token_accuracy": 0.17646803557872773,
      "num_tokens": 14867526.0,
      "step": 8570
    },
    {
      "entropy": 5.488861894607544,
      "epoch": 0.6671853724956234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004960944174711836,
      "loss": 5.1961,
      "mean_token_accuracy": 0.17473305016756058,
      "num_tokens": 14875770.0,
      "step": 8575
    },
    {
      "entropy": 5.506997108459473,
      "epoch": 0.667574401867341,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960892741986734,
      "loss": 5.2818,
      "mean_token_accuracy": 0.18370694667100906,
      "num_tokens": 14883872.0,
      "step": 8580
    },
    {
      "entropy": 5.532104063034057,
      "epoch": 0.6679634312390585,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960841275714853,
      "loss": 5.2978,
      "mean_token_accuracy": 0.17296381145715714,
      "num_tokens": 14891526.0,
      "step": 8585
    },
    {
      "entropy": 5.5480423927307125,
      "epoch": 0.6683524606107761,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960789775896975,
      "loss": 5.3519,
      "mean_token_accuracy": 0.17991099506616592,
      "num_tokens": 14900171.0,
      "step": 8590
    },
    {
      "entropy": 5.594220685958862,
      "epoch": 0.6687414899824937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960738242533881,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1777023196220398,
      "num_tokens": 14908537.0,
      "step": 8595
    },
    {
      "entropy": 5.508582925796508,
      "epoch": 0.6691305193542112,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960686675626353,
      "loss": 5.2544,
      "mean_token_accuracy": 0.17995063066482545,
      "num_tokens": 14917387.0,
      "step": 8600
    },
    {
      "entropy": 5.5501314163208,
      "epoch": 0.6695195487259288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960635075175173,
      "loss": 5.276,
      "mean_token_accuracy": 0.17990948408842086,
      "num_tokens": 14925945.0,
      "step": 8605
    },
    {
      "entropy": 5.539242887496949,
      "epoch": 0.6699085780976464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960583441181124,
      "loss": 5.2811,
      "mean_token_accuracy": 0.17707243263721467,
      "num_tokens": 14934002.0,
      "step": 8610
    },
    {
      "entropy": 5.532008934020996,
      "epoch": 0.670297607469364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960531773644992,
      "loss": 5.316,
      "mean_token_accuracy": 0.16889427751302719,
      "num_tokens": 14942658.0,
      "step": 8615
    },
    {
      "entropy": 5.613341426849365,
      "epoch": 0.6706866368410815,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960480072567557,
      "loss": 5.3829,
      "mean_token_accuracy": 0.17451770156621932,
      "num_tokens": 14952324.0,
      "step": 8620
    },
    {
      "entropy": 5.624604940414429,
      "epoch": 0.671075666212799,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960428337949604,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1732550621032715,
      "num_tokens": 14961964.0,
      "step": 8625
    },
    {
      "entropy": 5.65596866607666,
      "epoch": 0.6714646955845166,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960376569791921,
      "loss": 5.3486,
      "mean_token_accuracy": 0.1749127060174942,
      "num_tokens": 14970267.0,
      "step": 8630
    },
    {
      "entropy": 5.4941013813018795,
      "epoch": 0.6718537249562342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960324768095291,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1824185848236084,
      "num_tokens": 14978546.0,
      "step": 8635
    },
    {
      "entropy": 5.591194486618042,
      "epoch": 0.6722427543279518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960272932860502,
      "loss": 5.2791,
      "mean_token_accuracy": 0.17944162338972092,
      "num_tokens": 14986785.0,
      "step": 8640
    },
    {
      "entropy": 5.565518426895141,
      "epoch": 0.6726317836996694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960221064088338,
      "loss": 5.2121,
      "mean_token_accuracy": 0.17982794344425201,
      "num_tokens": 14994873.0,
      "step": 8645
    },
    {
      "entropy": 5.528623771667481,
      "epoch": 0.6730208130713868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004960169161779588,
      "loss": 5.2442,
      "mean_token_accuracy": 0.18907042294740678,
      "num_tokens": 15003108.0,
      "step": 8650
    },
    {
      "entropy": 5.401618480682373,
      "epoch": 0.6734098424431044,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960117225935038,
      "loss": 5.1494,
      "mean_token_accuracy": 0.18058359175920485,
      "num_tokens": 15011802.0,
      "step": 8655
    },
    {
      "entropy": 5.530126285552979,
      "epoch": 0.673798871814822,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960065256555477,
      "loss": 5.3056,
      "mean_token_accuracy": 0.17976801544427873,
      "num_tokens": 15020785.0,
      "step": 8660
    },
    {
      "entropy": 5.646392774581909,
      "epoch": 0.6741879011865396,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960013253641695,
      "loss": 5.3922,
      "mean_token_accuracy": 0.17093995064496995,
      "num_tokens": 15029218.0,
      "step": 8665
    },
    {
      "entropy": 5.60396032333374,
      "epoch": 0.6745769305582572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959961217194477,
      "loss": 5.3492,
      "mean_token_accuracy": 0.16516607105731965,
      "num_tokens": 15037843.0,
      "step": 8670
    },
    {
      "entropy": 5.530879878997803,
      "epoch": 0.6749659599299748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959909147214615,
      "loss": 5.2725,
      "mean_token_accuracy": 0.18173045814037322,
      "num_tokens": 15047053.0,
      "step": 8675
    },
    {
      "entropy": 5.571678161621094,
      "epoch": 0.6753549893016922,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959857043702902,
      "loss": 5.3923,
      "mean_token_accuracy": 0.17267663776874542,
      "num_tokens": 15056159.0,
      "step": 8680
    },
    {
      "entropy": 5.560159969329834,
      "epoch": 0.6757440186734098,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959804906660124,
      "loss": 5.3358,
      "mean_token_accuracy": 0.16832655519247056,
      "num_tokens": 15064853.0,
      "step": 8685
    },
    {
      "entropy": 5.582832288742066,
      "epoch": 0.6761330480451274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959752736087073,
      "loss": 5.3437,
      "mean_token_accuracy": 0.17521180510520934,
      "num_tokens": 15072671.0,
      "step": 8690
    },
    {
      "entropy": 5.525656604766846,
      "epoch": 0.676522077416845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959700531984543,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18609519004821778,
      "num_tokens": 15080787.0,
      "step": 8695
    },
    {
      "entropy": 5.537931632995606,
      "epoch": 0.6769111067885626,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959648294353324,
      "loss": 5.337,
      "mean_token_accuracy": 0.17343332022428512,
      "num_tokens": 15089906.0,
      "step": 8700
    },
    {
      "entropy": 5.553527593612671,
      "epoch": 0.6773001361602801,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959596023194208,
      "loss": 5.3828,
      "mean_token_accuracy": 0.17819689959287643,
      "num_tokens": 15099329.0,
      "step": 8705
    },
    {
      "entropy": 5.594941329956055,
      "epoch": 0.6776891655319977,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495954371850799,
      "loss": 5.3673,
      "mean_token_accuracy": 0.17315385341644288,
      "num_tokens": 15107744.0,
      "step": 8710
    },
    {
      "entropy": 5.612883567810059,
      "epoch": 0.6780781949037152,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959491380295463,
      "loss": 5.383,
      "mean_token_accuracy": 0.1757282093167305,
      "num_tokens": 15115918.0,
      "step": 8715
    },
    {
      "entropy": 5.57036247253418,
      "epoch": 0.6784672242754328,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959439008557422,
      "loss": 5.2432,
      "mean_token_accuracy": 0.18154032826423644,
      "num_tokens": 15124074.0,
      "step": 8720
    },
    {
      "entropy": 5.531103658676147,
      "epoch": 0.6788562536471504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495938660329466,
      "loss": 5.2554,
      "mean_token_accuracy": 0.18637112081050872,
      "num_tokens": 15132402.0,
      "step": 8725
    },
    {
      "entropy": 5.5278421401977536,
      "epoch": 0.6792452830188679,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959334164507973,
      "loss": 5.381,
      "mean_token_accuracy": 0.17353248447179795,
      "num_tokens": 15141640.0,
      "step": 8730
    },
    {
      "entropy": 5.59519329071045,
      "epoch": 0.6796343123905855,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959281692198155,
      "loss": 5.261,
      "mean_token_accuracy": 0.18473790287971498,
      "num_tokens": 15149919.0,
      "step": 8735
    },
    {
      "entropy": 5.491541719436645,
      "epoch": 0.6800233417623031,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959229186366007,
      "loss": 5.1076,
      "mean_token_accuracy": 0.19363221228122712,
      "num_tokens": 15157536.0,
      "step": 8740
    },
    {
      "entropy": 5.456545400619507,
      "epoch": 0.6804123711340206,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495917664701232,
      "loss": 5.3419,
      "mean_token_accuracy": 0.17189399525523186,
      "num_tokens": 15165848.0,
      "step": 8745
    },
    {
      "entropy": 5.569458913803101,
      "epoch": 0.6808014005057382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959124074137894,
      "loss": 5.324,
      "mean_token_accuracy": 0.17281854897737503,
      "num_tokens": 15174799.0,
      "step": 8750
    },
    {
      "entropy": 5.636410856246949,
      "epoch": 0.6811904298774557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959071467743526,
      "loss": 5.4293,
      "mean_token_accuracy": 0.17031558603048325,
      "num_tokens": 15183695.0,
      "step": 8755
    },
    {
      "entropy": 5.664930534362793,
      "epoch": 0.6815794592491733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959018827830016,
      "loss": 5.4388,
      "mean_token_accuracy": 0.17168623954057693,
      "num_tokens": 15193620.0,
      "step": 8760
    },
    {
      "entropy": 5.611712694168091,
      "epoch": 0.6819684886208909,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495896615439816,
      "loss": 5.2593,
      "mean_token_accuracy": 0.1811678811907768,
      "num_tokens": 15201535.0,
      "step": 8765
    },
    {
      "entropy": 5.654981660842895,
      "epoch": 0.6823575179926085,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958913447448759,
      "loss": 5.406,
      "mean_token_accuracy": 0.1682290881872177,
      "num_tokens": 15209678.0,
      "step": 8770
    },
    {
      "entropy": 5.486663866043091,
      "epoch": 0.682746547364326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958860706982613,
      "loss": 5.2933,
      "mean_token_accuracy": 0.17819326519966125,
      "num_tokens": 15218579.0,
      "step": 8775
    },
    {
      "entropy": 5.437908411026001,
      "epoch": 0.6831355767360435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495880793300052,
      "loss": 5.2003,
      "mean_token_accuracy": 0.1872469514608383,
      "num_tokens": 15227225.0,
      "step": 8780
    },
    {
      "entropy": 5.575411701202393,
      "epoch": 0.6835246061077611,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958755125503284,
      "loss": 5.3169,
      "mean_token_accuracy": 0.17595237642526626,
      "num_tokens": 15235900.0,
      "step": 8785
    },
    {
      "entropy": 5.615043497085571,
      "epoch": 0.6839136354794787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958702284491703,
      "loss": 5.2745,
      "mean_token_accuracy": 0.1787177413702011,
      "num_tokens": 15244538.0,
      "step": 8790
    },
    {
      "entropy": 5.476218605041504,
      "epoch": 0.6843026648511963,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004958649409966581,
      "loss": 5.2536,
      "mean_token_accuracy": 0.18225927352905275,
      "num_tokens": 15252902.0,
      "step": 8795
    },
    {
      "entropy": 5.539601230621338,
      "epoch": 0.6846916942229139,
      "grad_norm": 1.015625,
      "learning_rate": 0.000495859650192872,
      "loss": 5.4068,
      "mean_token_accuracy": 0.16932478845119475,
      "num_tokens": 15261514.0,
      "step": 8800
    },
    {
      "entropy": 5.595240354537964,
      "epoch": 0.6850807235946313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004958543560378922,
      "loss": 5.3335,
      "mean_token_accuracy": 0.17627801597118378,
      "num_tokens": 15270140.0,
      "step": 8805
    },
    {
      "entropy": 5.552701330184936,
      "epoch": 0.6854697529663489,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004958490585317991,
      "loss": 5.2879,
      "mean_token_accuracy": 0.17693503946065903,
      "num_tokens": 15279551.0,
      "step": 8810
    },
    {
      "entropy": 5.494368410110473,
      "epoch": 0.6858587823380665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495843757674673,
      "loss": 5.2693,
      "mean_token_accuracy": 0.1851952090859413,
      "num_tokens": 15288767.0,
      "step": 8815
    },
    {
      "entropy": 5.5630992412567135,
      "epoch": 0.6862478117097841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958384534665945,
      "loss": 5.2907,
      "mean_token_accuracy": 0.1815930113196373,
      "num_tokens": 15297713.0,
      "step": 8820
    },
    {
      "entropy": 5.571929407119751,
      "epoch": 0.6866368410815017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004958331459076438,
      "loss": 5.2981,
      "mean_token_accuracy": 0.17549618035554887,
      "num_tokens": 15306139.0,
      "step": 8825
    },
    {
      "entropy": 5.621170234680176,
      "epoch": 0.6870258704532192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958278349979018,
      "loss": 5.3584,
      "mean_token_accuracy": 0.17428736686706542,
      "num_tokens": 15315015.0,
      "step": 8830
    },
    {
      "entropy": 5.564057683944702,
      "epoch": 0.6874148998249368,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958225207374488,
      "loss": 5.2324,
      "mean_token_accuracy": 0.1814328208565712,
      "num_tokens": 15323366.0,
      "step": 8835
    },
    {
      "entropy": 5.493619823455811,
      "epoch": 0.6878039291966543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958172031263655,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1869863450527191,
      "num_tokens": 15331135.0,
      "step": 8840
    },
    {
      "entropy": 5.50973072052002,
      "epoch": 0.6881929585683719,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958118821647326,
      "loss": 5.2526,
      "mean_token_accuracy": 0.18131258487701415,
      "num_tokens": 15339561.0,
      "step": 8845
    },
    {
      "entropy": 5.563919305801392,
      "epoch": 0.6885819879400895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004958065578526308,
      "loss": 5.146,
      "mean_token_accuracy": 0.18987538367509843,
      "num_tokens": 15348000.0,
      "step": 8850
    },
    {
      "entropy": 5.597847843170166,
      "epoch": 0.688971017311807,
      "grad_norm": 1.15625,
      "learning_rate": 0.000495801230190141,
      "loss": 5.3535,
      "mean_token_accuracy": 0.1803091585636139,
      "num_tokens": 15355916.0,
      "step": 8855
    },
    {
      "entropy": 5.417778778076172,
      "epoch": 0.6893600466835246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957958991773441,
      "loss": 5.1951,
      "mean_token_accuracy": 0.17953578978776932,
      "num_tokens": 15364561.0,
      "step": 8860
    },
    {
      "entropy": 5.5490752220153805,
      "epoch": 0.6897490760552422,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004957905648143206,
      "loss": 5.2359,
      "mean_token_accuracy": 0.17897392362356185,
      "num_tokens": 15372582.0,
      "step": 8865
    },
    {
      "entropy": 5.545386743545532,
      "epoch": 0.6901381054269597,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957852271011519,
      "loss": 5.3297,
      "mean_token_accuracy": 0.1711472824215889,
      "num_tokens": 15381684.0,
      "step": 8870
    },
    {
      "entropy": 5.6327790260314945,
      "epoch": 0.6905271347986773,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957798860379187,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1704866975545883,
      "num_tokens": 15390858.0,
      "step": 8875
    },
    {
      "entropy": 5.549315643310547,
      "epoch": 0.6909161641703949,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957745416247022,
      "loss": 5.175,
      "mean_token_accuracy": 0.18461504131555556,
      "num_tokens": 15399439.0,
      "step": 8880
    },
    {
      "entropy": 5.481848478317261,
      "epoch": 0.6913051935421124,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957691938615833,
      "loss": 5.2553,
      "mean_token_accuracy": 0.18324231952428818,
      "num_tokens": 15407784.0,
      "step": 8885
    },
    {
      "entropy": 5.619164657592774,
      "epoch": 0.69169422291383,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957638427486434,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1737549901008606,
      "num_tokens": 15415985.0,
      "step": 8890
    },
    {
      "entropy": 5.524836540222168,
      "epoch": 0.6920832522855476,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957584882859636,
      "loss": 5.2655,
      "mean_token_accuracy": 0.18107877522706986,
      "num_tokens": 15424433.0,
      "step": 8895
    },
    {
      "entropy": 5.529126119613648,
      "epoch": 0.6924722816572652,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957531304736251,
      "loss": 5.3518,
      "mean_token_accuracy": 0.17498830556869507,
      "num_tokens": 15433041.0,
      "step": 8900
    },
    {
      "entropy": 5.493073415756226,
      "epoch": 0.6928613110289827,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957477693117091,
      "loss": 5.1882,
      "mean_token_accuracy": 0.17712959349155427,
      "num_tokens": 15441159.0,
      "step": 8905
    },
    {
      "entropy": 5.6728880405426025,
      "epoch": 0.6932503404007002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957424048002971,
      "loss": 5.503,
      "mean_token_accuracy": 0.17157228887081147,
      "num_tokens": 15450687.0,
      "step": 8910
    },
    {
      "entropy": 5.490257120132446,
      "epoch": 0.6936393697724178,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957370369394706,
      "loss": 5.2117,
      "mean_token_accuracy": 0.1827506020665169,
      "num_tokens": 15460450.0,
      "step": 8915
    },
    {
      "entropy": 5.500311231613159,
      "epoch": 0.6940283991441354,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004957316657293107,
      "loss": 5.2392,
      "mean_token_accuracy": 0.18319796472787858,
      "num_tokens": 15468435.0,
      "step": 8920
    },
    {
      "entropy": 5.488282489776611,
      "epoch": 0.694417428515853,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004957262911698992,
      "loss": 5.2068,
      "mean_token_accuracy": 0.18449322432279586,
      "num_tokens": 15476232.0,
      "step": 8925
    },
    {
      "entropy": 5.606828832626343,
      "epoch": 0.6948064578875706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004957209132613175,
      "loss": 5.3058,
      "mean_token_accuracy": 0.181530599296093,
      "num_tokens": 15484569.0,
      "step": 8930
    },
    {
      "entropy": 5.541490125656128,
      "epoch": 0.695195487259288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957155320036473,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1850089207291603,
      "num_tokens": 15492583.0,
      "step": 8935
    },
    {
      "entropy": 5.5534101009368895,
      "epoch": 0.6955845166310056,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957101473969702,
      "loss": 5.3658,
      "mean_token_accuracy": 0.17595504075288773,
      "num_tokens": 15501722.0,
      "step": 8940
    },
    {
      "entropy": 5.581349658966064,
      "epoch": 0.6959735460027232,
      "grad_norm": 1.109375,
      "learning_rate": 0.000495704759441368,
      "loss": 5.306,
      "mean_token_accuracy": 0.17931180596351623,
      "num_tokens": 15509682.0,
      "step": 8945
    },
    {
      "entropy": 5.620917654037475,
      "epoch": 0.6963625753744408,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004956993681369221,
      "loss": 5.2986,
      "mean_token_accuracy": 0.18323905318975447,
      "num_tokens": 15518971.0,
      "step": 8950
    },
    {
      "entropy": 5.558902025222778,
      "epoch": 0.6967516047461584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956939734837148,
      "loss": 5.2752,
      "mean_token_accuracy": 0.17688121646642685,
      "num_tokens": 15527385.0,
      "step": 8955
    },
    {
      "entropy": 5.582807779312134,
      "epoch": 0.6971406341178759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956885754818277,
      "loss": 5.3996,
      "mean_token_accuracy": 0.17462659776210784,
      "num_tokens": 15535325.0,
      "step": 8960
    },
    {
      "entropy": 5.604357385635376,
      "epoch": 0.6975296634895934,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956831741313427,
      "loss": 5.369,
      "mean_token_accuracy": 0.17730585783720015,
      "num_tokens": 15544098.0,
      "step": 8965
    },
    {
      "entropy": 5.5868736743927006,
      "epoch": 0.697918692861311,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956777694323418,
      "loss": 5.2427,
      "mean_token_accuracy": 0.17640165984630585,
      "num_tokens": 15552547.0,
      "step": 8970
    },
    {
      "entropy": 5.513116931915283,
      "epoch": 0.6983077222330286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495672361384907,
      "loss": 5.2732,
      "mean_token_accuracy": 0.17443608194589616,
      "num_tokens": 15561044.0,
      "step": 8975
    },
    {
      "entropy": 5.55181188583374,
      "epoch": 0.6986967516047462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956669499891202,
      "loss": 5.3358,
      "mean_token_accuracy": 0.16730567812919617,
      "num_tokens": 15569379.0,
      "step": 8980
    },
    {
      "entropy": 5.604987478256225,
      "epoch": 0.6990857809764637,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956615352450639,
      "loss": 5.3403,
      "mean_token_accuracy": 0.1733030527830124,
      "num_tokens": 15578535.0,
      "step": 8985
    },
    {
      "entropy": 5.511005640029907,
      "epoch": 0.6994748103481813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956561171528198,
      "loss": 5.1889,
      "mean_token_accuracy": 0.18649003952741622,
      "num_tokens": 15586972.0,
      "step": 8990
    },
    {
      "entropy": 5.460616207122802,
      "epoch": 0.6998638397198989,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956506957124704,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1901437148451805,
      "num_tokens": 15595092.0,
      "step": 8995
    },
    {
      "entropy": 5.612773513793945,
      "epoch": 0.7002528690916164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495645270924098,
      "loss": 5.4611,
      "mean_token_accuracy": 0.17564321607351302,
      "num_tokens": 15604348.0,
      "step": 9000
    },
    {
      "epoch": 0.7002528690916164,
      "eval_entropy": 5.389798035222366,
      "eval_loss": 5.328912734985352,
      "eval_mean_token_accuracy": 0.18488086865384837,
      "eval_num_tokens": 15604348.0,
      "eval_runtime": 21.6858,
      "eval_samples_per_second": 1916.373,
      "eval_steps_per_second": 239.558,
      "step": 9000
    },
    {
      "entropy": 5.556111764907837,
      "epoch": 0.700641898463334,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956398427877847,
      "loss": 5.3005,
      "mean_token_accuracy": 0.17604861706495284,
      "num_tokens": 15612597.0,
      "step": 9005
    },
    {
      "entropy": 5.540850639343262,
      "epoch": 0.7010309278350515,
      "grad_norm": 1.0,
      "learning_rate": 0.000495634411303613,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1795080214738846,
      "num_tokens": 15621555.0,
      "step": 9010
    },
    {
      "entropy": 5.554760408401489,
      "epoch": 0.7014199572067691,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956289764716654,
      "loss": 5.2341,
      "mean_token_accuracy": 0.17902394235134125,
      "num_tokens": 15630518.0,
      "step": 9015
    },
    {
      "entropy": 5.5721453666687015,
      "epoch": 0.7018089865784867,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956235382920241,
      "loss": 5.41,
      "mean_token_accuracy": 0.16584509015083312,
      "num_tokens": 15639126.0,
      "step": 9020
    },
    {
      "entropy": 5.493175649642945,
      "epoch": 0.7021980159502043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004956180967647717,
      "loss": 5.2275,
      "mean_token_accuracy": 0.1841141924262047,
      "num_tokens": 15647436.0,
      "step": 9025
    },
    {
      "entropy": 5.530618476867676,
      "epoch": 0.7025870453219218,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956126518899911,
      "loss": 5.2626,
      "mean_token_accuracy": 0.18042101860046386,
      "num_tokens": 15656439.0,
      "step": 9030
    },
    {
      "entropy": 5.552502059936524,
      "epoch": 0.7029760746936393,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004956072036677644,
      "loss": 5.2988,
      "mean_token_accuracy": 0.17644090354442596,
      "num_tokens": 15665119.0,
      "step": 9035
    },
    {
      "entropy": 5.566791677474976,
      "epoch": 0.7033651040653569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956017520981746,
      "loss": 5.3522,
      "mean_token_accuracy": 0.17432471066713334,
      "num_tokens": 15673430.0,
      "step": 9040
    },
    {
      "entropy": 5.643865871429443,
      "epoch": 0.7037541334370745,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955962971813044,
      "loss": 5.4061,
      "mean_token_accuracy": 0.16947294473648072,
      "num_tokens": 15682739.0,
      "step": 9045
    },
    {
      "entropy": 5.558068943023682,
      "epoch": 0.7041431628087921,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004955908389172364,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1732873484492302,
      "num_tokens": 15691206.0,
      "step": 9050
    },
    {
      "entropy": 5.621509456634522,
      "epoch": 0.7045321921805097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955853773060536,
      "loss": 5.3898,
      "mean_token_accuracy": 0.17490919977426528,
      "num_tokens": 15700238.0,
      "step": 9055
    },
    {
      "entropy": 5.600115966796875,
      "epoch": 0.7049212215522271,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955799123478388,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1805904373526573,
      "num_tokens": 15708654.0,
      "step": 9060
    },
    {
      "entropy": 5.606193733215332,
      "epoch": 0.7053102509239447,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955744440426748,
      "loss": 5.4156,
      "mean_token_accuracy": 0.17848946005105973,
      "num_tokens": 15717204.0,
      "step": 9065
    },
    {
      "entropy": 5.4877345085144045,
      "epoch": 0.7056992802956623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955689723906448,
      "loss": 5.1556,
      "mean_token_accuracy": 0.1923369437456131,
      "num_tokens": 15725016.0,
      "step": 9070
    },
    {
      "entropy": 5.649702692031861,
      "epoch": 0.7060883096673799,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955634973918318,
      "loss": 5.5062,
      "mean_token_accuracy": 0.17271387353539466,
      "num_tokens": 15734714.0,
      "step": 9075
    },
    {
      "entropy": 5.596701908111572,
      "epoch": 0.7064773390390975,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955580190463186,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17548872381448746,
      "num_tokens": 15742867.0,
      "step": 9080
    },
    {
      "entropy": 5.573616409301758,
      "epoch": 0.7068663684108151,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004955525373541886,
      "loss": 5.3359,
      "mean_token_accuracy": 0.17494852393865584,
      "num_tokens": 15752059.0,
      "step": 9085
    },
    {
      "entropy": 5.5120330333709715,
      "epoch": 0.7072553977825325,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955470523155249,
      "loss": 5.3605,
      "mean_token_accuracy": 0.17450406402349472,
      "num_tokens": 15761783.0,
      "step": 9090
    },
    {
      "entropy": 5.525177240371704,
      "epoch": 0.7076444271542501,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955415639304107,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1823938637971878,
      "num_tokens": 15770451.0,
      "step": 9095
    },
    {
      "entropy": 5.549926280975342,
      "epoch": 0.7080334565259677,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955360721989292,
      "loss": 5.3088,
      "mean_token_accuracy": 0.18095069825649263,
      "num_tokens": 15779529.0,
      "step": 9100
    },
    {
      "entropy": 5.561652946472168,
      "epoch": 0.7084224858976853,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955305771211641,
      "loss": 5.2761,
      "mean_token_accuracy": 0.1786726161837578,
      "num_tokens": 15788427.0,
      "step": 9105
    },
    {
      "entropy": 5.561768960952759,
      "epoch": 0.7088115152694029,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955250786971982,
      "loss": 5.3107,
      "mean_token_accuracy": 0.17775446325540542,
      "num_tokens": 15796530.0,
      "step": 9110
    },
    {
      "entropy": 5.49405255317688,
      "epoch": 0.7092005446411204,
      "grad_norm": 1.125,
      "learning_rate": 0.0004955195769271154,
      "loss": 5.2599,
      "mean_token_accuracy": 0.1851181998848915,
      "num_tokens": 15805058.0,
      "step": 9115
    },
    {
      "entropy": 5.65776515007019,
      "epoch": 0.709589574012838,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955140718109991,
      "loss": 5.3609,
      "mean_token_accuracy": 0.17976813912391662,
      "num_tokens": 15813729.0,
      "step": 9120
    },
    {
      "entropy": 5.577265214920044,
      "epoch": 0.7099786033845555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955085633489326,
      "loss": 5.2097,
      "mean_token_accuracy": 0.18789018839597701,
      "num_tokens": 15822305.0,
      "step": 9125
    },
    {
      "entropy": 5.553754806518555,
      "epoch": 0.7103676327562731,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955030515409997,
      "loss": 5.3513,
      "mean_token_accuracy": 0.17451984584331512,
      "num_tokens": 15830842.0,
      "step": 9130
    },
    {
      "entropy": 5.469909429550171,
      "epoch": 0.7107566621279907,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954975363872838,
      "loss": 5.1965,
      "mean_token_accuracy": 0.17950347810983658,
      "num_tokens": 15839790.0,
      "step": 9135
    },
    {
      "entropy": 5.5804030418396,
      "epoch": 0.7111456914997082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954920178878689,
      "loss": 5.3998,
      "mean_token_accuracy": 0.16854211539030076,
      "num_tokens": 15849307.0,
      "step": 9140
    },
    {
      "entropy": 5.647359418869018,
      "epoch": 0.7115347208714258,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004954864960428385,
      "loss": 5.3326,
      "mean_token_accuracy": 0.1800499364733696,
      "num_tokens": 15858688.0,
      "step": 9145
    },
    {
      "entropy": 5.576810455322265,
      "epoch": 0.7119237502431434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954809708522765,
      "loss": 5.2099,
      "mean_token_accuracy": 0.17958054095506668,
      "num_tokens": 15867574.0,
      "step": 9150
    },
    {
      "entropy": 5.537421703338623,
      "epoch": 0.712312779614861,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954754423162667,
      "loss": 5.235,
      "mean_token_accuracy": 0.18324693739414216,
      "num_tokens": 15875871.0,
      "step": 9155
    },
    {
      "entropy": 5.529123449325562,
      "epoch": 0.7127018089865785,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495469910434893,
      "loss": 5.303,
      "mean_token_accuracy": 0.17866899222135543,
      "num_tokens": 15884386.0,
      "step": 9160
    },
    {
      "entropy": 5.587394571304321,
      "epoch": 0.713090838358296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954643752082392,
      "loss": 5.2933,
      "mean_token_accuracy": 0.17422557026147842,
      "num_tokens": 15893086.0,
      "step": 9165
    },
    {
      "entropy": 5.517653369903565,
      "epoch": 0.7134798677300136,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954588366363896,
      "loss": 5.3217,
      "mean_token_accuracy": 0.17549112886190416,
      "num_tokens": 15902266.0,
      "step": 9170
    },
    {
      "entropy": 5.702410888671875,
      "epoch": 0.7138688971017312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954532947194279,
      "loss": 5.448,
      "mean_token_accuracy": 0.16685094088315963,
      "num_tokens": 15911133.0,
      "step": 9175
    },
    {
      "entropy": 5.529084730148315,
      "epoch": 0.7142579264734488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954477494574384,
      "loss": 5.2194,
      "mean_token_accuracy": 0.18191736340522766,
      "num_tokens": 15919455.0,
      "step": 9180
    },
    {
      "entropy": 5.49851655960083,
      "epoch": 0.7146469558451664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954422008505052,
      "loss": 5.2822,
      "mean_token_accuracy": 0.18629485964775086,
      "num_tokens": 15927592.0,
      "step": 9185
    },
    {
      "entropy": 5.491498279571533,
      "epoch": 0.7150359852168838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954366488987125,
      "loss": 5.221,
      "mean_token_accuracy": 0.1816892832517624,
      "num_tokens": 15935979.0,
      "step": 9190
    },
    {
      "entropy": 5.5549968719482425,
      "epoch": 0.7154250145886014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954310936021444,
      "loss": 5.3768,
      "mean_token_accuracy": 0.17618247270584106,
      "num_tokens": 15945921.0,
      "step": 9195
    },
    {
      "entropy": 5.626093864440918,
      "epoch": 0.715814043960319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004954255349608853,
      "loss": 5.3012,
      "mean_token_accuracy": 0.18052035868167876,
      "num_tokens": 15954875.0,
      "step": 9200
    },
    {
      "entropy": 5.56747088432312,
      "epoch": 0.7162030733320366,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954199729750198,
      "loss": 5.3087,
      "mean_token_accuracy": 0.17858943343162537,
      "num_tokens": 15963391.0,
      "step": 9205
    },
    {
      "entropy": 5.630508947372436,
      "epoch": 0.7165921027037542,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004954144076446318,
      "loss": 5.427,
      "mean_token_accuracy": 0.16947771459817887,
      "num_tokens": 15972366.0,
      "step": 9210
    },
    {
      "entropy": 5.638298368453979,
      "epoch": 0.7169811320754716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004954088389698061,
      "loss": 5.3947,
      "mean_token_accuracy": 0.177106873691082,
      "num_tokens": 15981196.0,
      "step": 9215
    },
    {
      "entropy": 5.48401198387146,
      "epoch": 0.7173701614471892,
      "grad_norm": 0.96875,
      "learning_rate": 0.000495403266950627,
      "loss": 5.3291,
      "mean_token_accuracy": 0.17288484573364257,
      "num_tokens": 15990811.0,
      "step": 9220
    },
    {
      "entropy": 5.557605028152466,
      "epoch": 0.7177591908189068,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953976915871792,
      "loss": 5.317,
      "mean_token_accuracy": 0.17596275806427003,
      "num_tokens": 15999594.0,
      "step": 9225
    },
    {
      "entropy": 5.687898397445679,
      "epoch": 0.7181482201906244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953921128795472,
      "loss": 5.4282,
      "mean_token_accuracy": 0.16423819810152054,
      "num_tokens": 16008179.0,
      "step": 9230
    },
    {
      "entropy": 5.582005643844605,
      "epoch": 0.718537249562342,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953865308278156,
      "loss": 5.3464,
      "mean_token_accuracy": 0.17871031463146209,
      "num_tokens": 16016583.0,
      "step": 9235
    },
    {
      "entropy": 5.609802484512329,
      "epoch": 0.7189262789340595,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953809454320693,
      "loss": 5.356,
      "mean_token_accuracy": 0.17320025712251663,
      "num_tokens": 16025516.0,
      "step": 9240
    },
    {
      "entropy": 5.472203969955444,
      "epoch": 0.7193153083057771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953753566923929,
      "loss": 5.2189,
      "mean_token_accuracy": 0.1882613107562065,
      "num_tokens": 16033536.0,
      "step": 9245
    },
    {
      "entropy": 5.519918298721313,
      "epoch": 0.7197043376774946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953697646088712,
      "loss": 5.1493,
      "mean_token_accuracy": 0.1862490639090538,
      "num_tokens": 16041918.0,
      "step": 9250
    },
    {
      "entropy": 5.49515471458435,
      "epoch": 0.7200933670492122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953641691815891,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1890513315796852,
      "num_tokens": 16050806.0,
      "step": 9255
    },
    {
      "entropy": 5.549209785461426,
      "epoch": 0.7204823964209298,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953585704106315,
      "loss": 5.3574,
      "mean_token_accuracy": 0.1688122883439064,
      "num_tokens": 16060140.0,
      "step": 9260
    },
    {
      "entropy": 5.590003824234008,
      "epoch": 0.7208714257926473,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953529682960832,
      "loss": 5.2188,
      "mean_token_accuracy": 0.18012620657682418,
      "num_tokens": 16068182.0,
      "step": 9265
    },
    {
      "entropy": 5.591183280944824,
      "epoch": 0.7212604551643649,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953473628380295,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1735826015472412,
      "num_tokens": 16076798.0,
      "step": 9270
    },
    {
      "entropy": 5.423460483551025,
      "epoch": 0.7216494845360825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953417540365553,
      "loss": 5.1083,
      "mean_token_accuracy": 0.18438050895929337,
      "num_tokens": 16084739.0,
      "step": 9275
    },
    {
      "entropy": 5.559713363647461,
      "epoch": 0.7220385139078,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953361418917455,
      "loss": 5.2661,
      "mean_token_accuracy": 0.18010660707950593,
      "num_tokens": 16093434.0,
      "step": 9280
    },
    {
      "entropy": 5.543549537658691,
      "epoch": 0.7224275432795176,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004953305264036856,
      "loss": 5.3158,
      "mean_token_accuracy": 0.17837266623973846,
      "num_tokens": 16102163.0,
      "step": 9285
    },
    {
      "entropy": 5.489007902145386,
      "epoch": 0.7228165726512352,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953249075724607,
      "loss": 5.29,
      "mean_token_accuracy": 0.18584892600774766,
      "num_tokens": 16110761.0,
      "step": 9290
    },
    {
      "entropy": 5.545398187637329,
      "epoch": 0.7232056020229527,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953192853981559,
      "loss": 5.325,
      "mean_token_accuracy": 0.1777096062898636,
      "num_tokens": 16119774.0,
      "step": 9295
    },
    {
      "entropy": 5.5385864734649655,
      "epoch": 0.7235946313946703,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004953136598808566,
      "loss": 5.196,
      "mean_token_accuracy": 0.18555736243724824,
      "num_tokens": 16128232.0,
      "step": 9300
    },
    {
      "entropy": 5.5455903053283695,
      "epoch": 0.7239836607663879,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953080310206481,
      "loss": 5.2331,
      "mean_token_accuracy": 0.1827569469809532,
      "num_tokens": 16137370.0,
      "step": 9305
    },
    {
      "entropy": 5.507047891616821,
      "epoch": 0.7243726901381055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004953023988176162,
      "loss": 5.2536,
      "mean_token_accuracy": 0.1793384850025177,
      "num_tokens": 16145791.0,
      "step": 9310
    },
    {
      "entropy": 5.635753631591797,
      "epoch": 0.724761719509823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952967632718458,
      "loss": 5.3666,
      "mean_token_accuracy": 0.17886020988225937,
      "num_tokens": 16153713.0,
      "step": 9315
    },
    {
      "entropy": 5.504679250717163,
      "epoch": 0.7251507488815405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952911243834227,
      "loss": 5.2166,
      "mean_token_accuracy": 0.17856226563453675,
      "num_tokens": 16162362.0,
      "step": 9320
    },
    {
      "entropy": 5.557938098907471,
      "epoch": 0.7255397782532581,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952854821524324,
      "loss": 5.289,
      "mean_token_accuracy": 0.1809445485472679,
      "num_tokens": 16171094.0,
      "step": 9325
    },
    {
      "entropy": 5.59300479888916,
      "epoch": 0.7259288076249757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952798365789606,
      "loss": 5.3487,
      "mean_token_accuracy": 0.17876584231853485,
      "num_tokens": 16180282.0,
      "step": 9330
    },
    {
      "entropy": 5.5803542137146,
      "epoch": 0.7263178369966933,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952741876630928,
      "loss": 5.2435,
      "mean_token_accuracy": 0.1806937649846077,
      "num_tokens": 16188497.0,
      "step": 9335
    },
    {
      "entropy": 5.607930040359497,
      "epoch": 0.7267068663684109,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952685354049148,
      "loss": 5.4144,
      "mean_token_accuracy": 0.16779697239398955,
      "num_tokens": 16197822.0,
      "step": 9340
    },
    {
      "entropy": 5.591944837570191,
      "epoch": 0.7270958957401283,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952628798045124,
      "loss": 5.3089,
      "mean_token_accuracy": 0.17668944597244263,
      "num_tokens": 16207770.0,
      "step": 9345
    },
    {
      "entropy": 5.5045722961425785,
      "epoch": 0.7274849251118459,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952572208619714,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1829061135649681,
      "num_tokens": 16216624.0,
      "step": 9350
    },
    {
      "entropy": 5.483622121810913,
      "epoch": 0.7278739544835635,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952515585773778,
      "loss": 5.1996,
      "mean_token_accuracy": 0.17681387364864348,
      "num_tokens": 16225585.0,
      "step": 9355
    },
    {
      "entropy": 5.52840633392334,
      "epoch": 0.7282629838552811,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952458929508171,
      "loss": 5.1684,
      "mean_token_accuracy": 0.18971401453018188,
      "num_tokens": 16233988.0,
      "step": 9360
    },
    {
      "entropy": 5.581288576126099,
      "epoch": 0.7286520132269987,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952402239823757,
      "loss": 5.3464,
      "mean_token_accuracy": 0.17401268780231477,
      "num_tokens": 16242291.0,
      "step": 9365
    },
    {
      "entropy": 5.472625684738159,
      "epoch": 0.7290410425987162,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004952345516721393,
      "loss": 5.2122,
      "mean_token_accuracy": 0.18057327717542648,
      "num_tokens": 16251102.0,
      "step": 9370
    },
    {
      "entropy": 5.479851818084716,
      "epoch": 0.7294300719704337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952288760201942,
      "loss": 5.211,
      "mean_token_accuracy": 0.18755067735910416,
      "num_tokens": 16259136.0,
      "step": 9375
    },
    {
      "entropy": 5.559380006790161,
      "epoch": 0.7298191013421513,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952231970266265,
      "loss": 5.3283,
      "mean_token_accuracy": 0.17745242565870284,
      "num_tokens": 16267958.0,
      "step": 9380
    },
    {
      "entropy": 5.667397785186767,
      "epoch": 0.7302081307138689,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952175146915223,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1704097494482994,
      "num_tokens": 16277616.0,
      "step": 9385
    },
    {
      "entropy": 5.65295000076294,
      "epoch": 0.7305971600855865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004952118290149679,
      "loss": 5.3033,
      "mean_token_accuracy": 0.18273016065359116,
      "num_tokens": 16286454.0,
      "step": 9390
    },
    {
      "entropy": 5.566510534286499,
      "epoch": 0.730986189457304,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004952061399970492,
      "loss": 5.2741,
      "mean_token_accuracy": 0.17336290329694748,
      "num_tokens": 16295722.0,
      "step": 9395
    },
    {
      "entropy": 5.566493225097656,
      "epoch": 0.7313752188290216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495200447637853,
      "loss": 5.3381,
      "mean_token_accuracy": 0.17226790636777878,
      "num_tokens": 16304180.0,
      "step": 9400
    },
    {
      "entropy": 5.531685256958008,
      "epoch": 0.7317642482007392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951947519374655,
      "loss": 5.225,
      "mean_token_accuracy": 0.18427305519580842,
      "num_tokens": 16312766.0,
      "step": 9405
    },
    {
      "entropy": 5.545846557617187,
      "epoch": 0.7321532775724567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951890528959731,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1751926675438881,
      "num_tokens": 16321174.0,
      "step": 9410
    },
    {
      "entropy": 5.498131275177002,
      "epoch": 0.7325423069441743,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004951833505134623,
      "loss": 5.2399,
      "mean_token_accuracy": 0.17664798051118852,
      "num_tokens": 16329870.0,
      "step": 9415
    },
    {
      "entropy": 5.5950346946716305,
      "epoch": 0.7329313363158918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951776447900196,
      "loss": 5.2883,
      "mean_token_accuracy": 0.17546151727437972,
      "num_tokens": 16338998.0,
      "step": 9420
    },
    {
      "entropy": 5.59651255607605,
      "epoch": 0.7333203656876094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004951719357257317,
      "loss": 5.2742,
      "mean_token_accuracy": 0.17708998173475266,
      "num_tokens": 16347313.0,
      "step": 9425
    },
    {
      "entropy": 5.56943187713623,
      "epoch": 0.733709395059327,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495166223320685,
      "loss": 5.3988,
      "mean_token_accuracy": 0.17598551213741304,
      "num_tokens": 16356565.0,
      "step": 9430
    },
    {
      "entropy": 5.5626061916351315,
      "epoch": 0.7340984244310446,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951605075749662,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1750992491841316,
      "num_tokens": 16365747.0,
      "step": 9435
    },
    {
      "entropy": 5.606553077697754,
      "epoch": 0.7344874538027621,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951547884886623,
      "loss": 5.3413,
      "mean_token_accuracy": 0.1672863632440567,
      "num_tokens": 16374650.0,
      "step": 9440
    },
    {
      "entropy": 5.567455911636353,
      "epoch": 0.7348764831744796,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951490660618598,
      "loss": 5.2837,
      "mean_token_accuracy": 0.17836343348026276,
      "num_tokens": 16384119.0,
      "step": 9445
    },
    {
      "entropy": 5.544294929504394,
      "epoch": 0.7352655125461972,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951433402946457,
      "loss": 5.2417,
      "mean_token_accuracy": 0.18445164561271668,
      "num_tokens": 16392606.0,
      "step": 9450
    },
    {
      "entropy": 5.481019306182861,
      "epoch": 0.7356545419179148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951376111871068,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18546178191900253,
      "num_tokens": 16401006.0,
      "step": 9455
    },
    {
      "entropy": 5.523701953887939,
      "epoch": 0.7360435712896324,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951318787393299,
      "loss": 5.2553,
      "mean_token_accuracy": 0.18083978742361068,
      "num_tokens": 16409401.0,
      "step": 9460
    },
    {
      "entropy": 5.602548837661743,
      "epoch": 0.73643260066135,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951261429514023,
      "loss": 5.3174,
      "mean_token_accuracy": 0.17583308815956117,
      "num_tokens": 16418771.0,
      "step": 9465
    },
    {
      "entropy": 5.5785486698150635,
      "epoch": 0.7368216300330674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951204038234106,
      "loss": 5.3619,
      "mean_token_accuracy": 0.1764096811413765,
      "num_tokens": 16427624.0,
      "step": 9470
    },
    {
      "entropy": 5.541777753829956,
      "epoch": 0.737210659404785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951146613554424,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1854480504989624,
      "num_tokens": 16435669.0,
      "step": 9475
    },
    {
      "entropy": 5.582617092132568,
      "epoch": 0.7375996887765026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951089155475843,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18146839141845703,
      "num_tokens": 16444283.0,
      "step": 9480
    },
    {
      "entropy": 5.4931660175323485,
      "epoch": 0.7379887181482202,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495103166399924,
      "loss": 5.2616,
      "mean_token_accuracy": 0.17441291362047195,
      "num_tokens": 16452741.0,
      "step": 9485
    },
    {
      "entropy": 5.522454881668091,
      "epoch": 0.7383777475199378,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950974139125484,
      "loss": 5.3088,
      "mean_token_accuracy": 0.1850725769996643,
      "num_tokens": 16461386.0,
      "step": 9490
    },
    {
      "entropy": 5.627692127227784,
      "epoch": 0.7387667768916554,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950916580855448,
      "loss": 5.3251,
      "mean_token_accuracy": 0.17463397830724717,
      "num_tokens": 16469901.0,
      "step": 9495
    },
    {
      "entropy": 5.560376739501953,
      "epoch": 0.7391558062633728,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004950858989190007,
      "loss": 5.2499,
      "mean_token_accuracy": 0.1811629593372345,
      "num_tokens": 16478076.0,
      "step": 9500
    },
    {
      "entropy": 5.480900526046753,
      "epoch": 0.7395448356350904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950801364130032,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18097128719091415,
      "num_tokens": 16486568.0,
      "step": 9505
    },
    {
      "entropy": 5.5960197925567625,
      "epoch": 0.739933865006808,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950743705676401,
      "loss": 5.3449,
      "mean_token_accuracy": 0.17725475132465363,
      "num_tokens": 16495299.0,
      "step": 9510
    },
    {
      "entropy": 5.539008760452271,
      "epoch": 0.7403228943785256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950686013829987,
      "loss": 5.2531,
      "mean_token_accuracy": 0.17483892142772675,
      "num_tokens": 16504576.0,
      "step": 9515
    },
    {
      "entropy": 5.574979066848755,
      "epoch": 0.7407119237502432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950628288591665,
      "loss": 5.3242,
      "mean_token_accuracy": 0.17738772332668304,
      "num_tokens": 16512715.0,
      "step": 9520
    },
    {
      "entropy": 5.613685178756714,
      "epoch": 0.7411009531219607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950570529962311,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1810723900794983,
      "num_tokens": 16521375.0,
      "step": 9525
    },
    {
      "entropy": 5.55305233001709,
      "epoch": 0.7414899824936783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950512737942803,
      "loss": 5.2493,
      "mean_token_accuracy": 0.18598097711801528,
      "num_tokens": 16529772.0,
      "step": 9530
    },
    {
      "entropy": 5.611551809310913,
      "epoch": 0.7418790118653958,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950454912534015,
      "loss": 5.2967,
      "mean_token_accuracy": 0.18408222794532775,
      "num_tokens": 16538365.0,
      "step": 9535
    },
    {
      "entropy": 5.6260138034820555,
      "epoch": 0.7422680412371134,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950397053736827,
      "loss": 5.4118,
      "mean_token_accuracy": 0.16625734269618989,
      "num_tokens": 16547668.0,
      "step": 9540
    },
    {
      "entropy": 5.631437683105469,
      "epoch": 0.742657070608831,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950339161552117,
      "loss": 5.3923,
      "mean_token_accuracy": 0.16748718023300171,
      "num_tokens": 16557060.0,
      "step": 9545
    },
    {
      "entropy": 5.593369626998902,
      "epoch": 0.7430460999805485,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950281235980761,
      "loss": 5.3649,
      "mean_token_accuracy": 0.18361901491880417,
      "num_tokens": 16565814.0,
      "step": 9550
    },
    {
      "entropy": 5.5311542510986325,
      "epoch": 0.7434351293522661,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495022327702364,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18616349399089813,
      "num_tokens": 16574536.0,
      "step": 9555
    },
    {
      "entropy": 5.522435569763184,
      "epoch": 0.7438241587239837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950165284681631,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19001736789941787,
      "num_tokens": 16583349.0,
      "step": 9560
    },
    {
      "entropy": 5.485368776321411,
      "epoch": 0.7442131880957013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004950107258955618,
      "loss": 5.1509,
      "mean_token_accuracy": 0.18268264830112457,
      "num_tokens": 16591374.0,
      "step": 9565
    },
    {
      "entropy": 5.503294038772583,
      "epoch": 0.7446022174674188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004950049199846479,
      "loss": 5.2609,
      "mean_token_accuracy": 0.17910871505737305,
      "num_tokens": 16599684.0,
      "step": 9570
    },
    {
      "entropy": 5.556149101257324,
      "epoch": 0.7449912468391363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004949991107355095,
      "loss": 5.2553,
      "mean_token_accuracy": 0.17982286363840103,
      "num_tokens": 16607618.0,
      "step": 9575
    },
    {
      "entropy": 5.535337591171265,
      "epoch": 0.7453802762108539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949932981482347,
      "loss": 5.3675,
      "mean_token_accuracy": 0.17090606391429902,
      "num_tokens": 16616259.0,
      "step": 9580
    },
    {
      "entropy": 5.541644048690796,
      "epoch": 0.7457693055825715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949874822229118,
      "loss": 5.225,
      "mean_token_accuracy": 0.1834418684244156,
      "num_tokens": 16624972.0,
      "step": 9585
    },
    {
      "entropy": 5.61938967704773,
      "epoch": 0.7461583349542891,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949816629596288,
      "loss": 5.3308,
      "mean_token_accuracy": 0.17839625775814055,
      "num_tokens": 16633517.0,
      "step": 9590
    },
    {
      "entropy": 5.598763275146484,
      "epoch": 0.7465473643260067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949758403584746,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1768048346042633,
      "num_tokens": 16642695.0,
      "step": 9595
    },
    {
      "entropy": 5.459473657608032,
      "epoch": 0.7469363936977241,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949700144195368,
      "loss": 5.2168,
      "mean_token_accuracy": 0.1797614425420761,
      "num_tokens": 16651251.0,
      "step": 9600
    },
    {
      "entropy": 5.521040964126587,
      "epoch": 0.7473254230694417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949641851429043,
      "loss": 5.2773,
      "mean_token_accuracy": 0.17668697834014893,
      "num_tokens": 16660570.0,
      "step": 9605
    },
    {
      "entropy": 5.5326752185821535,
      "epoch": 0.7477144524411593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949583525286654,
      "loss": 5.312,
      "mean_token_accuracy": 0.17187793254852296,
      "num_tokens": 16668678.0,
      "step": 9610
    },
    {
      "entropy": 5.43628339767456,
      "epoch": 0.7481034818128769,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949525165769085,
      "loss": 5.1444,
      "mean_token_accuracy": 0.18645231425762177,
      "num_tokens": 16677207.0,
      "step": 9615
    },
    {
      "entropy": 5.582274103164673,
      "epoch": 0.7484925111845945,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004949466772877224,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1755437359213829,
      "num_tokens": 16686015.0,
      "step": 9620
    },
    {
      "entropy": 5.54477105140686,
      "epoch": 0.748881540556312,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949408346611955,
      "loss": 5.1845,
      "mean_token_accuracy": 0.18544931560754777,
      "num_tokens": 16694989.0,
      "step": 9625
    },
    {
      "entropy": 5.584083795547485,
      "epoch": 0.7492705699280295,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004949349886974165,
      "loss": 5.3611,
      "mean_token_accuracy": 0.17635638415813445,
      "num_tokens": 16703888.0,
      "step": 9630
    },
    {
      "entropy": 5.55716118812561,
      "epoch": 0.7496595992997471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949291393964741,
      "loss": 5.2835,
      "mean_token_accuracy": 0.17723703235387803,
      "num_tokens": 16711880.0,
      "step": 9635
    },
    {
      "entropy": 5.583466053009033,
      "epoch": 0.7500486286714647,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494923286758457,
      "loss": 5.2772,
      "mean_token_accuracy": 0.18172440379858018,
      "num_tokens": 16720858.0,
      "step": 9640
    },
    {
      "entropy": 5.550045967102051,
      "epoch": 0.7504376580431823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949174307834541,
      "loss": 5.3488,
      "mean_token_accuracy": 0.17593863606452942,
      "num_tokens": 16729223.0,
      "step": 9645
    },
    {
      "entropy": 5.578950548171997,
      "epoch": 0.7508266874148998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949115714715543,
      "loss": 5.316,
      "mean_token_accuracy": 0.17637102603912352,
      "num_tokens": 16737854.0,
      "step": 9650
    },
    {
      "entropy": 5.5745062828063965,
      "epoch": 0.7512157167866174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949057088228465,
      "loss": 5.3249,
      "mean_token_accuracy": 0.17119787484407425,
      "num_tokens": 16747081.0,
      "step": 9655
    },
    {
      "entropy": 5.498596286773681,
      "epoch": 0.751604746158335,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948998428374194,
      "loss": 5.1653,
      "mean_token_accuracy": 0.18802380561828613,
      "num_tokens": 16756410.0,
      "step": 9660
    },
    {
      "entropy": 5.494810676574707,
      "epoch": 0.7519937755300525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948939735153622,
      "loss": 5.1923,
      "mean_token_accuracy": 0.18529070615768434,
      "num_tokens": 16764774.0,
      "step": 9665
    },
    {
      "entropy": 5.528837108612061,
      "epoch": 0.7523828049017701,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948881008567641,
      "loss": 5.3482,
      "mean_token_accuracy": 0.17671961337327957,
      "num_tokens": 16773484.0,
      "step": 9670
    },
    {
      "entropy": 5.576663970947266,
      "epoch": 0.7527718342734876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948822248617139,
      "loss": 5.1586,
      "mean_token_accuracy": 0.191119584441185,
      "num_tokens": 16781167.0,
      "step": 9675
    },
    {
      "entropy": 5.464426851272583,
      "epoch": 0.7531608636452052,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494876345530301,
      "loss": 5.212,
      "mean_token_accuracy": 0.18175776600837706,
      "num_tokens": 16790838.0,
      "step": 9680
    },
    {
      "entropy": 5.524459981918335,
      "epoch": 0.7535498930169228,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948704628626145,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1849897548556328,
      "num_tokens": 16799179.0,
      "step": 9685
    },
    {
      "entropy": 5.606106233596802,
      "epoch": 0.7539389223886404,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948645768587437,
      "loss": 5.1909,
      "mean_token_accuracy": 0.17714202255010605,
      "num_tokens": 16807516.0,
      "step": 9690
    },
    {
      "entropy": 5.601245403289795,
      "epoch": 0.7543279517603579,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004948586875187778,
      "loss": 5.4131,
      "mean_token_accuracy": 0.17605143338441848,
      "num_tokens": 16816601.0,
      "step": 9695
    },
    {
      "entropy": 5.533262538909912,
      "epoch": 0.7547169811320755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948527948428064,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17988608181476592,
      "num_tokens": 16825498.0,
      "step": 9700
    },
    {
      "entropy": 5.593303298950195,
      "epoch": 0.755106010503793,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948468988309187,
      "loss": 5.3737,
      "mean_token_accuracy": 0.17876372784376143,
      "num_tokens": 16834833.0,
      "step": 9705
    },
    {
      "entropy": 5.560336494445801,
      "epoch": 0.7554950398755106,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948409994832043,
      "loss": 5.2511,
      "mean_token_accuracy": 0.18087394088506697,
      "num_tokens": 16843476.0,
      "step": 9710
    },
    {
      "entropy": 5.517852830886841,
      "epoch": 0.7558840692472282,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948350967997526,
      "loss": 5.1705,
      "mean_token_accuracy": 0.18963382989168168,
      "num_tokens": 16851859.0,
      "step": 9715
    },
    {
      "entropy": 5.580209112167358,
      "epoch": 0.7562730986189458,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004948291907806532,
      "loss": 5.2378,
      "mean_token_accuracy": 0.18185634762048722,
      "num_tokens": 16860642.0,
      "step": 9720
    },
    {
      "entropy": 5.591513061523438,
      "epoch": 0.7566621279906633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948232814259957,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1745381847023964,
      "num_tokens": 16869232.0,
      "step": 9725
    },
    {
      "entropy": 5.492669534683228,
      "epoch": 0.7570511573623808,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948173687358699,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1836826354265213,
      "num_tokens": 16877880.0,
      "step": 9730
    },
    {
      "entropy": 5.456797504425049,
      "epoch": 0.7574401867340984,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948114527103652,
      "loss": 5.2193,
      "mean_token_accuracy": 0.18095895498991013,
      "num_tokens": 16887060.0,
      "step": 9735
    },
    {
      "entropy": 5.574520063400269,
      "epoch": 0.757829216105816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948055333495717,
      "loss": 5.2626,
      "mean_token_accuracy": 0.17601558417081833,
      "num_tokens": 16895980.0,
      "step": 9740
    },
    {
      "entropy": 5.662924909591675,
      "epoch": 0.7582182454775336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947996106535791,
      "loss": 5.4047,
      "mean_token_accuracy": 0.17123472839593887,
      "num_tokens": 16904334.0,
      "step": 9745
    },
    {
      "entropy": 5.501442003250122,
      "epoch": 0.7586072748492512,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947936846224773,
      "loss": 5.2864,
      "mean_token_accuracy": 0.18169119507074355,
      "num_tokens": 16913436.0,
      "step": 9750
    },
    {
      "entropy": 5.503163862228393,
      "epoch": 0.7589963042209686,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947877552563562,
      "loss": 5.2727,
      "mean_token_accuracy": 0.17634390145540238,
      "num_tokens": 16923214.0,
      "step": 9755
    },
    {
      "entropy": 5.579230499267578,
      "epoch": 0.7593853335926862,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947818225553056,
      "loss": 5.2044,
      "mean_token_accuracy": 0.18720707297325134,
      "num_tokens": 16932523.0,
      "step": 9760
    },
    {
      "entropy": 5.601711845397949,
      "epoch": 0.7597743629644038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947758865194156,
      "loss": 5.3313,
      "mean_token_accuracy": 0.17638456970453262,
      "num_tokens": 16940649.0,
      "step": 9765
    },
    {
      "entropy": 5.550697326660156,
      "epoch": 0.7601633923361214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947699471487766,
      "loss": 5.2834,
      "mean_token_accuracy": 0.18467652052640915,
      "num_tokens": 16949373.0,
      "step": 9770
    },
    {
      "entropy": 5.604041671752929,
      "epoch": 0.760552421707839,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947640044434782,
      "loss": 5.2943,
      "mean_token_accuracy": 0.18377356231212616,
      "num_tokens": 16957787.0,
      "step": 9775
    },
    {
      "entropy": 5.525785446166992,
      "epoch": 0.7609414510795565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947580584036109,
      "loss": 5.2154,
      "mean_token_accuracy": 0.18299616128206253,
      "num_tokens": 16967589.0,
      "step": 9780
    },
    {
      "entropy": 5.551060009002685,
      "epoch": 0.761330480451274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004947521090292649,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18813629150390626,
      "num_tokens": 16976719.0,
      "step": 9785
    },
    {
      "entropy": 5.5405426025390625,
      "epoch": 0.7617195098229916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004947461563205304,
      "loss": 5.3578,
      "mean_token_accuracy": 0.18054259270429612,
      "num_tokens": 16984805.0,
      "step": 9790
    },
    {
      "entropy": 5.530938386917114,
      "epoch": 0.7621085391947092,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947402002774977,
      "loss": 5.2936,
      "mean_token_accuracy": 0.175130957365036,
      "num_tokens": 16993298.0,
      "step": 9795
    },
    {
      "entropy": 5.506849193572998,
      "epoch": 0.7624975685664268,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947342409002572,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19266698211431504,
      "num_tokens": 17002170.0,
      "step": 9800
    },
    {
      "entropy": 5.521035671234131,
      "epoch": 0.7628865979381443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947282781888994,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19202028065919877,
      "num_tokens": 17011218.0,
      "step": 9805
    },
    {
      "entropy": 5.520613718032837,
      "epoch": 0.7632756273098619,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004947223121435147,
      "loss": 5.1636,
      "mean_token_accuracy": 0.1823309525847435,
      "num_tokens": 17019628.0,
      "step": 9810
    },
    {
      "entropy": 5.545638942718506,
      "epoch": 0.7636646566815795,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004947163427641936,
      "loss": 5.3798,
      "mean_token_accuracy": 0.17455653548240663,
      "num_tokens": 17028342.0,
      "step": 9815
    },
    {
      "entropy": 5.6043336391448975,
      "epoch": 0.764053686053297,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947103700510268,
      "loss": 5.3258,
      "mean_token_accuracy": 0.176914082467556,
      "num_tokens": 17037844.0,
      "step": 9820
    },
    {
      "entropy": 5.538714170455933,
      "epoch": 0.7644427154250146,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947043940041047,
      "loss": 5.1678,
      "mean_token_accuracy": 0.18721466660499572,
      "num_tokens": 17046430.0,
      "step": 9825
    },
    {
      "entropy": 5.522770977020263,
      "epoch": 0.7648317447967321,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004946984146235183,
      "loss": 5.2463,
      "mean_token_accuracy": 0.17984112799167634,
      "num_tokens": 17055757.0,
      "step": 9830
    },
    {
      "entropy": 5.562282133102417,
      "epoch": 0.7652207741684497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004946924319093579,
      "loss": 5.3668,
      "mean_token_accuracy": 0.1770143076777458,
      "num_tokens": 17063944.0,
      "step": 9835
    },
    {
      "entropy": 5.570206069946289,
      "epoch": 0.7656098035401673,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946864458617148,
      "loss": 5.2588,
      "mean_token_accuracy": 0.17300845831632614,
      "num_tokens": 17072895.0,
      "step": 9840
    },
    {
      "entropy": 5.521096086502075,
      "epoch": 0.7659988329118849,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946804564806793,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1775191456079483,
      "num_tokens": 17081828.0,
      "step": 9845
    },
    {
      "entropy": 5.536095046997071,
      "epoch": 0.7663878622836025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946744637663427,
      "loss": 5.2203,
      "mean_token_accuracy": 0.18323799669742585,
      "num_tokens": 17090386.0,
      "step": 9850
    },
    {
      "entropy": 5.663650846481323,
      "epoch": 0.7667768916553199,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946684677187956,
      "loss": 5.4161,
      "mean_token_accuracy": 0.1748966947197914,
      "num_tokens": 17099429.0,
      "step": 9855
    },
    {
      "entropy": 5.554952335357666,
      "epoch": 0.7671659210270375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946624683381292,
      "loss": 5.2989,
      "mean_token_accuracy": 0.17905012965202333,
      "num_tokens": 17107868.0,
      "step": 9860
    },
    {
      "entropy": 5.5565272808074955,
      "epoch": 0.7675549503987551,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946564656244345,
      "loss": 5.3356,
      "mean_token_accuracy": 0.17248685657978058,
      "num_tokens": 17116936.0,
      "step": 9865
    },
    {
      "entropy": 5.590495014190674,
      "epoch": 0.7679439797704727,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946504595778026,
      "loss": 5.264,
      "mean_token_accuracy": 0.17216774672269822,
      "num_tokens": 17125475.0,
      "step": 9870
    },
    {
      "entropy": 5.595300626754761,
      "epoch": 0.7683330091421903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946444501983246,
      "loss": 5.2723,
      "mean_token_accuracy": 0.18637385070323945,
      "num_tokens": 17134041.0,
      "step": 9875
    },
    {
      "entropy": 5.534225416183472,
      "epoch": 0.7687220385139077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946384374860916,
      "loss": 5.2857,
      "mean_token_accuracy": 0.1820813775062561,
      "num_tokens": 17142088.0,
      "step": 9880
    },
    {
      "entropy": 5.5372497081756595,
      "epoch": 0.7691110678856253,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004946324214411949,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1841478690505028,
      "num_tokens": 17150185.0,
      "step": 9885
    },
    {
      "entropy": 5.470294761657715,
      "epoch": 0.7695000972573429,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946264020637259,
      "loss": 5.0628,
      "mean_token_accuracy": 0.1939447656273842,
      "num_tokens": 17158437.0,
      "step": 9890
    },
    {
      "entropy": 5.440548849105835,
      "epoch": 0.7698891266290605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946203793537757,
      "loss": 5.244,
      "mean_token_accuracy": 0.1837555557489395,
      "num_tokens": 17167811.0,
      "step": 9895
    },
    {
      "entropy": 5.562580537796021,
      "epoch": 0.7702781560007781,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946143533114358,
      "loss": 5.3101,
      "mean_token_accuracy": 0.17614355236291884,
      "num_tokens": 17176982.0,
      "step": 9900
    },
    {
      "entropy": 5.608755016326905,
      "epoch": 0.7706671853724957,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946083239367976,
      "loss": 5.2377,
      "mean_token_accuracy": 0.17974676787853242,
      "num_tokens": 17185633.0,
      "step": 9905
    },
    {
      "entropy": 5.4947271823883055,
      "epoch": 0.7710562147442132,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946022912299527,
      "loss": 5.285,
      "mean_token_accuracy": 0.17526287883520125,
      "num_tokens": 17194996.0,
      "step": 9910
    },
    {
      "entropy": 5.6518114566802975,
      "epoch": 0.7714452441159307,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945962551909926,
      "loss": 5.3927,
      "mean_token_accuracy": 0.17043281644582747,
      "num_tokens": 17204159.0,
      "step": 9915
    },
    {
      "entropy": 5.51574559211731,
      "epoch": 0.7718342734876483,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945902158200089,
      "loss": 5.2703,
      "mean_token_accuracy": 0.1770494043827057,
      "num_tokens": 17212930.0,
      "step": 9920
    },
    {
      "entropy": 5.624597072601318,
      "epoch": 0.7722233028593659,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945841731170931,
      "loss": 5.3393,
      "mean_token_accuracy": 0.17416706383228303,
      "num_tokens": 17221884.0,
      "step": 9925
    },
    {
      "entropy": 5.518691301345825,
      "epoch": 0.7726123322310835,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945781270823369,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18571600914001465,
      "num_tokens": 17230525.0,
      "step": 9930
    },
    {
      "entropy": 5.439535808563233,
      "epoch": 0.773001361602801,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945720777158323,
      "loss": 5.1666,
      "mean_token_accuracy": 0.18698474764823914,
      "num_tokens": 17239240.0,
      "step": 9935
    },
    {
      "entropy": 5.619084358215332,
      "epoch": 0.7733903909745186,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004945660250176708,
      "loss": 5.2884,
      "mean_token_accuracy": 0.1752843976020813,
      "num_tokens": 17248174.0,
      "step": 9940
    },
    {
      "entropy": 5.574867486953735,
      "epoch": 0.7737794203462361,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945599689879443,
      "loss": 5.2803,
      "mean_token_accuracy": 0.17735401540994644,
      "num_tokens": 17256935.0,
      "step": 9945
    },
    {
      "entropy": 5.5688135623931885,
      "epoch": 0.7741684497179537,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945539096267448,
      "loss": 5.2913,
      "mean_token_accuracy": 0.18053762316703797,
      "num_tokens": 17265831.0,
      "step": 9950
    },
    {
      "entropy": 5.469559907913208,
      "epoch": 0.7745574790896713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945478469341642,
      "loss": 5.1775,
      "mean_token_accuracy": 0.188815376162529,
      "num_tokens": 17273851.0,
      "step": 9955
    },
    {
      "entropy": 5.547821998596191,
      "epoch": 0.7749465084613888,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945417809102944,
      "loss": 5.2865,
      "mean_token_accuracy": 0.17088476270437242,
      "num_tokens": 17283547.0,
      "step": 9960
    },
    {
      "entropy": 5.490760421752929,
      "epoch": 0.7753355378331064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945357115552275,
      "loss": 5.1435,
      "mean_token_accuracy": 0.18845568150281905,
      "num_tokens": 17291958.0,
      "step": 9965
    },
    {
      "entropy": 5.469471740722656,
      "epoch": 0.775724567204824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004945296388690557,
      "loss": 5.2974,
      "mean_token_accuracy": 0.1850552275776863,
      "num_tokens": 17301374.0,
      "step": 9970
    },
    {
      "entropy": 5.597596454620361,
      "epoch": 0.7761135965765416,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494523562851871,
      "loss": 5.4476,
      "mean_token_accuracy": 0.17390787452459336,
      "num_tokens": 17310649.0,
      "step": 9975
    },
    {
      "entropy": 5.664722537994384,
      "epoch": 0.7765026259482591,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004945174835037655,
      "loss": 5.3408,
      "mean_token_accuracy": 0.1781105414032936,
      "num_tokens": 17319156.0,
      "step": 9980
    },
    {
      "entropy": 5.5528970718383786,
      "epoch": 0.7768916553199766,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004945114008248319,
      "loss": 5.2616,
      "mean_token_accuracy": 0.18411820977926255,
      "num_tokens": 17327906.0,
      "step": 9985
    },
    {
      "entropy": 5.530083465576172,
      "epoch": 0.7772806846916942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945053148151619,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1852427065372467,
      "num_tokens": 17337042.0,
      "step": 9990
    },
    {
      "entropy": 5.544866991043091,
      "epoch": 0.7776697140634118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944992254748482,
      "loss": 5.2979,
      "mean_token_accuracy": 0.17376251220703126,
      "num_tokens": 17346208.0,
      "step": 9995
    },
    {
      "entropy": 5.609505224227905,
      "epoch": 0.7780587434351294,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944931328039832,
      "loss": 5.3232,
      "mean_token_accuracy": 0.18415360450744628,
      "num_tokens": 17354965.0,
      "step": 10000
    },
    {
      "entropy": 5.57321515083313,
      "epoch": 0.778447772806847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944870368026592,
      "loss": 5.284,
      "mean_token_accuracy": 0.17819664627313614,
      "num_tokens": 17364129.0,
      "step": 10005
    },
    {
      "entropy": 5.573775291442871,
      "epoch": 0.7788368021785644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944809374709689,
      "loss": 5.2878,
      "mean_token_accuracy": 0.17610488533973695,
      "num_tokens": 17372928.0,
      "step": 10010
    },
    {
      "entropy": 5.612327718734742,
      "epoch": 0.779225831550282,
      "grad_norm": 1.125,
      "learning_rate": 0.0004944748348090045,
      "loss": 5.3089,
      "mean_token_accuracy": 0.17743565291166305,
      "num_tokens": 17381076.0,
      "step": 10015
    },
    {
      "entropy": 5.590665245056153,
      "epoch": 0.7796148609219996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004944687288168589,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1801182821393013,
      "num_tokens": 17389614.0,
      "step": 10020
    },
    {
      "entropy": 5.6176183223724365,
      "epoch": 0.7800038902937172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004944626194946246,
      "loss": 5.414,
      "mean_token_accuracy": 0.1738124206662178,
      "num_tokens": 17398718.0,
      "step": 10025
    },
    {
      "entropy": 5.554498147964478,
      "epoch": 0.7803929196654348,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944565068423945,
      "loss": 5.203,
      "mean_token_accuracy": 0.18256326615810395,
      "num_tokens": 17408504.0,
      "step": 10030
    },
    {
      "entropy": 5.4566785335540775,
      "epoch": 0.7807819490371523,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004944503908602612,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18209146857261657,
      "num_tokens": 17417826.0,
      "step": 10035
    },
    {
      "entropy": 5.590363836288452,
      "epoch": 0.7811709784088698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944442715483174,
      "loss": 5.3467,
      "mean_token_accuracy": 0.17133993804454803,
      "num_tokens": 17426581.0,
      "step": 10040
    },
    {
      "entropy": 5.6116899967193605,
      "epoch": 0.7815600077805874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004944381489066561,
      "loss": 5.3081,
      "mean_token_accuracy": 0.17582398653030396,
      "num_tokens": 17435320.0,
      "step": 10045
    },
    {
      "entropy": 5.534014320373535,
      "epoch": 0.781949037152305,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944320229353702,
      "loss": 5.247,
      "mean_token_accuracy": 0.1847595140337944,
      "num_tokens": 17443707.0,
      "step": 10050
    },
    {
      "entropy": 5.437810373306275,
      "epoch": 0.7823380665240226,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944258936345526,
      "loss": 5.133,
      "mean_token_accuracy": 0.18285618871450424,
      "num_tokens": 17452410.0,
      "step": 10055
    },
    {
      "entropy": 5.519442367553711,
      "epoch": 0.7827270958957401,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004944197610042963,
      "loss": 5.224,
      "mean_token_accuracy": 0.1846740022301674,
      "num_tokens": 17460000.0,
      "step": 10060
    },
    {
      "entropy": 5.542432737350464,
      "epoch": 0.7831161252674577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004944136250446944,
      "loss": 5.2107,
      "mean_token_accuracy": 0.18191176801919937,
      "num_tokens": 17469179.0,
      "step": 10065
    },
    {
      "entropy": 5.5194908618927006,
      "epoch": 0.7835051546391752,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944074857558399,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1840083658695221,
      "num_tokens": 17478048.0,
      "step": 10070
    },
    {
      "entropy": 5.458236980438232,
      "epoch": 0.7838941840108928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944013431378261,
      "loss": 5.131,
      "mean_token_accuracy": 0.18433062583208085,
      "num_tokens": 17485959.0,
      "step": 10075
    },
    {
      "entropy": 5.512459135055542,
      "epoch": 0.7842832133826104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943951971907461,
      "loss": 5.222,
      "mean_token_accuracy": 0.1845149204134941,
      "num_tokens": 17495158.0,
      "step": 10080
    },
    {
      "entropy": 5.522275066375732,
      "epoch": 0.7846722427543279,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943890479146933,
      "loss": 5.2476,
      "mean_token_accuracy": 0.18333764970302582,
      "num_tokens": 17503939.0,
      "step": 10085
    },
    {
      "entropy": 5.52121205329895,
      "epoch": 0.7850612721260455,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943828953097607,
      "loss": 5.2873,
      "mean_token_accuracy": 0.17532044351100923,
      "num_tokens": 17513391.0,
      "step": 10090
    },
    {
      "entropy": 5.599463081359863,
      "epoch": 0.7854503014977631,
      "grad_norm": 1.0,
      "learning_rate": 0.000494376739376042,
      "loss": 5.3201,
      "mean_token_accuracy": 0.17570871859788895,
      "num_tokens": 17522922.0,
      "step": 10095
    },
    {
      "entropy": 5.580346012115479,
      "epoch": 0.7858393308694807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943705801136304,
      "loss": 5.3136,
      "mean_token_accuracy": 0.17581601142883302,
      "num_tokens": 17531920.0,
      "step": 10100
    },
    {
      "entropy": 5.646449947357178,
      "epoch": 0.7862283602411982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943644175226193,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1789740428328514,
      "num_tokens": 17539816.0,
      "step": 10105
    },
    {
      "entropy": 5.647852277755737,
      "epoch": 0.7866173896129158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943582516031025,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18268545418977739,
      "num_tokens": 17548584.0,
      "step": 10110
    },
    {
      "entropy": 5.513976001739502,
      "epoch": 0.7870064189846333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943520823551732,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18332510590553283,
      "num_tokens": 17557095.0,
      "step": 10115
    },
    {
      "entropy": 5.569630289077759,
      "epoch": 0.7873954483563509,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943459097789252,
      "loss": 5.2692,
      "mean_token_accuracy": 0.18324509412050247,
      "num_tokens": 17565937.0,
      "step": 10120
    },
    {
      "entropy": 5.516526746749878,
      "epoch": 0.7877844777280685,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943397338744523,
      "loss": 5.3523,
      "mean_token_accuracy": 0.17756358534097672,
      "num_tokens": 17574948.0,
      "step": 10125
    },
    {
      "entropy": 5.619197940826416,
      "epoch": 0.7881735070997861,
      "grad_norm": 0.90625,
      "learning_rate": 0.000494333554641848,
      "loss": 5.3654,
      "mean_token_accuracy": 0.1765134796500206,
      "num_tokens": 17584307.0,
      "step": 10130
    },
    {
      "entropy": 5.545857238769531,
      "epoch": 0.7885625364715037,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494327372081206,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1844295084476471,
      "num_tokens": 17592828.0,
      "step": 10135
    },
    {
      "entropy": 5.581251811981201,
      "epoch": 0.7889515658432211,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943211861926204,
      "loss": 5.2854,
      "mean_token_accuracy": 0.18676233440637588,
      "num_tokens": 17601110.0,
      "step": 10140
    },
    {
      "entropy": 5.5625608444213865,
      "epoch": 0.7893405952149387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943149969761848,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18348842412233352,
      "num_tokens": 17610221.0,
      "step": 10145
    },
    {
      "entropy": 5.540854406356812,
      "epoch": 0.7897296245866563,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004943088044319932,
      "loss": 5.174,
      "mean_token_accuracy": 0.18292491137981415,
      "num_tokens": 17618343.0,
      "step": 10150
    },
    {
      "entropy": 5.546886253356933,
      "epoch": 0.7901186539583739,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943026085601395,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1819344252347946,
      "num_tokens": 17626885.0,
      "step": 10155
    },
    {
      "entropy": 5.536163425445556,
      "epoch": 0.7905076833300915,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942964093607178,
      "loss": 5.36,
      "mean_token_accuracy": 0.17669307738542556,
      "num_tokens": 17635891.0,
      "step": 10160
    },
    {
      "entropy": 5.57954740524292,
      "epoch": 0.790896712701809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004942902068338222,
      "loss": 5.1719,
      "mean_token_accuracy": 0.18569800853729249,
      "num_tokens": 17643815.0,
      "step": 10165
    },
    {
      "entropy": 5.565684032440186,
      "epoch": 0.7912857420735265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942840009795466,
      "loss": 5.2398,
      "mean_token_accuracy": 0.1842930570244789,
      "num_tokens": 17652414.0,
      "step": 10170
    },
    {
      "entropy": 5.547009086608886,
      "epoch": 0.7916747714452441,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942777917979855,
      "loss": 5.3034,
      "mean_token_accuracy": 0.17551807165145875,
      "num_tokens": 17661382.0,
      "step": 10175
    },
    {
      "entropy": 5.531071758270263,
      "epoch": 0.7920638008169617,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942715792892328,
      "loss": 5.252,
      "mean_token_accuracy": 0.17437173128128053,
      "num_tokens": 17669745.0,
      "step": 10180
    },
    {
      "entropy": 5.560464191436767,
      "epoch": 0.7924528301886793,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494265363453383,
      "loss": 5.2042,
      "mean_token_accuracy": 0.18362361043691636,
      "num_tokens": 17678646.0,
      "step": 10185
    },
    {
      "entropy": 5.574765539169311,
      "epoch": 0.7928418595603968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942591442905302,
      "loss": 5.342,
      "mean_token_accuracy": 0.17223089784383774,
      "num_tokens": 17687197.0,
      "step": 10190
    },
    {
      "entropy": 5.568060207366943,
      "epoch": 0.7932308889321144,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942529218007689,
      "loss": 5.3322,
      "mean_token_accuracy": 0.17627218663692473,
      "num_tokens": 17696528.0,
      "step": 10195
    },
    {
      "entropy": 5.532153463363647,
      "epoch": 0.7936199183038319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942466959841936,
      "loss": 5.2109,
      "mean_token_accuracy": 0.18019879162311553,
      "num_tokens": 17704688.0,
      "step": 10200
    },
    {
      "entropy": 5.437143087387085,
      "epoch": 0.7940089476755495,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942404668408985,
      "loss": 5.1121,
      "mean_token_accuracy": 0.197529399394989,
      "num_tokens": 17713039.0,
      "step": 10205
    },
    {
      "entropy": 5.585537385940552,
      "epoch": 0.7943979770472671,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942342343709783,
      "loss": 5.4013,
      "mean_token_accuracy": 0.17179318517446518,
      "num_tokens": 17721771.0,
      "step": 10210
    },
    {
      "entropy": 5.5868367671966555,
      "epoch": 0.7947870064189846,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942279985745275,
      "loss": 5.2341,
      "mean_token_accuracy": 0.1795405477285385,
      "num_tokens": 17729982.0,
      "step": 10215
    },
    {
      "entropy": 5.548766279220581,
      "epoch": 0.7951760357907022,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942217594516409,
      "loss": 5.2484,
      "mean_token_accuracy": 0.1787489965558052,
      "num_tokens": 17738397.0,
      "step": 10220
    },
    {
      "entropy": 5.618678569793701,
      "epoch": 0.7955650651624198,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494215517002413,
      "loss": 5.359,
      "mean_token_accuracy": 0.17360880672931672,
      "num_tokens": 17747421.0,
      "step": 10225
    },
    {
      "entropy": 5.576296138763428,
      "epoch": 0.7959540945341373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942092712269384,
      "loss": 5.2894,
      "mean_token_accuracy": 0.17363293170928956,
      "num_tokens": 17756629.0,
      "step": 10230
    },
    {
      "entropy": 5.558225202560425,
      "epoch": 0.7963431239058549,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942030221253122,
      "loss": 5.2961,
      "mean_token_accuracy": 0.18580178320407867,
      "num_tokens": 17764595.0,
      "step": 10235
    },
    {
      "entropy": 5.547394037246704,
      "epoch": 0.7967321532775724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941967696976289,
      "loss": 5.182,
      "mean_token_accuracy": 0.18426274955272676,
      "num_tokens": 17773522.0,
      "step": 10240
    },
    {
      "entropy": 5.536650896072388,
      "epoch": 0.79712118264929,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941905139439835,
      "loss": 5.3027,
      "mean_token_accuracy": 0.183289997279644,
      "num_tokens": 17782850.0,
      "step": 10245
    },
    {
      "entropy": 5.470347118377686,
      "epoch": 0.7975102120210076,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941842548644712,
      "loss": 5.158,
      "mean_token_accuracy": 0.18935055434703826,
      "num_tokens": 17790638.0,
      "step": 10250
    },
    {
      "entropy": 5.573339891433716,
      "epoch": 0.7978992413927252,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941779924591864,
      "loss": 5.3018,
      "mean_token_accuracy": 0.17207662612199784,
      "num_tokens": 17799992.0,
      "step": 10255
    },
    {
      "entropy": 5.505545663833618,
      "epoch": 0.7982882707644428,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941717267282244,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1826378509402275,
      "num_tokens": 17808643.0,
      "step": 10260
    },
    {
      "entropy": 5.541860151290893,
      "epoch": 0.7986773001361602,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941654576716806,
      "loss": 5.2106,
      "mean_token_accuracy": 0.18334080129861832,
      "num_tokens": 17816551.0,
      "step": 10265
    },
    {
      "entropy": 5.516403532028198,
      "epoch": 0.7990663295078778,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941591852896495,
      "loss": 5.2391,
      "mean_token_accuracy": 0.18402206003665925,
      "num_tokens": 17825692.0,
      "step": 10270
    },
    {
      "entropy": 5.572368001937866,
      "epoch": 0.7994553588795954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004941529095822268,
      "loss": 5.3479,
      "mean_token_accuracy": 0.17121150195598603,
      "num_tokens": 17833920.0,
      "step": 10275
    },
    {
      "entropy": 5.572353267669678,
      "epoch": 0.799844388251313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004941466305495074,
      "loss": 5.3154,
      "mean_token_accuracy": 0.17272062599658966,
      "num_tokens": 17842339.0,
      "step": 10280
    },
    {
      "entropy": 5.507529592514038,
      "epoch": 0.8002334176230306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941403481915867,
      "loss": 5.221,
      "mean_token_accuracy": 0.18210890889167786,
      "num_tokens": 17851273.0,
      "step": 10285
    },
    {
      "entropy": 5.511989450454712,
      "epoch": 0.800622446994748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004941340625085601,
      "loss": 5.2588,
      "mean_token_accuracy": 0.18199315816164016,
      "num_tokens": 17860131.0,
      "step": 10290
    },
    {
      "entropy": 5.451030826568603,
      "epoch": 0.8010114763664656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941277735005228,
      "loss": 5.183,
      "mean_token_accuracy": 0.18900831192731857,
      "num_tokens": 17868670.0,
      "step": 10295
    },
    {
      "entropy": 5.55652003288269,
      "epoch": 0.8014005057381832,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941214811675702,
      "loss": 5.1796,
      "mean_token_accuracy": 0.18667263090610503,
      "num_tokens": 17876972.0,
      "step": 10300
    },
    {
      "entropy": 5.5093494892120365,
      "epoch": 0.8017895351099008,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941151855097982,
      "loss": 5.2754,
      "mean_token_accuracy": 0.17642293125391006,
      "num_tokens": 17885708.0,
      "step": 10305
    },
    {
      "entropy": 5.601290845870972,
      "epoch": 0.8021785644816184,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941088865273018,
      "loss": 5.291,
      "mean_token_accuracy": 0.17429496645927428,
      "num_tokens": 17894003.0,
      "step": 10310
    },
    {
      "entropy": 5.5281816005706785,
      "epoch": 0.802567593853336,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004941025842201769,
      "loss": 5.1275,
      "mean_token_accuracy": 0.19140205681324005,
      "num_tokens": 17903113.0,
      "step": 10315
    },
    {
      "entropy": 5.514891910552978,
      "epoch": 0.8029566232250535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940962785885189,
      "loss": 5.2165,
      "mean_token_accuracy": 0.17935004234313964,
      "num_tokens": 17912222.0,
      "step": 10320
    },
    {
      "entropy": 5.55033392906189,
      "epoch": 0.803345652596771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940899696324237,
      "loss": 5.2994,
      "mean_token_accuracy": 0.17741072028875352,
      "num_tokens": 17920900.0,
      "step": 10325
    },
    {
      "entropy": 5.60826849937439,
      "epoch": 0.8037346819684886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940836573519868,
      "loss": 5.3233,
      "mean_token_accuracy": 0.17978237718343734,
      "num_tokens": 17929221.0,
      "step": 10330
    },
    {
      "entropy": 5.515179872512817,
      "epoch": 0.8041237113402062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940773417473043,
      "loss": 5.235,
      "mean_token_accuracy": 0.18211790919303894,
      "num_tokens": 17938019.0,
      "step": 10335
    },
    {
      "entropy": 5.500888681411743,
      "epoch": 0.8045127407119238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940710228184717,
      "loss": 5.2083,
      "mean_token_accuracy": 0.18903656899929047,
      "num_tokens": 17946972.0,
      "step": 10340
    },
    {
      "entropy": 5.590160894393921,
      "epoch": 0.8049017700836413,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494064700565585,
      "loss": 5.3162,
      "mean_token_accuracy": 0.17665481865406035,
      "num_tokens": 17956487.0,
      "step": 10345
    },
    {
      "entropy": 5.478394937515259,
      "epoch": 0.8052907994553589,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940583749887403,
      "loss": 5.2167,
      "mean_token_accuracy": 0.18098169267177583,
      "num_tokens": 17965838.0,
      "step": 10350
    },
    {
      "entropy": 5.621855020523071,
      "epoch": 0.8056798288270764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940520460880333,
      "loss": 5.3214,
      "mean_token_accuracy": 0.17822395712137223,
      "num_tokens": 17974492.0,
      "step": 10355
    },
    {
      "entropy": 5.5938249111175535,
      "epoch": 0.806068858198794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940457138635601,
      "loss": 5.2904,
      "mean_token_accuracy": 0.18426005989313127,
      "num_tokens": 17983056.0,
      "step": 10360
    },
    {
      "entropy": 5.501426029205322,
      "epoch": 0.8064578875705116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940393783154169,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1829010561108589,
      "num_tokens": 17990822.0,
      "step": 10365
    },
    {
      "entropy": 5.47200608253479,
      "epoch": 0.8068469169422291,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004940330394436999,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19085012078285218,
      "num_tokens": 17999284.0,
      "step": 10370
    },
    {
      "entropy": 5.548887825012207,
      "epoch": 0.8072359463139467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494026697248505,
      "loss": 5.3193,
      "mean_token_accuracy": 0.18064596951007844,
      "num_tokens": 18007585.0,
      "step": 10375
    },
    {
      "entropy": 5.448264741897583,
      "epoch": 0.8076249756856643,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940203517299287,
      "loss": 5.1755,
      "mean_token_accuracy": 0.18118565976619722,
      "num_tokens": 18016332.0,
      "step": 10380
    },
    {
      "entropy": 5.569855976104736,
      "epoch": 0.8080140050573819,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494014002888067,
      "loss": 5.2932,
      "mean_token_accuracy": 0.1763343408703804,
      "num_tokens": 18025524.0,
      "step": 10385
    },
    {
      "entropy": 5.552229404449463,
      "epoch": 0.8084030344290994,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940076507230166,
      "loss": 5.2792,
      "mean_token_accuracy": 0.1820867657661438,
      "num_tokens": 18033951.0,
      "step": 10390
    },
    {
      "entropy": 5.680067157745361,
      "epoch": 0.8087920638008169,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004940012952348735,
      "loss": 5.3969,
      "mean_token_accuracy": 0.17616181969642639,
      "num_tokens": 18042258.0,
      "step": 10395
    },
    {
      "entropy": 5.593752193450928,
      "epoch": 0.8091810931725345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004939949364237345,
      "loss": 5.1729,
      "mean_token_accuracy": 0.18507616072893143,
      "num_tokens": 18050630.0,
      "step": 10400
    },
    {
      "entropy": 5.519416952133179,
      "epoch": 0.8095701225442521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004939885742896958,
      "loss": 5.2703,
      "mean_token_accuracy": 0.18009306788444518,
      "num_tokens": 18059141.0,
      "step": 10405
    },
    {
      "entropy": 5.5265296459197994,
      "epoch": 0.8099591519159697,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493982208832854,
      "loss": 5.2749,
      "mean_token_accuracy": 0.17932132631540298,
      "num_tokens": 18067831.0,
      "step": 10410
    },
    {
      "entropy": 5.637141418457031,
      "epoch": 0.8103481812876873,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939758400533058,
      "loss": 5.3654,
      "mean_token_accuracy": 0.17379961013793946,
      "num_tokens": 18076629.0,
      "step": 10415
    },
    {
      "entropy": 5.549745702743531,
      "epoch": 0.8107372106594047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939694679511478,
      "loss": 5.2133,
      "mean_token_accuracy": 0.18306296020746232,
      "num_tokens": 18085930.0,
      "step": 10420
    },
    {
      "entropy": 5.551803255081177,
      "epoch": 0.8111262400311223,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939630925264765,
      "loss": 5.2515,
      "mean_token_accuracy": 0.18021342754364014,
      "num_tokens": 18094947.0,
      "step": 10425
    },
    {
      "entropy": 5.555429410934448,
      "epoch": 0.8115152694028399,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000493956713779389,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1799653187394142,
      "num_tokens": 18103838.0,
      "step": 10430
    },
    {
      "entropy": 5.554565286636352,
      "epoch": 0.8119042987745575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939503317099817,
      "loss": 5.177,
      "mean_token_accuracy": 0.1844314604997635,
      "num_tokens": 18112587.0,
      "step": 10435
    },
    {
      "entropy": 5.527368497848511,
      "epoch": 0.8122933281462751,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004939439463183517,
      "loss": 5.1996,
      "mean_token_accuracy": 0.18574080765247344,
      "num_tokens": 18121094.0,
      "step": 10440
    },
    {
      "entropy": 5.594413471221924,
      "epoch": 0.8126823575179926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004939375576045958,
      "loss": 5.3006,
      "mean_token_accuracy": 0.18318731486797332,
      "num_tokens": 18129699.0,
      "step": 10445
    },
    {
      "entropy": 5.570493602752686,
      "epoch": 0.8130713868897101,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939311655688109,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1860767811536789,
      "num_tokens": 18138466.0,
      "step": 10450
    },
    {
      "entropy": 5.527794408798218,
      "epoch": 0.8134604162614277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939247702110941,
      "loss": 5.3087,
      "mean_token_accuracy": 0.1832980751991272,
      "num_tokens": 18147436.0,
      "step": 10455
    },
    {
      "entropy": 5.56290225982666,
      "epoch": 0.8138494456331453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004939183715315423,
      "loss": 5.2612,
      "mean_token_accuracy": 0.18561258018016816,
      "num_tokens": 18156228.0,
      "step": 10460
    },
    {
      "entropy": 5.56533784866333,
      "epoch": 0.8142384750048629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004939119695302526,
      "loss": 5.2736,
      "mean_token_accuracy": 0.17960922569036483,
      "num_tokens": 18164571.0,
      "step": 10465
    },
    {
      "entropy": 5.551176643371582,
      "epoch": 0.8146275043765804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004939055642073224,
      "loss": 5.2899,
      "mean_token_accuracy": 0.17826141268014908,
      "num_tokens": 18173322.0,
      "step": 10470
    },
    {
      "entropy": 5.590441417694092,
      "epoch": 0.815016533748298,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938991555628484,
      "loss": 5.3691,
      "mean_token_accuracy": 0.17837988436222077,
      "num_tokens": 18182306.0,
      "step": 10475
    },
    {
      "entropy": 5.642142868041992,
      "epoch": 0.8154055631200156,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004938927435969283,
      "loss": 5.3275,
      "mean_token_accuracy": 0.17471859157085418,
      "num_tokens": 18190567.0,
      "step": 10480
    },
    {
      "entropy": 5.636589860916137,
      "epoch": 0.8157945924917331,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004938863283096592,
      "loss": 5.358,
      "mean_token_accuracy": 0.1726572796702385,
      "num_tokens": 18199678.0,
      "step": 10485
    },
    {
      "entropy": 5.484126138687134,
      "epoch": 0.8161836218634507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938799097011384,
      "loss": 5.0973,
      "mean_token_accuracy": 0.18642169684171678,
      "num_tokens": 18208350.0,
      "step": 10490
    },
    {
      "entropy": 5.579818677902222,
      "epoch": 0.8165726512351682,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938734877714634,
      "loss": 5.2205,
      "mean_token_accuracy": 0.17800672203302384,
      "num_tokens": 18216618.0,
      "step": 10495
    },
    {
      "entropy": 5.56155366897583,
      "epoch": 0.8169616806068858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938670625207317,
      "loss": 5.2438,
      "mean_token_accuracy": 0.18664785325527192,
      "num_tokens": 18224633.0,
      "step": 10500
    },
    {
      "entropy": 5.540253305435181,
      "epoch": 0.8173507099786034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004938606339490406,
      "loss": 5.1886,
      "mean_token_accuracy": 0.18040821105241775,
      "num_tokens": 18233212.0,
      "step": 10505
    },
    {
      "entropy": 5.608698081970215,
      "epoch": 0.817739739350321,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938542020564877,
      "loss": 5.3636,
      "mean_token_accuracy": 0.17857760190963745,
      "num_tokens": 18241796.0,
      "step": 10510
    },
    {
      "entropy": 5.5558243751525875,
      "epoch": 0.8181287687220385,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004938477668431706,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1828357234597206,
      "num_tokens": 18251416.0,
      "step": 10515
    },
    {
      "entropy": 5.548902797698974,
      "epoch": 0.8185177980937561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938413283091871,
      "loss": 5.1576,
      "mean_token_accuracy": 0.18625995218753816,
      "num_tokens": 18259880.0,
      "step": 10520
    },
    {
      "entropy": 5.478700160980225,
      "epoch": 0.8189068274654736,
      "grad_norm": 1.125,
      "learning_rate": 0.0004938348864546347,
      "loss": 5.2436,
      "mean_token_accuracy": 0.18840804994106292,
      "num_tokens": 18268217.0,
      "step": 10525
    },
    {
      "entropy": 5.490329313278198,
      "epoch": 0.8192958568371912,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938284412796111,
      "loss": 5.301,
      "mean_token_accuracy": 0.18139395713806153,
      "num_tokens": 18276556.0,
      "step": 10530
    },
    {
      "entropy": 5.542969799041748,
      "epoch": 0.8196848862089088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938219927842144,
      "loss": 5.2535,
      "mean_token_accuracy": 0.18201657980680466,
      "num_tokens": 18285199.0,
      "step": 10535
    },
    {
      "entropy": 5.543492555618286,
      "epoch": 0.8200739155806264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938155409685422,
      "loss": 5.221,
      "mean_token_accuracy": 0.18483659029006957,
      "num_tokens": 18294004.0,
      "step": 10540
    },
    {
      "entropy": 5.479832887649536,
      "epoch": 0.820462944952344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938090858326923,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1871393457055092,
      "num_tokens": 18301761.0,
      "step": 10545
    },
    {
      "entropy": 5.584003448486328,
      "epoch": 0.8208519743240614,
      "grad_norm": 1.0625,
      "learning_rate": 0.000493802627376763,
      "loss": 5.349,
      "mean_token_accuracy": 0.17411619573831558,
      "num_tokens": 18309555.0,
      "step": 10550
    },
    {
      "entropy": 5.5634064197540285,
      "epoch": 0.821241003695779,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937961656008518,
      "loss": 5.3124,
      "mean_token_accuracy": 0.17790265679359435,
      "num_tokens": 18318011.0,
      "step": 10555
    },
    {
      "entropy": 5.587578916549683,
      "epoch": 0.8216300330674966,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937897005050573,
      "loss": 5.3019,
      "mean_token_accuracy": 0.17786704897880554,
      "num_tokens": 18326108.0,
      "step": 10560
    },
    {
      "entropy": 5.5615644454956055,
      "epoch": 0.8220190624392142,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937832320894772,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1778743878006935,
      "num_tokens": 18334777.0,
      "step": 10565
    },
    {
      "entropy": 5.5428698539733885,
      "epoch": 0.8224080918109318,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937767603542098,
      "loss": 5.223,
      "mean_token_accuracy": 0.17921229600906372,
      "num_tokens": 18343796.0,
      "step": 10570
    },
    {
      "entropy": 5.549758958816528,
      "epoch": 0.8227971211826492,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937702852993534,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1769055426120758,
      "num_tokens": 18352396.0,
      "step": 10575
    },
    {
      "entropy": 5.534566402435303,
      "epoch": 0.8231861505543668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493763806925006,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1870693102478981,
      "num_tokens": 18360914.0,
      "step": 10580
    },
    {
      "entropy": 5.5200659275054935,
      "epoch": 0.8235751799260844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937573252312661,
      "loss": 5.2379,
      "mean_token_accuracy": 0.18408682346343994,
      "num_tokens": 18369281.0,
      "step": 10585
    },
    {
      "entropy": 5.510223770141602,
      "epoch": 0.823964209297802,
      "grad_norm": 1.109375,
      "learning_rate": 0.000493750840218232,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19342626482248307,
      "num_tokens": 18377305.0,
      "step": 10590
    },
    {
      "entropy": 5.5342694282531735,
      "epoch": 0.8243532386695196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004937443518860021,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1735801115632057,
      "num_tokens": 18386563.0,
      "step": 10595
    },
    {
      "entropy": 5.528779554367065,
      "epoch": 0.8247422680412371,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937378602346747,
      "loss": 5.2364,
      "mean_token_accuracy": 0.1893967166543007,
      "num_tokens": 18395150.0,
      "step": 10600
    },
    {
      "entropy": 5.559446954727173,
      "epoch": 0.8251312974129547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004937313652643485,
      "loss": 5.3637,
      "mean_token_accuracy": 0.17537704557180406,
      "num_tokens": 18403790.0,
      "step": 10605
    },
    {
      "entropy": 5.588216972351074,
      "epoch": 0.8255203267846722,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493724866975122,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18539801836013795,
      "num_tokens": 18411936.0,
      "step": 10610
    },
    {
      "entropy": 5.4975566387176515,
      "epoch": 0.8259093561563898,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937183653670937,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1907249301671982,
      "num_tokens": 18421146.0,
      "step": 10615
    },
    {
      "entropy": 5.54138445854187,
      "epoch": 0.8262983855281074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937118604403623,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18405063152313234,
      "num_tokens": 18429498.0,
      "step": 10620
    },
    {
      "entropy": 5.594035053253174,
      "epoch": 0.8266874148998249,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937053521950266,
      "loss": 5.3016,
      "mean_token_accuracy": 0.1795467182993889,
      "num_tokens": 18438600.0,
      "step": 10625
    },
    {
      "entropy": 5.529293155670166,
      "epoch": 0.8270764442715425,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004936988406311854,
      "loss": 5.1834,
      "mean_token_accuracy": 0.18156871795654297,
      "num_tokens": 18447397.0,
      "step": 10630
    },
    {
      "entropy": 5.523371791839599,
      "epoch": 0.8274654736432601,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936923257489373,
      "loss": 5.2629,
      "mean_token_accuracy": 0.17718516588211058,
      "num_tokens": 18456677.0,
      "step": 10635
    },
    {
      "entropy": 5.508723735809326,
      "epoch": 0.8278545030149776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936858075483811,
      "loss": 5.1952,
      "mean_token_accuracy": 0.18520419001579286,
      "num_tokens": 18465605.0,
      "step": 10640
    },
    {
      "entropy": 5.508184766769409,
      "epoch": 0.8282435323866952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493679286029616,
      "loss": 5.2287,
      "mean_token_accuracy": 0.1800902873277664,
      "num_tokens": 18474419.0,
      "step": 10645
    },
    {
      "entropy": 5.64747462272644,
      "epoch": 0.8286325617584127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004936727611927407,
      "loss": 5.354,
      "mean_token_accuracy": 0.17717713862657547,
      "num_tokens": 18482646.0,
      "step": 10650
    },
    {
      "entropy": 5.54002046585083,
      "epoch": 0.8290215911301303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936662330378546,
      "loss": 5.2057,
      "mean_token_accuracy": 0.1830547034740448,
      "num_tokens": 18490899.0,
      "step": 10655
    },
    {
      "entropy": 5.526831150054932,
      "epoch": 0.8294106205018479,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936597015650561,
      "loss": 5.2091,
      "mean_token_accuracy": 0.18813900351524354,
      "num_tokens": 18499238.0,
      "step": 10660
    },
    {
      "entropy": 5.520270872116089,
      "epoch": 0.8297996498735655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004936531667744448,
      "loss": 5.1793,
      "mean_token_accuracy": 0.17968463003635407,
      "num_tokens": 18507668.0,
      "step": 10665
    },
    {
      "entropy": 5.588473844528198,
      "epoch": 0.8301886792452831,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004936466286661197,
      "loss": 5.2449,
      "mean_token_accuracy": 0.18253855854272844,
      "num_tokens": 18516366.0,
      "step": 10670
    },
    {
      "entropy": 5.5107598304748535,
      "epoch": 0.8305777086170005,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049364008724018,
      "loss": 5.2216,
      "mean_token_accuracy": 0.18302127122879028,
      "num_tokens": 18525012.0,
      "step": 10675
    },
    {
      "entropy": 5.5812681198120115,
      "epoch": 0.8309667379887181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493633542496725,
      "loss": 5.3062,
      "mean_token_accuracy": 0.18337736427783966,
      "num_tokens": 18533277.0,
      "step": 10680
    },
    {
      "entropy": 5.57138500213623,
      "epoch": 0.8313557673604357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004936269944358539,
      "loss": 5.3108,
      "mean_token_accuracy": 0.17496419250965117,
      "num_tokens": 18541738.0,
      "step": 10685
    },
    {
      "entropy": 5.58394193649292,
      "epoch": 0.8317447967321533,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936204430576664,
      "loss": 5.271,
      "mean_token_accuracy": 0.18091964572668076,
      "num_tokens": 18550305.0,
      "step": 10690
    },
    {
      "entropy": 5.544063138961792,
      "epoch": 0.8321338261038709,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936138883622614,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19183179289102553,
      "num_tokens": 18558528.0,
      "step": 10695
    },
    {
      "entropy": 5.469418525695801,
      "epoch": 0.8325228554755884,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004936073303497387,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18760383278131484,
      "num_tokens": 18567166.0,
      "step": 10700
    },
    {
      "entropy": 5.494637823104858,
      "epoch": 0.8329118848473059,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936007690201976,
      "loss": 5.1935,
      "mean_token_accuracy": 0.18928715586662292,
      "num_tokens": 18575986.0,
      "step": 10705
    },
    {
      "entropy": 5.599856281280518,
      "epoch": 0.8333009142190235,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493594204373738,
      "loss": 5.2465,
      "mean_token_accuracy": 0.17514331936836242,
      "num_tokens": 18585446.0,
      "step": 10710
    },
    {
      "entropy": 5.553297996520996,
      "epoch": 0.8336899435907411,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935876364104591,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18375732898712158,
      "num_tokens": 18594236.0,
      "step": 10715
    },
    {
      "entropy": 5.539151382446289,
      "epoch": 0.8340789729624587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935810651304608,
      "loss": 5.2727,
      "mean_token_accuracy": 0.1795378729701042,
      "num_tokens": 18603229.0,
      "step": 10720
    },
    {
      "entropy": 5.518246507644653,
      "epoch": 0.8344680023341763,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935744905338427,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19666348099708558,
      "num_tokens": 18611783.0,
      "step": 10725
    },
    {
      "entropy": 5.610187101364136,
      "epoch": 0.8348570317058938,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935679126207046,
      "loss": 5.3275,
      "mean_token_accuracy": 0.17979582101106645,
      "num_tokens": 18620901.0,
      "step": 10730
    },
    {
      "entropy": 5.523731851577759,
      "epoch": 0.8352460610776113,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004935613313911463,
      "loss": 5.2326,
      "mean_token_accuracy": 0.17777478843927383,
      "num_tokens": 18629447.0,
      "step": 10735
    },
    {
      "entropy": 5.581677579879761,
      "epoch": 0.8356350904493289,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935547468452678,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18409804850816727,
      "num_tokens": 18637895.0,
      "step": 10740
    },
    {
      "entropy": 5.576303434371948,
      "epoch": 0.8360241198210465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004935481589831688,
      "loss": 5.3899,
      "mean_token_accuracy": 0.17418855279684067,
      "num_tokens": 18646417.0,
      "step": 10745
    },
    {
      "entropy": 5.575652599334717,
      "epoch": 0.8364131491927641,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935415678049492,
      "loss": 5.3088,
      "mean_token_accuracy": 0.17884970754384993,
      "num_tokens": 18654814.0,
      "step": 10750
    },
    {
      "entropy": 5.595830678939819,
      "epoch": 0.8368021785644816,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935349733107094,
      "loss": 5.2989,
      "mean_token_accuracy": 0.18440109640359878,
      "num_tokens": 18663520.0,
      "step": 10755
    },
    {
      "entropy": 5.545376777648926,
      "epoch": 0.8371912079361992,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493528375500549,
      "loss": 5.3275,
      "mean_token_accuracy": 0.1782543271780014,
      "num_tokens": 18672033.0,
      "step": 10760
    },
    {
      "entropy": 5.54251537322998,
      "epoch": 0.8375802373079168,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935217743745685,
      "loss": 5.1263,
      "mean_token_accuracy": 0.191961570084095,
      "num_tokens": 18680690.0,
      "step": 10765
    },
    {
      "entropy": 5.522523355484009,
      "epoch": 0.8379692666796343,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004935151699328677,
      "loss": 5.2418,
      "mean_token_accuracy": 0.17692058384418488,
      "num_tokens": 18688683.0,
      "step": 10770
    },
    {
      "entropy": 5.484722566604614,
      "epoch": 0.8383582960513519,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935085621755471,
      "loss": 5.3234,
      "mean_token_accuracy": 0.17820271998643875,
      "num_tokens": 18697273.0,
      "step": 10775
    },
    {
      "entropy": 5.616714000701904,
      "epoch": 0.8387473254230694,
      "grad_norm": 1.046875,
      "learning_rate": 0.000493501951102707,
      "loss": 5.2921,
      "mean_token_accuracy": 0.17552858144044875,
      "num_tokens": 18705076.0,
      "step": 10780
    },
    {
      "entropy": 5.537790441513062,
      "epoch": 0.839136354794787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934953367144474,
      "loss": 5.1677,
      "mean_token_accuracy": 0.1860572099685669,
      "num_tokens": 18713596.0,
      "step": 10785
    },
    {
      "entropy": 5.485022783279419,
      "epoch": 0.8395253841665046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934887190108688,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1858276531100273,
      "num_tokens": 18722611.0,
      "step": 10790
    },
    {
      "entropy": 5.609499931335449,
      "epoch": 0.8399144135382222,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934820979920719,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1708538606762886,
      "num_tokens": 18731126.0,
      "step": 10795
    },
    {
      "entropy": 5.534915781021118,
      "epoch": 0.8403034429099397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004934754736581567,
      "loss": 5.2294,
      "mean_token_accuracy": 0.17858947664499283,
      "num_tokens": 18739815.0,
      "step": 10800
    },
    {
      "entropy": 5.478046035766601,
      "epoch": 0.8406924722816572,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493468846009224,
      "loss": 5.1545,
      "mean_token_accuracy": 0.19134612679481505,
      "num_tokens": 18748603.0,
      "step": 10805
    },
    {
      "entropy": 5.507309770584106,
      "epoch": 0.8410815016533748,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934622150453742,
      "loss": 5.2552,
      "mean_token_accuracy": 0.17772441059350969,
      "num_tokens": 18757819.0,
      "step": 10810
    },
    {
      "entropy": 5.602052545547485,
      "epoch": 0.8414705310250924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493455580766708,
      "loss": 5.266,
      "mean_token_accuracy": 0.17887115776538848,
      "num_tokens": 18766135.0,
      "step": 10815
    },
    {
      "entropy": 5.515768527984619,
      "epoch": 0.84185956039681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934489431733262,
      "loss": 5.1591,
      "mean_token_accuracy": 0.18530394583940507,
      "num_tokens": 18775202.0,
      "step": 10820
    },
    {
      "entropy": 5.508689069747925,
      "epoch": 0.8422485897685276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934423022653293,
      "loss": 5.2298,
      "mean_token_accuracy": 0.18810921162366867,
      "num_tokens": 18783466.0,
      "step": 10825
    },
    {
      "entropy": 5.543803453445435,
      "epoch": 0.842637619140245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934356580428182,
      "loss": 5.2305,
      "mean_token_accuracy": 0.1796533927321434,
      "num_tokens": 18791772.0,
      "step": 10830
    },
    {
      "entropy": 5.617976713180542,
      "epoch": 0.8430266485119626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004934290105058937,
      "loss": 5.3857,
      "mean_token_accuracy": 0.17383247017860412,
      "num_tokens": 18801094.0,
      "step": 10835
    },
    {
      "entropy": 5.553902578353882,
      "epoch": 0.8434156778836802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004934223596546565,
      "loss": 5.1615,
      "mean_token_accuracy": 0.18740533292293549,
      "num_tokens": 18809256.0,
      "step": 10840
    },
    {
      "entropy": 5.533125305175782,
      "epoch": 0.8438047072553978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934157054892077,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18046908229589462,
      "num_tokens": 18818109.0,
      "step": 10845
    },
    {
      "entropy": 5.549737119674683,
      "epoch": 0.8441937366271154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934090480096482,
      "loss": 5.203,
      "mean_token_accuracy": 0.1879824861884117,
      "num_tokens": 18826330.0,
      "step": 10850
    },
    {
      "entropy": 5.629470348358154,
      "epoch": 0.8445827659988329,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934023872160791,
      "loss": 5.4019,
      "mean_token_accuracy": 0.17526203095912934,
      "num_tokens": 18835280.0,
      "step": 10855
    },
    {
      "entropy": 5.499673795700073,
      "epoch": 0.8449717953705504,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004933957231086014,
      "loss": 5.1609,
      "mean_token_accuracy": 0.1841066911816597,
      "num_tokens": 18843300.0,
      "step": 10860
    },
    {
      "entropy": 5.463746643066406,
      "epoch": 0.845360824742268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004933890556873161,
      "loss": 5.2548,
      "mean_token_accuracy": 0.17553244531154633,
      "num_tokens": 18852257.0,
      "step": 10865
    },
    {
      "entropy": 5.56828932762146,
      "epoch": 0.8457498541139856,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004933823849523246,
      "loss": 5.2856,
      "mean_token_accuracy": 0.18161431699991226,
      "num_tokens": 18860621.0,
      "step": 10870
    },
    {
      "entropy": 5.5673925399780275,
      "epoch": 0.8461388834857032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933757109037279,
      "loss": 5.1902,
      "mean_token_accuracy": 0.18614138811826705,
      "num_tokens": 18869390.0,
      "step": 10875
    },
    {
      "entropy": 5.555185747146607,
      "epoch": 0.8465279128574207,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933690335416274,
      "loss": 5.2911,
      "mean_token_accuracy": 0.17937638312578202,
      "num_tokens": 18877942.0,
      "step": 10880
    },
    {
      "entropy": 5.503046321868896,
      "epoch": 0.8469169422291383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933623528661245,
      "loss": 5.2779,
      "mean_token_accuracy": 0.1784745156764984,
      "num_tokens": 18886752.0,
      "step": 10885
    },
    {
      "entropy": 5.5775063037872314,
      "epoch": 0.8473059716008559,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933556688773203,
      "loss": 5.2903,
      "mean_token_accuracy": 0.17868355214595794,
      "num_tokens": 18895708.0,
      "step": 10890
    },
    {
      "entropy": 5.567748641967773,
      "epoch": 0.8476950009725734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933489815753165,
      "loss": 5.2268,
      "mean_token_accuracy": 0.18473885357379913,
      "num_tokens": 18904610.0,
      "step": 10895
    },
    {
      "entropy": 5.520393228530883,
      "epoch": 0.848084030344291,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933422909602143,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1805527240037918,
      "num_tokens": 18912886.0,
      "step": 10900
    },
    {
      "entropy": 5.569060850143432,
      "epoch": 0.8484730597160085,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004933355970321155,
      "loss": 5.3132,
      "mean_token_accuracy": 0.17902422547340394,
      "num_tokens": 18921685.0,
      "step": 10905
    },
    {
      "entropy": 5.564014148712158,
      "epoch": 0.8488620890877261,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933288997911215,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1850571870803833,
      "num_tokens": 18929843.0,
      "step": 10910
    },
    {
      "entropy": 5.497473621368409,
      "epoch": 0.8492511184594437,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933221992373338,
      "loss": 5.2166,
      "mean_token_accuracy": 0.17965936958789824,
      "num_tokens": 18938402.0,
      "step": 10915
    },
    {
      "entropy": 5.563475131988525,
      "epoch": 0.8496401478311613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933154953708544,
      "loss": 5.3496,
      "mean_token_accuracy": 0.1703747108578682,
      "num_tokens": 18947643.0,
      "step": 10920
    },
    {
      "entropy": 5.616429376602173,
      "epoch": 0.8500291772028788,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933087881917848,
      "loss": 5.2793,
      "mean_token_accuracy": 0.17879072725772857,
      "num_tokens": 18956320.0,
      "step": 10925
    },
    {
      "entropy": 5.681483268737793,
      "epoch": 0.8504182065745964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933020777002268,
      "loss": 5.2558,
      "mean_token_accuracy": 0.1776057004928589,
      "num_tokens": 18964804.0,
      "step": 10930
    },
    {
      "entropy": 5.464506435394287,
      "epoch": 0.8508072359463139,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004932953638962823,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1821805253624916,
      "num_tokens": 18973061.0,
      "step": 10935
    },
    {
      "entropy": 5.506009244918824,
      "epoch": 0.8511962653180315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000493288646780053,
      "loss": 5.1873,
      "mean_token_accuracy": 0.18650328516960143,
      "num_tokens": 18981604.0,
      "step": 10940
    },
    {
      "entropy": 5.55337553024292,
      "epoch": 0.8515852946897491,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932819263516409,
      "loss": 5.2905,
      "mean_token_accuracy": 0.17652418315410615,
      "num_tokens": 18990194.0,
      "step": 10945
    },
    {
      "entropy": 5.532214689254761,
      "epoch": 0.8519743240614667,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932752026111481,
      "loss": 5.1958,
      "mean_token_accuracy": 0.181524658203125,
      "num_tokens": 18998803.0,
      "step": 10950
    },
    {
      "entropy": 5.625107955932617,
      "epoch": 0.8523633534331843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004932684755586765,
      "loss": 5.3102,
      "mean_token_accuracy": 0.17468704134225846,
      "num_tokens": 19007247.0,
      "step": 10955
    },
    {
      "entropy": 5.5398841381073,
      "epoch": 0.8527523828049017,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932617451943282,
      "loss": 5.233,
      "mean_token_accuracy": 0.18560332357883452,
      "num_tokens": 19015436.0,
      "step": 10960
    },
    {
      "entropy": 5.594821643829346,
      "epoch": 0.8531414121766193,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932550115182053,
      "loss": 5.329,
      "mean_token_accuracy": 0.17277351170778274,
      "num_tokens": 19023673.0,
      "step": 10965
    },
    {
      "entropy": 5.534565210342407,
      "epoch": 0.8535304415483369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049324827453041,
      "loss": 5.1617,
      "mean_token_accuracy": 0.1841248795390129,
      "num_tokens": 19031714.0,
      "step": 10970
    },
    {
      "entropy": 5.432816839218139,
      "epoch": 0.8539194709200545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932415342310446,
      "loss": 5.1636,
      "mean_token_accuracy": 0.1836773708462715,
      "num_tokens": 19039910.0,
      "step": 10975
    },
    {
      "entropy": 5.564401769638062,
      "epoch": 0.8543085002917721,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004932347906202111,
      "loss": 5.2955,
      "mean_token_accuracy": 0.18685130327939986,
      "num_tokens": 19048293.0,
      "step": 10980
    },
    {
      "entropy": 5.569058752059936,
      "epoch": 0.8546975296634896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000493228043698012,
      "loss": 5.2267,
      "mean_token_accuracy": 0.17633443474769592,
      "num_tokens": 19056209.0,
      "step": 10985
    },
    {
      "entropy": 5.5361498355865475,
      "epoch": 0.8550865590352071,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932212934645498,
      "loss": 5.131,
      "mean_token_accuracy": 0.18912424743175507,
      "num_tokens": 19063974.0,
      "step": 10990
    },
    {
      "entropy": 5.55227575302124,
      "epoch": 0.8554755884069247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932145399199268,
      "loss": 5.3794,
      "mean_token_accuracy": 0.17500831484794616,
      "num_tokens": 19072531.0,
      "step": 10995
    },
    {
      "entropy": 5.653836059570312,
      "epoch": 0.8558646177786423,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932077830642455,
      "loss": 5.3062,
      "mean_token_accuracy": 0.17791387736797332,
      "num_tokens": 19081299.0,
      "step": 11000
    },
    {
      "entropy": 5.5676939487457275,
      "epoch": 0.8562536471503599,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932010228976084,
      "loss": 5.2054,
      "mean_token_accuracy": 0.19290074110031127,
      "num_tokens": 19089599.0,
      "step": 11005
    },
    {
      "entropy": 5.505162048339844,
      "epoch": 0.8566426765220774,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493194259420118,
      "loss": 5.1904,
      "mean_token_accuracy": 0.18730612546205522,
      "num_tokens": 19097910.0,
      "step": 11010
    },
    {
      "entropy": 5.47117166519165,
      "epoch": 0.857031705893795,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493187492631877,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19171023964881898,
      "num_tokens": 19105939.0,
      "step": 11015
    },
    {
      "entropy": 5.477901887893677,
      "epoch": 0.8574207352655125,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931807225329882,
      "loss": 5.2035,
      "mean_token_accuracy": 0.18485299944877626,
      "num_tokens": 19115407.0,
      "step": 11020
    },
    {
      "entropy": 5.48896336555481,
      "epoch": 0.8578097646372301,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004931739491235541,
      "loss": 5.1254,
      "mean_token_accuracy": 0.19273362755775453,
      "num_tokens": 19123715.0,
      "step": 11025
    },
    {
      "entropy": 5.63070068359375,
      "epoch": 0.8581987940089477,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004931671724036777,
      "loss": 5.3584,
      "mean_token_accuracy": 0.17230788320302964,
      "num_tokens": 19132978.0,
      "step": 11030
    },
    {
      "entropy": 5.544386148452759,
      "epoch": 0.8585878233806652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931603923734616,
      "loss": 5.304,
      "mean_token_accuracy": 0.18017314970493317,
      "num_tokens": 19141822.0,
      "step": 11035
    },
    {
      "entropy": 5.48143048286438,
      "epoch": 0.8589768527523828,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004931536090330088,
      "loss": 5.1992,
      "mean_token_accuracy": 0.17714929580688477,
      "num_tokens": 19150612.0,
      "step": 11040
    },
    {
      "entropy": 5.528379964828491,
      "epoch": 0.8593658821241004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931468223824222,
      "loss": 5.154,
      "mean_token_accuracy": 0.187422776222229,
      "num_tokens": 19159266.0,
      "step": 11045
    },
    {
      "entropy": 5.66276741027832,
      "epoch": 0.859754911495818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931400324218048,
      "loss": 5.3327,
      "mean_token_accuracy": 0.18183230310678483,
      "num_tokens": 19167111.0,
      "step": 11050
    },
    {
      "entropy": 5.5265590190887455,
      "epoch": 0.8601439408675355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931332391512597,
      "loss": 5.299,
      "mean_token_accuracy": 0.17914700508117676,
      "num_tokens": 19177152.0,
      "step": 11055
    },
    {
      "entropy": 5.481976747512817,
      "epoch": 0.860532970239253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004931264425708897,
      "loss": 5.1277,
      "mean_token_accuracy": 0.18320668041706084,
      "num_tokens": 19185692.0,
      "step": 11060
    },
    {
      "entropy": 5.541192865371704,
      "epoch": 0.8609219996109706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931196426807983,
      "loss": 5.2936,
      "mean_token_accuracy": 0.18152756094932557,
      "num_tokens": 19194620.0,
      "step": 11065
    },
    {
      "entropy": 5.561440801620483,
      "epoch": 0.8613110289826882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004931128394810884,
      "loss": 5.1748,
      "mean_token_accuracy": 0.18811147212982177,
      "num_tokens": 19202569.0,
      "step": 11070
    },
    {
      "entropy": 5.5706174850463865,
      "epoch": 0.8617000583544058,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004931060329718634,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1782485216856003,
      "num_tokens": 19211117.0,
      "step": 11075
    },
    {
      "entropy": 5.602764034271241,
      "epoch": 0.8620890877261234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004930992231532265,
      "loss": 5.3058,
      "mean_token_accuracy": 0.1797836884856224,
      "num_tokens": 19220015.0,
      "step": 11080
    },
    {
      "entropy": 5.619830846786499,
      "epoch": 0.8624781170978408,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493092410025281,
      "loss": 5.3065,
      "mean_token_accuracy": 0.17686312794685363,
      "num_tokens": 19228344.0,
      "step": 11085
    },
    {
      "entropy": 5.50443320274353,
      "epoch": 0.8628671464695584,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004930855935881302,
      "loss": 5.1899,
      "mean_token_accuracy": 0.19230535328388215,
      "num_tokens": 19236776.0,
      "step": 11090
    },
    {
      "entropy": 5.623230743408203,
      "epoch": 0.863256175841276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930787738418777,
      "loss": 5.4134,
      "mean_token_accuracy": 0.1671035945415497,
      "num_tokens": 19246107.0,
      "step": 11095
    },
    {
      "entropy": 5.607535982131958,
      "epoch": 0.8636452052129936,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930719507866269,
      "loss": 5.2901,
      "mean_token_accuracy": 0.18587073683738708,
      "num_tokens": 19254549.0,
      "step": 11100
    },
    {
      "entropy": 5.572094249725342,
      "epoch": 0.8640342345847112,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930651244224814,
      "loss": 5.2949,
      "mean_token_accuracy": 0.1801662638783455,
      "num_tokens": 19263024.0,
      "step": 11105
    },
    {
      "entropy": 5.512962055206299,
      "epoch": 0.8644232639564287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004930582947495448,
      "loss": 5.208,
      "mean_token_accuracy": 0.18059032112360002,
      "num_tokens": 19271442.0,
      "step": 11110
    },
    {
      "entropy": 5.517531108856201,
      "epoch": 0.8648122933281462,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004930514617679206,
      "loss": 5.1614,
      "mean_token_accuracy": 0.19100626111030578,
      "num_tokens": 19279869.0,
      "step": 11115
    },
    {
      "entropy": 5.5686924934387205,
      "epoch": 0.8652013226998638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930446254777125,
      "loss": 5.3574,
      "mean_token_accuracy": 0.17958442270755767,
      "num_tokens": 19288726.0,
      "step": 11120
    },
    {
      "entropy": 5.583649730682373,
      "epoch": 0.8655903520715814,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930377858790242,
      "loss": 5.3107,
      "mean_token_accuracy": 0.1751727432012558,
      "num_tokens": 19297546.0,
      "step": 11125
    },
    {
      "entropy": 5.6224347114562985,
      "epoch": 0.865979381443299,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004930309429719595,
      "loss": 5.3092,
      "mean_token_accuracy": 0.17556509673595427,
      "num_tokens": 19306060.0,
      "step": 11130
    },
    {
      "entropy": 5.566957283020019,
      "epoch": 0.8663684108150166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930240967566224,
      "loss": 5.2567,
      "mean_token_accuracy": 0.18031541854143143,
      "num_tokens": 19314759.0,
      "step": 11135
    },
    {
      "entropy": 5.527809047698975,
      "epoch": 0.8667574401867341,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930172472331167,
      "loss": 5.2052,
      "mean_token_accuracy": 0.18215182721614837,
      "num_tokens": 19322922.0,
      "step": 11140
    },
    {
      "entropy": 5.4821202754974365,
      "epoch": 0.8671464695584516,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930103944015463,
      "loss": 5.1269,
      "mean_token_accuracy": 0.18408451080322266,
      "num_tokens": 19331946.0,
      "step": 11145
    },
    {
      "entropy": 5.600930023193359,
      "epoch": 0.8675354989301692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930035382620149,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1778065100312233,
      "num_tokens": 19340512.0,
      "step": 11150
    },
    {
      "entropy": 5.529328298568726,
      "epoch": 0.8679245283018868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492996678814627,
      "loss": 5.2405,
      "mean_token_accuracy": 0.1811877131462097,
      "num_tokens": 19349135.0,
      "step": 11155
    },
    {
      "entropy": 5.535189580917359,
      "epoch": 0.8683135576736044,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004929898160594865,
      "loss": 5.3284,
      "mean_token_accuracy": 0.17958931028842925,
      "num_tokens": 19358593.0,
      "step": 11160
    },
    {
      "entropy": 5.6465692043304445,
      "epoch": 0.8687025870453219,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004929829499966974,
      "loss": 5.3534,
      "mean_token_accuracy": 0.17606135904788972,
      "num_tokens": 19367366.0,
      "step": 11165
    },
    {
      "entropy": 5.592950296401978,
      "epoch": 0.8690916164170395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004929760806263641,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1734875425696373,
      "num_tokens": 19376872.0,
      "step": 11170
    },
    {
      "entropy": 5.615838670730591,
      "epoch": 0.8694806457887571,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929692079485906,
      "loss": 5.3295,
      "mean_token_accuracy": 0.17325059622526168,
      "num_tokens": 19385582.0,
      "step": 11175
    },
    {
      "entropy": 5.5427587032318115,
      "epoch": 0.8698696751604746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929623319634814,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1821684569120407,
      "num_tokens": 19394102.0,
      "step": 11180
    },
    {
      "entropy": 5.562895441055298,
      "epoch": 0.8702587045321922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929554526711407,
      "loss": 5.2562,
      "mean_token_accuracy": 0.17677873075008393,
      "num_tokens": 19403170.0,
      "step": 11185
    },
    {
      "entropy": 5.5495288372039795,
      "epoch": 0.8706477339039097,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929485700716729,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1873213306069374,
      "num_tokens": 19412396.0,
      "step": 11190
    },
    {
      "entropy": 5.550656509399414,
      "epoch": 0.8710367632756273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929416841651824,
      "loss": 5.2,
      "mean_token_accuracy": 0.18274609446525575,
      "num_tokens": 19421247.0,
      "step": 11195
    },
    {
      "entropy": 5.526799917221069,
      "epoch": 0.8714257926473449,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004929347949517739,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1788181945681572,
      "num_tokens": 19429412.0,
      "step": 11200
    },
    {
      "entropy": 5.582285690307617,
      "epoch": 0.8718148220190625,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929279024315516,
      "loss": 5.261,
      "mean_token_accuracy": 0.17433381527662278,
      "num_tokens": 19437551.0,
      "step": 11205
    },
    {
      "entropy": 5.462672519683838,
      "epoch": 0.87220385139078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929210066046204,
      "loss": 5.0975,
      "mean_token_accuracy": 0.18734758198261262,
      "num_tokens": 19446356.0,
      "step": 11210
    },
    {
      "entropy": 5.46998782157898,
      "epoch": 0.8725928807624975,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004929141074710847,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19531500339508057,
      "num_tokens": 19454142.0,
      "step": 11215
    },
    {
      "entropy": 5.617652940750122,
      "epoch": 0.8729819101342151,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004929072050310491,
      "loss": 5.3743,
      "mean_token_accuracy": 0.1781470462679863,
      "num_tokens": 19462849.0,
      "step": 11220
    },
    {
      "entropy": 5.5333860397338865,
      "epoch": 0.8733709395059327,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929002992846188,
      "loss": 5.1635,
      "mean_token_accuracy": 0.18983460515737532,
      "num_tokens": 19471287.0,
      "step": 11225
    },
    {
      "entropy": 5.52715311050415,
      "epoch": 0.8737599688776503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004928933902318981,
      "loss": 5.2262,
      "mean_token_accuracy": 0.18716436475515366,
      "num_tokens": 19480661.0,
      "step": 11230
    },
    {
      "entropy": 5.476283359527588,
      "epoch": 0.8741489982493679,
      "grad_norm": 1.125,
      "learning_rate": 0.000492886477872992,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1878957837820053,
      "num_tokens": 19489716.0,
      "step": 11235
    },
    {
      "entropy": 5.516184234619141,
      "epoch": 0.8745380276210853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004928795622080054,
      "loss": 5.271,
      "mean_token_accuracy": 0.17724187672138214,
      "num_tokens": 19499019.0,
      "step": 11240
    },
    {
      "entropy": 5.5679398536682125,
      "epoch": 0.8749270569928029,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004928726432370434,
      "loss": 5.2028,
      "mean_token_accuracy": 0.18291952461004257,
      "num_tokens": 19507281.0,
      "step": 11245
    },
    {
      "entropy": 5.585212564468383,
      "epoch": 0.8753160863645205,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928657209602107,
      "loss": 5.2724,
      "mean_token_accuracy": 0.18346301317214966,
      "num_tokens": 19516196.0,
      "step": 11250
    },
    {
      "entropy": 5.568840551376343,
      "epoch": 0.8757051157362381,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004928587953776125,
      "loss": 5.301,
      "mean_token_accuracy": 0.1767169713973999,
      "num_tokens": 19523966.0,
      "step": 11255
    },
    {
      "entropy": 5.544528913497925,
      "epoch": 0.8760941451079557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928518664893538,
      "loss": 5.1989,
      "mean_token_accuracy": 0.18428248316049575,
      "num_tokens": 19532848.0,
      "step": 11260
    },
    {
      "entropy": 5.5038737773895265,
      "epoch": 0.8764831744796732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928449342955397,
      "loss": 5.1579,
      "mean_token_accuracy": 0.18306698352098466,
      "num_tokens": 19540929.0,
      "step": 11265
    },
    {
      "entropy": 5.5189353942871096,
      "epoch": 0.8768722038513908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928379987962756,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1850006937980652,
      "num_tokens": 19549628.0,
      "step": 11270
    },
    {
      "entropy": 5.520605087280273,
      "epoch": 0.8772612332231083,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004928310599916666,
      "loss": 5.2004,
      "mean_token_accuracy": 0.17930981069803237,
      "num_tokens": 19559002.0,
      "step": 11275
    },
    {
      "entropy": 5.522569465637207,
      "epoch": 0.8776502625948259,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928241178818178,
      "loss": 5.211,
      "mean_token_accuracy": 0.185836298763752,
      "num_tokens": 19567757.0,
      "step": 11280
    },
    {
      "entropy": 5.536190605163574,
      "epoch": 0.8780392919665435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928171724668349,
      "loss": 5.1391,
      "mean_token_accuracy": 0.19180002361536025,
      "num_tokens": 19576595.0,
      "step": 11285
    },
    {
      "entropy": 5.529078960418701,
      "epoch": 0.878428321338261,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928102237468229,
      "loss": 5.2671,
      "mean_token_accuracy": 0.17619390189647674,
      "num_tokens": 19585237.0,
      "step": 11290
    },
    {
      "entropy": 5.504305791854859,
      "epoch": 0.8788173507099786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928032717218876,
      "loss": 5.1496,
      "mean_token_accuracy": 0.18973541706800462,
      "num_tokens": 19594474.0,
      "step": 11295
    },
    {
      "entropy": 5.515873193740845,
      "epoch": 0.8792063800816962,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927963163921343,
      "loss": 5.1986,
      "mean_token_accuracy": 0.18612133711576462,
      "num_tokens": 19603121.0,
      "step": 11300
    },
    {
      "entropy": 5.548409557342529,
      "epoch": 0.8795954094534137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004927893577576685,
      "loss": 5.2416,
      "mean_token_accuracy": 0.178195983171463,
      "num_tokens": 19611820.0,
      "step": 11305
    },
    {
      "entropy": 5.574872350692749,
      "epoch": 0.8799844388251313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004927823958185959,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1796655535697937,
      "num_tokens": 19620265.0,
      "step": 11310
    },
    {
      "entropy": 5.457470178604126,
      "epoch": 0.8803734681968488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492775430575022,
      "loss": 5.0919,
      "mean_token_accuracy": 0.18922037333250047,
      "num_tokens": 19628847.0,
      "step": 11315
    },
    {
      "entropy": 5.51951265335083,
      "epoch": 0.8807624975685664,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004927684620270527,
      "loss": 5.2556,
      "mean_token_accuracy": 0.18119800239801406,
      "num_tokens": 19637636.0,
      "step": 11320
    },
    {
      "entropy": 5.527203226089478,
      "epoch": 0.881151526940284,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927614901747935,
      "loss": 5.2369,
      "mean_token_accuracy": 0.1773467481136322,
      "num_tokens": 19646663.0,
      "step": 11325
    },
    {
      "entropy": 5.531182050704956,
      "epoch": 0.8815405563120016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004927545150183503,
      "loss": 5.2139,
      "mean_token_accuracy": 0.18406083434820175,
      "num_tokens": 19655382.0,
      "step": 11330
    },
    {
      "entropy": 5.52825083732605,
      "epoch": 0.8819295856837192,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492747536557829,
      "loss": 5.2201,
      "mean_token_accuracy": 0.18831291794776917,
      "num_tokens": 19663851.0,
      "step": 11335
    },
    {
      "entropy": 5.479758596420288,
      "epoch": 0.8823186150554367,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004927405547933353,
      "loss": 5.1738,
      "mean_token_accuracy": 0.1878650203347206,
      "num_tokens": 19672813.0,
      "step": 11340
    },
    {
      "entropy": 5.5052190780639645,
      "epoch": 0.8827076444271542,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927335697249753,
      "loss": 5.1695,
      "mean_token_accuracy": 0.18782930672168732,
      "num_tokens": 19682240.0,
      "step": 11345
    },
    {
      "entropy": 5.569630765914917,
      "epoch": 0.8830966737988718,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927265813528549,
      "loss": 5.2212,
      "mean_token_accuracy": 0.18088840693235397,
      "num_tokens": 19691080.0,
      "step": 11350
    },
    {
      "entropy": 5.5099883556365965,
      "epoch": 0.8834857031705894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927195896770804,
      "loss": 5.1779,
      "mean_token_accuracy": 0.1918042242527008,
      "num_tokens": 19700067.0,
      "step": 11355
    },
    {
      "entropy": 5.529403257369995,
      "epoch": 0.883874732542307,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004927125946977574,
      "loss": 5.1834,
      "mean_token_accuracy": 0.18812153488397598,
      "num_tokens": 19708236.0,
      "step": 11360
    },
    {
      "entropy": 5.560075139999389,
      "epoch": 0.8842637619140246,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927055964149923,
      "loss": 5.3031,
      "mean_token_accuracy": 0.18180075138807297,
      "num_tokens": 19717225.0,
      "step": 11365
    },
    {
      "entropy": 5.522982168197632,
      "epoch": 0.884652791285742,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926985948288914,
      "loss": 5.2625,
      "mean_token_accuracy": 0.1891829937696457,
      "num_tokens": 19725169.0,
      "step": 11370
    },
    {
      "entropy": 5.536978340148925,
      "epoch": 0.8850418206574596,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926915899395608,
      "loss": 5.2319,
      "mean_token_accuracy": 0.1837349683046341,
      "num_tokens": 19733937.0,
      "step": 11375
    },
    {
      "entropy": 5.510504961013794,
      "epoch": 0.8854308500291772,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926845817471068,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19284587651491164,
      "num_tokens": 19742661.0,
      "step": 11380
    },
    {
      "entropy": 5.487407159805298,
      "epoch": 0.8858198794008948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004926775702516358,
      "loss": 5.244,
      "mean_token_accuracy": 0.18677958250045776,
      "num_tokens": 19751314.0,
      "step": 11385
    },
    {
      "entropy": 5.523621511459351,
      "epoch": 0.8862089087726124,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004926705554532541,
      "loss": 5.1774,
      "mean_token_accuracy": 0.1917124480009079,
      "num_tokens": 19760144.0,
      "step": 11390
    },
    {
      "entropy": 5.505693769454956,
      "epoch": 0.8865979381443299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926635373520682,
      "loss": 5.1946,
      "mean_token_accuracy": 0.18730054050683975,
      "num_tokens": 19768574.0,
      "step": 11395
    },
    {
      "entropy": 5.585032606124878,
      "epoch": 0.8869869675160474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926565159481845,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1737936854362488,
      "num_tokens": 19778383.0,
      "step": 11400
    },
    {
      "entropy": 5.551654720306397,
      "epoch": 0.887375996887765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926494912417097,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1848152220249176,
      "num_tokens": 19787707.0,
      "step": 11405
    },
    {
      "entropy": 5.531039810180664,
      "epoch": 0.8877650262594826,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004926424632327503,
      "loss": 5.1094,
      "mean_token_accuracy": 0.19663903564214708,
      "num_tokens": 19795909.0,
      "step": 11410
    },
    {
      "entropy": 5.4795300483703615,
      "epoch": 0.8881540556312002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926354319214129,
      "loss": 5.1733,
      "mean_token_accuracy": 0.18454110771417617,
      "num_tokens": 19804094.0,
      "step": 11415
    },
    {
      "entropy": 5.460201930999756,
      "epoch": 0.8885430850029177,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926283973078041,
      "loss": 5.0873,
      "mean_token_accuracy": 0.18923551142215728,
      "num_tokens": 19813061.0,
      "step": 11420
    },
    {
      "entropy": 5.472792053222657,
      "epoch": 0.8889321143746353,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004926213593920309,
      "loss": 5.101,
      "mean_token_accuracy": 0.1912500724196434,
      "num_tokens": 19821432.0,
      "step": 11425
    },
    {
      "entropy": 5.5045366287231445,
      "epoch": 0.8893211437463528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004926143181742,
      "loss": 5.2396,
      "mean_token_accuracy": 0.18476988971233368,
      "num_tokens": 19830278.0,
      "step": 11430
    },
    {
      "entropy": 5.429034233093262,
      "epoch": 0.8897101731180704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004926072736544181,
      "loss": 5.1641,
      "mean_token_accuracy": 0.18787843286991118,
      "num_tokens": 19838872.0,
      "step": 11435
    },
    {
      "entropy": 5.632330846786499,
      "epoch": 0.890099202489788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926002258327922,
      "loss": 5.2996,
      "mean_token_accuracy": 0.17796659767627715,
      "num_tokens": 19848032.0,
      "step": 11440
    },
    {
      "entropy": 5.509681034088135,
      "epoch": 0.8904882318615055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925931747094292,
      "loss": 5.1606,
      "mean_token_accuracy": 0.19246817976236344,
      "num_tokens": 19856711.0,
      "step": 11445
    },
    {
      "entropy": 5.380377531051636,
      "epoch": 0.8908772612332231,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925861202844361,
      "loss": 5.1387,
      "mean_token_accuracy": 0.18535495549440384,
      "num_tokens": 19865643.0,
      "step": 11450
    },
    {
      "entropy": 5.535952854156494,
      "epoch": 0.8912662906049407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049257906255792,
      "loss": 5.1655,
      "mean_token_accuracy": 0.18515115231275558,
      "num_tokens": 19874029.0,
      "step": 11455
    },
    {
      "entropy": 5.645742321014405,
      "epoch": 0.8916553199766583,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492572001529988,
      "loss": 5.2919,
      "mean_token_accuracy": 0.17810761630535127,
      "num_tokens": 19883347.0,
      "step": 11460
    },
    {
      "entropy": 5.544638681411743,
      "epoch": 0.8920443493483758,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004925649372007469,
      "loss": 5.209,
      "mean_token_accuracy": 0.18567140847444535,
      "num_tokens": 19891822.0,
      "step": 11465
    },
    {
      "entropy": 5.475942897796631,
      "epoch": 0.8924333787200933,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925578695703045,
      "loss": 5.1903,
      "mean_token_accuracy": 0.18323710560798645,
      "num_tokens": 19899899.0,
      "step": 11470
    },
    {
      "entropy": 5.577597141265869,
      "epoch": 0.8928224080918109,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925507986387676,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18126303404569627,
      "num_tokens": 19908189.0,
      "step": 11475
    },
    {
      "entropy": 5.566147327423096,
      "epoch": 0.8932114374635285,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004925437244062436,
      "loss": 5.2064,
      "mean_token_accuracy": 0.18268861174583434,
      "num_tokens": 19916166.0,
      "step": 11480
    },
    {
      "entropy": 5.528143358230591,
      "epoch": 0.8936004668352461,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004925366468728397,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18684028089046478,
      "num_tokens": 19924222.0,
      "step": 11485
    },
    {
      "entropy": 5.52174859046936,
      "epoch": 0.8939894962069637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004925295660386635,
      "loss": 5.3024,
      "mean_token_accuracy": 0.17870192229747772,
      "num_tokens": 19931925.0,
      "step": 11490
    },
    {
      "entropy": 5.583100175857544,
      "epoch": 0.8943785255786811,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925224819038224,
      "loss": 5.2565,
      "mean_token_accuracy": 0.18239305317401885,
      "num_tokens": 19940432.0,
      "step": 11495
    },
    {
      "entropy": 5.591109132766723,
      "epoch": 0.8947675549503987,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004925153944684239,
      "loss": 5.2367,
      "mean_token_accuracy": 0.18197857439517975,
      "num_tokens": 19948013.0,
      "step": 11500
    },
    {
      "entropy": 5.552106237411499,
      "epoch": 0.8951565843221163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925083037325754,
      "loss": 5.1803,
      "mean_token_accuracy": 0.18869636058807374,
      "num_tokens": 19956161.0,
      "step": 11505
    },
    {
      "entropy": 5.570696830749512,
      "epoch": 0.8955456136938339,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925012096963847,
      "loss": 5.1691,
      "mean_token_accuracy": 0.19198454171419144,
      "num_tokens": 19964533.0,
      "step": 11510
    },
    {
      "entropy": 5.597382402420044,
      "epoch": 0.8959346430655515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924941123599593,
      "loss": 5.2119,
      "mean_token_accuracy": 0.18112147301435472,
      "num_tokens": 19973915.0,
      "step": 11515
    },
    {
      "entropy": 5.461422252655029,
      "epoch": 0.896323672437269,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004924870117234069,
      "loss": 5.2152,
      "mean_token_accuracy": 0.18430526405572892,
      "num_tokens": 19983422.0,
      "step": 11520
    },
    {
      "entropy": 5.562084293365478,
      "epoch": 0.8967127018089865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004924799077868353,
      "loss": 5.2086,
      "mean_token_accuracy": 0.18218649327754974,
      "num_tokens": 19992185.0,
      "step": 11525
    },
    {
      "entropy": 5.545620536804199,
      "epoch": 0.8971017311807041,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004924728005503522,
      "loss": 5.0783,
      "mean_token_accuracy": 0.18648597598075867,
      "num_tokens": 20000735.0,
      "step": 11530
    },
    {
      "entropy": 5.496080160140991,
      "epoch": 0.8974907605524217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924656900140655,
      "loss": 5.2555,
      "mean_token_accuracy": 0.18169461488723754,
      "num_tokens": 20009758.0,
      "step": 11535
    },
    {
      "entropy": 5.5143226146697994,
      "epoch": 0.8978797899241393,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004924585761780831,
      "loss": 5.1629,
      "mean_token_accuracy": 0.17769721448421477,
      "num_tokens": 20018369.0,
      "step": 11540
    },
    {
      "entropy": 5.59825382232666,
      "epoch": 0.8982688192958569,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492451459042513,
      "loss": 5.3005,
      "mean_token_accuracy": 0.17755273431539537,
      "num_tokens": 20027469.0,
      "step": 11545
    },
    {
      "entropy": 5.606056261062622,
      "epoch": 0.8986578486675744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492444338607463,
      "loss": 5.2245,
      "mean_token_accuracy": 0.17962005585432053,
      "num_tokens": 20035504.0,
      "step": 11550
    },
    {
      "entropy": 5.508863162994385,
      "epoch": 0.899046878039292,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924372148730413,
      "loss": 5.1739,
      "mean_token_accuracy": 0.19141211062669755,
      "num_tokens": 20045272.0,
      "step": 11555
    },
    {
      "entropy": 5.5456148147583,
      "epoch": 0.8994359074110095,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492430087839356,
      "loss": 5.1627,
      "mean_token_accuracy": 0.1815532147884369,
      "num_tokens": 20054041.0,
      "step": 11560
    },
    {
      "entropy": 5.543204641342163,
      "epoch": 0.8998249367827271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924229575065152,
      "loss": 5.1177,
      "mean_token_accuracy": 0.18746805787086487,
      "num_tokens": 20062337.0,
      "step": 11565
    },
    {
      "entropy": 5.539514589309692,
      "epoch": 0.9002139661544447,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492415823874627,
      "loss": 5.2739,
      "mean_token_accuracy": 0.18309991359710692,
      "num_tokens": 20070767.0,
      "step": 11570
    },
    {
      "entropy": 5.591424417495728,
      "epoch": 0.9006029955261622,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924086869437998,
      "loss": 5.2098,
      "mean_token_accuracy": 0.18047481924295425,
      "num_tokens": 20079791.0,
      "step": 11575
    },
    {
      "entropy": 5.481707239151001,
      "epoch": 0.9009920248978798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004924015467141417,
      "loss": 5.1125,
      "mean_token_accuracy": 0.19219427406787873,
      "num_tokens": 20088458.0,
      "step": 11580
    },
    {
      "entropy": 5.493997955322266,
      "epoch": 0.9013810542695974,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923944031857613,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1861426368355751,
      "num_tokens": 20097559.0,
      "step": 11585
    },
    {
      "entropy": 5.4814629554748535,
      "epoch": 0.901770083641315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004923872563587668,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19254623651504515,
      "num_tokens": 20105403.0,
      "step": 11590
    },
    {
      "entropy": 5.466583919525147,
      "epoch": 0.9021591130130325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004923801062332666,
      "loss": 5.2492,
      "mean_token_accuracy": 0.18395482003688812,
      "num_tokens": 20115073.0,
      "step": 11595
    },
    {
      "entropy": 5.479728555679321,
      "epoch": 0.90254814238475,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923729528093694,
      "loss": 5.2026,
      "mean_token_accuracy": 0.1841553971171379,
      "num_tokens": 20124064.0,
      "step": 11600
    },
    {
      "entropy": 5.583331441879272,
      "epoch": 0.9029371717564676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923657960871836,
      "loss": 5.2442,
      "mean_token_accuracy": 0.19219164103269576,
      "num_tokens": 20133364.0,
      "step": 11605
    },
    {
      "entropy": 5.494344425201416,
      "epoch": 0.9033262011281852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923586360668178,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1866675928235054,
      "num_tokens": 20142099.0,
      "step": 11610
    },
    {
      "entropy": 5.4702314853668215,
      "epoch": 0.9037152304999028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923514727483807,
      "loss": 5.142,
      "mean_token_accuracy": 0.18838441967964173,
      "num_tokens": 20150421.0,
      "step": 11615
    },
    {
      "entropy": 5.574986743927002,
      "epoch": 0.9041042598716204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492344306131981,
      "loss": 5.3034,
      "mean_token_accuracy": 0.18428612500429153,
      "num_tokens": 20159294.0,
      "step": 11620
    },
    {
      "entropy": 5.6521648406982425,
      "epoch": 0.9044932892433378,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004923371362177272,
      "loss": 5.3508,
      "mean_token_accuracy": 0.17535435110330583,
      "num_tokens": 20168082.0,
      "step": 11625
    },
    {
      "entropy": 5.544291162490845,
      "epoch": 0.9048823186150554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923299630057284,
      "loss": 5.1912,
      "mean_token_accuracy": 0.18422658890485763,
      "num_tokens": 20176933.0,
      "step": 11630
    },
    {
      "entropy": 5.5124529838562015,
      "epoch": 0.905271347986773,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004923227864960934,
      "loss": 5.2498,
      "mean_token_accuracy": 0.18039043694734574,
      "num_tokens": 20185507.0,
      "step": 11635
    },
    {
      "entropy": 5.534951019287109,
      "epoch": 0.9056603773584906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492315606688931,
      "loss": 5.2419,
      "mean_token_accuracy": 0.18117894232273102,
      "num_tokens": 20195206.0,
      "step": 11640
    },
    {
      "entropy": 5.536805629730225,
      "epoch": 0.9060494067302082,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923084235843501,
      "loss": 5.1733,
      "mean_token_accuracy": 0.1895822137594223,
      "num_tokens": 20203290.0,
      "step": 11645
    },
    {
      "entropy": 5.5393036842346195,
      "epoch": 0.9064384361019256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004923012371824598,
      "loss": 5.2688,
      "mean_token_accuracy": 0.18187354505062103,
      "num_tokens": 20211808.0,
      "step": 11650
    },
    {
      "entropy": 5.572345542907715,
      "epoch": 0.9068274654736432,
      "grad_norm": 1.0,
      "learning_rate": 0.000492294047483369,
      "loss": 5.2667,
      "mean_token_accuracy": 0.17910204082727432,
      "num_tokens": 20219980.0,
      "step": 11655
    },
    {
      "entropy": 5.5504203796386715,
      "epoch": 0.9072164948453608,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004922868544871869,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1864182770252228,
      "num_tokens": 20227880.0,
      "step": 11660
    },
    {
      "entropy": 5.497947359085083,
      "epoch": 0.9076055242170784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922796581940227,
      "loss": 5.229,
      "mean_token_accuracy": 0.17243314981460572,
      "num_tokens": 20237023.0,
      "step": 11665
    },
    {
      "entropy": 5.611249113082886,
      "epoch": 0.907994553588796,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004922724586039855,
      "loss": 5.2695,
      "mean_token_accuracy": 0.18479921221733092,
      "num_tokens": 20246046.0,
      "step": 11670
    },
    {
      "entropy": 5.616004896163941,
      "epoch": 0.9083835829605135,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004922652557171846,
      "loss": 5.2141,
      "mean_token_accuracy": 0.1874903380870819,
      "num_tokens": 20254392.0,
      "step": 11675
    },
    {
      "entropy": 5.444818639755249,
      "epoch": 0.9087726123322311,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922580495337292,
      "loss": 5.1491,
      "mean_token_accuracy": 0.19083634465932847,
      "num_tokens": 20263529.0,
      "step": 11680
    },
    {
      "entropy": 5.488363361358642,
      "epoch": 0.9091616417039486,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004922508400537288,
      "loss": 5.2219,
      "mean_token_accuracy": 0.18348418176174164,
      "num_tokens": 20272226.0,
      "step": 11685
    },
    {
      "entropy": 5.542445135116577,
      "epoch": 0.9095506710756662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922436272772926,
      "loss": 5.3149,
      "mean_token_accuracy": 0.17373220771551132,
      "num_tokens": 20281084.0,
      "step": 11690
    },
    {
      "entropy": 5.553008699417115,
      "epoch": 0.9099397004473838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004922364112045301,
      "loss": 5.3019,
      "mean_token_accuracy": 0.18025858253240584,
      "num_tokens": 20290137.0,
      "step": 11695
    },
    {
      "entropy": 5.545133972167969,
      "epoch": 0.9103287298191013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492229191835551,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19323284775018693,
      "num_tokens": 20298305.0,
      "step": 11700
    },
    {
      "entropy": 5.530434894561767,
      "epoch": 0.9107177591908189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004922219691704645,
      "loss": 5.2398,
      "mean_token_accuracy": 0.18429169952869415,
      "num_tokens": 20306289.0,
      "step": 11705
    },
    {
      "entropy": 5.603179264068603,
      "epoch": 0.9111067885625365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922147432093805,
      "loss": 5.3343,
      "mean_token_accuracy": 0.1802520915865898,
      "num_tokens": 20314826.0,
      "step": 11710
    },
    {
      "entropy": 5.663875675201416,
      "epoch": 0.911495817934254,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922075139524083,
      "loss": 5.3789,
      "mean_token_accuracy": 0.1761022612452507,
      "num_tokens": 20323244.0,
      "step": 11715
    },
    {
      "entropy": 5.552782726287842,
      "epoch": 0.9118848473059716,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492200281399658,
      "loss": 5.2543,
      "mean_token_accuracy": 0.17747358977794647,
      "num_tokens": 20331943.0,
      "step": 11720
    },
    {
      "entropy": 5.545907878875733,
      "epoch": 0.9122738766776891,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492193045551239,
      "loss": 5.2501,
      "mean_token_accuracy": 0.18026496320962906,
      "num_tokens": 20339940.0,
      "step": 11725
    },
    {
      "entropy": 5.582332372665405,
      "epoch": 0.9126629060494067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921858064072612,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19396570026874543,
      "num_tokens": 20348556.0,
      "step": 11730
    },
    {
      "entropy": 5.5467949390411375,
      "epoch": 0.9130519354211243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004921785639678347,
      "loss": 5.284,
      "mean_token_accuracy": 0.1855788990855217,
      "num_tokens": 20356909.0,
      "step": 11735
    },
    {
      "entropy": 5.522295665740967,
      "epoch": 0.9134409647928419,
      "grad_norm": 1.0,
      "learning_rate": 0.000492171318233069,
      "loss": 5.1802,
      "mean_token_accuracy": 0.18770511597394943,
      "num_tokens": 20365496.0,
      "step": 11740
    },
    {
      "entropy": 5.512374591827393,
      "epoch": 0.9138299941645595,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004921640692030742,
      "loss": 5.192,
      "mean_token_accuracy": 0.18125805705785752,
      "num_tokens": 20374763.0,
      "step": 11745
    },
    {
      "entropy": 5.5617804527282715,
      "epoch": 0.914219023536277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921568168779603,
      "loss": 5.2463,
      "mean_token_accuracy": 0.18501508980989456,
      "num_tokens": 20383861.0,
      "step": 11750
    },
    {
      "entropy": 5.54015588760376,
      "epoch": 0.9146080529079945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921495612578374,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19265497624874114,
      "num_tokens": 20392004.0,
      "step": 11755
    },
    {
      "entropy": 5.566636085510254,
      "epoch": 0.9149970822797121,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921423023428156,
      "loss": 5.2262,
      "mean_token_accuracy": 0.17809344977140426,
      "num_tokens": 20401078.0,
      "step": 11760
    },
    {
      "entropy": 5.676298666000366,
      "epoch": 0.9153861116514297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921350401330049,
      "loss": 5.363,
      "mean_token_accuracy": 0.17461358308792113,
      "num_tokens": 20409347.0,
      "step": 11765
    },
    {
      "entropy": 5.572146987915039,
      "epoch": 0.9157751410231473,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921277746285155,
      "loss": 5.181,
      "mean_token_accuracy": 0.18716677725315095,
      "num_tokens": 20417862.0,
      "step": 11770
    },
    {
      "entropy": 5.502587556838989,
      "epoch": 0.9161641703948649,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921205058294579,
      "loss": 5.1463,
      "mean_token_accuracy": 0.18311374336481095,
      "num_tokens": 20426556.0,
      "step": 11775
    },
    {
      "entropy": 5.554633855819702,
      "epoch": 0.9165531997665823,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492113233735942,
      "loss": 5.274,
      "mean_token_accuracy": 0.17725925594568254,
      "num_tokens": 20435336.0,
      "step": 11780
    },
    {
      "entropy": 5.576182079315186,
      "epoch": 0.9169422291382999,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921059583480785,
      "loss": 5.2452,
      "mean_token_accuracy": 0.18238117545843124,
      "num_tokens": 20443884.0,
      "step": 11785
    },
    {
      "entropy": 5.522510528564453,
      "epoch": 0.9173312585100175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004920986796659775,
      "loss": 5.1359,
      "mean_token_accuracy": 0.19682417213916778,
      "num_tokens": 20451953.0,
      "step": 11790
    },
    {
      "entropy": 5.54691219329834,
      "epoch": 0.9177202878817351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920913976897498,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1833428680896759,
      "num_tokens": 20461480.0,
      "step": 11795
    },
    {
      "entropy": 5.584488487243652,
      "epoch": 0.9181093172534527,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920841124195055,
      "loss": 5.2438,
      "mean_token_accuracy": 0.1795766457915306,
      "num_tokens": 20470324.0,
      "step": 11800
    },
    {
      "entropy": 5.589950799942017,
      "epoch": 0.9184983466251702,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004920768238553554,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1787121295928955,
      "num_tokens": 20478795.0,
      "step": 11805
    },
    {
      "entropy": 5.560394668579102,
      "epoch": 0.9188873759968877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920695319974099,
      "loss": 5.1903,
      "mean_token_accuracy": 0.18176875114440919,
      "num_tokens": 20487922.0,
      "step": 11810
    },
    {
      "entropy": 5.5067990779876705,
      "epoch": 0.9192764053686053,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004920622368457798,
      "loss": 5.1975,
      "mean_token_accuracy": 0.1936303734779358,
      "num_tokens": 20495883.0,
      "step": 11815
    },
    {
      "entropy": 5.522235488891601,
      "epoch": 0.9196654347403229,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004920549384005759,
      "loss": 5.2036,
      "mean_token_accuracy": 0.18583762496709824,
      "num_tokens": 20504403.0,
      "step": 11820
    },
    {
      "entropy": 5.495119857788086,
      "epoch": 0.9200544641120405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920476366619086,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19050845205783845,
      "num_tokens": 20512603.0,
      "step": 11825
    },
    {
      "entropy": 5.473958349227905,
      "epoch": 0.920443493483758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492040331629889,
      "loss": 5.1275,
      "mean_token_accuracy": 0.18999258577823638,
      "num_tokens": 20520764.0,
      "step": 11830
    },
    {
      "entropy": 5.463704442977905,
      "epoch": 0.9208325228554756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920330233046277,
      "loss": 5.0985,
      "mean_token_accuracy": 0.1880857989192009,
      "num_tokens": 20529648.0,
      "step": 11835
    },
    {
      "entropy": 5.51321325302124,
      "epoch": 0.9212215522271932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920257116862359,
      "loss": 5.163,
      "mean_token_accuracy": 0.1889480784535408,
      "num_tokens": 20538489.0,
      "step": 11840
    },
    {
      "entropy": 5.546636533737183,
      "epoch": 0.9216105815989107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004920183967748241,
      "loss": 5.1991,
      "mean_token_accuracy": 0.1878497987985611,
      "num_tokens": 20548042.0,
      "step": 11845
    },
    {
      "entropy": 5.436558675765991,
      "epoch": 0.9219996109706283,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920110785705037,
      "loss": 5.1087,
      "mean_token_accuracy": 0.1985417440533638,
      "num_tokens": 20556559.0,
      "step": 11850
    },
    {
      "entropy": 5.576743412017822,
      "epoch": 0.9223886403423458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920037570733857,
      "loss": 5.3405,
      "mean_token_accuracy": 0.1739198684692383,
      "num_tokens": 20566162.0,
      "step": 11855
    },
    {
      "entropy": 5.5404211521148685,
      "epoch": 0.9227776697140634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004919964322835809,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1928445130586624,
      "num_tokens": 20574575.0,
      "step": 11860
    },
    {
      "entropy": 5.4385003566741945,
      "epoch": 0.923166699085781,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004919891042012006,
      "loss": 5.1272,
      "mean_token_accuracy": 0.18770445883274078,
      "num_tokens": 20583394.0,
      "step": 11865
    },
    {
      "entropy": 5.483497285842896,
      "epoch": 0.9235557284574986,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919817728263562,
      "loss": 5.2105,
      "mean_token_accuracy": 0.18159369379281998,
      "num_tokens": 20592413.0,
      "step": 11870
    },
    {
      "entropy": 5.569636392593384,
      "epoch": 0.9239447578292161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919744381591586,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1906165897846222,
      "num_tokens": 20601403.0,
      "step": 11875
    },
    {
      "entropy": 5.538726663589477,
      "epoch": 0.9243337872009336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919671001997193,
      "loss": 5.1762,
      "mean_token_accuracy": 0.18856113404035568,
      "num_tokens": 20610342.0,
      "step": 11880
    },
    {
      "entropy": 5.490780687332153,
      "epoch": 0.9247228165726512,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004919597589481497,
      "loss": 5.2173,
      "mean_token_accuracy": 0.18497272133827208,
      "num_tokens": 20619120.0,
      "step": 11885
    },
    {
      "entropy": 5.466904354095459,
      "epoch": 0.9251118459443688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491952414404561,
      "loss": 5.1606,
      "mean_token_accuracy": 0.18369911164045333,
      "num_tokens": 20628713.0,
      "step": 11890
    },
    {
      "entropy": 5.580748081207275,
      "epoch": 0.9255008753160864,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919450665690646,
      "loss": 5.1597,
      "mean_token_accuracy": 0.18222711682319642,
      "num_tokens": 20636338.0,
      "step": 11895
    },
    {
      "entropy": 5.579671859741211,
      "epoch": 0.925889904687804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919377154417724,
      "loss": 5.1742,
      "mean_token_accuracy": 0.18615632355213166,
      "num_tokens": 20644186.0,
      "step": 11900
    },
    {
      "entropy": 5.51911997795105,
      "epoch": 0.9262789340595214,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004919303610227954,
      "loss": 5.2634,
      "mean_token_accuracy": 0.18841380923986434,
      "num_tokens": 20653672.0,
      "step": 11905
    },
    {
      "entropy": 5.488134717941284,
      "epoch": 0.926667963431239,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919230033122457,
      "loss": 5.1478,
      "mean_token_accuracy": 0.1854111894965172,
      "num_tokens": 20662936.0,
      "step": 11910
    },
    {
      "entropy": 5.518908405303955,
      "epoch": 0.9270569928029566,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004919156423102345,
      "loss": 5.2568,
      "mean_token_accuracy": 0.18243345767259597,
      "num_tokens": 20671265.0,
      "step": 11915
    },
    {
      "entropy": 5.613058280944824,
      "epoch": 0.9274460221746742,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004919082780168736,
      "loss": 5.2967,
      "mean_token_accuracy": 0.17161622494459153,
      "num_tokens": 20680049.0,
      "step": 11920
    },
    {
      "entropy": 5.517641925811768,
      "epoch": 0.9278350515463918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919009104322751,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1864025816321373,
      "num_tokens": 20688411.0,
      "step": 11925
    },
    {
      "entropy": 5.548143529891968,
      "epoch": 0.9282240809181093,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918935395565503,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1809676021337509,
      "num_tokens": 20696012.0,
      "step": 11930
    },
    {
      "entropy": 5.503843879699707,
      "epoch": 0.9286131102898268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918861653898113,
      "loss": 5.1957,
      "mean_token_accuracy": 0.18054557740688323,
      "num_tokens": 20704135.0,
      "step": 11935
    },
    {
      "entropy": 5.63679838180542,
      "epoch": 0.9290021396615444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004918787879321701,
      "loss": 5.2541,
      "mean_token_accuracy": 0.17987410575151444,
      "num_tokens": 20712529.0,
      "step": 11940
    },
    {
      "entropy": 5.462137937545776,
      "epoch": 0.929391169033262,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918714071837384,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19069624841213226,
      "num_tokens": 20721079.0,
      "step": 11945
    },
    {
      "entropy": 5.549027156829834,
      "epoch": 0.9297801984049796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004918640231446282,
      "loss": 5.2716,
      "mean_token_accuracy": 0.1769583523273468,
      "num_tokens": 20730079.0,
      "step": 11950
    },
    {
      "entropy": 5.618971157073974,
      "epoch": 0.9301692277766972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918566358149517,
      "loss": 5.3079,
      "mean_token_accuracy": 0.18117557018995284,
      "num_tokens": 20738200.0,
      "step": 11955
    },
    {
      "entropy": 5.527298355102539,
      "epoch": 0.9305582571484147,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491849245194821,
      "loss": 5.1883,
      "mean_token_accuracy": 0.18633153140544892,
      "num_tokens": 20746663.0,
      "step": 11960
    },
    {
      "entropy": 5.5764788627624515,
      "epoch": 0.9309472865201323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918418512843479,
      "loss": 5.3183,
      "mean_token_accuracy": 0.17480258643627167,
      "num_tokens": 20755990.0,
      "step": 11965
    },
    {
      "entropy": 5.601742506027222,
      "epoch": 0.9313363158918498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004918344540836451,
      "loss": 5.2167,
      "mean_token_accuracy": 0.1803658574819565,
      "num_tokens": 20764985.0,
      "step": 11970
    },
    {
      "entropy": 5.541550254821777,
      "epoch": 0.9317253452635674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918270535928244,
      "loss": 5.2209,
      "mean_token_accuracy": 0.18549701571464539,
      "num_tokens": 20773415.0,
      "step": 11975
    },
    {
      "entropy": 5.542168283462525,
      "epoch": 0.932114374635285,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004918196498119984,
      "loss": 5.2735,
      "mean_token_accuracy": 0.1800215870141983,
      "num_tokens": 20782611.0,
      "step": 11980
    },
    {
      "entropy": 5.65689058303833,
      "epoch": 0.9325034040070025,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004918122427412793,
      "loss": 5.3108,
      "mean_token_accuracy": 0.17959070801734925,
      "num_tokens": 20790859.0,
      "step": 11985
    },
    {
      "entropy": 5.610575008392334,
      "epoch": 0.9328924333787201,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004918048323807795,
      "loss": 5.2126,
      "mean_token_accuracy": 0.18499375879764557,
      "num_tokens": 20798994.0,
      "step": 11990
    },
    {
      "entropy": 5.533700275421142,
      "epoch": 0.9332814627504377,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917974187306114,
      "loss": 5.1613,
      "mean_token_accuracy": 0.18504722267389298,
      "num_tokens": 20807343.0,
      "step": 11995
    },
    {
      "entropy": 5.557862138748169,
      "epoch": 0.9336704921221552,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917900017908875,
      "loss": 5.1642,
      "mean_token_accuracy": 0.1907860204577446,
      "num_tokens": 20815703.0,
      "step": 12000
    },
    {
      "epoch": 0.9336704921221552,
      "eval_entropy": 5.344935700432169,
      "eval_loss": 5.24092960357666,
      "eval_mean_token_accuracy": 0.19120996321217157,
      "eval_num_tokens": 20815703.0,
      "eval_runtime": 21.784,
      "eval_samples_per_second": 1907.735,
      "eval_steps_per_second": 238.478,
      "step": 12000
    },
    {
      "entropy": 5.539333724975586,
      "epoch": 0.9340595214938728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917825815617205,
      "loss": 5.1612,
      "mean_token_accuracy": 0.18685321658849716,
      "num_tokens": 20823901.0,
      "step": 12005
    },
    {
      "entropy": 5.590603351593018,
      "epoch": 0.9344485508655903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917751580432229,
      "loss": 5.3046,
      "mean_token_accuracy": 0.17925916463136674,
      "num_tokens": 20832933.0,
      "step": 12010
    },
    {
      "entropy": 5.565207862854004,
      "epoch": 0.9348375802373079,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917677312355072,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19245655238628387,
      "num_tokens": 20841090.0,
      "step": 12015
    },
    {
      "entropy": 5.557745027542114,
      "epoch": 0.9352266096090255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917603011386863,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18042668849229812,
      "num_tokens": 20848705.0,
      "step": 12020
    },
    {
      "entropy": 5.5556652545928955,
      "epoch": 0.9356156389807431,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917528677528727,
      "loss": 5.2269,
      "mean_token_accuracy": 0.179660427570343,
      "num_tokens": 20857149.0,
      "step": 12025
    },
    {
      "entropy": 5.593832492828369,
      "epoch": 0.9360046683524607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917454310781795,
      "loss": 5.2913,
      "mean_token_accuracy": 0.183056016266346,
      "num_tokens": 20866035.0,
      "step": 12030
    },
    {
      "entropy": 5.54173731803894,
      "epoch": 0.9363936977241781,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004917379911147193,
      "loss": 5.2,
      "mean_token_accuracy": 0.18876660019159316,
      "num_tokens": 20874590.0,
      "step": 12035
    },
    {
      "entropy": 5.593485164642334,
      "epoch": 0.9367827270958957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004917305478626049,
      "loss": 5.2299,
      "mean_token_accuracy": 0.18825214654207229,
      "num_tokens": 20883872.0,
      "step": 12040
    },
    {
      "entropy": 5.499920177459717,
      "epoch": 0.9371717564676133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917231013219495,
      "loss": 5.1352,
      "mean_token_accuracy": 0.19150085896253585,
      "num_tokens": 20892515.0,
      "step": 12045
    },
    {
      "entropy": 5.529827070236206,
      "epoch": 0.9375607858393309,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491715651492866,
      "loss": 5.2308,
      "mean_token_accuracy": 0.18484504520893097,
      "num_tokens": 20901337.0,
      "step": 12050
    },
    {
      "entropy": 5.526296615600586,
      "epoch": 0.9379498152110485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004917081983754675,
      "loss": 5.2277,
      "mean_token_accuracy": 0.18402009308338166,
      "num_tokens": 20910034.0,
      "step": 12055
    },
    {
      "entropy": 5.598119640350342,
      "epoch": 0.938338844582766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917007419698669,
      "loss": 5.2581,
      "mean_token_accuracy": 0.18303631991147995,
      "num_tokens": 20918832.0,
      "step": 12060
    },
    {
      "entropy": 5.4956066608428955,
      "epoch": 0.9387278739544835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916932822761776,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19157517403364183,
      "num_tokens": 20927734.0,
      "step": 12065
    },
    {
      "entropy": 5.543244457244873,
      "epoch": 0.9391169033262011,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004916858192945126,
      "loss": 5.1174,
      "mean_token_accuracy": 0.1897149235010147,
      "num_tokens": 20935888.0,
      "step": 12070
    },
    {
      "entropy": 5.5040624141693115,
      "epoch": 0.9395059326979187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916783530249852,
      "loss": 5.1154,
      "mean_token_accuracy": 0.18593257814645767,
      "num_tokens": 20944577.0,
      "step": 12075
    },
    {
      "entropy": 5.510186719894409,
      "epoch": 0.9398949620696363,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916708834677086,
      "loss": 5.2261,
      "mean_token_accuracy": 0.181659696996212,
      "num_tokens": 20952527.0,
      "step": 12080
    },
    {
      "entropy": 5.598353004455566,
      "epoch": 0.9402839914413538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916634106227962,
      "loss": 5.2704,
      "mean_token_accuracy": 0.18085542172193528,
      "num_tokens": 20961882.0,
      "step": 12085
    },
    {
      "entropy": 5.650716781616211,
      "epoch": 0.9406730208130714,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916559344903614,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18876594603061675,
      "num_tokens": 20970782.0,
      "step": 12090
    },
    {
      "entropy": 5.562792205810547,
      "epoch": 0.941062050184789,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916484550705176,
      "loss": 5.2499,
      "mean_token_accuracy": 0.1796768546104431,
      "num_tokens": 20979927.0,
      "step": 12095
    },
    {
      "entropy": 5.520222759246826,
      "epoch": 0.9414510795565065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916409723633785,
      "loss": 5.1696,
      "mean_token_accuracy": 0.18996877670288087,
      "num_tokens": 20988626.0,
      "step": 12100
    },
    {
      "entropy": 5.565795230865478,
      "epoch": 0.9418401089282241,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004916334863690573,
      "loss": 5.2828,
      "mean_token_accuracy": 0.17252252697944642,
      "num_tokens": 20997065.0,
      "step": 12105
    },
    {
      "entropy": 5.455416250228882,
      "epoch": 0.9422291382999416,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004916259970876678,
      "loss": 5.0346,
      "mean_token_accuracy": 0.1853836178779602,
      "num_tokens": 21005611.0,
      "step": 12110
    },
    {
      "entropy": 5.418533182144165,
      "epoch": 0.9426181676716592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916185045193237,
      "loss": 5.0488,
      "mean_token_accuracy": 0.19643110483884813,
      "num_tokens": 21014162.0,
      "step": 12115
    },
    {
      "entropy": 5.503196096420288,
      "epoch": 0.9430071970433768,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004916110086641384,
      "loss": 5.2703,
      "mean_token_accuracy": 0.18162964880466462,
      "num_tokens": 21022780.0,
      "step": 12120
    },
    {
      "entropy": 5.575558137893677,
      "epoch": 0.9433962264150944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004916035095222259,
      "loss": 5.1858,
      "mean_token_accuracy": 0.18557081669569014,
      "num_tokens": 21031101.0,
      "step": 12125
    },
    {
      "entropy": 5.4627422332763675,
      "epoch": 0.9437852557868119,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004915960070936999,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1908903867006302,
      "num_tokens": 21040212.0,
      "step": 12130
    },
    {
      "entropy": 5.515059280395508,
      "epoch": 0.9441742851585294,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915885013786742,
      "loss": 5.1421,
      "mean_token_accuracy": 0.1901004806160927,
      "num_tokens": 21048643.0,
      "step": 12135
    },
    {
      "entropy": 5.543565320968628,
      "epoch": 0.944563314530247,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915809923772628,
      "loss": 5.2118,
      "mean_token_accuracy": 0.1857934921979904,
      "num_tokens": 21057070.0,
      "step": 12140
    },
    {
      "entropy": 5.5097757339477536,
      "epoch": 0.9449523439019646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915734800895796,
      "loss": 5.2025,
      "mean_token_accuracy": 0.185279943048954,
      "num_tokens": 21065102.0,
      "step": 12145
    },
    {
      "entropy": 5.563621950149536,
      "epoch": 0.9453413732736822,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004915659645157383,
      "loss": 5.304,
      "mean_token_accuracy": 0.1851240798830986,
      "num_tokens": 21073968.0,
      "step": 12150
    },
    {
      "entropy": 5.609937620162964,
      "epoch": 0.9457304026453998,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915584456558535,
      "loss": 5.3462,
      "mean_token_accuracy": 0.17858497351408004,
      "num_tokens": 21083363.0,
      "step": 12155
    },
    {
      "entropy": 5.607294368743896,
      "epoch": 0.9461194320171173,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004915509235100388,
      "loss": 5.3237,
      "mean_token_accuracy": 0.17933631539344788,
      "num_tokens": 21091691.0,
      "step": 12160
    },
    {
      "entropy": 5.513938617706299,
      "epoch": 0.9465084613888348,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915433980784086,
      "loss": 5.1462,
      "mean_token_accuracy": 0.17888393700122834,
      "num_tokens": 21100512.0,
      "step": 12165
    },
    {
      "entropy": 5.493751955032349,
      "epoch": 0.9468974907605524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004915358693610769,
      "loss": 5.2542,
      "mean_token_accuracy": 0.18847894966602324,
      "num_tokens": 21109658.0,
      "step": 12170
    },
    {
      "entropy": 5.505000829696655,
      "epoch": 0.94728652013227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915283373581581,
      "loss": 5.1964,
      "mean_token_accuracy": 0.18724531531333924,
      "num_tokens": 21118710.0,
      "step": 12175
    },
    {
      "entropy": 5.548243665695191,
      "epoch": 0.9476755495039876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915208020697664,
      "loss": 5.1678,
      "mean_token_accuracy": 0.190705007314682,
      "num_tokens": 21127672.0,
      "step": 12180
    },
    {
      "entropy": 5.6351231098175045,
      "epoch": 0.9480645788757052,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915132634960162,
      "loss": 5.3377,
      "mean_token_accuracy": 0.17378480285406112,
      "num_tokens": 21135688.0,
      "step": 12185
    },
    {
      "entropy": 5.598523139953613,
      "epoch": 0.9484536082474226,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004915057216370218,
      "loss": 5.1896,
      "mean_token_accuracy": 0.18350013345479965,
      "num_tokens": 21143557.0,
      "step": 12190
    },
    {
      "entropy": 5.55081090927124,
      "epoch": 0.9488426376191402,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004914981764928977,
      "loss": 5.2341,
      "mean_token_accuracy": 0.18359110206365586,
      "num_tokens": 21152056.0,
      "step": 12195
    },
    {
      "entropy": 5.482287836074829,
      "epoch": 0.9492316669908578,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914906280637584,
      "loss": 5.1855,
      "mean_token_accuracy": 0.18782722651958467,
      "num_tokens": 21160590.0,
      "step": 12200
    },
    {
      "entropy": 5.52836012840271,
      "epoch": 0.9496206963625754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914830763497183,
      "loss": 5.2694,
      "mean_token_accuracy": 0.17721753269433976,
      "num_tokens": 21169333.0,
      "step": 12205
    },
    {
      "entropy": 5.519309663772583,
      "epoch": 0.950009725734293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004914755213508922,
      "loss": 5.0301,
      "mean_token_accuracy": 0.19665799736976625,
      "num_tokens": 21178222.0,
      "step": 12210
    },
    {
      "entropy": 5.49220495223999,
      "epoch": 0.9503987551060105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914679630673945,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1919606015086174,
      "num_tokens": 21187029.0,
      "step": 12215
    },
    {
      "entropy": 5.600217294692993,
      "epoch": 0.950787784477728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914604014993401,
      "loss": 5.2356,
      "mean_token_accuracy": 0.18390725255012513,
      "num_tokens": 21195994.0,
      "step": 12220
    },
    {
      "entropy": 5.542682838439942,
      "epoch": 0.9511768138494456,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004914528366468436,
      "loss": 5.1855,
      "mean_token_accuracy": 0.18452553302049637,
      "num_tokens": 21204534.0,
      "step": 12225
    },
    {
      "entropy": 5.4876096725463865,
      "epoch": 0.9515658432211632,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004914452685100199,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20137422531843185,
      "num_tokens": 21212852.0,
      "step": 12230
    },
    {
      "entropy": 5.556637001037598,
      "epoch": 0.9519548725928808,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914376970889836,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1755823850631714,
      "num_tokens": 21221040.0,
      "step": 12235
    },
    {
      "entropy": 5.49406213760376,
      "epoch": 0.9523439019645983,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914301223838497,
      "loss": 5.0824,
      "mean_token_accuracy": 0.1865405797958374,
      "num_tokens": 21230227.0,
      "step": 12240
    },
    {
      "entropy": 5.572826814651489,
      "epoch": 0.9527329313363159,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004914225443947334,
      "loss": 5.3065,
      "mean_token_accuracy": 0.17179667800664902,
      "num_tokens": 21238496.0,
      "step": 12245
    },
    {
      "entropy": 5.498713159561158,
      "epoch": 0.9531219607080335,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914149631217494,
      "loss": 5.1683,
      "mean_token_accuracy": 0.18232299834489823,
      "num_tokens": 21246806.0,
      "step": 12250
    },
    {
      "entropy": 5.509085035324096,
      "epoch": 0.953510990079751,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914073785650127,
      "loss": 5.1536,
      "mean_token_accuracy": 0.18383436799049377,
      "num_tokens": 21255841.0,
      "step": 12255
    },
    {
      "entropy": 5.601258993148804,
      "epoch": 0.9539000194514686,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913997907246385,
      "loss": 5.2045,
      "mean_token_accuracy": 0.1773657813668251,
      "num_tokens": 21264823.0,
      "step": 12260
    },
    {
      "entropy": 5.530019950866699,
      "epoch": 0.9542890488231861,
      "grad_norm": 0.984375,
      "learning_rate": 0.000491392199600742,
      "loss": 5.2244,
      "mean_token_accuracy": 0.18132253587245942,
      "num_tokens": 21273382.0,
      "step": 12265
    },
    {
      "entropy": 5.514166307449341,
      "epoch": 0.9546780781949037,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913846051934382,
      "loss": 5.2299,
      "mean_token_accuracy": 0.18331311494112015,
      "num_tokens": 21282216.0,
      "step": 12270
    },
    {
      "entropy": 5.567118835449219,
      "epoch": 0.9550671075666213,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004913770075028425,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1815290629863739,
      "num_tokens": 21290208.0,
      "step": 12275
    },
    {
      "entropy": 5.631261491775513,
      "epoch": 0.9554561369383389,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004913694065290701,
      "loss": 5.2797,
      "mean_token_accuracy": 0.17556531429290773,
      "num_tokens": 21298864.0,
      "step": 12280
    },
    {
      "entropy": 5.439242029190064,
      "epoch": 0.9558451663100564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913618022722363,
      "loss": 5.1179,
      "mean_token_accuracy": 0.1903344064950943,
      "num_tokens": 21307484.0,
      "step": 12285
    },
    {
      "entropy": 5.491755819320678,
      "epoch": 0.9562341956817739,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004913541947324566,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18501874059438705,
      "num_tokens": 21316160.0,
      "step": 12290
    },
    {
      "entropy": 5.519949579238892,
      "epoch": 0.9566232250534915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913465839098463,
      "loss": 5.0924,
      "mean_token_accuracy": 0.19052087664604186,
      "num_tokens": 21324777.0,
      "step": 12295
    },
    {
      "entropy": 5.578331565856933,
      "epoch": 0.9570122544252091,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913389698045209,
      "loss": 5.2174,
      "mean_token_accuracy": 0.1854473501443863,
      "num_tokens": 21332994.0,
      "step": 12300
    },
    {
      "entropy": 5.48744387626648,
      "epoch": 0.9574012837969267,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491331352416596,
      "loss": 5.152,
      "mean_token_accuracy": 0.18614051192998887,
      "num_tokens": 21341457.0,
      "step": 12305
    },
    {
      "entropy": 5.551470232009888,
      "epoch": 0.9577903131686443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913237317461872,
      "loss": 5.1957,
      "mean_token_accuracy": 0.17776983380317687,
      "num_tokens": 21349487.0,
      "step": 12310
    },
    {
      "entropy": 5.492322158813477,
      "epoch": 0.9581793425403617,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049131610779341,
      "loss": 5.2213,
      "mean_token_accuracy": 0.18082193285226822,
      "num_tokens": 21358411.0,
      "step": 12315
    },
    {
      "entropy": 5.57997088432312,
      "epoch": 0.9585683719120793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913084805583803,
      "loss": 5.2725,
      "mean_token_accuracy": 0.18279048949480056,
      "num_tokens": 21367257.0,
      "step": 12320
    },
    {
      "entropy": 5.612073993682861,
      "epoch": 0.9589574012837969,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004913008500412136,
      "loss": 5.2816,
      "mean_token_accuracy": 0.17700959891080856,
      "num_tokens": 21376361.0,
      "step": 12325
    },
    {
      "entropy": 5.595424509048462,
      "epoch": 0.9593464306555145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912932162420259,
      "loss": 5.2577,
      "mean_token_accuracy": 0.18183140605688095,
      "num_tokens": 21384658.0,
      "step": 12330
    },
    {
      "entropy": 5.510373592376709,
      "epoch": 0.9597354600272321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004912855791609327,
      "loss": 5.1322,
      "mean_token_accuracy": 0.1854855313897133,
      "num_tokens": 21393390.0,
      "step": 12335
    },
    {
      "entropy": 5.584775400161743,
      "epoch": 0.9601244893989496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004912779387980502,
      "loss": 5.255,
      "mean_token_accuracy": 0.1805700972676277,
      "num_tokens": 21401591.0,
      "step": 12340
    },
    {
      "entropy": 5.415788412094116,
      "epoch": 0.9605135187706672,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912702951534942,
      "loss": 5.0317,
      "mean_token_accuracy": 0.19965379685163498,
      "num_tokens": 21409933.0,
      "step": 12345
    },
    {
      "entropy": 5.531234216690064,
      "epoch": 0.9609025481423847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912626482273809,
      "loss": 5.1114,
      "mean_token_accuracy": 0.18745438605546952,
      "num_tokens": 21418665.0,
      "step": 12350
    },
    {
      "entropy": 5.5606427669525145,
      "epoch": 0.9612915775141023,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912549980198258,
      "loss": 5.2846,
      "mean_token_accuracy": 0.18365840315818788,
      "num_tokens": 21427532.0,
      "step": 12355
    },
    {
      "entropy": 5.6447347640991214,
      "epoch": 0.9616806068858199,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004912473445309455,
      "loss": 5.2209,
      "mean_token_accuracy": 0.18302836567163466,
      "num_tokens": 21436101.0,
      "step": 12360
    },
    {
      "entropy": 5.583501625061035,
      "epoch": 0.9620696362575375,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491239687760856,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1888975739479065,
      "num_tokens": 21444905.0,
      "step": 12365
    },
    {
      "entropy": 5.588873767852784,
      "epoch": 0.962458665629255,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004912320277096732,
      "loss": 5.2947,
      "mean_token_accuracy": 0.175302354991436,
      "num_tokens": 21453498.0,
      "step": 12370
    },
    {
      "entropy": 5.560341024398804,
      "epoch": 0.9628476950009726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912243643775137,
      "loss": 5.1811,
      "mean_token_accuracy": 0.19413408190011977,
      "num_tokens": 21461884.0,
      "step": 12375
    },
    {
      "entropy": 5.486970710754394,
      "epoch": 0.9632367243726901,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912166977644936,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19314700216054917,
      "num_tokens": 21469694.0,
      "step": 12380
    },
    {
      "entropy": 5.505225419998169,
      "epoch": 0.9636257537444077,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004912090278707293,
      "loss": 5.2424,
      "mean_token_accuracy": 0.18412844687700272,
      "num_tokens": 21478988.0,
      "step": 12385
    },
    {
      "entropy": 5.653407335281372,
      "epoch": 0.9640147831161253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491201354696337,
      "loss": 5.2834,
      "mean_token_accuracy": 0.18375380635261535,
      "num_tokens": 21488336.0,
      "step": 12390
    },
    {
      "entropy": 5.552691698074341,
      "epoch": 0.9644038124878428,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004911936782414334,
      "loss": 5.21,
      "mean_token_accuracy": 0.1889895647764206,
      "num_tokens": 21496792.0,
      "step": 12395
    },
    {
      "entropy": 5.524241018295288,
      "epoch": 0.9647928418595604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004911859985061348,
      "loss": 5.2217,
      "mean_token_accuracy": 0.17647273391485213,
      "num_tokens": 21505796.0,
      "step": 12400
    },
    {
      "entropy": 5.517217540740967,
      "epoch": 0.965181871231278,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911783154905577,
      "loss": 5.2023,
      "mean_token_accuracy": 0.18912200033664703,
      "num_tokens": 21514904.0,
      "step": 12405
    },
    {
      "entropy": 5.603016376495361,
      "epoch": 0.9655709006029956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004911706291948188,
      "loss": 5.3784,
      "mean_token_accuracy": 0.17836669236421585,
      "num_tokens": 21524231.0,
      "step": 12410
    },
    {
      "entropy": 5.573432683944702,
      "epoch": 0.9659599299747131,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911629396190348,
      "loss": 5.1622,
      "mean_token_accuracy": 0.18772236108779908,
      "num_tokens": 21532852.0,
      "step": 12415
    },
    {
      "entropy": 5.592479038238525,
      "epoch": 0.9663489593464306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911552467633221,
      "loss": 5.2286,
      "mean_token_accuracy": 0.1844874456524849,
      "num_tokens": 21542562.0,
      "step": 12420
    },
    {
      "entropy": 5.525732088088989,
      "epoch": 0.9667379887181482,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004911475506277976,
      "loss": 5.2385,
      "mean_token_accuracy": 0.17982458770275117,
      "num_tokens": 21551503.0,
      "step": 12425
    },
    {
      "entropy": 5.4624053001403805,
      "epoch": 0.9671270180898658,
      "grad_norm": 1.0,
      "learning_rate": 0.000491139851212578,
      "loss": 5.0396,
      "mean_token_accuracy": 0.202691887319088,
      "num_tokens": 21560712.0,
      "step": 12430
    },
    {
      "entropy": 5.505254459381104,
      "epoch": 0.9675160474615834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911321485177802,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19221523702144622,
      "num_tokens": 21569811.0,
      "step": 12435
    },
    {
      "entropy": 5.491925954818726,
      "epoch": 0.967905076833301,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911244425435212,
      "loss": 5.2313,
      "mean_token_accuracy": 0.18152496665716172,
      "num_tokens": 21578423.0,
      "step": 12440
    },
    {
      "entropy": 5.530194234848023,
      "epoch": 0.9682941062050184,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911167332899176,
      "loss": 5.1861,
      "mean_token_accuracy": 0.18658440113067626,
      "num_tokens": 21587509.0,
      "step": 12445
    },
    {
      "entropy": 5.572329521179199,
      "epoch": 0.968683135576736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911090207570867,
      "loss": 5.1729,
      "mean_token_accuracy": 0.19146473109722137,
      "num_tokens": 21596054.0,
      "step": 12450
    },
    {
      "entropy": 5.588417959213257,
      "epoch": 0.9690721649484536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911013049451453,
      "loss": 5.2193,
      "mean_token_accuracy": 0.18457165211439133,
      "num_tokens": 21604872.0,
      "step": 12455
    },
    {
      "entropy": 5.599390029907227,
      "epoch": 0.9694611943201712,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910935858542106,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1784984514117241,
      "num_tokens": 21613773.0,
      "step": 12460
    },
    {
      "entropy": 5.4978679656982425,
      "epoch": 0.9698502236918888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004910858634843997,
      "loss": 5.1018,
      "mean_token_accuracy": 0.1890167698264122,
      "num_tokens": 21622071.0,
      "step": 12465
    },
    {
      "entropy": 5.5291248798370365,
      "epoch": 0.9702392530636063,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910781378358297,
      "loss": 5.1166,
      "mean_token_accuracy": 0.18756571859121324,
      "num_tokens": 21630698.0,
      "step": 12470
    },
    {
      "entropy": 5.592549467086792,
      "epoch": 0.9706282824353238,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910704089086178,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19187143743038176,
      "num_tokens": 21638446.0,
      "step": 12475
    },
    {
      "entropy": 5.5418932914733885,
      "epoch": 0.9710173118070414,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004910626767028815,
      "loss": 5.2034,
      "mean_token_accuracy": 0.18510141372680664,
      "num_tokens": 21647214.0,
      "step": 12480
    },
    {
      "entropy": 5.582453489303589,
      "epoch": 0.971406341178759,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004910549412187379,
      "loss": 5.3091,
      "mean_token_accuracy": 0.1761996865272522,
      "num_tokens": 21656047.0,
      "step": 12485
    },
    {
      "entropy": 5.62225308418274,
      "epoch": 0.9717953705504766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004910472024563045,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1827482298016548,
      "num_tokens": 21664374.0,
      "step": 12490
    },
    {
      "entropy": 5.625817823410034,
      "epoch": 0.9721843999221941,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910394604156987,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1825633242726326,
      "num_tokens": 21673334.0,
      "step": 12495
    },
    {
      "entropy": 5.553891038894653,
      "epoch": 0.9725734292939117,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004910317150970379,
      "loss": 5.1419,
      "mean_token_accuracy": 0.19343614876270293,
      "num_tokens": 21681392.0,
      "step": 12500
    },
    {
      "entropy": 5.55709547996521,
      "epoch": 0.9729624586656292,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910239665004397,
      "loss": 5.1598,
      "mean_token_accuracy": 0.18340199589729309,
      "num_tokens": 21690039.0,
      "step": 12505
    },
    {
      "entropy": 5.503588247299194,
      "epoch": 0.9733514880373468,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910162146260216,
      "loss": 5.1345,
      "mean_token_accuracy": 0.1927665263414383,
      "num_tokens": 21699006.0,
      "step": 12510
    },
    {
      "entropy": 5.510761451721192,
      "epoch": 0.9737405174090644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004910084594739013,
      "loss": 5.1415,
      "mean_token_accuracy": 0.18670180737972258,
      "num_tokens": 21707730.0,
      "step": 12515
    },
    {
      "entropy": 5.528998613357544,
      "epoch": 0.9741295467807819,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004910007010441964,
      "loss": 5.2105,
      "mean_token_accuracy": 0.18834204971790314,
      "num_tokens": 21715671.0,
      "step": 12520
    },
    {
      "entropy": 5.570176410675049,
      "epoch": 0.9745185761524995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004909929393370248,
      "loss": 5.2717,
      "mean_token_accuracy": 0.1823419839143753,
      "num_tokens": 21724000.0,
      "step": 12525
    },
    {
      "entropy": 5.626552104949951,
      "epoch": 0.9749076055242171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909851743525041,
      "loss": 5.2792,
      "mean_token_accuracy": 0.18159958273172377,
      "num_tokens": 21732915.0,
      "step": 12530
    },
    {
      "entropy": 5.698964405059814,
      "epoch": 0.9752966348959347,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490977406090752,
      "loss": 5.3776,
      "mean_token_accuracy": 0.17277083843946456,
      "num_tokens": 21742495.0,
      "step": 12535
    },
    {
      "entropy": 5.565761423110962,
      "epoch": 0.9756856642676522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004909696345518867,
      "loss": 5.1848,
      "mean_token_accuracy": 0.1832648128271103,
      "num_tokens": 21751198.0,
      "step": 12540
    },
    {
      "entropy": 5.503470325469971,
      "epoch": 0.9760746936393697,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004909618597360258,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19248865842819213,
      "num_tokens": 21759804.0,
      "step": 12545
    },
    {
      "entropy": 5.483126401901245,
      "epoch": 0.9764637230110873,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909540816432876,
      "loss": 5.0838,
      "mean_token_accuracy": 0.1929810792207718,
      "num_tokens": 21768541.0,
      "step": 12550
    },
    {
      "entropy": 5.526939058303833,
      "epoch": 0.9768527523828049,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004909463002737897,
      "loss": 5.1519,
      "mean_token_accuracy": 0.19035652428865432,
      "num_tokens": 21777909.0,
      "step": 12555
    },
    {
      "entropy": 5.570001649856567,
      "epoch": 0.9772417817545225,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909385156276506,
      "loss": 5.2326,
      "mean_token_accuracy": 0.18371648341417313,
      "num_tokens": 21786844.0,
      "step": 12560
    },
    {
      "entropy": 5.526794385910034,
      "epoch": 0.9776308111262401,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004909307277049881,
      "loss": 5.1177,
      "mean_token_accuracy": 0.18790711909532548,
      "num_tokens": 21796157.0,
      "step": 12565
    },
    {
      "entropy": 5.5055419921875,
      "epoch": 0.9780198404979576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004909229365059205,
      "loss": 5.2303,
      "mean_token_accuracy": 0.18127598464488984,
      "num_tokens": 21804447.0,
      "step": 12570
    },
    {
      "entropy": 5.552242946624756,
      "epoch": 0.9784088698696751,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909151420305661,
      "loss": 5.2426,
      "mean_token_accuracy": 0.18326650857925414,
      "num_tokens": 21813569.0,
      "step": 12575
    },
    {
      "entropy": 5.5926836967468265,
      "epoch": 0.9787978992413927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000490907344279043,
      "loss": 5.231,
      "mean_token_accuracy": 0.18444594293832778,
      "num_tokens": 21822194.0,
      "step": 12580
    },
    {
      "entropy": 5.574924373626709,
      "epoch": 0.9791869286131103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004908995432514698,
      "loss": 5.2384,
      "mean_token_accuracy": 0.18146923184394836,
      "num_tokens": 21830012.0,
      "step": 12585
    },
    {
      "entropy": 5.486664915084839,
      "epoch": 0.9795759579848279,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004908917389479645,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1896378830075264,
      "num_tokens": 21838177.0,
      "step": 12590
    },
    {
      "entropy": 5.496251249313355,
      "epoch": 0.9799649873565455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908839313686456,
      "loss": 5.1923,
      "mean_token_accuracy": 0.1851340040564537,
      "num_tokens": 21847005.0,
      "step": 12595
    },
    {
      "entropy": 5.584972620010376,
      "epoch": 0.980354016728263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908761205136319,
      "loss": 5.1256,
      "mean_token_accuracy": 0.18699826449155807,
      "num_tokens": 21854363.0,
      "step": 12600
    },
    {
      "entropy": 5.597399377822876,
      "epoch": 0.9807430460999805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004908683063830414,
      "loss": 5.1544,
      "mean_token_accuracy": 0.18721675276756286,
      "num_tokens": 21862285.0,
      "step": 12605
    },
    {
      "entropy": 5.528862190246582,
      "epoch": 0.9811320754716981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908604889769932,
      "loss": 5.2228,
      "mean_token_accuracy": 0.18969684988260269,
      "num_tokens": 21870632.0,
      "step": 12610
    },
    {
      "entropy": 5.599672126770019,
      "epoch": 0.9815211048434157,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908526682956054,
      "loss": 5.3218,
      "mean_token_accuracy": 0.18269346952438353,
      "num_tokens": 21879251.0,
      "step": 12615
    },
    {
      "entropy": 5.515309047698975,
      "epoch": 0.9819101342151333,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908448443389972,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1889792114496231,
      "num_tokens": 21887582.0,
      "step": 12620
    },
    {
      "entropy": 5.545315980911255,
      "epoch": 0.9822991635868508,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908370171072869,
      "loss": 5.1853,
      "mean_token_accuracy": 0.18147605806589126,
      "num_tokens": 21896418.0,
      "step": 12625
    },
    {
      "entropy": 5.514461946487427,
      "epoch": 0.9826881929585684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908291866005933,
      "loss": 5.1281,
      "mean_token_accuracy": 0.18880711495876312,
      "num_tokens": 21905035.0,
      "step": 12630
    },
    {
      "entropy": 5.553278589248658,
      "epoch": 0.9830772223302859,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004908213528190355,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19049129635095596,
      "num_tokens": 21913858.0,
      "step": 12635
    },
    {
      "entropy": 5.569316101074219,
      "epoch": 0.9834662517020035,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004908135157627321,
      "loss": 5.2159,
      "mean_token_accuracy": 0.1872427299618721,
      "num_tokens": 21922375.0,
      "step": 12640
    },
    {
      "entropy": 5.585264205932617,
      "epoch": 0.9838552810737211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908056754318022,
      "loss": 5.2353,
      "mean_token_accuracy": 0.17861511409282685,
      "num_tokens": 21931216.0,
      "step": 12645
    },
    {
      "entropy": 5.5599353313446045,
      "epoch": 0.9842443104454386,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004907978318263646,
      "loss": 5.2165,
      "mean_token_accuracy": 0.18171363919973374,
      "num_tokens": 21939683.0,
      "step": 12650
    },
    {
      "entropy": 5.580514144897461,
      "epoch": 0.9846333398171562,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907899849465383,
      "loss": 5.2629,
      "mean_token_accuracy": 0.18881067931652068,
      "num_tokens": 21948288.0,
      "step": 12655
    },
    {
      "entropy": 5.529097175598144,
      "epoch": 0.9850223691888738,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907821347924424,
      "loss": 5.1936,
      "mean_token_accuracy": 0.18810323178768157,
      "num_tokens": 21957870.0,
      "step": 12660
    },
    {
      "entropy": 5.609283018112182,
      "epoch": 0.9854113985605913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004907742813641963,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1858854278922081,
      "num_tokens": 21965844.0,
      "step": 12665
    },
    {
      "entropy": 5.477437305450439,
      "epoch": 0.9858004279323089,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907664246619187,
      "loss": 5.147,
      "mean_token_accuracy": 0.1858885794878006,
      "num_tokens": 21974364.0,
      "step": 12670
    },
    {
      "entropy": 5.481590747833252,
      "epoch": 0.9861894573040264,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907585646857293,
      "loss": 5.0948,
      "mean_token_accuracy": 0.1894430011510849,
      "num_tokens": 21982618.0,
      "step": 12675
    },
    {
      "entropy": 5.561007738113403,
      "epoch": 0.986578486675744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004907507014357467,
      "loss": 5.2402,
      "mean_token_accuracy": 0.18591585904359817,
      "num_tokens": 21991823.0,
      "step": 12680
    },
    {
      "entropy": 5.572273397445679,
      "epoch": 0.9869675160474616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907428349120909,
      "loss": 5.1427,
      "mean_token_accuracy": 0.18160430043935777,
      "num_tokens": 22000773.0,
      "step": 12685
    },
    {
      "entropy": 5.607509613037109,
      "epoch": 0.9873565454191792,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004907349651148809,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1756952613592148,
      "num_tokens": 22009895.0,
      "step": 12690
    },
    {
      "entropy": 5.600894927978516,
      "epoch": 0.9877455747908968,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907270920442361,
      "loss": 5.2368,
      "mean_token_accuracy": 0.17951168566942216,
      "num_tokens": 22017624.0,
      "step": 12695
    },
    {
      "entropy": 5.616975402832031,
      "epoch": 0.9881346041626142,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907192157002762,
      "loss": 5.3479,
      "mean_token_accuracy": 0.1795940190553665,
      "num_tokens": 22026700.0,
      "step": 12700
    },
    {
      "entropy": 5.55825514793396,
      "epoch": 0.9885236335343318,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907113360831204,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19315554350614547,
      "num_tokens": 22034623.0,
      "step": 12705
    },
    {
      "entropy": 5.545919513702392,
      "epoch": 0.9889126629060494,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907034531928886,
      "loss": 5.2179,
      "mean_token_accuracy": 0.18369993269443513,
      "num_tokens": 22042871.0,
      "step": 12710
    },
    {
      "entropy": 5.484036254882812,
      "epoch": 0.989301692277767,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906955670297001,
      "loss": 5.0888,
      "mean_token_accuracy": 0.1927085191011429,
      "num_tokens": 22051818.0,
      "step": 12715
    },
    {
      "entropy": 5.634485292434692,
      "epoch": 0.9896907216494846,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906876775936746,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1796593874692917,
      "num_tokens": 22060464.0,
      "step": 12720
    },
    {
      "entropy": 5.553441667556763,
      "epoch": 0.990079751021202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004906797848849321,
      "loss": 5.2212,
      "mean_token_accuracy": 0.17699324786663057,
      "num_tokens": 22069037.0,
      "step": 12725
    },
    {
      "entropy": 5.545280885696411,
      "epoch": 0.9904687803929196,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490671888903592,
      "loss": 5.1827,
      "mean_token_accuracy": 0.18326786607503892,
      "num_tokens": 22077010.0,
      "step": 12730
    },
    {
      "entropy": 5.501655578613281,
      "epoch": 0.9908578097646372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004906639896497744,
      "loss": 5.2385,
      "mean_token_accuracy": 0.18212108165025712,
      "num_tokens": 22086165.0,
      "step": 12735
    },
    {
      "entropy": 5.507706499099731,
      "epoch": 0.9912468391363548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004906560871235988,
      "loss": 5.0883,
      "mean_token_accuracy": 0.18723068833351136,
      "num_tokens": 22094798.0,
      "step": 12740
    },
    {
      "entropy": 5.61138129234314,
      "epoch": 0.9916358685080724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906481813251854,
      "loss": 5.2564,
      "mean_token_accuracy": 0.18288826048374177,
      "num_tokens": 22103959.0,
      "step": 12745
    },
    {
      "entropy": 5.583155298233033,
      "epoch": 0.9920248978797899,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906402722546542,
      "loss": 5.2343,
      "mean_token_accuracy": 0.1878318041563034,
      "num_tokens": 22112719.0,
      "step": 12750
    },
    {
      "entropy": 5.523591136932373,
      "epoch": 0.9924139272515075,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490632359912125,
      "loss": 5.1037,
      "mean_token_accuracy": 0.18953030109405516,
      "num_tokens": 22121145.0,
      "step": 12755
    },
    {
      "entropy": 5.537987518310547,
      "epoch": 0.992802956623225,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490624444297718,
      "loss": 5.2239,
      "mean_token_accuracy": 0.18051352053880693,
      "num_tokens": 22130606.0,
      "step": 12760
    },
    {
      "entropy": 5.55096926689148,
      "epoch": 0.9931919859949426,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906165254115533,
      "loss": 5.2282,
      "mean_token_accuracy": 0.17990311980247498,
      "num_tokens": 22140237.0,
      "step": 12765
    },
    {
      "entropy": 5.544623136520386,
      "epoch": 0.9935810153666602,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000490608603253751,
      "loss": 5.1995,
      "mean_token_accuracy": 0.17970874905586243,
      "num_tokens": 22149734.0,
      "step": 12770
    },
    {
      "entropy": 5.5156396389007565,
      "epoch": 0.9939700447383778,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906006778244312,
      "loss": 5.2618,
      "mean_token_accuracy": 0.18165697455406188,
      "num_tokens": 22159346.0,
      "step": 12775
    },
    {
      "entropy": 5.455259084701538,
      "epoch": 0.9943590741100953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004905927491237145,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18956034779548644,
      "num_tokens": 22167624.0,
      "step": 12780
    },
    {
      "entropy": 5.6506531715393065,
      "epoch": 0.9947481034818129,
      "grad_norm": 1.125,
      "learning_rate": 0.0004905848171517208,
      "loss": 5.3197,
      "mean_token_accuracy": 0.1811376094818115,
      "num_tokens": 22176124.0,
      "step": 12785
    },
    {
      "entropy": 5.540608310699463,
      "epoch": 0.9951371328535304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004905768819085706,
      "loss": 5.2278,
      "mean_token_accuracy": 0.18458505868911743,
      "num_tokens": 22184434.0,
      "step": 12790
    },
    {
      "entropy": 5.537732219696045,
      "epoch": 0.995526162225248,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905689433943846,
      "loss": 5.2133,
      "mean_token_accuracy": 0.18687775880098342,
      "num_tokens": 22193431.0,
      "step": 12795
    },
    {
      "entropy": 5.536033821105957,
      "epoch": 0.9959151915969656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004905610016092828,
      "loss": 5.1693,
      "mean_token_accuracy": 0.18484147191047667,
      "num_tokens": 22201634.0,
      "step": 12800
    },
    {
      "entropy": 5.499694204330444,
      "epoch": 0.9963042209686831,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905530565533859,
      "loss": 5.2203,
      "mean_token_accuracy": 0.1857329100370407,
      "num_tokens": 22210864.0,
      "step": 12805
    },
    {
      "entropy": 5.571992874145508,
      "epoch": 0.9966932503404007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004905451082268146,
      "loss": 5.2261,
      "mean_token_accuracy": 0.18939330875873567,
      "num_tokens": 22219800.0,
      "step": 12810
    },
    {
      "entropy": 5.565126562118531,
      "epoch": 0.9970822797121183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905371566296891,
      "loss": 5.2598,
      "mean_token_accuracy": 0.18318777680397033,
      "num_tokens": 22228826.0,
      "step": 12815
    },
    {
      "entropy": 5.5704811096191404,
      "epoch": 0.9974713090838359,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905292017621305,
      "loss": 5.2071,
      "mean_token_accuracy": 0.18567290306091308,
      "num_tokens": 22237076.0,
      "step": 12820
    },
    {
      "entropy": 5.531954193115235,
      "epoch": 0.9978603384555534,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905212436242593,
      "loss": 5.1709,
      "mean_token_accuracy": 0.18910606801509858,
      "num_tokens": 22245984.0,
      "step": 12825
    },
    {
      "entropy": 5.523936986923218,
      "epoch": 0.9982493678272709,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004905132822161962,
      "loss": 5.2307,
      "mean_token_accuracy": 0.18591742366552352,
      "num_tokens": 22254122.0,
      "step": 12830
    },
    {
      "entropy": 5.529425859451294,
      "epoch": 0.9986383971989885,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004905053175380621,
      "loss": 5.2462,
      "mean_token_accuracy": 0.17681712806224822,
      "num_tokens": 22263027.0,
      "step": 12835
    },
    {
      "entropy": 5.521054887771607,
      "epoch": 0.9990274265707061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904973495899778,
      "loss": 5.1666,
      "mean_token_accuracy": 0.18843168914318084,
      "num_tokens": 22272175.0,
      "step": 12840
    },
    {
      "entropy": 5.564848470687866,
      "epoch": 0.9994164559424237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004904893783720642,
      "loss": 5.2302,
      "mean_token_accuracy": 0.18553511798381805,
      "num_tokens": 22281127.0,
      "step": 12845
    },
    {
      "entropy": 5.600316762924194,
      "epoch": 0.9998054853141413,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004904814038844424,
      "loss": 5.3276,
      "mean_token_accuracy": 0.17699699848890305,
      "num_tokens": 22289696.0,
      "step": 12850
    },
    {
      "entropy": 5.570947435167101,
      "epoch": 1.000155611748687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904734261272332,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1872315224674013,
      "num_tokens": 22298101.0,
      "step": 12855
    },
    {
      "entropy": 5.566927242279053,
      "epoch": 1.0005446411204046,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004904654451005576,
      "loss": 5.123,
      "mean_token_accuracy": 0.18937319964170457,
      "num_tokens": 22305872.0,
      "step": 12860
    },
    {
      "entropy": 5.5282454013824465,
      "epoch": 1.0009336704921221,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004904574608045369,
      "loss": 5.1391,
      "mean_token_accuracy": 0.18944088220596314,
      "num_tokens": 22314743.0,
      "step": 12865
    },
    {
      "entropy": 5.594086933135986,
      "epoch": 1.0013226998638398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004904494732392923,
      "loss": 5.1582,
      "mean_token_accuracy": 0.18941092044115065,
      "num_tokens": 22324158.0,
      "step": 12870
    },
    {
      "entropy": 5.50680832862854,
      "epoch": 1.0017117292355573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904414824049448,
      "loss": 5.0604,
      "mean_token_accuracy": 0.19448835402727127,
      "num_tokens": 22333263.0,
      "step": 12875
    },
    {
      "entropy": 5.497782754898071,
      "epoch": 1.0021007586072748,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004904334883016156,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18971726000308992,
      "num_tokens": 22341053.0,
      "step": 12880
    },
    {
      "entropy": 5.503505611419678,
      "epoch": 1.0024897879789925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904254909294261,
      "loss": 5.0366,
      "mean_token_accuracy": 0.19540570229291915,
      "num_tokens": 22348924.0,
      "step": 12885
    },
    {
      "entropy": 5.5107889652252195,
      "epoch": 1.00287881735071,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004904174902884978,
      "loss": 5.1182,
      "mean_token_accuracy": 0.18857326805591584,
      "num_tokens": 22356738.0,
      "step": 12890
    },
    {
      "entropy": 5.547894430160523,
      "epoch": 1.0032678467224276,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904094863789519,
      "loss": 5.1509,
      "mean_token_accuracy": 0.18756003230810164,
      "num_tokens": 22365461.0,
      "step": 12895
    },
    {
      "entropy": 5.601795101165772,
      "epoch": 1.0036568760941451,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049040147920091,
      "loss": 5.1087,
      "mean_token_accuracy": 0.18588180989027023,
      "num_tokens": 22374895.0,
      "step": 12900
    },
    {
      "entropy": 5.556158399581909,
      "epoch": 1.0040459054658626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903934687544933,
      "loss": 5.1441,
      "mean_token_accuracy": 0.18870716392993928,
      "num_tokens": 22384276.0,
      "step": 12905
    },
    {
      "entropy": 5.582968282699585,
      "epoch": 1.0044349348375803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903854550398237,
      "loss": 5.1022,
      "mean_token_accuracy": 0.18820783942937852,
      "num_tokens": 22392415.0,
      "step": 12910
    },
    {
      "entropy": 5.574676084518432,
      "epoch": 1.0048239642092978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903774380570226,
      "loss": 5.1983,
      "mean_token_accuracy": 0.1800116553902626,
      "num_tokens": 22402054.0,
      "step": 12915
    },
    {
      "entropy": 5.520833683013916,
      "epoch": 1.0052129935810155,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903694178062117,
      "loss": 5.0742,
      "mean_token_accuracy": 0.18684456497430801,
      "num_tokens": 22410505.0,
      "step": 12920
    },
    {
      "entropy": 5.446739435195923,
      "epoch": 1.005602022952733,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903613942875126,
      "loss": 5.0316,
      "mean_token_accuracy": 0.1939515694975853,
      "num_tokens": 22419718.0,
      "step": 12925
    },
    {
      "entropy": 5.534476566314697,
      "epoch": 1.0059910523244504,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903533675010472,
      "loss": 5.068,
      "mean_token_accuracy": 0.2003800392150879,
      "num_tokens": 22427787.0,
      "step": 12930
    },
    {
      "entropy": 5.562624883651734,
      "epoch": 1.006380081696168,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004903453374469373,
      "loss": 5.1275,
      "mean_token_accuracy": 0.18749895989894866,
      "num_tokens": 22436385.0,
      "step": 12935
    },
    {
      "entropy": 5.477677631378174,
      "epoch": 1.0067691110678856,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004903373041253045,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19734283685684204,
      "num_tokens": 22444477.0,
      "step": 12940
    },
    {
      "entropy": 5.531784868240356,
      "epoch": 1.0071581404396033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004903292675362709,
      "loss": 5.0534,
      "mean_token_accuracy": 0.19506693929433822,
      "num_tokens": 22452412.0,
      "step": 12945
    },
    {
      "entropy": 5.480631399154663,
      "epoch": 1.0075471698113208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004903212276799584,
      "loss": 5.0428,
      "mean_token_accuracy": 0.20085077583789826,
      "num_tokens": 22461570.0,
      "step": 12950
    },
    {
      "entropy": 5.595657873153686,
      "epoch": 1.0079361991830382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903131845564889,
      "loss": 5.1544,
      "mean_token_accuracy": 0.1855711206793785,
      "num_tokens": 22469835.0,
      "step": 12955
    },
    {
      "entropy": 5.509905433654785,
      "epoch": 1.008325228554756,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004903051381659847,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19227265417575837,
      "num_tokens": 22478771.0,
      "step": 12960
    },
    {
      "entropy": 5.4361261367797855,
      "epoch": 1.0087142579264734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902970885085677,
      "loss": 5.0277,
      "mean_token_accuracy": 0.1993486315011978,
      "num_tokens": 22487138.0,
      "step": 12965
    },
    {
      "entropy": 5.473689126968384,
      "epoch": 1.009103287298191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049028903558436,
      "loss": 5.0955,
      "mean_token_accuracy": 0.19316711127758027,
      "num_tokens": 22495384.0,
      "step": 12970
    },
    {
      "entropy": 5.627360773086548,
      "epoch": 1.0094923166699086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902809793934838,
      "loss": 5.1869,
      "mean_token_accuracy": 0.1792323723435402,
      "num_tokens": 22503521.0,
      "step": 12975
    },
    {
      "entropy": 5.583735036849975,
      "epoch": 1.009881346041626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004902729199360615,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1965876966714859,
      "num_tokens": 22512155.0,
      "step": 12980
    },
    {
      "entropy": 5.59104642868042,
      "epoch": 1.0102703754133437,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004902648572122153,
      "loss": 5.1687,
      "mean_token_accuracy": 0.18694278597831726,
      "num_tokens": 22519844.0,
      "step": 12985
    },
    {
      "entropy": 5.490280294418335,
      "epoch": 1.0106594047850612,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902567912220674,
      "loss": 5.1123,
      "mean_token_accuracy": 0.18605371415615082,
      "num_tokens": 22529123.0,
      "step": 12990
    },
    {
      "entropy": 5.588703155517578,
      "epoch": 1.011048434156779,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902487219657404,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19163298308849336,
      "num_tokens": 22537632.0,
      "step": 12995
    },
    {
      "entropy": 5.545936059951782,
      "epoch": 1.0114374635284964,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902406494433566,
      "loss": 5.0418,
      "mean_token_accuracy": 0.1861702486872673,
      "num_tokens": 22546370.0,
      "step": 13000
    },
    {
      "entropy": 5.516361236572266,
      "epoch": 1.0118264929002139,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004902325736550386,
      "loss": 5.1246,
      "mean_token_accuracy": 0.1919540986418724,
      "num_tokens": 22554527.0,
      "step": 13005
    },
    {
      "entropy": 5.538467264175415,
      "epoch": 1.0122155222719316,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902244946009088,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1949560597538948,
      "num_tokens": 22563803.0,
      "step": 13010
    },
    {
      "entropy": 5.562205600738525,
      "epoch": 1.012604551643649,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902164122810899,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19037722945213317,
      "num_tokens": 22572333.0,
      "step": 13015
    },
    {
      "entropy": 5.459561347961426,
      "epoch": 1.0129935810153667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004902083266957045,
      "loss": 5.0213,
      "mean_token_accuracy": 0.20114053040742874,
      "num_tokens": 22580376.0,
      "step": 13020
    },
    {
      "entropy": 5.504111480712891,
      "epoch": 1.0133826103870842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902002378448753,
      "loss": 5.1692,
      "mean_token_accuracy": 0.18777503669261933,
      "num_tokens": 22589074.0,
      "step": 13025
    },
    {
      "entropy": 5.517139339447022,
      "epoch": 1.0137716397588017,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490192145728725,
      "loss": 4.9898,
      "mean_token_accuracy": 0.19804249405860902,
      "num_tokens": 22596905.0,
      "step": 13030
    },
    {
      "entropy": 5.461124849319458,
      "epoch": 1.0141606691305194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901840503473764,
      "loss": 5.1232,
      "mean_token_accuracy": 0.18640264123678207,
      "num_tokens": 22605556.0,
      "step": 13035
    },
    {
      "entropy": 5.5397154808044435,
      "epoch": 1.0145496985022369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901759517009522,
      "loss": 5.0884,
      "mean_token_accuracy": 0.19742010533809662,
      "num_tokens": 22613443.0,
      "step": 13040
    },
    {
      "entropy": 5.448122787475586,
      "epoch": 1.0149387278739546,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901678497895755,
      "loss": 5.031,
      "mean_token_accuracy": 0.19067345708608627,
      "num_tokens": 22622287.0,
      "step": 13045
    },
    {
      "entropy": 5.486344480514527,
      "epoch": 1.015327757245672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901597446133692,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1972237393260002,
      "num_tokens": 22630961.0,
      "step": 13050
    },
    {
      "entropy": 5.546230983734131,
      "epoch": 1.0157167866173895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901516361724564,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19114070534706115,
      "num_tokens": 22639325.0,
      "step": 13055
    },
    {
      "entropy": 5.57363920211792,
      "epoch": 1.0161058159891072,
      "grad_norm": 1.0,
      "learning_rate": 0.0004901435244669597,
      "loss": 5.1301,
      "mean_token_accuracy": 0.1863555818796158,
      "num_tokens": 22649028.0,
      "step": 13060
    },
    {
      "entropy": 5.521204805374145,
      "epoch": 1.0164948453608247,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004901354094970025,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1956435263156891,
      "num_tokens": 22657742.0,
      "step": 13065
    },
    {
      "entropy": 5.550022315979004,
      "epoch": 1.0168838747325424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004901272912627081,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1880657345056534,
      "num_tokens": 22665807.0,
      "step": 13070
    },
    {
      "entropy": 5.452243185043335,
      "epoch": 1.0172729041042599,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901191697641992,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20027755498886107,
      "num_tokens": 22674182.0,
      "step": 13075
    },
    {
      "entropy": 5.450249338150025,
      "epoch": 1.0176619334759773,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004901110450015994,
      "loss": 5.1258,
      "mean_token_accuracy": 0.18456079959869384,
      "num_tokens": 22683129.0,
      "step": 13080
    },
    {
      "entropy": 5.573884296417236,
      "epoch": 1.018050962847695,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901029169750319,
      "loss": 5.1753,
      "mean_token_accuracy": 0.19053862243890762,
      "num_tokens": 22692394.0,
      "step": 13085
    },
    {
      "entropy": 5.495051145553589,
      "epoch": 1.0184399922194125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004900947856846201,
      "loss": 4.9741,
      "mean_token_accuracy": 0.19873649179935454,
      "num_tokens": 22700830.0,
      "step": 13090
    },
    {
      "entropy": 5.604632806777954,
      "epoch": 1.0188290215911302,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490086651130487,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19330211281776427,
      "num_tokens": 22709796.0,
      "step": 13095
    },
    {
      "entropy": 5.4647266387939455,
      "epoch": 1.0192180509628477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900785133127566,
      "loss": 5.0712,
      "mean_token_accuracy": 0.1953159302473068,
      "num_tokens": 22717550.0,
      "step": 13100
    },
    {
      "entropy": 5.507863855361938,
      "epoch": 1.0196070803345652,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490070372231552,
      "loss": 5.1308,
      "mean_token_accuracy": 0.1866422951221466,
      "num_tokens": 22726630.0,
      "step": 13105
    },
    {
      "entropy": 5.598577880859375,
      "epoch": 1.0199961097062828,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900622278869968,
      "loss": 5.1299,
      "mean_token_accuracy": 0.1922068029642105,
      "num_tokens": 22735364.0,
      "step": 13110
    },
    {
      "entropy": 5.517694759368896,
      "epoch": 1.0203851390780003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900540802792146,
      "loss": 5.009,
      "mean_token_accuracy": 0.19232653230428695,
      "num_tokens": 22743910.0,
      "step": 13115
    },
    {
      "entropy": 5.473497533798218,
      "epoch": 1.020774168449718,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490045929408329,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19305328279733658,
      "num_tokens": 22753061.0,
      "step": 13120
    },
    {
      "entropy": 5.54612545967102,
      "epoch": 1.0211631978214355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900377752744637,
      "loss": 5.0904,
      "mean_token_accuracy": 0.19209272563457488,
      "num_tokens": 22761401.0,
      "step": 13125
    },
    {
      "entropy": 5.586156034469605,
      "epoch": 1.021552227193153,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900296178777426,
      "loss": 5.1661,
      "mean_token_accuracy": 0.18434716612100602,
      "num_tokens": 22769307.0,
      "step": 13130
    },
    {
      "entropy": 5.5873126029968265,
      "epoch": 1.0219412565648707,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490021457218289,
      "loss": 5.2136,
      "mean_token_accuracy": 0.1899673029780388,
      "num_tokens": 22778159.0,
      "step": 13135
    },
    {
      "entropy": 5.46327052116394,
      "epoch": 1.0223302859365881,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004900132932962272,
      "loss": 4.9486,
      "mean_token_accuracy": 0.19610001742839814,
      "num_tokens": 22785872.0,
      "step": 13140
    },
    {
      "entropy": 5.537806367874145,
      "epoch": 1.0227193153083058,
      "grad_norm": 1.109375,
      "learning_rate": 0.000490005126111681,
      "loss": 5.123,
      "mean_token_accuracy": 0.19678149819374086,
      "num_tokens": 22794531.0,
      "step": 13145
    },
    {
      "entropy": 5.54573769569397,
      "epoch": 1.0231083446800233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000489996955664774,
      "loss": 5.1454,
      "mean_token_accuracy": 0.18805739134550095,
      "num_tokens": 22803246.0,
      "step": 13150
    },
    {
      "entropy": 5.502979040145874,
      "epoch": 1.023497374051741,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004899887819556306,
      "loss": 5.081,
      "mean_token_accuracy": 0.1919192299246788,
      "num_tokens": 22811316.0,
      "step": 13155
    },
    {
      "entropy": 5.521839046478272,
      "epoch": 1.0238864034234585,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899806049843745,
      "loss": 5.0778,
      "mean_token_accuracy": 0.18884354829788208,
      "num_tokens": 22820754.0,
      "step": 13160
    },
    {
      "entropy": 5.569931507110596,
      "epoch": 1.024275432795176,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004899724247511299,
      "loss": 5.176,
      "mean_token_accuracy": 0.18050881773233413,
      "num_tokens": 22829185.0,
      "step": 13165
    },
    {
      "entropy": 5.604481363296509,
      "epoch": 1.0246644621668937,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489964241256021,
      "loss": 5.1757,
      "mean_token_accuracy": 0.18873506635427476,
      "num_tokens": 22838379.0,
      "step": 13170
    },
    {
      "entropy": 5.504822444915772,
      "epoch": 1.0250534915386111,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899560544991718,
      "loss": 5.1359,
      "mean_token_accuracy": 0.18978343307971954,
      "num_tokens": 22847186.0,
      "step": 13175
    },
    {
      "entropy": 5.509841251373291,
      "epoch": 1.0254425209103288,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899478644807068,
      "loss": 5.1588,
      "mean_token_accuracy": 0.19042373597621917,
      "num_tokens": 22856882.0,
      "step": 13180
    },
    {
      "entropy": 5.531721305847168,
      "epoch": 1.0258315502820463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899396712007498,
      "loss": 5.057,
      "mean_token_accuracy": 0.19042808711528778,
      "num_tokens": 22865747.0,
      "step": 13185
    },
    {
      "entropy": 5.408338928222657,
      "epoch": 1.0262205796537638,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004899314746594256,
      "loss": 4.9172,
      "mean_token_accuracy": 0.20211312025785447,
      "num_tokens": 22873847.0,
      "step": 13190
    },
    {
      "entropy": 5.4934186935424805,
      "epoch": 1.0266096090254815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899232748568584,
      "loss": 5.0577,
      "mean_token_accuracy": 0.1899716481566429,
      "num_tokens": 22882333.0,
      "step": 13195
    },
    {
      "entropy": 5.520166110992432,
      "epoch": 1.026998638397199,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004899150717931724,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19532584697008132,
      "num_tokens": 22891344.0,
      "step": 13200
    },
    {
      "entropy": 5.534826707839966,
      "epoch": 1.0273876677689167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899068654684924,
      "loss": 5.1456,
      "mean_token_accuracy": 0.18749040216207505,
      "num_tokens": 22899783.0,
      "step": 13205
    },
    {
      "entropy": 5.508007287979126,
      "epoch": 1.0277766971406341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898986558829428,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19265253245830535,
      "num_tokens": 22907274.0,
      "step": 13210
    },
    {
      "entropy": 5.538780403137207,
      "epoch": 1.0281657265123516,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004898904430366479,
      "loss": 5.1805,
      "mean_token_accuracy": 0.18277360051870345,
      "num_tokens": 22915744.0,
      "step": 13215
    },
    {
      "entropy": 5.588352537155151,
      "epoch": 1.0285547558840693,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898822269297328,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1881759434938431,
      "num_tokens": 22924244.0,
      "step": 13220
    },
    {
      "entropy": 5.5746777057647705,
      "epoch": 1.0289437852557868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898740075623218,
      "loss": 5.1495,
      "mean_token_accuracy": 0.19108967781066893,
      "num_tokens": 22932676.0,
      "step": 13225
    },
    {
      "entropy": 5.519115591049195,
      "epoch": 1.0293328146275045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004898657849345399,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19305302202701569,
      "num_tokens": 22941327.0,
      "step": 13230
    },
    {
      "entropy": 5.504632759094238,
      "epoch": 1.029721843999222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004898575590465116,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19412790685892106,
      "num_tokens": 22949540.0,
      "step": 13235
    },
    {
      "entropy": 5.571454381942749,
      "epoch": 1.0301108733709394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898493298983619,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1841746151447296,
      "num_tokens": 22958375.0,
      "step": 13240
    },
    {
      "entropy": 5.582480335235596,
      "epoch": 1.0304999027426571,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004898410974902155,
      "loss": 5.1198,
      "mean_token_accuracy": 0.1899153023958206,
      "num_tokens": 22966594.0,
      "step": 13245
    },
    {
      "entropy": 5.59900689125061,
      "epoch": 1.0308889321143746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898328618221974,
      "loss": 5.1851,
      "mean_token_accuracy": 0.17713644504547119,
      "num_tokens": 22975208.0,
      "step": 13250
    },
    {
      "entropy": 5.556050157546997,
      "epoch": 1.0312779614860923,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004898246228944327,
      "loss": 5.1365,
      "mean_token_accuracy": 0.18322844803333282,
      "num_tokens": 22983563.0,
      "step": 13255
    },
    {
      "entropy": 5.474011850357056,
      "epoch": 1.0316669908578098,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898163807070461,
      "loss": 5.045,
      "mean_token_accuracy": 0.19628906548023223,
      "num_tokens": 22992866.0,
      "step": 13260
    },
    {
      "entropy": 5.508804988861084,
      "epoch": 1.0320560202295272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004898081352601629,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19585597068071364,
      "num_tokens": 23001536.0,
      "step": 13265
    },
    {
      "entropy": 5.492587232589722,
      "epoch": 1.032445049601245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897998865539082,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19023662507534028,
      "num_tokens": 23009712.0,
      "step": 13270
    },
    {
      "entropy": 5.486681318283081,
      "epoch": 1.0328340789729624,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004897916345884069,
      "loss": 4.9991,
      "mean_token_accuracy": 0.19824901670217515,
      "num_tokens": 23017970.0,
      "step": 13275
    },
    {
      "entropy": 5.500798749923706,
      "epoch": 1.0332231083446801,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004897833793637847,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19602361172437668,
      "num_tokens": 23026231.0,
      "step": 13280
    },
    {
      "entropy": 5.586293888092041,
      "epoch": 1.0336121377163976,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897751208801665,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1916987344622612,
      "num_tokens": 23034962.0,
      "step": 13285
    },
    {
      "entropy": 5.521128988265991,
      "epoch": 1.034001167088115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897668591376777,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18796224445104598,
      "num_tokens": 23045105.0,
      "step": 13290
    },
    {
      "entropy": 5.523331880569458,
      "epoch": 1.0343901964598328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897585941364436,
      "loss": 5.0959,
      "mean_token_accuracy": 0.18961455225944518,
      "num_tokens": 23054315.0,
      "step": 13295
    },
    {
      "entropy": 5.581834506988526,
      "epoch": 1.0347792258315502,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897503258765896,
      "loss": 5.1295,
      "mean_token_accuracy": 0.18837738633155823,
      "num_tokens": 23063089.0,
      "step": 13300
    },
    {
      "entropy": 5.547751426696777,
      "epoch": 1.035168255203268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004897420543582413,
      "loss": 5.131,
      "mean_token_accuracy": 0.1804849848151207,
      "num_tokens": 23071827.0,
      "step": 13305
    },
    {
      "entropy": 5.469673919677734,
      "epoch": 1.0355572845749854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000489733779581524,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1970272183418274,
      "num_tokens": 23080090.0,
      "step": 13310
    },
    {
      "entropy": 5.529992628097534,
      "epoch": 1.035946313946703,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004897255015465635,
      "loss": 5.1499,
      "mean_token_accuracy": 0.18443081974983216,
      "num_tokens": 23089201.0,
      "step": 13315
    },
    {
      "entropy": 5.553167629241943,
      "epoch": 1.0363353433184206,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897172202534853,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19346014112234117,
      "num_tokens": 23097708.0,
      "step": 13320
    },
    {
      "entropy": 5.566406297683716,
      "epoch": 1.036724372690138,
      "grad_norm": 1.046875,
      "learning_rate": 0.000489708935702415,
      "loss": 5.1994,
      "mean_token_accuracy": 0.17937947511672975,
      "num_tokens": 23107554.0,
      "step": 13325
    },
    {
      "entropy": 5.511495876312256,
      "epoch": 1.0371134020618558,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897006478934784,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19319655746221542,
      "num_tokens": 23116781.0,
      "step": 13330
    },
    {
      "entropy": 5.450475072860717,
      "epoch": 1.0375024314335732,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896923568268011,
      "loss": 5.0409,
      "mean_token_accuracy": 0.19122885167598724,
      "num_tokens": 23125490.0,
      "step": 13335
    },
    {
      "entropy": 5.550891923904419,
      "epoch": 1.0378914608052907,
      "grad_norm": 1.15625,
      "learning_rate": 0.000489684062502509,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1895194470882416,
      "num_tokens": 23133693.0,
      "step": 13340
    },
    {
      "entropy": 5.409361886978149,
      "epoch": 1.0382804901770084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004896757649207281,
      "loss": 4.98,
      "mean_token_accuracy": 0.19491089433431624,
      "num_tokens": 23142888.0,
      "step": 13345
    },
    {
      "entropy": 5.5186511993408205,
      "epoch": 1.0386695195487259,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489667464081584,
      "loss": 4.9968,
      "mean_token_accuracy": 0.19732685983181,
      "num_tokens": 23151413.0,
      "step": 13350
    },
    {
      "entropy": 5.520129728317261,
      "epoch": 1.0390585489204436,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004896591599852029,
      "loss": 5.0228,
      "mean_token_accuracy": 0.19055050760507583,
      "num_tokens": 23159545.0,
      "step": 13355
    },
    {
      "entropy": 5.542135095596313,
      "epoch": 1.039447578292161,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004896508526317107,
      "loss": 5.0923,
      "mean_token_accuracy": 0.19229323863983155,
      "num_tokens": 23167963.0,
      "step": 13360
    },
    {
      "entropy": 5.519417572021484,
      "epoch": 1.0398366076638785,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004896425420212334,
      "loss": 5.096,
      "mean_token_accuracy": 0.18962192088365554,
      "num_tokens": 23176505.0,
      "step": 13365
    },
    {
      "entropy": 5.476184415817261,
      "epoch": 1.0402256370355962,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004896342281538973,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19449099898338318,
      "num_tokens": 23184776.0,
      "step": 13370
    },
    {
      "entropy": 5.584010219573974,
      "epoch": 1.0406146664073137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004896259110298283,
      "loss": 5.1527,
      "mean_token_accuracy": 0.19312323331832887,
      "num_tokens": 23193284.0,
      "step": 13375
    },
    {
      "entropy": 5.609039258956909,
      "epoch": 1.0410036957790314,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004896175906491528,
      "loss": 5.1033,
      "mean_token_accuracy": 0.18922057449817659,
      "num_tokens": 23201664.0,
      "step": 13380
    },
    {
      "entropy": 5.530050992965698,
      "epoch": 1.0413927251507489,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896092670119968,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19316325038671495,
      "num_tokens": 23210831.0,
      "step": 13385
    },
    {
      "entropy": 5.481321811676025,
      "epoch": 1.0417817545224664,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004896009401184869,
      "loss": 5.0875,
      "mean_token_accuracy": 0.18993585109710692,
      "num_tokens": 23218945.0,
      "step": 13390
    },
    {
      "entropy": 5.501623487472534,
      "epoch": 1.042170783894184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895926099687493,
      "loss": 5.1051,
      "mean_token_accuracy": 0.18563676625490189,
      "num_tokens": 23227922.0,
      "step": 13395
    },
    {
      "entropy": 5.517203712463379,
      "epoch": 1.0425598132659015,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004895842765629104,
      "loss": 5.0465,
      "mean_token_accuracy": 0.19482695907354355,
      "num_tokens": 23237726.0,
      "step": 13400
    },
    {
      "entropy": 5.528455829620361,
      "epoch": 1.0429488426376192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895759399010966,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18443805128335952,
      "num_tokens": 23246237.0,
      "step": 13405
    },
    {
      "entropy": 5.458320951461792,
      "epoch": 1.0433378720093367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895675999834345,
      "loss": 5.0574,
      "mean_token_accuracy": 0.1902329295873642,
      "num_tokens": 23254883.0,
      "step": 13410
    },
    {
      "entropy": 5.583170700073242,
      "epoch": 1.0437269013810542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004895592568100505,
      "loss": 5.1936,
      "mean_token_accuracy": 0.18406160473823546,
      "num_tokens": 23263270.0,
      "step": 13415
    },
    {
      "entropy": 5.520224142074585,
      "epoch": 1.0441159307527719,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004895509103810714,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18978043347597123,
      "num_tokens": 23271823.0,
      "step": 13420
    },
    {
      "entropy": 5.528475332260132,
      "epoch": 1.0445049601244893,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895425606966237,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19269345551729203,
      "num_tokens": 23280382.0,
      "step": 13425
    },
    {
      "entropy": 5.560562562942505,
      "epoch": 1.044893989496207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000489534207756834,
      "loss": 5.1508,
      "mean_token_accuracy": 0.18927441239356996,
      "num_tokens": 23288107.0,
      "step": 13430
    },
    {
      "entropy": 5.482879304885865,
      "epoch": 1.0452830188679245,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895258515618293,
      "loss": 5.0621,
      "mean_token_accuracy": 0.19035191535949708,
      "num_tokens": 23295661.0,
      "step": 13435
    },
    {
      "entropy": 5.599353504180908,
      "epoch": 1.045672048239642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895174921117362,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1922643169760704,
      "num_tokens": 23304862.0,
      "step": 13440
    },
    {
      "entropy": 5.612282943725586,
      "epoch": 1.0460610776113597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895091294066816,
      "loss": 5.1348,
      "mean_token_accuracy": 0.19007511138916017,
      "num_tokens": 23312992.0,
      "step": 13445
    },
    {
      "entropy": 5.576588678359985,
      "epoch": 1.0464501069830772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004895007634467924,
      "loss": 5.1994,
      "mean_token_accuracy": 0.18481669574975967,
      "num_tokens": 23321942.0,
      "step": 13450
    },
    {
      "entropy": 5.515787935256958,
      "epoch": 1.0468391363547949,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894923942321955,
      "loss": 5.0924,
      "mean_token_accuracy": 0.18557893931865693,
      "num_tokens": 23330608.0,
      "step": 13455
    },
    {
      "entropy": 5.49301962852478,
      "epoch": 1.0472281657265123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004894840217630179,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19485004991292953,
      "num_tokens": 23339792.0,
      "step": 13460
    },
    {
      "entropy": 5.55698595046997,
      "epoch": 1.0476171950982298,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004894756460393868,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19522764384746552,
      "num_tokens": 23347813.0,
      "step": 13465
    },
    {
      "entropy": 5.437216520309448,
      "epoch": 1.0480062244699475,
      "grad_norm": 1.125,
      "learning_rate": 0.000489467267061429,
      "loss": 5.056,
      "mean_token_accuracy": 0.19620070308446885,
      "num_tokens": 23356110.0,
      "step": 13470
    },
    {
      "entropy": 5.529351329803466,
      "epoch": 1.048395253841665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004894588848292718,
      "loss": 5.1082,
      "mean_token_accuracy": 0.18862626254558562,
      "num_tokens": 23364249.0,
      "step": 13475
    },
    {
      "entropy": 5.567995834350586,
      "epoch": 1.0487842832133827,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894504993430425,
      "loss": 5.1286,
      "mean_token_accuracy": 0.1856210336089134,
      "num_tokens": 23373072.0,
      "step": 13480
    },
    {
      "entropy": 5.460308647155761,
      "epoch": 1.0491733125851002,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489442110602868,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18919277787208558,
      "num_tokens": 23381752.0,
      "step": 13485
    },
    {
      "entropy": 5.517527675628662,
      "epoch": 1.0495623419568176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004894337186088759,
      "loss": 5.1152,
      "mean_token_accuracy": 0.18922088891267777,
      "num_tokens": 23389872.0,
      "step": 13490
    },
    {
      "entropy": 5.481854724884033,
      "epoch": 1.0499513713285353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894253233611934,
      "loss": 5.0031,
      "mean_token_accuracy": 0.19523075222969055,
      "num_tokens": 23398226.0,
      "step": 13495
    },
    {
      "entropy": 5.505682039260864,
      "epoch": 1.0503404007002528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000489416924859948,
      "loss": 5.0467,
      "mean_token_accuracy": 0.19352334439754487,
      "num_tokens": 23406776.0,
      "step": 13500
    },
    {
      "entropy": 5.526216602325439,
      "epoch": 1.0507294300719705,
      "grad_norm": 1.046875,
      "learning_rate": 0.000489408523105267,
      "loss": 5.0972,
      "mean_token_accuracy": 0.190244422852993,
      "num_tokens": 23415034.0,
      "step": 13505
    },
    {
      "entropy": 5.51914610862732,
      "epoch": 1.051118459443688,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004894001180972779,
      "loss": 5.085,
      "mean_token_accuracy": 0.19196409583091736,
      "num_tokens": 23423826.0,
      "step": 13510
    },
    {
      "entropy": 5.446078491210938,
      "epoch": 1.0515074888154055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893917098361082,
      "loss": 5.0212,
      "mean_token_accuracy": 0.19590983539819717,
      "num_tokens": 23432166.0,
      "step": 13515
    },
    {
      "entropy": 5.512406253814698,
      "epoch": 1.0518965181871232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893832983218856,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1878315880894661,
      "num_tokens": 23441065.0,
      "step": 13520
    },
    {
      "entropy": 5.569066429138184,
      "epoch": 1.0522855475588406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004893748835547377,
      "loss": 5.1603,
      "mean_token_accuracy": 0.1871422678232193,
      "num_tokens": 23450514.0,
      "step": 13525
    },
    {
      "entropy": 5.566702365875244,
      "epoch": 1.0526745769305583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893664655347921,
      "loss": 5.1922,
      "mean_token_accuracy": 0.188896282017231,
      "num_tokens": 23459354.0,
      "step": 13530
    },
    {
      "entropy": 5.572860336303711,
      "epoch": 1.0530636063022758,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893580442621767,
      "loss": 5.1441,
      "mean_token_accuracy": 0.18836628794670104,
      "num_tokens": 23468646.0,
      "step": 13535
    },
    {
      "entropy": 5.487472820281982,
      "epoch": 1.0534526356739935,
      "grad_norm": 1.09375,
      "learning_rate": 0.000489349619737019,
      "loss": 5.0201,
      "mean_token_accuracy": 0.19957176446914673,
      "num_tokens": 23477042.0,
      "step": 13540
    },
    {
      "entropy": 5.435347223281861,
      "epoch": 1.053841665045711,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489341191959447,
      "loss": 5.0206,
      "mean_token_accuracy": 0.1936554104089737,
      "num_tokens": 23485760.0,
      "step": 13545
    },
    {
      "entropy": 5.4848167419433596,
      "epoch": 1.0542306944174284,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004893327609295887,
      "loss": 5.1018,
      "mean_token_accuracy": 0.19138745069503785,
      "num_tokens": 23494425.0,
      "step": 13550
    },
    {
      "entropy": 5.550305509567261,
      "epoch": 1.0546197237891461,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004893243266475719,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18880690336227418,
      "num_tokens": 23502667.0,
      "step": 13555
    },
    {
      "entropy": 5.54180006980896,
      "epoch": 1.0550087531608636,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004893158891135245,
      "loss": 5.119,
      "mean_token_accuracy": 0.19139880686998367,
      "num_tokens": 23511706.0,
      "step": 13560
    },
    {
      "entropy": 5.567579078674316,
      "epoch": 1.055397782532581,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004893074483275746,
      "loss": 5.1484,
      "mean_token_accuracy": 0.18930281102657318,
      "num_tokens": 23520678.0,
      "step": 13565
    },
    {
      "entropy": 5.519304895401001,
      "epoch": 1.0557868119042988,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004892990042898503,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19876897037029267,
      "num_tokens": 23530246.0,
      "step": 13570
    },
    {
      "entropy": 5.493484115600586,
      "epoch": 1.0561758412760163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004892905570004798,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1881159260869026,
      "num_tokens": 23539128.0,
      "step": 13575
    },
    {
      "entropy": 5.518405485153198,
      "epoch": 1.056564870647734,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489282106459591,
      "loss": 5.1833,
      "mean_token_accuracy": 0.1884162664413452,
      "num_tokens": 23548062.0,
      "step": 13580
    },
    {
      "entropy": 5.557589912414551,
      "epoch": 1.0569539000194514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004892736526673124,
      "loss": 5.2147,
      "mean_token_accuracy": 0.1793177217245102,
      "num_tokens": 23557148.0,
      "step": 13585
    },
    {
      "entropy": 5.481782865524292,
      "epoch": 1.0573429293911691,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004892651956237721,
      "loss": 5.0499,
      "mean_token_accuracy": 0.19795615524053572,
      "num_tokens": 23565949.0,
      "step": 13590
    },
    {
      "entropy": 5.520569801330566,
      "epoch": 1.0577319587628866,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004892567353290986,
      "loss": 5.1014,
      "mean_token_accuracy": 0.1934608742594719,
      "num_tokens": 23574216.0,
      "step": 13595
    },
    {
      "entropy": 5.502425861358643,
      "epoch": 1.058120988134604,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892482717834201,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19388829320669174,
      "num_tokens": 23582362.0,
      "step": 13600
    },
    {
      "entropy": 5.560519599914551,
      "epoch": 1.0585100175063218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004892398049868651,
      "loss": 5.1594,
      "mean_token_accuracy": 0.18354610800743104,
      "num_tokens": 23591737.0,
      "step": 13605
    },
    {
      "entropy": 5.490676975250244,
      "epoch": 1.0588990468780393,
      "grad_norm": 1.078125,
      "learning_rate": 0.000489231334939562,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19327364414930343,
      "num_tokens": 23600302.0,
      "step": 13610
    },
    {
      "entropy": 5.500968360900879,
      "epoch": 1.059288076249757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004892228616416395,
      "loss": 5.0961,
      "mean_token_accuracy": 0.18980936110019683,
      "num_tokens": 23609189.0,
      "step": 13615
    },
    {
      "entropy": 5.5319067478179935,
      "epoch": 1.0596771056214744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892143850932259,
      "loss": 5.1157,
      "mean_token_accuracy": 0.18929604291915894,
      "num_tokens": 23618174.0,
      "step": 13620
    },
    {
      "entropy": 5.481690263748169,
      "epoch": 1.060066134993192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048920590529445,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1978852465748787,
      "num_tokens": 23626908.0,
      "step": 13625
    },
    {
      "entropy": 5.532262134552002,
      "epoch": 1.0604551643649096,
      "grad_norm": 1.125,
      "learning_rate": 0.0004891974222454406,
      "loss": 5.0928,
      "mean_token_accuracy": 0.1863192781805992,
      "num_tokens": 23635379.0,
      "step": 13630
    },
    {
      "entropy": 5.529179573059082,
      "epoch": 1.060844193736627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891889359463261,
      "loss": 5.1171,
      "mean_token_accuracy": 0.18847307711839675,
      "num_tokens": 23644249.0,
      "step": 13635
    },
    {
      "entropy": 5.484790658950805,
      "epoch": 1.0612332231083448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004891804463972354,
      "loss": 5.1008,
      "mean_token_accuracy": 0.18955399692058564,
      "num_tokens": 23652438.0,
      "step": 13640
    },
    {
      "entropy": 5.528585386276245,
      "epoch": 1.0616222524800623,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004891719535982974,
      "loss": 5.014,
      "mean_token_accuracy": 0.1927129179239273,
      "num_tokens": 23660238.0,
      "step": 13645
    },
    {
      "entropy": 5.534489345550537,
      "epoch": 1.0620112818517797,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004891634575496408,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19337467849254608,
      "num_tokens": 23669213.0,
      "step": 13650
    },
    {
      "entropy": 5.502006912231446,
      "epoch": 1.0624003112234974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004891549582513946,
      "loss": 5.2033,
      "mean_token_accuracy": 0.18002655804157258,
      "num_tokens": 23677941.0,
      "step": 13655
    },
    {
      "entropy": 5.54436297416687,
      "epoch": 1.062789340595215,
      "grad_norm": 1.4375,
      "learning_rate": 0.000489146455703688,
      "loss": 5.0857,
      "mean_token_accuracy": 0.195508873462677,
      "num_tokens": 23686407.0,
      "step": 13660
    },
    {
      "entropy": 5.571394634246826,
      "epoch": 1.0631783699669326,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004891379499066495,
      "loss": 4.9981,
      "mean_token_accuracy": 0.19812499433755876,
      "num_tokens": 23694434.0,
      "step": 13665
    },
    {
      "entropy": 5.545417451858521,
      "epoch": 1.06356739933865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891294408604087,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19098694175481795,
      "num_tokens": 23702949.0,
      "step": 13670
    },
    {
      "entropy": 5.520367050170899,
      "epoch": 1.0639564287103676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004891209285650942,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18994503170251847,
      "num_tokens": 23710888.0,
      "step": 13675
    },
    {
      "entropy": 5.497464275360107,
      "epoch": 1.0643454580820852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004891124130208355,
      "loss": 5.049,
      "mean_token_accuracy": 0.20407042056322097,
      "num_tokens": 23718590.0,
      "step": 13680
    },
    {
      "entropy": 5.479694271087647,
      "epoch": 1.0647344874538027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004891038942277618,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19078048020601274,
      "num_tokens": 23726920.0,
      "step": 13685
    },
    {
      "entropy": 5.545700550079346,
      "epoch": 1.0651235168255204,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890953721860022,
      "loss": 5.1105,
      "mean_token_accuracy": 0.18301011025905609,
      "num_tokens": 23735984.0,
      "step": 13690
    },
    {
      "entropy": 5.616705989837646,
      "epoch": 1.065512546197238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890868468956862,
      "loss": 5.1735,
      "mean_token_accuracy": 0.18324353992938996,
      "num_tokens": 23745144.0,
      "step": 13695
    },
    {
      "entropy": 5.4905389785766605,
      "epoch": 1.0659015755689554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000489078318356943,
      "loss": 5.1052,
      "mean_token_accuracy": 0.18754354119300842,
      "num_tokens": 23753865.0,
      "step": 13700
    },
    {
      "entropy": 5.539251613616943,
      "epoch": 1.066290604940673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890697865699021,
      "loss": 5.1249,
      "mean_token_accuracy": 0.18820196390151978,
      "num_tokens": 23762275.0,
      "step": 13705
    },
    {
      "entropy": 5.540315437316894,
      "epoch": 1.0666796343123905,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004890612515346929,
      "loss": 5.059,
      "mean_token_accuracy": 0.1941820800304413,
      "num_tokens": 23770309.0,
      "step": 13710
    },
    {
      "entropy": 5.520441675186158,
      "epoch": 1.0670686636841082,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004890527132514448,
      "loss": 5.0897,
      "mean_token_accuracy": 0.18692035973072052,
      "num_tokens": 23778595.0,
      "step": 13715
    },
    {
      "entropy": 5.471551036834716,
      "epoch": 1.0674576930558257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004890441717202876,
      "loss": 5.0251,
      "mean_token_accuracy": 0.19196586608886718,
      "num_tokens": 23787122.0,
      "step": 13720
    },
    {
      "entropy": 5.537634372711182,
      "epoch": 1.0678467224275432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004890356269413507,
      "loss": 5.1219,
      "mean_token_accuracy": 0.18820492923259735,
      "num_tokens": 23795677.0,
      "step": 13725
    },
    {
      "entropy": 5.52154688835144,
      "epoch": 1.068235751799261,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489027078914764,
      "loss": 5.0499,
      "mean_token_accuracy": 0.19325429648160936,
      "num_tokens": 23803752.0,
      "step": 13730
    },
    {
      "entropy": 5.475596857070923,
      "epoch": 1.0686247811709784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004890185276406569,
      "loss": 5.0365,
      "mean_token_accuracy": 0.19341997653245926,
      "num_tokens": 23812545.0,
      "step": 13735
    },
    {
      "entropy": 5.528506135940551,
      "epoch": 1.069013810542696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004890099731191592,
      "loss": 5.1836,
      "mean_token_accuracy": 0.18966481387615203,
      "num_tokens": 23820971.0,
      "step": 13740
    },
    {
      "entropy": 5.569648599624633,
      "epoch": 1.0694028399144135,
      "grad_norm": 1.09375,
      "learning_rate": 0.000489001415350401,
      "loss": 5.1986,
      "mean_token_accuracy": 0.18508488982915877,
      "num_tokens": 23828987.0,
      "step": 13745
    },
    {
      "entropy": 5.510803890228272,
      "epoch": 1.069791869286131,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889928543345118,
      "loss": 5.0453,
      "mean_token_accuracy": 0.19472482055425644,
      "num_tokens": 23837399.0,
      "step": 13750
    },
    {
      "entropy": 5.562136173248291,
      "epoch": 1.0701808986578487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004889842900716217,
      "loss": 5.1713,
      "mean_token_accuracy": 0.18426414132118224,
      "num_tokens": 23846654.0,
      "step": 13755
    },
    {
      "entropy": 5.56233434677124,
      "epoch": 1.0705699280295662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004889757225618606,
      "loss": 5.1318,
      "mean_token_accuracy": 0.18338098675012587,
      "num_tokens": 23854961.0,
      "step": 13760
    },
    {
      "entropy": 5.567819309234619,
      "epoch": 1.0709589574012839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004889671518053584,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1829996943473816,
      "num_tokens": 23864093.0,
      "step": 13765
    },
    {
      "entropy": 5.566392469406128,
      "epoch": 1.0713479867730014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004889585778022453,
      "loss": 5.1887,
      "mean_token_accuracy": 0.19020708352327348,
      "num_tokens": 23872287.0,
      "step": 13770
    },
    {
      "entropy": 5.534811067581177,
      "epoch": 1.0717370161447188,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004889500005526514,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19526924788951874,
      "num_tokens": 23880747.0,
      "step": 13775
    },
    {
      "entropy": 5.486981439590454,
      "epoch": 1.0721260455164365,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004889414200567067,
      "loss": 5.1055,
      "mean_token_accuracy": 0.18355966061353685,
      "num_tokens": 23889374.0,
      "step": 13780
    },
    {
      "entropy": 5.609719848632812,
      "epoch": 1.072515074888154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889328363145415,
      "loss": 5.234,
      "mean_token_accuracy": 0.1825035646557808,
      "num_tokens": 23898090.0,
      "step": 13785
    },
    {
      "entropy": 5.487572479248047,
      "epoch": 1.0729041042598717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004889242493262859,
      "loss": 5.0385,
      "mean_token_accuracy": 0.19667478054761886,
      "num_tokens": 23906263.0,
      "step": 13790
    },
    {
      "entropy": 5.541490173339843,
      "epoch": 1.0732931336315892,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004889156590920704,
      "loss": 5.0638,
      "mean_token_accuracy": 0.19306663274765015,
      "num_tokens": 23914292.0,
      "step": 13795
    },
    {
      "entropy": 5.405428218841553,
      "epoch": 1.0736821630033067,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889070656120253,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19119308292865753,
      "num_tokens": 23923264.0,
      "step": 13800
    },
    {
      "entropy": 5.425667190551758,
      "epoch": 1.0740711923750244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004888984688862809,
      "loss": 5.0231,
      "mean_token_accuracy": 0.19385481625795364,
      "num_tokens": 23932340.0,
      "step": 13805
    },
    {
      "entropy": 5.6034478664398195,
      "epoch": 1.0744602217467418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888898689149677,
      "loss": 5.1931,
      "mean_token_accuracy": 0.18485065251588823,
      "num_tokens": 23940843.0,
      "step": 13810
    },
    {
      "entropy": 5.498393535614014,
      "epoch": 1.0748492511184595,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004888812656982162,
      "loss": 5.1886,
      "mean_token_accuracy": 0.18606418818235398,
      "num_tokens": 23949127.0,
      "step": 13815
    },
    {
      "entropy": 5.51581072807312,
      "epoch": 1.075238280490177,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888726592361569,
      "loss": 5.2607,
      "mean_token_accuracy": 0.18363263010978698,
      "num_tokens": 23958218.0,
      "step": 13820
    },
    {
      "entropy": 5.608628940582276,
      "epoch": 1.0756273098618945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004888640495289204,
      "loss": 5.0433,
      "mean_token_accuracy": 0.1902975246310234,
      "num_tokens": 23966330.0,
      "step": 13825
    },
    {
      "entropy": 5.58606538772583,
      "epoch": 1.0760163392336122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004888554365766374,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19019514918327332,
      "num_tokens": 23975719.0,
      "step": 13830
    },
    {
      "entropy": 5.454965353012085,
      "epoch": 1.0764053686053296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004888468203794385,
      "loss": 5.056,
      "mean_token_accuracy": 0.19712476283311844,
      "num_tokens": 23984115.0,
      "step": 13835
    },
    {
      "entropy": 5.525501108169555,
      "epoch": 1.0767943979770473,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004888382009374545,
      "loss": 5.1039,
      "mean_token_accuracy": 0.18318722695112227,
      "num_tokens": 23993242.0,
      "step": 13840
    },
    {
      "entropy": 5.502299213409424,
      "epoch": 1.0771834273487648,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004888295782508161,
      "loss": 5.1299,
      "mean_token_accuracy": 0.19632060676813126,
      "num_tokens": 24001812.0,
      "step": 13845
    },
    {
      "entropy": 5.490289926528931,
      "epoch": 1.0775724567204823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888209523196542,
      "loss": 5.0503,
      "mean_token_accuracy": 0.19435303211212157,
      "num_tokens": 24010112.0,
      "step": 13850
    },
    {
      "entropy": 5.523157596588135,
      "epoch": 1.0779614860922,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004888123231440997,
      "loss": 5.1812,
      "mean_token_accuracy": 0.18417596817016602,
      "num_tokens": 24018067.0,
      "step": 13855
    },
    {
      "entropy": 5.510356330871582,
      "epoch": 1.0783505154639175,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888036907242834,
      "loss": 5.0869,
      "mean_token_accuracy": 0.18788596242666245,
      "num_tokens": 24026771.0,
      "step": 13860
    },
    {
      "entropy": 5.510292196273804,
      "epoch": 1.0787395448356352,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004887950550603366,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1890761062502861,
      "num_tokens": 24035322.0,
      "step": 13865
    },
    {
      "entropy": 5.62044243812561,
      "epoch": 1.0791285742073526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004887864161523901,
      "loss": 5.1896,
      "mean_token_accuracy": 0.1821748599410057,
      "num_tokens": 24043871.0,
      "step": 13870
    },
    {
      "entropy": 5.522692108154297,
      "epoch": 1.0795176035790701,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004887777740005749,
      "loss": 5.077,
      "mean_token_accuracy": 0.1976902648806572,
      "num_tokens": 24052807.0,
      "step": 13875
    },
    {
      "entropy": 5.466462421417236,
      "epoch": 1.0799066329507878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004887691286050224,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19799642860889435,
      "num_tokens": 24061201.0,
      "step": 13880
    },
    {
      "entropy": 5.520284986495971,
      "epoch": 1.0802956623225053,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004887604799658635,
      "loss": 5.1241,
      "mean_token_accuracy": 0.19045908153057098,
      "num_tokens": 24070521.0,
      "step": 13885
    },
    {
      "entropy": 5.5125054836273195,
      "epoch": 1.080684691694223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004887518280832295,
      "loss": 4.9854,
      "mean_token_accuracy": 0.1991085186600685,
      "num_tokens": 24079941.0,
      "step": 13890
    },
    {
      "entropy": 5.52765007019043,
      "epoch": 1.0810737210659405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004887431729572519,
      "loss": 5.1109,
      "mean_token_accuracy": 0.193418750166893,
      "num_tokens": 24088696.0,
      "step": 13895
    },
    {
      "entropy": 5.477832651138305,
      "epoch": 1.081462750437658,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004887345145880617,
      "loss": 5.023,
      "mean_token_accuracy": 0.19367682486772536,
      "num_tokens": 24098285.0,
      "step": 13900
    },
    {
      "entropy": 5.466937065124512,
      "epoch": 1.0818517798093756,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004887258529757904,
      "loss": 5.004,
      "mean_token_accuracy": 0.19981971979141236,
      "num_tokens": 24107267.0,
      "step": 13905
    },
    {
      "entropy": 5.531937122344971,
      "epoch": 1.082240809181093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004887171881205696,
      "loss": 5.084,
      "mean_token_accuracy": 0.19397533237934111,
      "num_tokens": 24115755.0,
      "step": 13910
    },
    {
      "entropy": 5.511270236968994,
      "epoch": 1.0826298385528108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004887085200225306,
      "loss": 5.0282,
      "mean_token_accuracy": 0.198336361348629,
      "num_tokens": 24124207.0,
      "step": 13915
    },
    {
      "entropy": 5.520353078842163,
      "epoch": 1.0830188679245283,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004886998486818049,
      "loss": 5.1182,
      "mean_token_accuracy": 0.18697736114263536,
      "num_tokens": 24133771.0,
      "step": 13920
    },
    {
      "entropy": 5.495987892150879,
      "epoch": 1.083407897296246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886911740985242,
      "loss": 4.9972,
      "mean_token_accuracy": 0.19933753311634064,
      "num_tokens": 24142186.0,
      "step": 13925
    },
    {
      "entropy": 5.448917579650879,
      "epoch": 1.0837969266679635,
      "grad_norm": 1.0,
      "learning_rate": 0.00048868249627282,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19557779431343078,
      "num_tokens": 24151073.0,
      "step": 13930
    },
    {
      "entropy": 5.5553436279296875,
      "epoch": 1.084185956039681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488673815204824,
      "loss": 5.12,
      "mean_token_accuracy": 0.1866609275341034,
      "num_tokens": 24160092.0,
      "step": 13935
    },
    {
      "entropy": 5.532994174957276,
      "epoch": 1.0845749854113986,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886651308946681,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19020156860351561,
      "num_tokens": 24167982.0,
      "step": 13940
    },
    {
      "entropy": 5.50459361076355,
      "epoch": 1.084964014783116,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886564433424839,
      "loss": 5.1973,
      "mean_token_accuracy": 0.18243169486522676,
      "num_tokens": 24176753.0,
      "step": 13945
    },
    {
      "entropy": 5.500538301467896,
      "epoch": 1.0853530441548336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886477525484032,
      "loss": 5.0944,
      "mean_token_accuracy": 0.18765871822834015,
      "num_tokens": 24185923.0,
      "step": 13950
    },
    {
      "entropy": 5.508293962478637,
      "epoch": 1.0857420735265513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886390585125579,
      "loss": 4.9888,
      "mean_token_accuracy": 0.19479743093252183,
      "num_tokens": 24194614.0,
      "step": 13955
    },
    {
      "entropy": 5.638990974426269,
      "epoch": 1.0861311028982688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004886303612350799,
      "loss": 5.1832,
      "mean_token_accuracy": 0.18542629182338716,
      "num_tokens": 24203623.0,
      "step": 13960
    },
    {
      "entropy": 5.4925895690917965,
      "epoch": 1.0865201322699864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886216607161013,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20337249785661698,
      "num_tokens": 24211936.0,
      "step": 13965
    },
    {
      "entropy": 5.5153755187988285,
      "epoch": 1.086909161641704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886129569557538,
      "loss": 5.025,
      "mean_token_accuracy": 0.195184426009655,
      "num_tokens": 24219944.0,
      "step": 13970
    },
    {
      "entropy": 5.510942554473877,
      "epoch": 1.0872981910134216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886042499541699,
      "loss": 5.1479,
      "mean_token_accuracy": 0.1842548057436943,
      "num_tokens": 24228911.0,
      "step": 13975
    },
    {
      "entropy": 5.561695909500122,
      "epoch": 1.087687220385139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004885955397114813,
      "loss": 5.138,
      "mean_token_accuracy": 0.18774104565382005,
      "num_tokens": 24237997.0,
      "step": 13980
    },
    {
      "entropy": 5.528283786773682,
      "epoch": 1.0880762497568566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004885868262278205,
      "loss": 5.0966,
      "mean_token_accuracy": 0.1936579316854477,
      "num_tokens": 24247099.0,
      "step": 13985
    },
    {
      "entropy": 5.500274753570556,
      "epoch": 1.0884652791285743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885781095033195,
      "loss": 5.0585,
      "mean_token_accuracy": 0.1931614398956299,
      "num_tokens": 24255749.0,
      "step": 13990
    },
    {
      "entropy": 5.543603420257568,
      "epoch": 1.0888543085002917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004885693895381108,
      "loss": 5.1707,
      "mean_token_accuracy": 0.18787231892347336,
      "num_tokens": 24263891.0,
      "step": 13995
    },
    {
      "entropy": 5.481777477264404,
      "epoch": 1.0892433378720092,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004885606663323263,
      "loss": 5.0286,
      "mean_token_accuracy": 0.19781796932220458,
      "num_tokens": 24272660.0,
      "step": 14000
    },
    {
      "entropy": 5.522601556777954,
      "epoch": 1.089632367243727,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004885519398860987,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2010129302740097,
      "num_tokens": 24280922.0,
      "step": 14005
    },
    {
      "entropy": 5.49679970741272,
      "epoch": 1.0900213966154444,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885432101995604,
      "loss": 5.137,
      "mean_token_accuracy": 0.18826484829187393,
      "num_tokens": 24289630.0,
      "step": 14010
    },
    {
      "entropy": 5.424733257293701,
      "epoch": 1.090410425987162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885344772728436,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20214839428663253,
      "num_tokens": 24298310.0,
      "step": 14015
    },
    {
      "entropy": 5.439482021331787,
      "epoch": 1.0907994553588796,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004885257411060812,
      "loss": 4.9583,
      "mean_token_accuracy": 0.19974494576454163,
      "num_tokens": 24306839.0,
      "step": 14020
    },
    {
      "entropy": 5.494307184219361,
      "epoch": 1.0911884847305973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004885170016994053,
      "loss": 5.1582,
      "mean_token_accuracy": 0.18334333151578902,
      "num_tokens": 24315571.0,
      "step": 14025
    },
    {
      "entropy": 5.551033544540405,
      "epoch": 1.0915775141023147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004885082590529489,
      "loss": 5.1173,
      "mean_token_accuracy": 0.18992575258016586,
      "num_tokens": 24325137.0,
      "step": 14030
    },
    {
      "entropy": 5.584569787979126,
      "epoch": 1.0919665434740322,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004884995131668445,
      "loss": 5.1448,
      "mean_token_accuracy": 0.18923401534557344,
      "num_tokens": 24333993.0,
      "step": 14035
    },
    {
      "entropy": 5.587527322769165,
      "epoch": 1.09235557284575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884907640412248,
      "loss": 5.2145,
      "mean_token_accuracy": 0.18532725423574448,
      "num_tokens": 24342488.0,
      "step": 14040
    },
    {
      "entropy": 5.5090875148773195,
      "epoch": 1.0927446022174674,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884820116762225,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20387738943099976,
      "num_tokens": 24350250.0,
      "step": 14045
    },
    {
      "entropy": 5.491398191452026,
      "epoch": 1.093133631589185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884732560719706,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1878104627132416,
      "num_tokens": 24360375.0,
      "step": 14050
    },
    {
      "entropy": 5.509702062606811,
      "epoch": 1.0935226609609026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004884644972286017,
      "loss": 5.1955,
      "mean_token_accuracy": 0.17976798117160797,
      "num_tokens": 24368410.0,
      "step": 14055
    },
    {
      "entropy": 5.541974592208862,
      "epoch": 1.09391169033262,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004884557351462488,
      "loss": 5.0601,
      "mean_token_accuracy": 0.18985181152820588,
      "num_tokens": 24377618.0,
      "step": 14060
    },
    {
      "entropy": 5.567614793777466,
      "epoch": 1.0943007197043377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884469698250449,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19346515238285064,
      "num_tokens": 24385844.0,
      "step": 14065
    },
    {
      "entropy": 5.542163848876953,
      "epoch": 1.0946897490760552,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488438201265123,
      "loss": 5.2029,
      "mean_token_accuracy": 0.18775518834590912,
      "num_tokens": 24394411.0,
      "step": 14070
    },
    {
      "entropy": 5.553159189224243,
      "epoch": 1.095078778447773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004884294294666161,
      "loss": 5.1294,
      "mean_token_accuracy": 0.18309123814105988,
      "num_tokens": 24402791.0,
      "step": 14075
    },
    {
      "entropy": 5.495318078994751,
      "epoch": 1.0954678078194904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884206544296573,
      "loss": 5.0503,
      "mean_token_accuracy": 0.19533255845308303,
      "num_tokens": 24411268.0,
      "step": 14080
    },
    {
      "entropy": 5.491215753555298,
      "epoch": 1.0958568371912079,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884118761543797,
      "loss": 5.0305,
      "mean_token_accuracy": 0.19949003010988237,
      "num_tokens": 24419417.0,
      "step": 14085
    },
    {
      "entropy": 5.546337699890136,
      "epoch": 1.0962458665629256,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004884030946409167,
      "loss": 5.0755,
      "mean_token_accuracy": 0.19459165632724762,
      "num_tokens": 24427417.0,
      "step": 14090
    },
    {
      "entropy": 5.599336719512939,
      "epoch": 1.096634895934643,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004883943098894013,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1938049465417862,
      "num_tokens": 24436061.0,
      "step": 14095
    },
    {
      "entropy": 5.514726066589356,
      "epoch": 1.0970239253063607,
      "grad_norm": 1.109375,
      "learning_rate": 0.000488385521899967,
      "loss": 5.1476,
      "mean_token_accuracy": 0.18713939636945726,
      "num_tokens": 24444313.0,
      "step": 14100
    },
    {
      "entropy": 5.514452648162842,
      "epoch": 1.0974129546780782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000488376730672747,
      "loss": 5.0743,
      "mean_token_accuracy": 0.18727093935012817,
      "num_tokens": 24452971.0,
      "step": 14105
    },
    {
      "entropy": 5.551153135299683,
      "epoch": 1.0978019840497957,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883679362078746,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1925017550587654,
      "num_tokens": 24461858.0,
      "step": 14110
    },
    {
      "entropy": 5.557491302490234,
      "epoch": 1.0981910134215134,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883591385054836,
      "loss": 5.0785,
      "mean_token_accuracy": 0.192092627286911,
      "num_tokens": 24470694.0,
      "step": 14115
    },
    {
      "entropy": 5.529823589324951,
      "epoch": 1.0985800427932308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004883503375657072,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1922311305999756,
      "num_tokens": 24479668.0,
      "step": 14120
    },
    {
      "entropy": 5.561577653884887,
      "epoch": 1.0989690721649485,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883415333886789,
      "loss": 5.1202,
      "mean_token_accuracy": 0.18860125690698623,
      "num_tokens": 24487604.0,
      "step": 14125
    },
    {
      "entropy": 5.573582363128662,
      "epoch": 1.099358101536666,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004883327259745325,
      "loss": 5.0542,
      "mean_token_accuracy": 0.19323680102825164,
      "num_tokens": 24496847.0,
      "step": 14130
    },
    {
      "entropy": 5.596796894073487,
      "epoch": 1.0997471309083835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883239153234015,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1843985930085182,
      "num_tokens": 24505265.0,
      "step": 14135
    },
    {
      "entropy": 5.554666566848755,
      "epoch": 1.1001361602801012,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883151014354197,
      "loss": 5.1437,
      "mean_token_accuracy": 0.18750548511743545,
      "num_tokens": 24513772.0,
      "step": 14140
    },
    {
      "entropy": 5.519763469696045,
      "epoch": 1.1005251896518187,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883062843107207,
      "loss": 4.9834,
      "mean_token_accuracy": 0.1968691736459732,
      "num_tokens": 24521882.0,
      "step": 14145
    },
    {
      "entropy": 5.4848716259002686,
      "epoch": 1.1009142190235364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882974639494383,
      "loss": 5.0219,
      "mean_token_accuracy": 0.1945369452238083,
      "num_tokens": 24530876.0,
      "step": 14150
    },
    {
      "entropy": 5.395925569534302,
      "epoch": 1.1013032483952538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004882886403517065,
      "loss": 4.9856,
      "mean_token_accuracy": 0.19821424633264542,
      "num_tokens": 24540829.0,
      "step": 14155
    },
    {
      "entropy": 5.50280237197876,
      "epoch": 1.1016922777669713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004882798135176589,
      "loss": 5.088,
      "mean_token_accuracy": 0.19001563042402267,
      "num_tokens": 24549310.0,
      "step": 14160
    },
    {
      "entropy": 5.5018387794494625,
      "epoch": 1.102081307138689,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048827098344742973,
      "loss": 5.0411,
      "mean_token_accuracy": 0.19774002879858016,
      "num_tokens": 24558812.0,
      "step": 14165
    },
    {
      "entropy": 5.435721445083618,
      "epoch": 1.1024703365104065,
      "grad_norm": 1.125,
      "learning_rate": 0.0004882621501411528,
      "loss": 4.9819,
      "mean_token_accuracy": 0.19857096523046494,
      "num_tokens": 24567253.0,
      "step": 14170
    },
    {
      "entropy": 5.537141180038452,
      "epoch": 1.1028593658821242,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004882533135989622,
      "loss": 5.1763,
      "mean_token_accuracy": 0.18676015585660935,
      "num_tokens": 24576090.0,
      "step": 14175
    },
    {
      "entropy": 5.558816766738891,
      "epoch": 1.1032483952538417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000488244473820992,
      "loss": 5.1129,
      "mean_token_accuracy": 0.18958090841770173,
      "num_tokens": 24584826.0,
      "step": 14180
    },
    {
      "entropy": 5.5177881717681885,
      "epoch": 1.1036374246255591,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048823563080737634,
      "loss": 5.034,
      "mean_token_accuracy": 0.19506576508283616,
      "num_tokens": 24593363.0,
      "step": 14185
    },
    {
      "entropy": 5.480960512161255,
      "epoch": 1.1040264539972768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882267845582493,
      "loss": 5.1193,
      "mean_token_accuracy": 0.19619522392749786,
      "num_tokens": 24602391.0,
      "step": 14190
    },
    {
      "entropy": 5.5346204280853275,
      "epoch": 1.1044154833689943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048821793507374526,
      "loss": 5.1115,
      "mean_token_accuracy": 0.19084785282611846,
      "num_tokens": 24610986.0,
      "step": 14195
    },
    {
      "entropy": 5.446318674087524,
      "epoch": 1.104804512740712,
      "grad_norm": 1.125,
      "learning_rate": 0.0004882090823539984,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19422779530286788,
      "num_tokens": 24620304.0,
      "step": 14200
    },
    {
      "entropy": 5.562791919708252,
      "epoch": 1.1051935421124295,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882002263991431,
      "loss": 5.1226,
      "mean_token_accuracy": 0.18928630352020265,
      "num_tokens": 24628867.0,
      "step": 14205
    },
    {
      "entropy": 5.5842968940734865,
      "epoch": 1.105582571484147,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048819136720931364,
      "loss": 5.0919,
      "mean_token_accuracy": 0.19019657671451567,
      "num_tokens": 24637428.0,
      "step": 14210
    },
    {
      "entropy": 5.553181123733521,
      "epoch": 1.1059716008558647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048818250478464457,
      "loss": 5.0984,
      "mean_token_accuracy": 0.18389081805944443,
      "num_tokens": 24646065.0,
      "step": 14215
    },
    {
      "entropy": 5.577740955352783,
      "epoch": 1.1063606302275821,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004881736391252703,
      "loss": 5.0819,
      "mean_token_accuracy": 0.1892639711499214,
      "num_tokens": 24654118.0,
      "step": 14220
    },
    {
      "entropy": 5.494794130325317,
      "epoch": 1.1067496595992998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004881647702313253,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1982915908098221,
      "num_tokens": 24662127.0,
      "step": 14225
    },
    {
      "entropy": 5.41578426361084,
      "epoch": 1.1071386889710173,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048815589810294427,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19056423455476762,
      "num_tokens": 24671116.0,
      "step": 14230
    },
    {
      "entropy": 5.507541561126709,
      "epoch": 1.1075277183427348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048814702274026174,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1913096383213997,
      "num_tokens": 24679471.0,
      "step": 14235
    },
    {
      "entropy": 5.562342119216919,
      "epoch": 1.1079167477144525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004881381441434124,
      "loss": 5.1544,
      "mean_token_accuracy": 0.18864265829324722,
      "num_tokens": 24688575.0,
      "step": 14240
    },
    {
      "entropy": 5.494248342514038,
      "epoch": 1.10830577708617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000488129262312531,
      "loss": 5.048,
      "mean_token_accuracy": 0.2001872718334198,
      "num_tokens": 24696350.0,
      "step": 14245
    },
    {
      "entropy": 5.573805284500122,
      "epoch": 1.1086948064578876,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048812037724775217,
      "loss": 5.1692,
      "mean_token_accuracy": 0.19086775928735733,
      "num_tokens": 24705176.0,
      "step": 14250
    },
    {
      "entropy": 5.578050661087036,
      "epoch": 1.1090838358296051,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004881114889492109,
      "loss": 5.1382,
      "mean_token_accuracy": 0.19075945466756822,
      "num_tokens": 24714067.0,
      "step": 14255
    },
    {
      "entropy": 5.524376583099365,
      "epoch": 1.1094728652013226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048810259741704197,
      "loss": 5.0547,
      "mean_token_accuracy": 0.19222726076841354,
      "num_tokens": 24722915.0,
      "step": 14260
    },
    {
      "entropy": 5.53132734298706,
      "epoch": 1.1098618945730403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004880937026513803,
      "loss": 5.0924,
      "mean_token_accuracy": 0.1966213896870613,
      "num_tokens": 24731359.0,
      "step": 14265
    },
    {
      "entropy": 5.512762928009034,
      "epoch": 1.1102509239447578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004880848046523609,
      "loss": 5.1163,
      "mean_token_accuracy": 0.18796170353889466,
      "num_tokens": 24739537.0,
      "step": 14270
    },
    {
      "entropy": 5.535584163665772,
      "epoch": 1.1106399533164755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048807590342011875,
      "loss": 5.1893,
      "mean_token_accuracy": 0.1853242114186287,
      "num_tokens": 24749257.0,
      "step": 14275
    },
    {
      "entropy": 5.547473621368408,
      "epoch": 1.111028982688193,
      "grad_norm": 1.125,
      "learning_rate": 0.00048806699895478895,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19717821031808852,
      "num_tokens": 24757715.0,
      "step": 14280
    },
    {
      "entropy": 5.569555854797363,
      "epoch": 1.1114180120599104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004880580912565065,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19177425354719163,
      "num_tokens": 24765974.0,
      "step": 14285
    },
    {
      "entropy": 5.572491836547852,
      "epoch": 1.1118070414316281,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004880491803254067,
      "loss": 5.1092,
      "mean_token_accuracy": 0.1915055364370346,
      "num_tokens": 24774329.0,
      "step": 14290
    },
    {
      "entropy": 5.581033658981323,
      "epoch": 1.1121960708033456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004880402661616246,
      "loss": 5.1389,
      "mean_token_accuracy": 0.18946656882762908,
      "num_tokens": 24782509.0,
      "step": 14295
    },
    {
      "entropy": 5.474265098571777,
      "epoch": 1.1125851001750633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004880313487652956,
      "loss": 5.0067,
      "mean_token_accuracy": 0.19869548529386521,
      "num_tokens": 24790871.0,
      "step": 14300
    },
    {
      "entropy": 5.543883848190307,
      "epoch": 1.1129741295467808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048802242813655503,
      "loss": 5.0773,
      "mean_token_accuracy": 0.19381869584321976,
      "num_tokens": 24799870.0,
      "step": 14305
    },
    {
      "entropy": 5.465011501312256,
      "epoch": 1.1133631589184985,
      "grad_norm": 1.109375,
      "learning_rate": 0.000488013504275538,
      "loss": 5.0125,
      "mean_token_accuracy": 0.19025528579950332,
      "num_tokens": 24809337.0,
      "step": 14310
    },
    {
      "entropy": 5.431163692474366,
      "epoch": 1.113752188290216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004880045771823802,
      "loss": 5.0057,
      "mean_token_accuracy": 0.1957953691482544,
      "num_tokens": 24817724.0,
      "step": 14315
    },
    {
      "entropy": 5.535935258865356,
      "epoch": 1.1141412176619334,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048799564685721695,
      "loss": 5.1214,
      "mean_token_accuracy": 0.19439117908477782,
      "num_tokens": 24826299.0,
      "step": 14320
    },
    {
      "entropy": 5.48707070350647,
      "epoch": 1.1145302470336511,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004879867133001837,
      "loss": 5.0793,
      "mean_token_accuracy": 0.195002917945385,
      "num_tokens": 24835371.0,
      "step": 14325
    },
    {
      "entropy": 5.538400602340698,
      "epoch": 1.1149192764053686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004879777765114162,
      "loss": 5.0665,
      "mean_token_accuracy": 0.1913912519812584,
      "num_tokens": 24843230.0,
      "step": 14330
    },
    {
      "entropy": 5.510708332061768,
      "epoch": 1.115308305777086,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048796883649104996,
      "loss": 5.0203,
      "mean_token_accuracy": 0.19591016918420792,
      "num_tokens": 24852100.0,
      "step": 14335
    },
    {
      "entropy": 5.552013158798218,
      "epoch": 1.1156973351488038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004879598932392206,
      "loss": 5.1574,
      "mean_token_accuracy": 0.18482344895601271,
      "num_tokens": 24860591.0,
      "step": 14340
    },
    {
      "entropy": 5.51492748260498,
      "epoch": 1.1160863645205212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048795094675606364,
      "loss": 4.9946,
      "mean_token_accuracy": 0.1979513257741928,
      "num_tokens": 24868909.0,
      "step": 14345
    },
    {
      "entropy": 5.554353284835815,
      "epoch": 1.116475393892239,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879419970417152,
      "loss": 5.1306,
      "mean_token_accuracy": 0.18582761585712432,
      "num_tokens": 24877816.0,
      "step": 14350
    },
    {
      "entropy": 5.497121858596802,
      "epoch": 1.1168644232639564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048793304409631077,
      "loss": 4.9659,
      "mean_token_accuracy": 0.1990955501794815,
      "num_tokens": 24885871.0,
      "step": 14355
    },
    {
      "entropy": 5.46532301902771,
      "epoch": 1.117253452635674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004879240879199863,
      "loss": 5.1,
      "mean_token_accuracy": 0.19749603867530824,
      "num_tokens": 24894138.0,
      "step": 14360
    },
    {
      "entropy": 5.471949625015259,
      "epoch": 1.1176424820073916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048791512851287786,
      "loss": 5.019,
      "mean_token_accuracy": 0.19310873597860337,
      "num_tokens": 24903174.0,
      "step": 14365
    },
    {
      "entropy": 5.552790880203247,
      "epoch": 1.118031511379109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048790616587512107,
      "loss": 5.0366,
      "mean_token_accuracy": 0.19963193088769912,
      "num_tokens": 24911365.0,
      "step": 14370
    },
    {
      "entropy": 5.463595390319824,
      "epoch": 1.1184205407508268,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004878972000068521,
      "loss": 4.9731,
      "mean_token_accuracy": 0.1970108851790428,
      "num_tokens": 24919222.0,
      "step": 14375
    },
    {
      "entropy": 5.49577841758728,
      "epoch": 1.1188095701225442,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048788823090820707,
      "loss": 5.1415,
      "mean_token_accuracy": 0.1958930015563965,
      "num_tokens": 24927548.0,
      "step": 14380
    },
    {
      "entropy": 5.601397943496704,
      "epoch": 1.1191985994942617,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048787925857932194,
      "loss": 5.209,
      "mean_token_accuracy": 0.18764911592006683,
      "num_tokens": 24935888.0,
      "step": 14385
    },
    {
      "entropy": 5.59907078742981,
      "epoch": 1.1195876288659794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048787028302033287,
      "loss": 5.1799,
      "mean_token_accuracy": 0.1897195339202881,
      "num_tokens": 24944799.0,
      "step": 14390
    },
    {
      "entropy": 5.481232452392578,
      "epoch": 1.1199766582376969,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048786130423137606,
      "loss": 4.9976,
      "mean_token_accuracy": 0.198268386721611,
      "num_tokens": 24952748.0,
      "step": 14395
    },
    {
      "entropy": 5.533926057815552,
      "epoch": 1.1203656876094146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878523222125879,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18734532296657563,
      "num_tokens": 24961293.0,
      "step": 14400
    },
    {
      "entropy": 5.504284906387329,
      "epoch": 1.120754716981132,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004878433369641044,
      "loss": 5.1328,
      "mean_token_accuracy": 0.18297696411609649,
      "num_tokens": 24969192.0,
      "step": 14405
    },
    {
      "entropy": 5.4895867824554445,
      "epoch": 1.1211437463528497,
      "grad_norm": 1.125,
      "learning_rate": 0.0004878343484860621,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19968957901000978,
      "num_tokens": 24976872.0,
      "step": 14410
    },
    {
      "entropy": 5.529160118103027,
      "epoch": 1.1215327757245672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004878253567785972,
      "loss": 5.0849,
      "mean_token_accuracy": 0.19520883858203888,
      "num_tokens": 24984706.0,
      "step": 14415
    },
    {
      "entropy": 5.547358560562134,
      "epoch": 1.1219218050962847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048781636184184644,
      "loss": 5.1134,
      "mean_token_accuracy": 0.19159133732318878,
      "num_tokens": 24993873.0,
      "step": 14420
    },
    {
      "entropy": 5.509705352783203,
      "epoch": 1.1223108344680024,
      "grad_norm": 1.140625,
      "learning_rate": 0.000487807363675946,
      "loss": 5.0921,
      "mean_token_accuracy": 0.1895897015929222,
      "num_tokens": 25002232.0,
      "step": 14425
    },
    {
      "entropy": 5.584260892868042,
      "epoch": 1.1226998638397199,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004877983622810326,
      "loss": 5.061,
      "mean_token_accuracy": 0.1898328498005867,
      "num_tokens": 25010530.0,
      "step": 14430
    },
    {
      "entropy": 5.56055703163147,
      "epoch": 1.1230888932114376,
      "grad_norm": 1.125,
      "learning_rate": 0.0004877893576572427,
      "loss": 5.0864,
      "mean_token_accuracy": 0.1903713434934616,
      "num_tokens": 25018740.0,
      "step": 14435
    },
    {
      "entropy": 5.508065414428711,
      "epoch": 1.123477922583155,
      "grad_norm": 1.0625,
      "learning_rate": 0.000487780349804713,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1983823448419571,
      "num_tokens": 25026727.0,
      "step": 14440
    },
    {
      "entropy": 5.44137921333313,
      "epoch": 1.1238669519548725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004877713387235802,
      "loss": 4.9846,
      "mean_token_accuracy": 0.19495244920253754,
      "num_tokens": 25034496.0,
      "step": 14445
    },
    {
      "entropy": 5.482018232345581,
      "epoch": 1.1242559813265902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487762324413981,
      "loss": 5.0776,
      "mean_token_accuracy": 0.1911689519882202,
      "num_tokens": 25043608.0,
      "step": 14450
    },
    {
      "entropy": 5.516173362731934,
      "epoch": 1.1246450106983077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048775330687605213,
      "loss": 4.9934,
      "mean_token_accuracy": 0.1898203283548355,
      "num_tokens": 25052674.0,
      "step": 14455
    },
    {
      "entropy": 5.643764448165894,
      "epoch": 1.1250340400700254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004877442861099305,
      "loss": 5.1418,
      "mean_token_accuracy": 0.19075909554958342,
      "num_tokens": 25060881.0,
      "step": 14460
    },
    {
      "entropy": 5.497550773620605,
      "epoch": 1.1254230694417429,
      "grad_norm": 1.125,
      "learning_rate": 0.0004877352621157529,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1964103415608406,
      "num_tokens": 25069319.0,
      "step": 14465
    },
    {
      "entropy": 5.584656429290772,
      "epoch": 1.1258120988134603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048772623489365635,
      "loss": 5.189,
      "mean_token_accuracy": 0.18591021746397018,
      "num_tokens": 25077685.0,
      "step": 14470
    },
    {
      "entropy": 5.548371458053589,
      "epoch": 1.126201128185178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004877172044437777,
      "loss": 5.1805,
      "mean_token_accuracy": 0.18713123202323914,
      "num_tokens": 25086978.0,
      "step": 14475
    },
    {
      "entropy": 5.602936792373657,
      "epoch": 1.1265901575568955,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048770817076625416,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1916776955127716,
      "num_tokens": 25095656.0,
      "step": 14480
    },
    {
      "entropy": 5.568920135498047,
      "epoch": 1.126979186928613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048769913386122253,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19699035584926605,
      "num_tokens": 25104458.0,
      "step": 14485
    },
    {
      "entropy": 5.4849803924560545,
      "epoch": 1.1273682163003307,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004876900937288202,
      "loss": 5.0457,
      "mean_token_accuracy": 0.19261066913604735,
      "num_tokens": 25112302.0,
      "step": 14490
    },
    {
      "entropy": 5.535676288604736,
      "epoch": 1.1277572456720482,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048768105036918426,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19813272356987,
      "num_tokens": 25120463.0,
      "step": 14495
    },
    {
      "entropy": 5.435155057907105,
      "epoch": 1.1281462750437659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048767200378245187,
      "loss": 5.077,
      "mean_token_accuracy": 0.18878049105405809,
      "num_tokens": 25128866.0,
      "step": 14500
    },
    {
      "entropy": 5.489331007003784,
      "epoch": 1.1285353044154833,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048766295396876025,
      "loss": 5.0897,
      "mean_token_accuracy": 0.19369358122348784,
      "num_tokens": 25136979.0,
      "step": 14505
    },
    {
      "entropy": 5.506005382537841,
      "epoch": 1.128924333787201,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048765390092824683,
      "loss": 4.9885,
      "mean_token_accuracy": 0.19709485918283462,
      "num_tokens": 25145901.0,
      "step": 14510
    },
    {
      "entropy": 5.545856523513794,
      "epoch": 1.1293133631589185,
      "grad_norm": 1.140625,
      "learning_rate": 0.000487644844661049,
      "loss": 5.04,
      "mean_token_accuracy": 0.1975775897502899,
      "num_tokens": 25155203.0,
      "step": 14515
    },
    {
      "entropy": 5.4466136455535885,
      "epoch": 1.129702392530636,
      "grad_norm": 1.0,
      "learning_rate": 0.000487635785167304,
      "loss": 5.0024,
      "mean_token_accuracy": 0.19326263815164565,
      "num_tokens": 25163529.0,
      "step": 14520
    },
    {
      "entropy": 5.4899390697479244,
      "epoch": 1.1300914219023537,
      "grad_norm": 1.0,
      "learning_rate": 0.00048762672244714953,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1813514217734337,
      "num_tokens": 25172970.0,
      "step": 14525
    },
    {
      "entropy": 5.57536768913269,
      "epoch": 1.1304804512740712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004876176565007229,
      "loss": 5.1238,
      "mean_token_accuracy": 0.18822799026966094,
      "num_tokens": 25182244.0,
      "step": 14530
    },
    {
      "entropy": 5.6030515193939205,
      "epoch": 1.1308694806457888,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048760858732816174,
      "loss": 5.1514,
      "mean_token_accuracy": 0.1858677849173546,
      "num_tokens": 25191589.0,
      "step": 14535
    },
    {
      "entropy": 5.468666648864746,
      "epoch": 1.1312585100175063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004875995149296037,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19213001430034637,
      "num_tokens": 25200156.0,
      "step": 14540
    },
    {
      "entropy": 5.51993145942688,
      "epoch": 1.1316475393892238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004875904393051864,
      "loss": 5.0902,
      "mean_token_accuracy": 0.19721226543188095,
      "num_tokens": 25208890.0,
      "step": 14545
    },
    {
      "entropy": 5.481405639648438,
      "epoch": 1.1320365687609415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004875813604550476,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19486669301986695,
      "num_tokens": 25218013.0,
      "step": 14550
    },
    {
      "entropy": 5.552138757705689,
      "epoch": 1.132425598132659,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048757227837932494,
      "loss": 5.0593,
      "mean_token_accuracy": 0.19654567837715148,
      "num_tokens": 25226501.0,
      "step": 14555
    },
    {
      "entropy": 5.590053749084473,
      "epoch": 1.1328146275043767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004875631930781563,
      "loss": 5.2439,
      "mean_token_accuracy": 0.17747534364461898,
      "num_tokens": 25235731.0,
      "step": 14560
    },
    {
      "entropy": 5.492092895507812,
      "epoch": 1.1332036568760941,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048755410455167967,
      "loss": 5.0663,
      "mean_token_accuracy": 0.1915368616580963,
      "num_tokens": 25243659.0,
      "step": 14565
    },
    {
      "entropy": 5.49891619682312,
      "epoch": 1.1335926862478116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004875450128000326,
      "loss": 5.0712,
      "mean_token_accuracy": 0.18761184364557265,
      "num_tokens": 25253013.0,
      "step": 14570
    },
    {
      "entropy": 5.55296835899353,
      "epoch": 1.1339817156195293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004875359178233534,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1903602510690689,
      "num_tokens": 25261491.0,
      "step": 14575
    },
    {
      "entropy": 5.390804100036621,
      "epoch": 1.1343707449912468,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048752681962177985,
      "loss": 4.958,
      "mean_token_accuracy": 0.20374990552663802,
      "num_tokens": 25270323.0,
      "step": 14580
    },
    {
      "entropy": 5.497735023498535,
      "epoch": 1.1347597743629645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048751771819545013,
      "loss": 5.0431,
      "mean_token_accuracy": 0.19499655961990356,
      "num_tokens": 25280070.0,
      "step": 14585
    },
    {
      "entropy": 5.582213020324707,
      "epoch": 1.135148803734682,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048750861354450236,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18367218375205993,
      "num_tokens": 25288849.0,
      "step": 14590
    },
    {
      "entropy": 5.525808238983155,
      "epoch": 1.1355378331063994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004874995056690746,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19305911511182786,
      "num_tokens": 25297198.0,
      "step": 14595
    },
    {
      "entropy": 5.52663221359253,
      "epoch": 1.1359268624781171,
      "grad_norm": 1.09375,
      "learning_rate": 0.000487490394569305,
      "loss": 5.167,
      "mean_token_accuracy": 0.19226615279912948,
      "num_tokens": 25306783.0,
      "step": 14600
    },
    {
      "entropy": 5.5615002632141115,
      "epoch": 1.1363158918498346,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004874812802453319,
      "loss": 5.1873,
      "mean_token_accuracy": 0.18378738462924957,
      "num_tokens": 25316042.0,
      "step": 14605
    },
    {
      "entropy": 5.561297082901001,
      "epoch": 1.1367049212215523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004874721626972936,
      "loss": 5.1417,
      "mean_token_accuracy": 0.18660765290260314,
      "num_tokens": 25324969.0,
      "step": 14610
    },
    {
      "entropy": 5.522599697113037,
      "epoch": 1.1370939505932698,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874630419253284,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1939899817109108,
      "num_tokens": 25334050.0,
      "step": 14615
    },
    {
      "entropy": 5.497210931777954,
      "epoch": 1.1374829799649873,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048745391792957474,
      "loss": 5.0218,
      "mean_token_accuracy": 0.19273800253868104,
      "num_tokens": 25341517.0,
      "step": 14620
    },
    {
      "entropy": 5.521051263809204,
      "epoch": 1.137872009336705,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048744479071017097,
      "loss": 5.1982,
      "mean_token_accuracy": 0.18015148341655732,
      "num_tokens": 25350252.0,
      "step": 14625
    },
    {
      "entropy": 5.506719064712525,
      "epoch": 1.1382610387084224,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874356602672556,
      "loss": 5.0303,
      "mean_token_accuracy": 0.19943147599697114,
      "num_tokens": 25359503.0,
      "step": 14630
    },
    {
      "entropy": 5.50998740196228,
      "epoch": 1.1386500680801401,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004874265266009673,
      "loss": 5.0482,
      "mean_token_accuracy": 0.1946001723408699,
      "num_tokens": 25368253.0,
      "step": 14635
    },
    {
      "entropy": 5.573533153533935,
      "epoch": 1.1390390974518576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004874173897114445,
      "loss": 5.1383,
      "mean_token_accuracy": 0.18823929131031036,
      "num_tokens": 25376468.0,
      "step": 14640
    },
    {
      "entropy": 5.545216035842896,
      "epoch": 1.1394281268235753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048740824959882603,
      "loss": 5.1117,
      "mean_token_accuracy": 0.1903836101293564,
      "num_tokens": 25385760.0,
      "step": 14645
    },
    {
      "entropy": 5.499283361434936,
      "epoch": 1.1398171561952928,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004873991062632504,
      "loss": 5.1061,
      "mean_token_accuracy": 0.19080909192562104,
      "num_tokens": 25394485.0,
      "step": 14650
    },
    {
      "entropy": 5.566388416290283,
      "epoch": 1.1402061855670103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048738995970485635,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19277772754430772,
      "num_tokens": 25403485.0,
      "step": 14655
    },
    {
      "entropy": 5.5904418468475345,
      "epoch": 1.140595214938728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004873808099237827,
      "loss": 5.0684,
      "mean_token_accuracy": 0.1894974336028099,
      "num_tokens": 25412254.0,
      "step": 14660
    },
    {
      "entropy": 5.496951055526734,
      "epoch": 1.1409842443104454,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004873716569201684,
      "loss": 4.962,
      "mean_token_accuracy": 0.20020188689231871,
      "num_tokens": 25420353.0,
      "step": 14665
    },
    {
      "entropy": 5.473997735977173,
      "epoch": 1.141373273682163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048736250069415213,
      "loss": 5.2298,
      "mean_token_accuracy": 0.18009419590234757,
      "num_tokens": 25429832.0,
      "step": 14670
    },
    {
      "entropy": 5.492134857177734,
      "epoch": 1.1417623030538806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873533412458729,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1943511724472046,
      "num_tokens": 25438484.0,
      "step": 14675
    },
    {
      "entropy": 5.558124208450318,
      "epoch": 1.142151332425598,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048734417857546986,
      "loss": 5.1542,
      "mean_token_accuracy": 0.1868046596646309,
      "num_tokens": 25447602.0,
      "step": 14680
    },
    {
      "entropy": 5.526629161834717,
      "epoch": 1.1425403617973158,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004873350126830818,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2005312994122505,
      "num_tokens": 25455904.0,
      "step": 14685
    },
    {
      "entropy": 5.501600980758667,
      "epoch": 1.1429293911690332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004873258435688479,
      "loss": 5.1076,
      "mean_token_accuracy": 0.18971978574991227,
      "num_tokens": 25464170.0,
      "step": 14690
    },
    {
      "entropy": 5.480167007446289,
      "epoch": 1.143318420540751,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004873166712329073,
      "loss": 5.0058,
      "mean_token_accuracy": 0.1964828848838806,
      "num_tokens": 25472940.0,
      "step": 14695
    },
    {
      "entropy": 5.54087381362915,
      "epoch": 1.1437074499124684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048730749567539914,
      "loss": 5.052,
      "mean_token_accuracy": 0.19378894716501235,
      "num_tokens": 25481328.0,
      "step": 14700
    },
    {
      "entropy": 5.533055448532105,
      "epoch": 1.144096479284186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048729831689646257,
      "loss": 5.0665,
      "mean_token_accuracy": 0.19425686299800873,
      "num_tokens": 25489547.0,
      "step": 14705
    },
    {
      "entropy": 5.445548486709595,
      "epoch": 1.1444855086559036,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048728913489623705,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19614889919757844,
      "num_tokens": 25498433.0,
      "step": 14710
    },
    {
      "entropy": 5.532954549789428,
      "epoch": 1.144874538027621,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004872799496748618,
      "loss": 5.055,
      "mean_token_accuracy": 0.19378120601177215,
      "num_tokens": 25506646.0,
      "step": 14715
    },
    {
      "entropy": 5.468440103530884,
      "epoch": 1.1452635673993385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004872707612324761,
      "loss": 5.001,
      "mean_token_accuracy": 0.19625376164913177,
      "num_tokens": 25515431.0,
      "step": 14720
    },
    {
      "entropy": 5.602016973495483,
      "epoch": 1.1456525967710562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004872615695692196,
      "loss": 5.1749,
      "mean_token_accuracy": 0.18625230491161346,
      "num_tokens": 25524368.0,
      "step": 14725
    },
    {
      "entropy": 5.612739086151123,
      "epoch": 1.1460416261427737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004872523746852315,
      "loss": 5.1667,
      "mean_token_accuracy": 0.18316102921962737,
      "num_tokens": 25532694.0,
      "step": 14730
    },
    {
      "entropy": 5.47451639175415,
      "epoch": 1.1464306555144914,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048724317658065146,
      "loss": 5.1449,
      "mean_token_accuracy": 0.19039965867996217,
      "num_tokens": 25541807.0,
      "step": 14735
    },
    {
      "entropy": 5.582694101333618,
      "epoch": 1.146819684886209,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048723397525561916,
      "loss": 5.1101,
      "mean_token_accuracy": 0.18608397096395493,
      "num_tokens": 25549797.0,
      "step": 14740
    },
    {
      "entropy": 5.569544649124145,
      "epoch": 1.1472087142579266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048722477071027394,
      "loss": 5.1297,
      "mean_token_accuracy": 0.1851296082139015,
      "num_tokens": 25558373.0,
      "step": 14745
    },
    {
      "entropy": 5.517207479476928,
      "epoch": 1.147597743629644,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048721556294475574,
      "loss": 5.1073,
      "mean_token_accuracy": 0.19463250786066055,
      "num_tokens": 25566844.0,
      "step": 14750
    },
    {
      "entropy": 5.574475383758545,
      "epoch": 1.1479867730013615,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048720635195920403,
      "loss": 5.1724,
      "mean_token_accuracy": 0.18665237426757814,
      "num_tokens": 25575074.0,
      "step": 14755
    },
    {
      "entropy": 5.53858208656311,
      "epoch": 1.1483758023730792,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004871971377537588,
      "loss": 5.047,
      "mean_token_accuracy": 0.20292440801858902,
      "num_tokens": 25584526.0,
      "step": 14760
    },
    {
      "entropy": 5.586027002334594,
      "epoch": 1.1487648317447967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004871879203285597,
      "loss": 5.1472,
      "mean_token_accuracy": 0.18911695331335068,
      "num_tokens": 25592669.0,
      "step": 14765
    },
    {
      "entropy": 5.528279781341553,
      "epoch": 1.1491538611165142,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048717869968374655,
      "loss": 5.0824,
      "mean_token_accuracy": 0.18976491838693618,
      "num_tokens": 25600956.0,
      "step": 14770
    },
    {
      "entropy": 5.5058228969573975,
      "epoch": 1.1495428904882319,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048716947581945954,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1943865403532982,
      "num_tokens": 25608757.0,
      "step": 14775
    },
    {
      "entropy": 5.53881516456604,
      "epoch": 1.1499319198599494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004871602487358383,
      "loss": 5.103,
      "mean_token_accuracy": 0.18191706538200378,
      "num_tokens": 25618122.0,
      "step": 14780
    },
    {
      "entropy": 5.552245855331421,
      "epoch": 1.150320949231667,
      "grad_norm": 1.078125,
      "learning_rate": 0.000487151018433023,
      "loss": 5.1698,
      "mean_token_accuracy": 0.19088986963033677,
      "num_tokens": 25626692.0,
      "step": 14785
    },
    {
      "entropy": 5.565927457809448,
      "epoch": 1.1507099786033845,
      "grad_norm": 1.0,
      "learning_rate": 0.00048714178491115367,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1899450972676277,
      "num_tokens": 25635263.0,
      "step": 14790
    },
    {
      "entropy": 5.527678823471069,
      "epoch": 1.1510990079751022,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048713254817037046,
      "loss": 5.0016,
      "mean_token_accuracy": 0.1933281034231186,
      "num_tokens": 25643050.0,
      "step": 14795
    },
    {
      "entropy": 5.482155656814575,
      "epoch": 1.1514880373468197,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004871233082108135,
      "loss": 4.9746,
      "mean_token_accuracy": 0.1943673387169838,
      "num_tokens": 25652294.0,
      "step": 14800
    },
    {
      "entropy": 5.582728528976441,
      "epoch": 1.1518770667185372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871140650326228,
      "loss": 5.2355,
      "mean_token_accuracy": 0.18144925087690353,
      "num_tokens": 25661563.0,
      "step": 14805
    },
    {
      "entropy": 5.558869361877441,
      "epoch": 1.1522660960902549,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871048186359389,
      "loss": 5.1304,
      "mean_token_accuracy": 0.18788866549730301,
      "num_tokens": 25670353.0,
      "step": 14810
    },
    {
      "entropy": 5.597156715393067,
      "epoch": 1.1526551254619724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000487095569020902,
      "loss": 5.1809,
      "mean_token_accuracy": 0.18663873076438903,
      "num_tokens": 25678595.0,
      "step": 14815
    },
    {
      "entropy": 5.47860164642334,
      "epoch": 1.1530441548336898,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004870863161876524,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2055928647518158,
      "num_tokens": 25687901.0,
      "step": 14820
    },
    {
      "entropy": 5.49315299987793,
      "epoch": 1.1534331842054075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004870770601363305,
      "loss": 5.034,
      "mean_token_accuracy": 0.20184279680252076,
      "num_tokens": 25696600.0,
      "step": 14825
    },
    {
      "entropy": 5.424399566650391,
      "epoch": 1.153822213577125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048706780086707675,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20356038063764573,
      "num_tokens": 25705633.0,
      "step": 14830
    },
    {
      "entropy": 5.471970176696777,
      "epoch": 1.1542112429488427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004870585383800317,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2049849361181259,
      "num_tokens": 25713835.0,
      "step": 14835
    },
    {
      "entropy": 5.5152259349823,
      "epoch": 1.1546002723205602,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048704927267533585,
      "loss": 5.0316,
      "mean_token_accuracy": 0.19295763224363327,
      "num_tokens": 25721867.0,
      "step": 14840
    },
    {
      "entropy": 5.422769975662232,
      "epoch": 1.1549893016922779,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048704000375312975,
      "loss": 5.1142,
      "mean_token_accuracy": 0.18979655355215072,
      "num_tokens": 25730221.0,
      "step": 14845
    },
    {
      "entropy": 5.514937686920166,
      "epoch": 1.1553783310639953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048703073161355414,
      "loss": 5.0669,
      "mean_token_accuracy": 0.19476206749677658,
      "num_tokens": 25738427.0,
      "step": 14850
    },
    {
      "entropy": 5.5138355731964115,
      "epoch": 1.1557673604357128,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004870214562567497,
      "loss": 5.0396,
      "mean_token_accuracy": 0.19730647951364516,
      "num_tokens": 25746965.0,
      "step": 14855
    },
    {
      "entropy": 5.502241230010986,
      "epoch": 1.1561563898074305,
      "grad_norm": 1.125,
      "learning_rate": 0.000487012177682857,
      "loss": 5.0741,
      "mean_token_accuracy": 0.19906560480594634,
      "num_tokens": 25754999.0,
      "step": 14860
    },
    {
      "entropy": 5.539675855636597,
      "epoch": 1.156545419179148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870028958920171,
      "loss": 5.0965,
      "mean_token_accuracy": 0.18872542530298234,
      "num_tokens": 25763557.0,
      "step": 14865
    },
    {
      "entropy": 5.513170003890991,
      "epoch": 1.1569344485508655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004869936108843706,
      "loss": 5.1025,
      "mean_token_accuracy": 0.19070142805576323,
      "num_tokens": 25772967.0,
      "step": 14870
    },
    {
      "entropy": 5.561624383926391,
      "epoch": 1.1573234779225832,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048698432266005854,
      "loss": 5.1983,
      "mean_token_accuracy": 0.18754016906023024,
      "num_tokens": 25783570.0,
      "step": 14875
    },
    {
      "entropy": 5.514008665084839,
      "epoch": 1.1577125072943006,
      "grad_norm": 1.0,
      "learning_rate": 0.0004869750312192217,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1927780479192734,
      "num_tokens": 25793079.0,
      "step": 14880
    },
    {
      "entropy": 5.484844064712524,
      "epoch": 1.1581015366660183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048696573656200123,
      "loss": 4.9838,
      "mean_token_accuracy": 0.1981871247291565,
      "num_tokens": 25801821.0,
      "step": 14885
    },
    {
      "entropy": 5.407988882064819,
      "epoch": 1.1584905660377358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048695643868853806,
      "loss": 4.9721,
      "mean_token_accuracy": 0.19894556999206542,
      "num_tokens": 25810240.0,
      "step": 14890
    },
    {
      "entropy": 5.535338926315307,
      "epoch": 1.1588795954094535,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004869471375989733,
      "loss": 5.0182,
      "mean_token_accuracy": 0.19438907653093337,
      "num_tokens": 25818058.0,
      "step": 14895
    },
    {
      "entropy": 5.511289024353028,
      "epoch": 1.159268624781171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004869378332934481,
      "loss": 5.0931,
      "mean_token_accuracy": 0.19097632467746734,
      "num_tokens": 25827529.0,
      "step": 14900
    },
    {
      "entropy": 5.558517074584961,
      "epoch": 1.1596576541528885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048692852577210356,
      "loss": 5.1786,
      "mean_token_accuracy": 0.19371974021196364,
      "num_tokens": 25835765.0,
      "step": 14905
    },
    {
      "entropy": 5.466887092590332,
      "epoch": 1.1600466835246062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000486919215035081,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20165580362081528,
      "num_tokens": 25844402.0,
      "step": 14910
    },
    {
      "entropy": 5.449689102172852,
      "epoch": 1.1604357128963236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004869099010825217,
      "loss": 5.0133,
      "mean_token_accuracy": 0.1969584733247757,
      "num_tokens": 25853392.0,
      "step": 14915
    },
    {
      "entropy": 5.506316709518432,
      "epoch": 1.1608247422680413,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048690058391456684,
      "loss": 5.1119,
      "mean_token_accuracy": 0.1921757459640503,
      "num_tokens": 25861274.0,
      "step": 14920
    },
    {
      "entropy": 5.465091943740845,
      "epoch": 1.1612137716397588,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004868912635313579,
      "loss": 5.017,
      "mean_token_accuracy": 0.19121111631393434,
      "num_tokens": 25869570.0,
      "step": 14925
    },
    {
      "entropy": 5.492658567428589,
      "epoch": 1.1616028010114763,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004868819399330364,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20260507315397264,
      "num_tokens": 25877907.0,
      "step": 14930
    },
    {
      "entropy": 5.486486148834229,
      "epoch": 1.161991830383194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004868726131197436,
      "loss": 5.093,
      "mean_token_accuracy": 0.18671614676713943,
      "num_tokens": 25885817.0,
      "step": 14935
    },
    {
      "entropy": 5.486477613449097,
      "epoch": 1.1623808597549115,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004868632830916212,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19345428049564362,
      "num_tokens": 25894444.0,
      "step": 14940
    },
    {
      "entropy": 5.553924131393432,
      "epoch": 1.1627698891266292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004868539498488106,
      "loss": 5.1406,
      "mean_token_accuracy": 0.19124075025320053,
      "num_tokens": 25904116.0,
      "step": 14945
    },
    {
      "entropy": 5.657451200485229,
      "epoch": 1.1631589184983466,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004868446133914536,
      "loss": 5.2665,
      "mean_token_accuracy": 0.18250771909952163,
      "num_tokens": 25913889.0,
      "step": 14950
    },
    {
      "entropy": 5.574344539642334,
      "epoch": 1.163547947870064,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048683527371969185,
      "loss": 5.144,
      "mean_token_accuracy": 0.18110618591308594,
      "num_tokens": 25922674.0,
      "step": 14955
    },
    {
      "entropy": 5.468374824523925,
      "epoch": 1.1639369772417818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048682593083366685,
      "loss": 5.0135,
      "mean_token_accuracy": 0.2038408786058426,
      "num_tokens": 25930919.0,
      "step": 14960
    },
    {
      "entropy": 5.515384340286255,
      "epoch": 1.1643260066134993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048681658473352055,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19111277908086777,
      "num_tokens": 25939199.0,
      "step": 14965
    },
    {
      "entropy": 5.570595979690552,
      "epoch": 1.164715035985217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004868072354193947,
      "loss": 5.1304,
      "mean_token_accuracy": 0.19237570762634276,
      "num_tokens": 25947990.0,
      "step": 14970
    },
    {
      "entropy": 5.534336185455322,
      "epoch": 1.1651040653569344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867978828914312,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19478816539049149,
      "num_tokens": 25956479.0,
      "step": 14975
    },
    {
      "entropy": 5.478084421157837,
      "epoch": 1.165493094728652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867885271497719,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19069902449846268,
      "num_tokens": 25965487.0,
      "step": 14980
    },
    {
      "entropy": 5.547683763504028,
      "epoch": 1.1658821241003696,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004867791681945588,
      "loss": 5.0758,
      "mean_token_accuracy": 0.19630294144153596,
      "num_tokens": 25974247.0,
      "step": 14985
    },
    {
      "entropy": 5.5134974956512455,
      "epoch": 1.166271153472087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048676980602593395,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1907958447933197,
      "num_tokens": 25983203.0,
      "step": 14990
    },
    {
      "entropy": 5.490906143188477,
      "epoch": 1.1666601828438048,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048676044064403936,
      "loss": 5.1056,
      "mean_token_accuracy": 0.18912539929151534,
      "num_tokens": 25992635.0,
      "step": 14995
    },
    {
      "entropy": 5.538551568984985,
      "epoch": 1.1670492122155223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004867510720490171,
      "loss": 5.1748,
      "mean_token_accuracy": 0.18411308974027635,
      "num_tokens": 26001144.0,
      "step": 15000
    },
    {
      "epoch": 1.1670492122155223,
      "eval_entropy": 5.365386801527829,
      "eval_loss": 5.178375244140625,
      "eval_mean_token_accuracy": 0.19709104069789174,
      "eval_num_tokens": 26001144.0,
      "eval_runtime": 21.938,
      "eval_samples_per_second": 1894.335,
      "eval_steps_per_second": 236.803,
      "step": 15000
    },
    {
      "entropy": 5.537917184829712,
      "epoch": 1.1674382415872397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004867417002410094,
      "loss": 5.0497,
      "mean_token_accuracy": 0.19973308145999907,
      "num_tokens": 26009669.0,
      "step": 15005
    },
    {
      "entropy": 5.536905145645141,
      "epoch": 1.1678272709589574,
      "grad_norm": 1.125,
      "learning_rate": 0.00048673232522015845,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1951049044728279,
      "num_tokens": 26017909.0,
      "step": 15010
    },
    {
      "entropy": 5.55702805519104,
      "epoch": 1.168216300330675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004867229469866064,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19209877103567125,
      "num_tokens": 26026417.0,
      "step": 15015
    },
    {
      "entropy": 5.510050106048584,
      "epoch": 1.1686053297023926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004867135655404956,
      "loss": 5.1152,
      "mean_token_accuracy": 0.18961258977651596,
      "num_tokens": 26034879.0,
      "step": 15020
    },
    {
      "entropy": 5.583929824829101,
      "epoch": 1.16899435907411,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048670418088196844,
      "loss": 5.2206,
      "mean_token_accuracy": 0.18888526111841203,
      "num_tokens": 26044106.0,
      "step": 15025
    },
    {
      "entropy": 5.58198127746582,
      "epoch": 1.1693833884458276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004866947930111674,
      "loss": 5.157,
      "mean_token_accuracy": 0.18726833760738373,
      "num_tokens": 26052338.0,
      "step": 15030
    },
    {
      "entropy": 5.584809017181397,
      "epoch": 1.1697724178175453,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048668540192823467,
      "loss": 5.1519,
      "mean_token_accuracy": 0.1888921156525612,
      "num_tokens": 26061086.0,
      "step": 15035
    },
    {
      "entropy": 5.526157283782959,
      "epoch": 1.1701614471892627,
      "grad_norm": 1.046875,
      "learning_rate": 0.000486676007633313,
      "loss": 5.0437,
      "mean_token_accuracy": 0.19405822455883026,
      "num_tokens": 26070398.0,
      "step": 15040
    },
    {
      "entropy": 5.500879287719727,
      "epoch": 1.1705504765609804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004866666101265448,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19007622748613356,
      "num_tokens": 26078965.0,
      "step": 15045
    },
    {
      "entropy": 5.444583702087402,
      "epoch": 1.170939505932698,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048665720940807274,
      "loss": 5.1299,
      "mean_token_accuracy": 0.18771238178014754,
      "num_tokens": 26087775.0,
      "step": 15050
    },
    {
      "entropy": 5.535777473449707,
      "epoch": 1.1713285353044154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048664780547803935,
      "loss": 5.0371,
      "mean_token_accuracy": 0.19582536965608596,
      "num_tokens": 26096773.0,
      "step": 15055
    },
    {
      "entropy": 5.559192276000976,
      "epoch": 1.171717564676133,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004866383983365873,
      "loss": 5.2004,
      "mean_token_accuracy": 0.18374485224485398,
      "num_tokens": 26105446.0,
      "step": 15060
    },
    {
      "entropy": 5.538025379180908,
      "epoch": 1.1721065940478506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048662898798385956,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1946607992053032,
      "num_tokens": 26114251.0,
      "step": 15065
    },
    {
      "entropy": 5.644707250595093,
      "epoch": 1.1724956234195683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048661957441999875,
      "loss": 5.2213,
      "mean_token_accuracy": 0.18250214159488679,
      "num_tokens": 26123033.0,
      "step": 15070
    },
    {
      "entropy": 5.562620782852173,
      "epoch": 1.1728846527912857,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048661015764514765,
      "loss": 5.0005,
      "mean_token_accuracy": 0.19505824893712997,
      "num_tokens": 26131729.0,
      "step": 15075
    },
    {
      "entropy": 5.584011459350586,
      "epoch": 1.1732736821630034,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048660073765944927,
      "loss": 5.1383,
      "mean_token_accuracy": 0.18596526980400085,
      "num_tokens": 26140056.0,
      "step": 15080
    },
    {
      "entropy": 5.530083894729614,
      "epoch": 1.173662711534721,
      "grad_norm": 1.125,
      "learning_rate": 0.0004865913144630465,
      "loss": 5.1489,
      "mean_token_accuracy": 0.1988367572426796,
      "num_tokens": 26148947.0,
      "step": 15085
    },
    {
      "entropy": 5.582390546798706,
      "epoch": 1.1740517409064384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048658188805608227,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1842094600200653,
      "num_tokens": 26157485.0,
      "step": 15090
    },
    {
      "entropy": 5.547963809967041,
      "epoch": 1.174440770278156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004865724584386998,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18905287683010102,
      "num_tokens": 26166554.0,
      "step": 15095
    },
    {
      "entropy": 5.55906229019165,
      "epoch": 1.1748297996498736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004865630256110418,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18985276073217391,
      "num_tokens": 26174398.0,
      "step": 15100
    },
    {
      "entropy": 5.498431825637818,
      "epoch": 1.175218829021591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048655358957325174,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20197213143110276,
      "num_tokens": 26182885.0,
      "step": 15105
    },
    {
      "entropy": 5.477062559127807,
      "epoch": 1.1756078583933087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048654415032547277,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19032207876443863,
      "num_tokens": 26191400.0,
      "step": 15110
    },
    {
      "entropy": 5.50479154586792,
      "epoch": 1.1759968877650262,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048653470786784794,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20039502680301666,
      "num_tokens": 26200149.0,
      "step": 15115
    },
    {
      "entropy": 5.536980056762696,
      "epoch": 1.176385917136744,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004865252622005206,
      "loss": 5.0632,
      "mean_token_accuracy": 0.19969607293605804,
      "num_tokens": 26209069.0,
      "step": 15120
    },
    {
      "entropy": 5.527573442459106,
      "epoch": 1.1767749465084614,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048651581332363405,
      "loss": 5.1636,
      "mean_token_accuracy": 0.18913728147745132,
      "num_tokens": 26217740.0,
      "step": 15125
    },
    {
      "entropy": 5.466390323638916,
      "epoch": 1.177163975880179,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048650636123733176,
      "loss": 5.0123,
      "mean_token_accuracy": 0.19546746760606765,
      "num_tokens": 26226744.0,
      "step": 15130
    },
    {
      "entropy": 5.469076633453369,
      "epoch": 1.1775530052518965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004864969059417571,
      "loss": 5.1015,
      "mean_token_accuracy": 0.19212812334299087,
      "num_tokens": 26234510.0,
      "step": 15135
    },
    {
      "entropy": 5.4968016147613525,
      "epoch": 1.177942034623614,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004864874474370534,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20273134410381316,
      "num_tokens": 26244394.0,
      "step": 15140
    },
    {
      "entropy": 5.546889877319336,
      "epoch": 1.1783310639953317,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048647798572336445,
      "loss": 4.9643,
      "mean_token_accuracy": 0.20582231432199477,
      "num_tokens": 26252562.0,
      "step": 15145
    },
    {
      "entropy": 5.591189289093018,
      "epoch": 1.1787200933670492,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048646852080083355,
      "loss": 5.1621,
      "mean_token_accuracy": 0.18658907413482667,
      "num_tokens": 26261427.0,
      "step": 15150
    },
    {
      "entropy": 5.502444267272949,
      "epoch": 1.1791091227387667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864590526696045,
      "loss": 5.1638,
      "mean_token_accuracy": 0.18148885369300843,
      "num_tokens": 26269714.0,
      "step": 15155
    },
    {
      "entropy": 5.547440719604492,
      "epoch": 1.1794981521104844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048644958132982094,
      "loss": 5.1449,
      "mean_token_accuracy": 0.18680790960788726,
      "num_tokens": 26278846.0,
      "step": 15160
    },
    {
      "entropy": 5.514695453643799,
      "epoch": 1.1798871814822018,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048644010678162644,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1944452002644539,
      "num_tokens": 26287564.0,
      "step": 15165
    },
    {
      "entropy": 5.552087593078613,
      "epoch": 1.1802762108539195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004864306290251649,
      "loss": 5.1358,
      "mean_token_accuracy": 0.19395565539598464,
      "num_tokens": 26296204.0,
      "step": 15170
    },
    {
      "entropy": 5.529143190383911,
      "epoch": 1.180665240225637,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048642114806058014,
      "loss": 5.0588,
      "mean_token_accuracy": 0.19301970005035402,
      "num_tokens": 26304697.0,
      "step": 15175
    },
    {
      "entropy": 5.452383518218994,
      "epoch": 1.1810542695973547,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004864116638880159,
      "loss": 5.0586,
      "mean_token_accuracy": 0.18893171548843385,
      "num_tokens": 26313670.0,
      "step": 15180
    },
    {
      "entropy": 5.483880567550659,
      "epoch": 1.1814432989690722,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004864021765076162,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19663384705781936,
      "num_tokens": 26322080.0,
      "step": 15185
    },
    {
      "entropy": 5.567835092544556,
      "epoch": 1.1818323283407897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048639268591952494,
      "loss": 5.1124,
      "mean_token_accuracy": 0.18932854533195495,
      "num_tokens": 26330034.0,
      "step": 15190
    },
    {
      "entropy": 5.3948451519012455,
      "epoch": 1.1822213577125074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048638319212388614,
      "loss": 4.9396,
      "mean_token_accuracy": 0.19961443394422532,
      "num_tokens": 26338126.0,
      "step": 15195
    },
    {
      "entropy": 5.509631872177124,
      "epoch": 1.1826103870842248,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004863736951208438,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19103592336177827,
      "num_tokens": 26347005.0,
      "step": 15200
    },
    {
      "entropy": 5.58599328994751,
      "epoch": 1.1829994164559423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048636419491054217,
      "loss": 5.172,
      "mean_token_accuracy": 0.1891467198729515,
      "num_tokens": 26356165.0,
      "step": 15205
    },
    {
      "entropy": 5.588860654830933,
      "epoch": 1.18338844582766,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004863546914931252,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19238228648900985,
      "num_tokens": 26363814.0,
      "step": 15210
    },
    {
      "entropy": 5.532528924942016,
      "epoch": 1.1837774751993775,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004863451848687373,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20160799622535705,
      "num_tokens": 26371943.0,
      "step": 15215
    },
    {
      "entropy": 5.561339330673218,
      "epoch": 1.1841665045710952,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048633567503752253,
      "loss": 5.1389,
      "mean_token_accuracy": 0.18834577202796937,
      "num_tokens": 26382552.0,
      "step": 15220
    },
    {
      "entropy": 5.519689464569092,
      "epoch": 1.1845555339428127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004863261619996253,
      "loss": 5.0917,
      "mean_token_accuracy": 0.1896722823381424,
      "num_tokens": 26391008.0,
      "step": 15225
    },
    {
      "entropy": 5.428038835525513,
      "epoch": 1.1849445633145304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004863166457551899,
      "loss": 4.9085,
      "mean_token_accuracy": 0.20176288485527039,
      "num_tokens": 26399236.0,
      "step": 15230
    },
    {
      "entropy": 5.54332127571106,
      "epoch": 1.1853335926862478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004863071263043608,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1901614934206009,
      "num_tokens": 26407693.0,
      "step": 15235
    },
    {
      "entropy": 5.595538234710693,
      "epoch": 1.1857226220579653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004862976036472823,
      "loss": 5.1512,
      "mean_token_accuracy": 0.19004150629043579,
      "num_tokens": 26417082.0,
      "step": 15240
    },
    {
      "entropy": 5.603360319137574,
      "epoch": 1.186111651429683,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048628807778409907,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19141340255737305,
      "num_tokens": 26426328.0,
      "step": 15245
    },
    {
      "entropy": 5.554902791976929,
      "epoch": 1.1865006808014005,
      "grad_norm": 1.125,
      "learning_rate": 0.0004862785487149555,
      "loss": 5.1651,
      "mean_token_accuracy": 0.18587993681430817,
      "num_tokens": 26435186.0,
      "step": 15250
    },
    {
      "entropy": 5.518285179138184,
      "epoch": 1.186889710173118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862690164399963,
      "loss": 5.0194,
      "mean_token_accuracy": 0.19702616333961487,
      "num_tokens": 26444661.0,
      "step": 15255
    },
    {
      "entropy": 5.4932496547698975,
      "epoch": 1.1872787395448356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048625948095936593,
      "loss": 4.9808,
      "mean_token_accuracy": 0.19817111045122146,
      "num_tokens": 26453620.0,
      "step": 15260
    },
    {
      "entropy": 5.428113603591919,
      "epoch": 1.1876677689165531,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048624994227320923,
      "loss": 4.9083,
      "mean_token_accuracy": 0.20008229911327363,
      "num_tokens": 26462327.0,
      "step": 15265
    },
    {
      "entropy": 5.459143352508545,
      "epoch": 1.1880567982882708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048624040038167094,
      "loss": 5.0009,
      "mean_token_accuracy": 0.19366169422864915,
      "num_tokens": 26471104.0,
      "step": 15270
    },
    {
      "entropy": 5.554694700241089,
      "epoch": 1.1884458276599883,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048623085528489584,
      "loss": 5.0244,
      "mean_token_accuracy": 0.19665573090314864,
      "num_tokens": 26479263.0,
      "step": 15275
    },
    {
      "entropy": 5.544868469238281,
      "epoch": 1.188834857031706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048622130698302863,
      "loss": 5.1019,
      "mean_token_accuracy": 0.19141687154769899,
      "num_tokens": 26488036.0,
      "step": 15280
    },
    {
      "entropy": 5.519224071502686,
      "epoch": 1.1892238864034235,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048621175547621426,
      "loss": 5.0681,
      "mean_token_accuracy": 0.20257239490747453,
      "num_tokens": 26496702.0,
      "step": 15285
    },
    {
      "entropy": 5.553879880905152,
      "epoch": 1.189612915775141,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004862022007645978,
      "loss": 5.0926,
      "mean_token_accuracy": 0.1854624256491661,
      "num_tokens": 26505962.0,
      "step": 15290
    },
    {
      "entropy": 5.586704063415527,
      "epoch": 1.1900019451468586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048619264284832403,
      "loss": 5.2026,
      "mean_token_accuracy": 0.18470224887132644,
      "num_tokens": 26515502.0,
      "step": 15295
    },
    {
      "entropy": 5.6221137046813965,
      "epoch": 1.1903909745185761,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048618308172753804,
      "loss": 5.1306,
      "mean_token_accuracy": 0.1920813426375389,
      "num_tokens": 26524425.0,
      "step": 15300
    },
    {
      "entropy": 5.515850067138672,
      "epoch": 1.1907800038902936,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004861735174023849,
      "loss": 5.1534,
      "mean_token_accuracy": 0.18629422187805175,
      "num_tokens": 26533314.0,
      "step": 15305
    },
    {
      "entropy": 5.563751888275147,
      "epoch": 1.1911690332620113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048616394987300985,
      "loss": 5.1393,
      "mean_token_accuracy": 0.18802347928285598,
      "num_tokens": 26542455.0,
      "step": 15310
    },
    {
      "entropy": 5.596159172058106,
      "epoch": 1.1915580626337288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004861543791395579,
      "loss": 5.0575,
      "mean_token_accuracy": 0.19191517680883408,
      "num_tokens": 26551621.0,
      "step": 15315
    },
    {
      "entropy": 5.521759653091431,
      "epoch": 1.1919470920054465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004861448052021743,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20019185543060303,
      "num_tokens": 26559979.0,
      "step": 15320
    },
    {
      "entropy": 5.505478048324585,
      "epoch": 1.192336121377164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004861352280610044,
      "loss": 5.1022,
      "mean_token_accuracy": 0.18678427040576934,
      "num_tokens": 26568923.0,
      "step": 15325
    },
    {
      "entropy": 5.520033931732177,
      "epoch": 1.1927251507488816,
      "grad_norm": 1.125,
      "learning_rate": 0.00048612564771619346,
      "loss": 5.0398,
      "mean_token_accuracy": 0.1987572655081749,
      "num_tokens": 26576867.0,
      "step": 15330
    },
    {
      "entropy": 5.546887540817261,
      "epoch": 1.193114180120599,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048611606416788686,
      "loss": 5.1638,
      "mean_token_accuracy": 0.18710995614528655,
      "num_tokens": 26585418.0,
      "step": 15335
    },
    {
      "entropy": 5.597817707061767,
      "epoch": 1.1935032094923166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048610647741622996,
      "loss": 5.1588,
      "mean_token_accuracy": 0.19065070301294326,
      "num_tokens": 26594262.0,
      "step": 15340
    },
    {
      "entropy": 5.527783679962158,
      "epoch": 1.1938922388640343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004860968874613683,
      "loss": 5.0428,
      "mean_token_accuracy": 0.18844302743673325,
      "num_tokens": 26603086.0,
      "step": 15345
    },
    {
      "entropy": 5.547747945785522,
      "epoch": 1.1942812682357518,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004860872943034474,
      "loss": 5.112,
      "mean_token_accuracy": 0.19713467955589295,
      "num_tokens": 26611622.0,
      "step": 15350
    },
    {
      "entropy": 5.475402450561523,
      "epoch": 1.1946702976074695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004860776979426128,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20340909212827682,
      "num_tokens": 26620159.0,
      "step": 15355
    },
    {
      "entropy": 5.583709573745727,
      "epoch": 1.195059326979187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004860680983790101,
      "loss": 5.0618,
      "mean_token_accuracy": 0.18898301124572753,
      "num_tokens": 26628640.0,
      "step": 15360
    },
    {
      "entropy": 5.501337957382202,
      "epoch": 1.1954483563509044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004860584956127849,
      "loss": 5.0483,
      "mean_token_accuracy": 0.1919516459107399,
      "num_tokens": 26637250.0,
      "step": 15365
    },
    {
      "entropy": 5.527614545822144,
      "epoch": 1.195837385722622,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048604888964408306,
      "loss": 5.1172,
      "mean_token_accuracy": 0.19156016558408737,
      "num_tokens": 26645199.0,
      "step": 15370
    },
    {
      "entropy": 5.4766936779022215,
      "epoch": 1.1962264150943396,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004860392804730502,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1978173866868019,
      "num_tokens": 26653315.0,
      "step": 15375
    },
    {
      "entropy": 5.450601053237915,
      "epoch": 1.1966154444660573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048602966809983204,
      "loss": 5.0117,
      "mean_token_accuracy": 0.19976484924554824,
      "num_tokens": 26662063.0,
      "step": 15380
    },
    {
      "entropy": 5.501540803909302,
      "epoch": 1.1970044738377748,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004860200525245746,
      "loss": 5.083,
      "mean_token_accuracy": 0.195105741918087,
      "num_tokens": 26671097.0,
      "step": 15385
    },
    {
      "entropy": 5.5723248481750485,
      "epoch": 1.1973935032094922,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004860104337474239,
      "loss": 5.1317,
      "mean_token_accuracy": 0.1880512699484825,
      "num_tokens": 26678878.0,
      "step": 15390
    },
    {
      "entropy": 5.546806192398071,
      "epoch": 1.19778253258121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048600081176852555,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19119182080030442,
      "num_tokens": 26686743.0,
      "step": 15395
    },
    {
      "entropy": 5.429030323028565,
      "epoch": 1.1981715619529274,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048599118658802575,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20244764685630798,
      "num_tokens": 26695391.0,
      "step": 15400
    },
    {
      "entropy": 5.485673904418945,
      "epoch": 1.198560591324645,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004859815582060706,
      "loss": 5.0284,
      "mean_token_accuracy": 0.1951361984014511,
      "num_tokens": 26703980.0,
      "step": 15405
    },
    {
      "entropy": 5.5597252368927,
      "epoch": 1.1989496206963626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004859719266228061,
      "loss": 5.0748,
      "mean_token_accuracy": 0.187735615670681,
      "num_tokens": 26712514.0,
      "step": 15410
    },
    {
      "entropy": 5.5764378070831295,
      "epoch": 1.19933865006808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004859622918383784,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1921241745352745,
      "num_tokens": 26720576.0,
      "step": 15415
    },
    {
      "entropy": 5.555357980728149,
      "epoch": 1.1997276794397977,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004859526538529337,
      "loss": 5.1977,
      "mean_token_accuracy": 0.18107503801584243,
      "num_tokens": 26729319.0,
      "step": 15420
    },
    {
      "entropy": 5.53706202507019,
      "epoch": 1.2001167088115152,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004859430126666182,
      "loss": 5.0279,
      "mean_token_accuracy": 0.19322039633989335,
      "num_tokens": 26738073.0,
      "step": 15425
    },
    {
      "entropy": 5.540720129013062,
      "epoch": 1.200505738183233,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048593336827957824,
      "loss": 5.1304,
      "mean_token_accuracy": 0.19256405979394914,
      "num_tokens": 26746341.0,
      "step": 15430
    },
    {
      "entropy": 5.464187717437744,
      "epoch": 1.2008947675549504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048592372069196024,
      "loss": 5.0344,
      "mean_token_accuracy": 0.18831251412630082,
      "num_tokens": 26754844.0,
      "step": 15435
    },
    {
      "entropy": 5.575746154785156,
      "epoch": 1.2012837969266679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004859140699039104,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1884491980075836,
      "num_tokens": 26763942.0,
      "step": 15440
    },
    {
      "entropy": 5.471569633483886,
      "epoch": 1.2016728262983856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048590441591557526,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19693680703639985,
      "num_tokens": 26773468.0,
      "step": 15445
    },
    {
      "entropy": 5.496414709091186,
      "epoch": 1.202061855670103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048589475872710134,
      "loss": 5.0016,
      "mean_token_accuracy": 0.19465675503015517,
      "num_tokens": 26781707.0,
      "step": 15450
    },
    {
      "entropy": 5.543661880493164,
      "epoch": 1.2024508850418207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004858850983386351,
      "loss": 5.1055,
      "mean_token_accuracy": 0.18679908514022828,
      "num_tokens": 26789992.0,
      "step": 15455
    },
    {
      "entropy": 5.576060676574707,
      "epoch": 1.2028399144135382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858754347503231,
      "loss": 5.1474,
      "mean_token_accuracy": 0.18643400222063064,
      "num_tokens": 26798959.0,
      "step": 15460
    },
    {
      "entropy": 5.607064580917358,
      "epoch": 1.203228943785256,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048586576796231216,
      "loss": 5.1323,
      "mean_token_accuracy": 0.18768735826015473,
      "num_tokens": 26807498.0,
      "step": 15465
    },
    {
      "entropy": 5.537709760665893,
      "epoch": 1.2036179731569734,
      "grad_norm": 1.125,
      "learning_rate": 0.0004858560979747487,
      "loss": 5.0805,
      "mean_token_accuracy": 0.19618216753005982,
      "num_tokens": 26815931.0,
      "step": 15470
    },
    {
      "entropy": 5.483714389801025,
      "epoch": 1.2040070025286909,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048584642478777967,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20317025780677794,
      "num_tokens": 26824058.0,
      "step": 15475
    },
    {
      "entropy": 5.478203630447387,
      "epoch": 1.2043960319004086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858367484015517,
      "loss": 5.0391,
      "mean_token_accuracy": 0.19507866501808166,
      "num_tokens": 26833351.0,
      "step": 15480
    },
    {
      "entropy": 5.526114416122437,
      "epoch": 1.204785061272126,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048582706881621166,
      "loss": 5.1005,
      "mean_token_accuracy": 0.18726721554994583,
      "num_tokens": 26842263.0,
      "step": 15485
    },
    {
      "entropy": 5.50456919670105,
      "epoch": 1.2051740906438435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858173860319063,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20379763692617417,
      "num_tokens": 26850788.0,
      "step": 15490
    },
    {
      "entropy": 5.468435287475586,
      "epoch": 1.2055631200155612,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048580770004878277,
      "loss": 5.0582,
      "mean_token_accuracy": 0.19138407260179519,
      "num_tokens": 26859902.0,
      "step": 15495
    },
    {
      "entropy": 5.601291036605835,
      "epoch": 1.2059521493872787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857980108669879,
      "loss": 5.1343,
      "mean_token_accuracy": 0.18124507814645768,
      "num_tokens": 26868079.0,
      "step": 15500
    },
    {
      "entropy": 5.554708671569824,
      "epoch": 1.2063411787589964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048578831848666875,
      "loss": 5.1167,
      "mean_token_accuracy": 0.1846241757273674,
      "num_tokens": 26877403.0,
      "step": 15505
    },
    {
      "entropy": 5.547872114181518,
      "epoch": 1.2067302081307139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048577862290797235,
      "loss": 5.1269,
      "mean_token_accuracy": 0.18569187074899673,
      "num_tokens": 26885858.0,
      "step": 15510
    },
    {
      "entropy": 5.571923780441284,
      "epoch": 1.2071192375024316,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048576892413104586,
      "loss": 5.092,
      "mean_token_accuracy": 0.2029923602938652,
      "num_tokens": 26894525.0,
      "step": 15515
    },
    {
      "entropy": 5.605479574203491,
      "epoch": 1.207508266874149,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857592221560364,
      "loss": 5.1117,
      "mean_token_accuracy": 0.18453886210918427,
      "num_tokens": 26903197.0,
      "step": 15520
    },
    {
      "entropy": 5.57784481048584,
      "epoch": 1.2078972962458665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004857495169830912,
      "loss": 5.2084,
      "mean_token_accuracy": 0.18588347136974334,
      "num_tokens": 26911770.0,
      "step": 15525
    },
    {
      "entropy": 5.5144754409790036,
      "epoch": 1.2082863256175842,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004857398086123575,
      "loss": 4.9951,
      "mean_token_accuracy": 0.19654562920331956,
      "num_tokens": 26920753.0,
      "step": 15530
    },
    {
      "entropy": 5.532224559783936,
      "epoch": 1.2086753549893017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004857300970439827,
      "loss": 5.0933,
      "mean_token_accuracy": 0.18869938999414443,
      "num_tokens": 26930555.0,
      "step": 15535
    },
    {
      "entropy": 5.505837488174438,
      "epoch": 1.2090643843610192,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048572038227811404,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2005390927195549,
      "num_tokens": 26938841.0,
      "step": 15540
    },
    {
      "entropy": 5.499691152572632,
      "epoch": 1.2094534137327368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000485710664314899,
      "loss": 5.1689,
      "mean_token_accuracy": 0.19040360748767854,
      "num_tokens": 26947414.0,
      "step": 15545
    },
    {
      "entropy": 5.609728240966797,
      "epoch": 1.2098424431044543,
      "grad_norm": 1.15625,
      "learning_rate": 0.000485700943154485,
      "loss": 5.1322,
      "mean_token_accuracy": 0.1889816015958786,
      "num_tokens": 26955696.0,
      "step": 15550
    },
    {
      "entropy": 5.631604242324829,
      "epoch": 1.210231472476172,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004856912187970195,
      "loss": 5.1252,
      "mean_token_accuracy": 0.18986061960458755,
      "num_tokens": 26964030.0,
      "step": 15555
    },
    {
      "entropy": 5.497256278991699,
      "epoch": 1.2106205018478895,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004856814912426502,
      "loss": 5.0743,
      "mean_token_accuracy": 0.19458989053964615,
      "num_tokens": 26972825.0,
      "step": 15560
    },
    {
      "entropy": 5.583717012405396,
      "epoch": 1.2110095312196072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856717604915245,
      "loss": 5.1685,
      "mean_token_accuracy": 0.18714942783117294,
      "num_tokens": 26982417.0,
      "step": 15565
    },
    {
      "entropy": 5.6811261653900145,
      "epoch": 1.2113985605913247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856620265437902,
      "loss": 5.2974,
      "mean_token_accuracy": 0.17627157568931578,
      "num_tokens": 26992241.0,
      "step": 15570
    },
    {
      "entropy": 5.633410787582397,
      "epoch": 1.2117875899630421,
      "grad_norm": 1.109375,
      "learning_rate": 0.000485652289399595,
      "loss": 5.167,
      "mean_token_accuracy": 0.18984016627073289,
      "num_tokens": 27000515.0,
      "step": 15575
    },
    {
      "entropy": 5.615796375274658,
      "epoch": 1.2121766193347598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048564254905908655,
      "loss": 5.1906,
      "mean_token_accuracy": 0.18530738949775696,
      "num_tokens": 27009727.0,
      "step": 15580
    },
    {
      "entropy": 5.555057764053345,
      "epoch": 1.2125656487064773,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048563280552241266,
      "loss": 5.0994,
      "mean_token_accuracy": 0.18569034039974214,
      "num_tokens": 27020129.0,
      "step": 15585
    },
    {
      "entropy": 5.597054862976075,
      "epoch": 1.2129546780781948,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004856230587897212,
      "loss": 5.178,
      "mean_token_accuracy": 0.18824368715286255,
      "num_tokens": 27028660.0,
      "step": 15590
    },
    {
      "entropy": 5.535981798171997,
      "epoch": 1.2133437074499125,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004856133088611602,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1891512930393219,
      "num_tokens": 27037168.0,
      "step": 15595
    },
    {
      "entropy": 5.487007713317871,
      "epoch": 1.21373273682163,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004856035557368773,
      "loss": 4.9959,
      "mean_token_accuracy": 0.19277766197919846,
      "num_tokens": 27045937.0,
      "step": 15600
    },
    {
      "entropy": 5.560944843292236,
      "epoch": 1.2141217661933477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048559379941702074,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19251652359962462,
      "num_tokens": 27054187.0,
      "step": 15605
    },
    {
      "entropy": 5.502214622497559,
      "epoch": 1.2145107955650651,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048558403990173844,
      "loss": 5.0462,
      "mean_token_accuracy": 0.19605625718832015,
      "num_tokens": 27062718.0,
      "step": 15610
    },
    {
      "entropy": 5.513731384277344,
      "epoch": 1.2148998249367828,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048557427719117855,
      "loss": 5.0986,
      "mean_token_accuracy": 0.18849898278713226,
      "num_tokens": 27071946.0,
      "step": 15615
    },
    {
      "entropy": 5.525702619552613,
      "epoch": 1.2152888543085003,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004855645112854891,
      "loss": 5.1118,
      "mean_token_accuracy": 0.18231558352708815,
      "num_tokens": 27080989.0,
      "step": 15620
    },
    {
      "entropy": 5.608476686477661,
      "epoch": 1.2156778836802178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855547421848184,
      "loss": 5.1626,
      "mean_token_accuracy": 0.1853901505470276,
      "num_tokens": 27090073.0,
      "step": 15625
    },
    {
      "entropy": 5.552972936630249,
      "epoch": 1.2160669130519355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048554496988931456,
      "loss": 5.1602,
      "mean_token_accuracy": 0.1906544968485832,
      "num_tokens": 27098802.0,
      "step": 15630
    },
    {
      "entropy": 5.499906063079834,
      "epoch": 1.216455942423653,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048553519439912597,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18986087292432785,
      "num_tokens": 27107219.0,
      "step": 15635
    },
    {
      "entropy": 5.488286304473877,
      "epoch": 1.2168449717953704,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004855254157144009,
      "loss": 5.0561,
      "mean_token_accuracy": 0.19102613627910614,
      "num_tokens": 27115836.0,
      "step": 15640
    },
    {
      "entropy": 5.551061010360717,
      "epoch": 1.2172340011670881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855156338352877,
      "loss": 5.036,
      "mean_token_accuracy": 0.1913347825407982,
      "num_tokens": 27125146.0,
      "step": 15645
    },
    {
      "entropy": 5.54945740699768,
      "epoch": 1.2176230305388056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048550584876193493,
      "loss": 5.0708,
      "mean_token_accuracy": 0.20212605446577073,
      "num_tokens": 27133371.0,
      "step": 15650
    },
    {
      "entropy": 5.610711336135864,
      "epoch": 1.2180120599105233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048549606049449085,
      "loss": 5.1473,
      "mean_token_accuracy": 0.18392594307661056,
      "num_tokens": 27142271.0,
      "step": 15655
    },
    {
      "entropy": 5.514962911605835,
      "epoch": 1.2184010892822408,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004854862690331041,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20161230713129044,
      "num_tokens": 27150455.0,
      "step": 15660
    },
    {
      "entropy": 5.497422552108764,
      "epoch": 1.2187901186539585,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048547647437792327,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1872053012251854,
      "num_tokens": 27159127.0,
      "step": 15665
    },
    {
      "entropy": 5.551613759994507,
      "epoch": 1.219179148025676,
      "grad_norm": 1.0625,
      "learning_rate": 0.000485466676529097,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1880350723862648,
      "num_tokens": 27167909.0,
      "step": 15670
    },
    {
      "entropy": 5.564609670639038,
      "epoch": 1.2195681773973934,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048545687548677373,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19004094898700713,
      "num_tokens": 27177083.0,
      "step": 15675
    },
    {
      "entropy": 5.530086469650269,
      "epoch": 1.2199572067691111,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004854470712511025,
      "loss": 5.0334,
      "mean_token_accuracy": 0.19694966375827788,
      "num_tokens": 27185465.0,
      "step": 15680
    },
    {
      "entropy": 5.524537897109985,
      "epoch": 1.2203462361408286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048543726382223193,
      "loss": 5.1025,
      "mean_token_accuracy": 0.18814062625169753,
      "num_tokens": 27194294.0,
      "step": 15685
    },
    {
      "entropy": 5.486384773254395,
      "epoch": 1.220735265512546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048542745320031075,
      "loss": 5.1097,
      "mean_token_accuracy": 0.18863716274499892,
      "num_tokens": 27203949.0,
      "step": 15690
    },
    {
      "entropy": 5.590418434143066,
      "epoch": 1.2211242948842638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048541763938548795,
      "loss": 5.1489,
      "mean_token_accuracy": 0.18752705156803132,
      "num_tokens": 27213458.0,
      "step": 15695
    },
    {
      "entropy": 5.476040983200074,
      "epoch": 1.2215133242559812,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048540782237791244,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2018255189061165,
      "num_tokens": 27222328.0,
      "step": 15700
    },
    {
      "entropy": 5.500834941864014,
      "epoch": 1.221902353627699,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048539800217773293,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18911173343658447,
      "num_tokens": 27231016.0,
      "step": 15705
    },
    {
      "entropy": 5.525045108795166,
      "epoch": 1.2222913829994164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004853881787850988,
      "loss": 5.1459,
      "mean_token_accuracy": 0.197455695271492,
      "num_tokens": 27239785.0,
      "step": 15710
    },
    {
      "entropy": 5.603106689453125,
      "epoch": 1.2226804123711341,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048537835220015886,
      "loss": 5.0594,
      "mean_token_accuracy": 0.19688236713409424,
      "num_tokens": 27248576.0,
      "step": 15715
    },
    {
      "entropy": 5.552882242202759,
      "epoch": 1.2230694417428516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004853685224230623,
      "loss": 5.0225,
      "mean_token_accuracy": 0.19412660002708435,
      "num_tokens": 27257350.0,
      "step": 15720
    },
    {
      "entropy": 5.512541198730469,
      "epoch": 1.223458471114569,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048535868945395825,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1878654718399048,
      "num_tokens": 27266455.0,
      "step": 15725
    },
    {
      "entropy": 5.578158807754517,
      "epoch": 1.2238475004862868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048534885329299586,
      "loss": 5.1521,
      "mean_token_accuracy": 0.1926531031727791,
      "num_tokens": 27276423.0,
      "step": 15730
    },
    {
      "entropy": 5.578703117370606,
      "epoch": 1.2242365298580042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004853390139403245,
      "loss": 5.1195,
      "mean_token_accuracy": 0.1875462681055069,
      "num_tokens": 27285131.0,
      "step": 15735
    },
    {
      "entropy": 5.533165121078492,
      "epoch": 1.224625559229722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048532917139609334,
      "loss": 5.0784,
      "mean_token_accuracy": 0.19391572326421738,
      "num_tokens": 27293884.0,
      "step": 15740
    },
    {
      "entropy": 5.481677532196045,
      "epoch": 1.2250145886014394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004853193256604518,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19720928817987443,
      "num_tokens": 27302416.0,
      "step": 15745
    },
    {
      "entropy": 5.603465795516968,
      "epoch": 1.2254036179731569,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048530947673354924,
      "loss": 5.1148,
      "mean_token_accuracy": 0.18776411265134813,
      "num_tokens": 27311144.0,
      "step": 15750
    },
    {
      "entropy": 5.490378952026367,
      "epoch": 1.2257926473448746,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004852996246155351,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20270612090826035,
      "num_tokens": 27319911.0,
      "step": 15755
    },
    {
      "entropy": 5.468423748016358,
      "epoch": 1.226181676716592,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048528976930655896,
      "loss": 5.0291,
      "mean_token_accuracy": 0.19342004656791686,
      "num_tokens": 27329591.0,
      "step": 15760
    },
    {
      "entropy": 5.576354360580444,
      "epoch": 1.2265707060883098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048527991080677015,
      "loss": 5.1469,
      "mean_token_accuracy": 0.18709017634391784,
      "num_tokens": 27337915.0,
      "step": 15765
    },
    {
      "entropy": 5.477359771728516,
      "epoch": 1.2269597354600272,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048527004911631843,
      "loss": 5.0422,
      "mean_token_accuracy": 0.1962601959705353,
      "num_tokens": 27346799.0,
      "step": 15770
    },
    {
      "entropy": 5.588991975784301,
      "epoch": 1.2273487648317447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004852601842353534,
      "loss": 5.1389,
      "mean_token_accuracy": 0.19243317246437072,
      "num_tokens": 27355429.0,
      "step": 15775
    },
    {
      "entropy": 5.5116547584533695,
      "epoch": 1.2277377942034624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048525031616402476,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20260435193777085,
      "num_tokens": 27363895.0,
      "step": 15780
    },
    {
      "entropy": 5.510281229019165,
      "epoch": 1.2281268235751799,
      "grad_norm": 1.125,
      "learning_rate": 0.00048524044490248227,
      "loss": 5.0949,
      "mean_token_accuracy": 0.19600262343883515,
      "num_tokens": 27372593.0,
      "step": 15785
    },
    {
      "entropy": 5.467556715011597,
      "epoch": 1.2285158529468976,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048523057045087557,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20544480830430983,
      "num_tokens": 27381574.0,
      "step": 15790
    },
    {
      "entropy": 5.555751180648803,
      "epoch": 1.228904882318615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048522069280935466,
      "loss": 5.0761,
      "mean_token_accuracy": 0.19362273961305618,
      "num_tokens": 27390338.0,
      "step": 15795
    },
    {
      "entropy": 5.552414083480835,
      "epoch": 1.2292939116903325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004852108119780693,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19301716685295106,
      "num_tokens": 27398843.0,
      "step": 15800
    },
    {
      "entropy": 5.557084274291992,
      "epoch": 1.2296829410620502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004852009279571694,
      "loss": 5.1137,
      "mean_token_accuracy": 0.19164892733097078,
      "num_tokens": 27407700.0,
      "step": 15805
    },
    {
      "entropy": 5.504955148696899,
      "epoch": 1.2300719704337677,
      "grad_norm": 1.03125,
      "learning_rate": 0.000485191040746805,
      "loss": 5.0618,
      "mean_token_accuracy": 0.1960111916065216,
      "num_tokens": 27416559.0,
      "step": 15810
    },
    {
      "entropy": 5.508121585845947,
      "epoch": 1.2304609998054854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004851811503471262,
      "loss": 4.9851,
      "mean_token_accuracy": 0.19695171415805818,
      "num_tokens": 27424655.0,
      "step": 15815
    },
    {
      "entropy": 5.497120332717896,
      "epoch": 1.2308500291772029,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048517125675828294,
      "loss": 5.0702,
      "mean_token_accuracy": 0.19357269406318664,
      "num_tokens": 27434090.0,
      "step": 15820
    },
    {
      "entropy": 5.54311170578003,
      "epoch": 1.2312390585489204,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048516135998042536,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20452131927013398,
      "num_tokens": 27442555.0,
      "step": 15825
    },
    {
      "entropy": 5.50479474067688,
      "epoch": 1.231628087920638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004851514600137037,
      "loss": 5.0526,
      "mean_token_accuracy": 0.19500044733285904,
      "num_tokens": 27451180.0,
      "step": 15830
    },
    {
      "entropy": 5.583836698532105,
      "epoch": 1.2320171172923555,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048514155685826807,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1813034236431122,
      "num_tokens": 27460847.0,
      "step": 15835
    },
    {
      "entropy": 5.557675361633301,
      "epoch": 1.2324061466640732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004851316505142688,
      "loss": 5.0624,
      "mean_token_accuracy": 0.19568282812833787,
      "num_tokens": 27470102.0,
      "step": 15840
    },
    {
      "entropy": 5.547676277160645,
      "epoch": 1.2327951760357907,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048512174098185615,
      "loss": 5.1123,
      "mean_token_accuracy": 0.19469647258520126,
      "num_tokens": 27479066.0,
      "step": 15845
    },
    {
      "entropy": 5.486546516418457,
      "epoch": 1.2331842054075082,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048511182826118055,
      "loss": 5.0311,
      "mean_token_accuracy": 0.1919628843665123,
      "num_tokens": 27487681.0,
      "step": 15850
    },
    {
      "entropy": 5.594352054595947,
      "epoch": 1.2335732347792259,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048510191235239246,
      "loss": 5.0529,
      "mean_token_accuracy": 0.19886396676301957,
      "num_tokens": 27496067.0,
      "step": 15855
    },
    {
      "entropy": 5.487336730957031,
      "epoch": 1.2339622641509433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048509199325564217,
      "loss": 4.9776,
      "mean_token_accuracy": 0.199818916618824,
      "num_tokens": 27503801.0,
      "step": 15860
    },
    {
      "entropy": 5.401217269897461,
      "epoch": 1.234351293522661,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004850820709710803,
      "loss": 4.9646,
      "mean_token_accuracy": 0.1940133184194565,
      "num_tokens": 27512018.0,
      "step": 15865
    },
    {
      "entropy": 5.492809534072876,
      "epoch": 1.2347403228943785,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004850721454988574,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2061188891530037,
      "num_tokens": 27521632.0,
      "step": 15870
    },
    {
      "entropy": 5.630766487121582,
      "epoch": 1.235129352266096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048506221683912405,
      "loss": 5.2134,
      "mean_token_accuracy": 0.190043905377388,
      "num_tokens": 27530439.0,
      "step": 15875
    },
    {
      "entropy": 5.523299932479858,
      "epoch": 1.2355183816378137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004850522849920309,
      "loss": 5.0707,
      "mean_token_accuracy": 0.19871943145990373,
      "num_tokens": 27539079.0,
      "step": 15880
    },
    {
      "entropy": 5.505356311798096,
      "epoch": 1.2359074110095312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048504234995772863,
      "loss": 5.01,
      "mean_token_accuracy": 0.19357948899269103,
      "num_tokens": 27547701.0,
      "step": 15885
    },
    {
      "entropy": 5.551141309738159,
      "epoch": 1.2362964403812489,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004850324117363681,
      "loss": 5.0248,
      "mean_token_accuracy": 0.1950746789574623,
      "num_tokens": 27556094.0,
      "step": 15890
    },
    {
      "entropy": 5.557127046585083,
      "epoch": 1.2366854697529663,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048502247032809997,
      "loss": 5.091,
      "mean_token_accuracy": 0.19358098655939102,
      "num_tokens": 27564530.0,
      "step": 15895
    },
    {
      "entropy": 5.530200433731079,
      "epoch": 1.237074499124684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004850125257330751,
      "loss": 5.0304,
      "mean_token_accuracy": 0.19774871468544006,
      "num_tokens": 27573467.0,
      "step": 15900
    },
    {
      "entropy": 5.535633325576782,
      "epoch": 1.2374635284964015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048500257795144446,
      "loss": 5.1014,
      "mean_token_accuracy": 0.19193962216377258,
      "num_tokens": 27582170.0,
      "step": 15905
    },
    {
      "entropy": 5.59482159614563,
      "epoch": 1.237852557868119,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000484992626983359,
      "loss": 5.1526,
      "mean_token_accuracy": 0.19181772768497468,
      "num_tokens": 27590887.0,
      "step": 15910
    },
    {
      "entropy": 5.565155458450318,
      "epoch": 1.2382415872398367,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004849826728289696,
      "loss": 5.1379,
      "mean_token_accuracy": 0.19272060841321945,
      "num_tokens": 27599761.0,
      "step": 15915
    },
    {
      "entropy": 5.630237627029419,
      "epoch": 1.2386306166115542,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048497271548842737,
      "loss": 5.2164,
      "mean_token_accuracy": 0.18394846916198732,
      "num_tokens": 27608583.0,
      "step": 15920
    },
    {
      "entropy": 5.530681800842285,
      "epoch": 1.2390196459832716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004849627549618834,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2033820405602455,
      "num_tokens": 27616899.0,
      "step": 15925
    },
    {
      "entropy": 5.509924697875976,
      "epoch": 1.2394086753549893,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048495279124948886,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1959036886692047,
      "num_tokens": 27625975.0,
      "step": 15930
    },
    {
      "entropy": 5.6516149044036865,
      "epoch": 1.2397977047267068,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004849428243513948,
      "loss": 5.26,
      "mean_token_accuracy": 0.18386876583099365,
      "num_tokens": 27634806.0,
      "step": 15935
    },
    {
      "entropy": 5.554934310913086,
      "epoch": 1.2401867340984245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048493285426775255,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1940790519118309,
      "num_tokens": 27643641.0,
      "step": 15940
    },
    {
      "entropy": 5.532153987884522,
      "epoch": 1.240575763470142,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004849228809987135,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19593401700258256,
      "num_tokens": 27652592.0,
      "step": 15945
    },
    {
      "entropy": 5.52583498954773,
      "epoch": 1.2409647928418597,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004849129045444288,
      "loss": 5.0018,
      "mean_token_accuracy": 0.19533077329397203,
      "num_tokens": 27661205.0,
      "step": 15950
    },
    {
      "entropy": 5.489716577529907,
      "epoch": 1.2413538222135772,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004849029249050498,
      "loss": 5.0205,
      "mean_token_accuracy": 0.1954821303486824,
      "num_tokens": 27669074.0,
      "step": 15955
    },
    {
      "entropy": 5.551987218856811,
      "epoch": 1.2417428515852946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048489294208072805,
      "loss": 5.0574,
      "mean_token_accuracy": 0.19411068856716157,
      "num_tokens": 27676863.0,
      "step": 15960
    },
    {
      "entropy": 5.586891460418701,
      "epoch": 1.2421318809570123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048488295607161495,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19111596792936325,
      "num_tokens": 27686491.0,
      "step": 15965
    },
    {
      "entropy": 5.520835256576538,
      "epoch": 1.2425209103287298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004848729668778621,
      "loss": 5.1323,
      "mean_token_accuracy": 0.19250125885009767,
      "num_tokens": 27694964.0,
      "step": 15970
    },
    {
      "entropy": 5.554769277572632,
      "epoch": 1.2429099397004473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004848629744996211,
      "loss": 5.0451,
      "mean_token_accuracy": 0.1958884507417679,
      "num_tokens": 27703766.0,
      "step": 15975
    },
    {
      "entropy": 5.576978635787964,
      "epoch": 1.243298969072165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004848529789370434,
      "loss": 5.0621,
      "mean_token_accuracy": 0.1917141407728195,
      "num_tokens": 27712773.0,
      "step": 15980
    },
    {
      "entropy": 5.5655499458312985,
      "epoch": 1.2436879984438824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004848429801902808,
      "loss": 5.1631,
      "mean_token_accuracy": 0.19076957404613495,
      "num_tokens": 27721788.0,
      "step": 15985
    },
    {
      "entropy": 5.594088315963745,
      "epoch": 1.2440770278156001,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048483297825948485,
      "loss": 5.2075,
      "mean_token_accuracy": 0.18437306731939315,
      "num_tokens": 27731222.0,
      "step": 15990
    },
    {
      "entropy": 5.64353837966919,
      "epoch": 1.2444660571873176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048482297314480756,
      "loss": 5.2343,
      "mean_token_accuracy": 0.18508951514959335,
      "num_tokens": 27740623.0,
      "step": 15995
    },
    {
      "entropy": 5.603208732604981,
      "epoch": 1.2448550865590353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004848129648464006,
      "loss": 5.1225,
      "mean_token_accuracy": 0.18544287979602814,
      "num_tokens": 27749272.0,
      "step": 16000
    },
    {
      "entropy": 5.5229521751403805,
      "epoch": 1.2452441159307528,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004848029533644159,
      "loss": 5.1666,
      "mean_token_accuracy": 0.1891465350985527,
      "num_tokens": 27758435.0,
      "step": 16005
    },
    {
      "entropy": 5.625425338745117,
      "epoch": 1.2456331453024703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004847929386990052,
      "loss": 5.1292,
      "mean_token_accuracy": 0.18991667926311492,
      "num_tokens": 27767132.0,
      "step": 16010
    },
    {
      "entropy": 5.594956016540527,
      "epoch": 1.246022174674188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048478292085032065,
      "loss": 5.0701,
      "mean_token_accuracy": 0.19087830781936646,
      "num_tokens": 27775661.0,
      "step": 16015
    },
    {
      "entropy": 5.549164438247681,
      "epoch": 1.2464112040459054,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004847728998185142,
      "loss": 5.0847,
      "mean_token_accuracy": 0.18831372410058975,
      "num_tokens": 27783819.0,
      "step": 16020
    },
    {
      "entropy": 5.4139440059661865,
      "epoch": 1.246800233417623,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048476287560373786,
      "loss": 4.9038,
      "mean_token_accuracy": 0.1979458063840866,
      "num_tokens": 27792717.0,
      "step": 16025
    },
    {
      "entropy": 5.6446168422698975,
      "epoch": 1.2471892627893406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048475284820614375,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18445671051740647,
      "num_tokens": 27802314.0,
      "step": 16030
    },
    {
      "entropy": 5.595372104644776,
      "epoch": 1.247578292161058,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048474281762588407,
      "loss": 5.1427,
      "mean_token_accuracy": 0.18943841755390167,
      "num_tokens": 27810837.0,
      "step": 16035
    },
    {
      "entropy": 5.55095100402832,
      "epoch": 1.2479673215327758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847327838631109,
      "loss": 5.112,
      "mean_token_accuracy": 0.18820951133966446,
      "num_tokens": 27819743.0,
      "step": 16040
    },
    {
      "entropy": 5.575311374664307,
      "epoch": 1.2483563509044933,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004847227469179766,
      "loss": 5.1529,
      "mean_token_accuracy": 0.18688541352748872,
      "num_tokens": 27827982.0,
      "step": 16045
    },
    {
      "entropy": 5.582282161712646,
      "epoch": 1.248745380276211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004847127067906334,
      "loss": 5.0933,
      "mean_token_accuracy": 0.19153002500534058,
      "num_tokens": 27836539.0,
      "step": 16050
    },
    {
      "entropy": 5.574767017364502,
      "epoch": 1.2491344096479284,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048470266348123365,
      "loss": 5.1072,
      "mean_token_accuracy": 0.18847111910581588,
      "num_tokens": 27844987.0,
      "step": 16055
    },
    {
      "entropy": 5.528655576705932,
      "epoch": 1.249523439019646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004846926169899298,
      "loss": 5.0806,
      "mean_token_accuracy": 0.19213380068540573,
      "num_tokens": 27853517.0,
      "step": 16060
    },
    {
      "entropy": 5.520037317276001,
      "epoch": 1.2499124683913636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048468256731687426,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1991806611418724,
      "num_tokens": 27862107.0,
      "step": 16065
    },
    {
      "entropy": 5.535496664047241,
      "epoch": 1.250301497763081,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846725144622194,
      "loss": 5.0684,
      "mean_token_accuracy": 0.19231536686420442,
      "num_tokens": 27870611.0,
      "step": 16070
    },
    {
      "entropy": 5.539904069900513,
      "epoch": 1.2506905271347986,
      "grad_norm": 1.0625,
      "learning_rate": 0.000484662458426118,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19335533827543258,
      "num_tokens": 27879520.0,
      "step": 16075
    },
    {
      "entropy": 5.457077407836914,
      "epoch": 1.2510795565065163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048465239920872247,
      "loss": 4.994,
      "mean_token_accuracy": 0.1975537970662117,
      "num_tokens": 27888371.0,
      "step": 16080
    },
    {
      "entropy": 5.491286087036133,
      "epoch": 1.2514685858782337,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048464233681018545,
      "loss": 4.9997,
      "mean_token_accuracy": 0.198992620408535,
      "num_tokens": 27896518.0,
      "step": 16085
    },
    {
      "entropy": 5.543295669555664,
      "epoch": 1.2518576152499514,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846322712306596,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19242860823869706,
      "num_tokens": 27905397.0,
      "step": 16090
    },
    {
      "entropy": 5.520310974121093,
      "epoch": 1.252246644621669,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048462220247029783,
      "loss": 4.9803,
      "mean_token_accuracy": 0.1944587618112564,
      "num_tokens": 27914057.0,
      "step": 16095
    },
    {
      "entropy": 5.4607775688171385,
      "epoch": 1.2526356739933866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048461213052925265,
      "loss": 5.0103,
      "mean_token_accuracy": 0.19643427580595016,
      "num_tokens": 27922872.0,
      "step": 16100
    },
    {
      "entropy": 5.44712643623352,
      "epoch": 1.253024703365104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846020554076771,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1972901329398155,
      "num_tokens": 27931143.0,
      "step": 16105
    },
    {
      "entropy": 5.55177493095398,
      "epoch": 1.2534137327368216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845919771057239,
      "loss": 5.1455,
      "mean_token_accuracy": 0.18036955893039702,
      "num_tokens": 27939915.0,
      "step": 16110
    },
    {
      "entropy": 5.571521663665772,
      "epoch": 1.2538027621085392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845818956235462,
      "loss": 5.0214,
      "mean_token_accuracy": 0.19408749043941498,
      "num_tokens": 27948125.0,
      "step": 16115
    },
    {
      "entropy": 5.608725881576538,
      "epoch": 1.2541917914802567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048457181096129664,
      "loss": 5.2325,
      "mean_token_accuracy": 0.178395639359951,
      "num_tokens": 27957288.0,
      "step": 16120
    },
    {
      "entropy": 5.5557608127594,
      "epoch": 1.2545808208519742,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004845617231191285,
      "loss": 5.0675,
      "mean_token_accuracy": 0.1874172493815422,
      "num_tokens": 27965754.0,
      "step": 16125
    },
    {
      "entropy": 5.577339935302734,
      "epoch": 1.254969850223692,
      "grad_norm": 1.125,
      "learning_rate": 0.0004845516320971948,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19957368224859237,
      "num_tokens": 27974060.0,
      "step": 16130
    },
    {
      "entropy": 5.5170361518859865,
      "epoch": 1.2553588795954094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004845415378956486,
      "loss": 5.0488,
      "mean_token_accuracy": 0.1952352523803711,
      "num_tokens": 27983417.0,
      "step": 16135
    },
    {
      "entropy": 5.596839714050293,
      "epoch": 1.255747908967127,
      "grad_norm": 1.078125,
      "learning_rate": 0.000484531440514643,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19414311796426773,
      "num_tokens": 27992425.0,
      "step": 16140
    },
    {
      "entropy": 5.580276107788086,
      "epoch": 1.2561369383388445,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048452133995433136,
      "loss": 5.117,
      "mean_token_accuracy": 0.19307217597961426,
      "num_tokens": 28000711.0,
      "step": 16145
    },
    {
      "entropy": 5.508587265014649,
      "epoch": 1.2565259677105622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004845112362148668,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1945103660225868,
      "num_tokens": 28010193.0,
      "step": 16150
    },
    {
      "entropy": 5.534761381149292,
      "epoch": 1.2569149970822797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004845011292964028,
      "loss": 5.1328,
      "mean_token_accuracy": 0.19204179495573043,
      "num_tokens": 28020030.0,
      "step": 16155
    },
    {
      "entropy": 5.507098436355591,
      "epoch": 1.2573040264539972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048449101919909265,
      "loss": 5.003,
      "mean_token_accuracy": 0.19443392753601074,
      "num_tokens": 28028171.0,
      "step": 16160
    },
    {
      "entropy": 5.576512050628662,
      "epoch": 1.257693055825715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048448090592308955,
      "loss": 5.0788,
      "mean_token_accuracy": 0.19209689199924468,
      "num_tokens": 28036652.0,
      "step": 16165
    },
    {
      "entropy": 5.599478340148925,
      "epoch": 1.2580820851974324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004844707894685473,
      "loss": 5.0902,
      "mean_token_accuracy": 0.19421993792057038,
      "num_tokens": 28044691.0,
      "step": 16170
    },
    {
      "entropy": 5.524524068832397,
      "epoch": 1.2584711145691498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004844606698356192,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1873532310128212,
      "num_tokens": 28053186.0,
      "step": 16175
    },
    {
      "entropy": 5.597494888305664,
      "epoch": 1.2588601439408675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004844505470244588,
      "loss": 5.1473,
      "mean_token_accuracy": 0.19274385869503022,
      "num_tokens": 28061827.0,
      "step": 16180
    },
    {
      "entropy": 5.675650787353516,
      "epoch": 1.2592491733125852,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004844404210352197,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19142398983240128,
      "num_tokens": 28070191.0,
      "step": 16185
    },
    {
      "entropy": 5.586566352844239,
      "epoch": 1.2596382026843027,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048443029186805554,
      "loss": 5.1457,
      "mean_token_accuracy": 0.18605474084615709,
      "num_tokens": 28078502.0,
      "step": 16190
    },
    {
      "entropy": 5.532293891906738,
      "epoch": 1.2600272320560202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004844201595231201,
      "loss": 5.0561,
      "mean_token_accuracy": 0.19519804120063783,
      "num_tokens": 28086801.0,
      "step": 16195
    },
    {
      "entropy": 5.547823905944824,
      "epoch": 1.2604162614277379,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048441002400056706,
      "loss": 5.1036,
      "mean_token_accuracy": 0.19093164801597595,
      "num_tokens": 28095184.0,
      "step": 16200
    },
    {
      "entropy": 5.508849096298218,
      "epoch": 1.2608052907994554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004843998853005502,
      "loss": 5.0035,
      "mean_token_accuracy": 0.19692357033491134,
      "num_tokens": 28103571.0,
      "step": 16205
    },
    {
      "entropy": 5.6406043529510494,
      "epoch": 1.2611943201711728,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843897434232233,
      "loss": 5.271,
      "mean_token_accuracy": 0.1766631558537483,
      "num_tokens": 28113142.0,
      "step": 16210
    },
    {
      "entropy": 5.561850070953369,
      "epoch": 1.2615833495428905,
      "grad_norm": 1.125,
      "learning_rate": 0.00048437959836874047,
      "loss": 5.0677,
      "mean_token_accuracy": 0.19162545204162598,
      "num_tokens": 28121464.0,
      "step": 16215
    },
    {
      "entropy": 5.54154224395752,
      "epoch": 1.261972378914608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048436945013725544,
      "loss": 5.0435,
      "mean_token_accuracy": 0.19706882238388063,
      "num_tokens": 28130384.0,
      "step": 16220
    },
    {
      "entropy": 5.495460128784179,
      "epoch": 1.2623614082863255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048435929872892226,
      "loss": 5.073,
      "mean_token_accuracy": 0.19532164931297302,
      "num_tokens": 28139030.0,
      "step": 16225
    },
    {
      "entropy": 5.531817436218262,
      "epoch": 1.2627504376580432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004843491441438949,
      "loss": 4.9607,
      "mean_token_accuracy": 0.20064347088336945,
      "num_tokens": 28147228.0,
      "step": 16230
    },
    {
      "entropy": 5.566263914108276,
      "epoch": 1.2631394670297609,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048433898638232755,
      "loss": 5.0214,
      "mean_token_accuracy": 0.1959443673491478,
      "num_tokens": 28156262.0,
      "step": 16235
    },
    {
      "entropy": 5.58278522491455,
      "epoch": 1.2635284964014784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004843288254443742,
      "loss": 5.1261,
      "mean_token_accuracy": 0.1898317888379097,
      "num_tokens": 28165953.0,
      "step": 16240
    },
    {
      "entropy": 5.528287696838379,
      "epoch": 1.2639175257731958,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004843186613301892,
      "loss": 5.0965,
      "mean_token_accuracy": 0.19257779270410538,
      "num_tokens": 28174362.0,
      "step": 16245
    },
    {
      "entropy": 5.4810363292694095,
      "epoch": 1.2643065551449135,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004843084940399266,
      "loss": 4.9821,
      "mean_token_accuracy": 0.1989048331975937,
      "num_tokens": 28182657.0,
      "step": 16250
    },
    {
      "entropy": 5.542258167266846,
      "epoch": 1.264695584516631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842983235737408,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19014571458101273,
      "num_tokens": 28191605.0,
      "step": 16255
    },
    {
      "entropy": 5.522048616409302,
      "epoch": 1.2650846138883485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004842881499317861,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19313347041606904,
      "num_tokens": 28200619.0,
      "step": 16260
    },
    {
      "entropy": 5.615197229385376,
      "epoch": 1.2654736432600662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004842779731142168,
      "loss": 5.1031,
      "mean_token_accuracy": 0.1963601678609848,
      "num_tokens": 28208638.0,
      "step": 16265
    },
    {
      "entropy": 5.5232823371887205,
      "epoch": 1.2658626726317836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048426779312118735,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1896592378616333,
      "num_tokens": 28217221.0,
      "step": 16270
    },
    {
      "entropy": 5.545674133300781,
      "epoch": 1.2662517020035013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004842576099528522,
      "loss": 5.1029,
      "mean_token_accuracy": 0.19116133600473403,
      "num_tokens": 28225498.0,
      "step": 16275
    },
    {
      "entropy": 5.45075421333313,
      "epoch": 1.2666407313752188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004842474236093659,
      "loss": 5.0307,
      "mean_token_accuracy": 0.1960647389292717,
      "num_tokens": 28234264.0,
      "step": 16280
    },
    {
      "entropy": 5.539739370346069,
      "epoch": 1.2670297607469365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048423723409088306,
      "loss": 5.1449,
      "mean_token_accuracy": 0.1962319254875183,
      "num_tokens": 28242850.0,
      "step": 16285
    },
    {
      "entropy": 5.572647619247436,
      "epoch": 1.267418790118654,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004842270413975582,
      "loss": 5.1651,
      "mean_token_accuracy": 0.18833777904510499,
      "num_tokens": 28251700.0,
      "step": 16290
    },
    {
      "entropy": 5.5048723220825195,
      "epoch": 1.2678078194903715,
      "grad_norm": 1.0625,
      "learning_rate": 0.000484216845529546,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19177947640419007,
      "num_tokens": 28260448.0,
      "step": 16295
    },
    {
      "entropy": 5.540106010437012,
      "epoch": 1.2681968488620892,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048420664648700113,
      "loss": 5.1429,
      "mean_token_accuracy": 0.18850154131650926,
      "num_tokens": 28269829.0,
      "step": 16300
    },
    {
      "entropy": 5.569955635070801,
      "epoch": 1.2685858782338066,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004841964442700784,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19818419665098191,
      "num_tokens": 28278152.0,
      "step": 16305
    },
    {
      "entropy": 5.569440507888794,
      "epoch": 1.2689749076055241,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048418623887893264,
      "loss": 5.1645,
      "mean_token_accuracy": 0.1920655518770218,
      "num_tokens": 28287184.0,
      "step": 16310
    },
    {
      "entropy": 5.521697616577148,
      "epoch": 1.2693639369772418,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048417603031371867,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1947348177433014,
      "num_tokens": 28295644.0,
      "step": 16315
    },
    {
      "entropy": 5.588948154449463,
      "epoch": 1.2697529663489593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841658185745913,
      "loss": 5.0991,
      "mean_token_accuracy": 0.18904222249984742,
      "num_tokens": 28304718.0,
      "step": 16320
    },
    {
      "entropy": 5.540207099914551,
      "epoch": 1.270141995720677,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048415560366170564,
      "loss": 5.0592,
      "mean_token_accuracy": 0.19663500487804414,
      "num_tokens": 28313190.0,
      "step": 16325
    },
    {
      "entropy": 5.521539163589478,
      "epoch": 1.2705310250923945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004841453855752165,
      "loss": 5.122,
      "mean_token_accuracy": 0.1913640320301056,
      "num_tokens": 28322729.0,
      "step": 16330
    },
    {
      "entropy": 5.54754056930542,
      "epoch": 1.2709200544641122,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004841351643152791,
      "loss": 5.1772,
      "mean_token_accuracy": 0.1868587464094162,
      "num_tokens": 28331657.0,
      "step": 16335
    },
    {
      "entropy": 5.577137422561646,
      "epoch": 1.2713090838358296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004841249398820485,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19130265414714814,
      "num_tokens": 28340319.0,
      "step": 16340
    },
    {
      "entropy": 5.493728876113892,
      "epoch": 1.271698113207547,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004841147122756797,
      "loss": 5.0468,
      "mean_token_accuracy": 0.19203629493713378,
      "num_tokens": 28349303.0,
      "step": 16345
    },
    {
      "entropy": 5.573841714859009,
      "epoch": 1.2720871425792648,
      "grad_norm": 1.109375,
      "learning_rate": 0.000484104481496328,
      "loss": 5.1118,
      "mean_token_accuracy": 0.189174884557724,
      "num_tokens": 28357811.0,
      "step": 16350
    },
    {
      "entropy": 5.605975866317749,
      "epoch": 1.2724761719509823,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004840942475441486,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19020716696977616,
      "num_tokens": 28367155.0,
      "step": 16355
    },
    {
      "entropy": 5.502915096282959,
      "epoch": 1.2728652013226998,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840840104192969,
      "loss": 4.9267,
      "mean_token_accuracy": 0.20211808979511262,
      "num_tokens": 28375213.0,
      "step": 16360
    },
    {
      "entropy": 5.518007755279541,
      "epoch": 1.2732542306944175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000484073770121928,
      "loss": 5.1104,
      "mean_token_accuracy": 0.18614376485347747,
      "num_tokens": 28384085.0,
      "step": 16365
    },
    {
      "entropy": 5.475695991516114,
      "epoch": 1.273643260066135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840635266521975,
      "loss": 4.9842,
      "mean_token_accuracy": 0.19675394594669343,
      "num_tokens": 28392725.0,
      "step": 16370
    },
    {
      "entropy": 5.478953552246094,
      "epoch": 1.2740322894378526,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004840532800102607,
      "loss": 5.0836,
      "mean_token_accuracy": 0.1909507617354393,
      "num_tokens": 28401438.0,
      "step": 16375
    },
    {
      "entropy": 5.5330485820770265,
      "epoch": 1.27442131880957,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048404303019627314,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19418832212686538,
      "num_tokens": 28410703.0,
      "step": 16380
    },
    {
      "entropy": 5.527261877059937,
      "epoch": 1.2748103481812878,
      "grad_norm": 1.0,
      "learning_rate": 0.00048403277721039036,
      "loss": 4.971,
      "mean_token_accuracy": 0.20318710505962373,
      "num_tokens": 28419238.0,
      "step": 16385
    },
    {
      "entropy": 5.5728343486785885,
      "epoch": 1.2751993775530053,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004840225210527679,
      "loss": 5.0225,
      "mean_token_accuracy": 0.19700838774442672,
      "num_tokens": 28427124.0,
      "step": 16390
    },
    {
      "entropy": 5.537129497528076,
      "epoch": 1.2755884069247228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004840122617235613,
      "loss": 5.0766,
      "mean_token_accuracy": 0.19507914632558823,
      "num_tokens": 28435605.0,
      "step": 16395
    },
    {
      "entropy": 5.4511305809021,
      "epoch": 1.2759774362964404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004840019992229263,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20236364305019378,
      "num_tokens": 28443675.0,
      "step": 16400
    },
    {
      "entropy": 5.522972822189331,
      "epoch": 1.276366465668158,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048399173355101867,
      "loss": 5.0684,
      "mean_token_accuracy": 0.19576678723096846,
      "num_tokens": 28452337.0,
      "step": 16405
    },
    {
      "entropy": 5.532163143157959,
      "epoch": 1.2767554950398754,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048398146470799417,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19529873728752137,
      "num_tokens": 28461933.0,
      "step": 16410
    },
    {
      "entropy": 5.6129436016082765,
      "epoch": 1.277144524411593,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048397119269400846,
      "loss": 5.0915,
      "mean_token_accuracy": 0.19946979433298112,
      "num_tokens": 28469800.0,
      "step": 16415
    },
    {
      "entropy": 5.50711236000061,
      "epoch": 1.2775335537833106,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048396091750921766,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20323861837387086,
      "num_tokens": 28478149.0,
      "step": 16420
    },
    {
      "entropy": 5.503691339492798,
      "epoch": 1.2779225831550283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839506391537774,
      "loss": 5.1239,
      "mean_token_accuracy": 0.18725541234016418,
      "num_tokens": 28487141.0,
      "step": 16425
    },
    {
      "entropy": 5.461289024353027,
      "epoch": 1.2783116125267457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004839403576278438,
      "loss": 4.9876,
      "mean_token_accuracy": 0.19375622719526292,
      "num_tokens": 28495611.0,
      "step": 16430
    },
    {
      "entropy": 5.597225713729858,
      "epoch": 1.2787006418984634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004839300729315729,
      "loss": 5.1774,
      "mean_token_accuracy": 0.19214192479848863,
      "num_tokens": 28504525.0,
      "step": 16435
    },
    {
      "entropy": 5.548247909545898,
      "epoch": 1.279089671270181,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004839197850651206,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19279366880655288,
      "num_tokens": 28514208.0,
      "step": 16440
    },
    {
      "entropy": 5.570009851455689,
      "epoch": 1.2794787006418984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048390949402864317,
      "loss": 5.1181,
      "mean_token_accuracy": 0.1943153277039528,
      "num_tokens": 28522424.0,
      "step": 16445
    },
    {
      "entropy": 5.617263650894165,
      "epoch": 1.279867730013616,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048389919982229666,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19810546785593033,
      "num_tokens": 28530809.0,
      "step": 16450
    },
    {
      "entropy": 5.553160810470581,
      "epoch": 1.2802567593853336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048388890244623723,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1919565513730049,
      "num_tokens": 28539665.0,
      "step": 16455
    },
    {
      "entropy": 5.47309193611145,
      "epoch": 1.280645788757051,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004838786019006213,
      "loss": 4.8897,
      "mean_token_accuracy": 0.2034697189927101,
      "num_tokens": 28547958.0,
      "step": 16460
    },
    {
      "entropy": 5.529098272323608,
      "epoch": 1.2810348181287687,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048386829818560493,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20221673995256423,
      "num_tokens": 28556917.0,
      "step": 16465
    },
    {
      "entropy": 5.605680274963379,
      "epoch": 1.2814238475004862,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048385799130134464,
      "loss": 5.141,
      "mean_token_accuracy": 0.18426260501146316,
      "num_tokens": 28565807.0,
      "step": 16470
    },
    {
      "entropy": 5.5321591854095455,
      "epoch": 1.281812876872204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838476812479968,
      "loss": 5.0757,
      "mean_token_accuracy": 0.18770620077848435,
      "num_tokens": 28574321.0,
      "step": 16475
    },
    {
      "entropy": 5.4793863773345945,
      "epoch": 1.2822019062439214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048383736802571774,
      "loss": 5.025,
      "mean_token_accuracy": 0.1971733957529068,
      "num_tokens": 28582691.0,
      "step": 16480
    },
    {
      "entropy": 5.488288259506225,
      "epoch": 1.282590935615639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000483827051634664,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19521720111370086,
      "num_tokens": 28591039.0,
      "step": 16485
    },
    {
      "entropy": 5.580220890045166,
      "epoch": 1.2829799649873566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048381673207499224,
      "loss": 5.1287,
      "mean_token_accuracy": 0.18587790727615355,
      "num_tokens": 28600217.0,
      "step": 16490
    },
    {
      "entropy": 5.52021369934082,
      "epoch": 1.283368994359074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004838064093468588,
      "loss": 5.028,
      "mean_token_accuracy": 0.1940670058131218,
      "num_tokens": 28608275.0,
      "step": 16495
    },
    {
      "entropy": 5.501334381103516,
      "epoch": 1.2837580237307917,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004837960834504206,
      "loss": 5.0768,
      "mean_token_accuracy": 0.19270583540201186,
      "num_tokens": 28616829.0,
      "step": 16500
    },
    {
      "entropy": 5.5097144603729244,
      "epoch": 1.2841470531025092,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837857543858341,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1985089048743248,
      "num_tokens": 28625307.0,
      "step": 16505
    },
    {
      "entropy": 5.5988534450531,
      "epoch": 1.2845360824742267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004837754221532561,
      "loss": 5.1793,
      "mean_token_accuracy": 0.18543199449777603,
      "num_tokens": 28634858.0,
      "step": 16510
    },
    {
      "entropy": 5.5542519092559814,
      "epoch": 1.2849251118459444,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048376508675284334,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1900027424097061,
      "num_tokens": 28642934.0,
      "step": 16515
    },
    {
      "entropy": 5.507322788238525,
      "epoch": 1.2853141412176619,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048375474818475274,
      "loss": 5.0567,
      "mean_token_accuracy": 0.19527116268873215,
      "num_tokens": 28651410.0,
      "step": 16520
    },
    {
      "entropy": 5.476380777359009,
      "epoch": 1.2857031705893796,
      "grad_norm": 1.125,
      "learning_rate": 0.00048374440644914104,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20037445574998855,
      "num_tokens": 28659969.0,
      "step": 16525
    },
    {
      "entropy": 5.5409753799438475,
      "epoch": 1.286092199961097,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048373406154616523,
      "loss": 5.0921,
      "mean_token_accuracy": 0.19093319922685623,
      "num_tokens": 28668681.0,
      "step": 16530
    },
    {
      "entropy": 5.509018611907959,
      "epoch": 1.2864812293328147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048372371347598226,
      "loss": 5.0237,
      "mean_token_accuracy": 0.19172058254480362,
      "num_tokens": 28677369.0,
      "step": 16535
    },
    {
      "entropy": 5.521035289764404,
      "epoch": 1.2868702587045322,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048371336223874923,
      "loss": 5.0511,
      "mean_token_accuracy": 0.19819203913211822,
      "num_tokens": 28685843.0,
      "step": 16540
    },
    {
      "entropy": 5.599702024459839,
      "epoch": 1.2872592880762497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483703007834623,
      "loss": 5.1266,
      "mean_token_accuracy": 0.19013416320085524,
      "num_tokens": 28694382.0,
      "step": 16545
    },
    {
      "entropy": 5.573768949508667,
      "epoch": 1.2876483174479674,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004836926502637609,
      "loss": 5.1299,
      "mean_token_accuracy": 0.19363074004650116,
      "num_tokens": 28702335.0,
      "step": 16550
    },
    {
      "entropy": 5.417899751663208,
      "epoch": 1.2880373468196848,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048368228952632005,
      "loss": 4.8599,
      "mean_token_accuracy": 0.20722525119781493,
      "num_tokens": 28710302.0,
      "step": 16555
    },
    {
      "entropy": 5.554664039611817,
      "epoch": 1.2884263761914023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048367192562245756,
      "loss": 5.1016,
      "mean_token_accuracy": 0.19354615062475206,
      "num_tokens": 28719523.0,
      "step": 16560
    },
    {
      "entropy": 5.615047264099121,
      "epoch": 1.28881540556312,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048366155855233067,
      "loss": 5.1467,
      "mean_token_accuracy": 0.18935743868350982,
      "num_tokens": 28727919.0,
      "step": 16565
    },
    {
      "entropy": 5.571574306488037,
      "epoch": 1.2892044349348377,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004836511883160968,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1934323400259018,
      "num_tokens": 28735999.0,
      "step": 16570
    },
    {
      "entropy": 5.494749450683594,
      "epoch": 1.2895934643065552,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004836408149139133,
      "loss": 4.9159,
      "mean_token_accuracy": 0.1986479341983795,
      "num_tokens": 28745046.0,
      "step": 16575
    },
    {
      "entropy": 5.518639278411865,
      "epoch": 1.2899824936782727,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048363043834593743,
      "loss": 5.0628,
      "mean_token_accuracy": 0.19155598282814026,
      "num_tokens": 28753464.0,
      "step": 16580
    },
    {
      "entropy": 5.437093019485474,
      "epoch": 1.2903715230499904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004836200586123268,
      "loss": 4.993,
      "mean_token_accuracy": 0.20200058817863464,
      "num_tokens": 28762153.0,
      "step": 16585
    },
    {
      "entropy": 5.541484737396241,
      "epoch": 1.2907605524217078,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048360967571323875,
      "loss": 5.0393,
      "mean_token_accuracy": 0.1900661125779152,
      "num_tokens": 28771329.0,
      "step": 16590
    },
    {
      "entropy": 5.587620830535888,
      "epoch": 1.2911495817934253,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048359928964883094,
      "loss": 5.1447,
      "mean_token_accuracy": 0.19094182252883912,
      "num_tokens": 28779882.0,
      "step": 16595
    },
    {
      "entropy": 5.512536239624024,
      "epoch": 1.291538611165143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000483588900419261,
      "loss": 5.1175,
      "mean_token_accuracy": 0.184048992395401,
      "num_tokens": 28788757.0,
      "step": 16600
    },
    {
      "entropy": 5.5078154563903805,
      "epoch": 1.2919276405368605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004835785080246864,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20464495122432708,
      "num_tokens": 28797037.0,
      "step": 16605
    },
    {
      "entropy": 5.573528814315796,
      "epoch": 1.292316669908578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000483568112465265,
      "loss": 5.0778,
      "mean_token_accuracy": 0.19279929101467133,
      "num_tokens": 28805516.0,
      "step": 16610
    },
    {
      "entropy": 5.508523893356323,
      "epoch": 1.2927056992802957,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048355771374115436,
      "loss": 5.0534,
      "mean_token_accuracy": 0.1931885689496994,
      "num_tokens": 28813609.0,
      "step": 16615
    },
    {
      "entropy": 5.545459747314453,
      "epoch": 1.2930947286520134,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004835473118525125,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19255853146314622,
      "num_tokens": 28822156.0,
      "step": 16620
    },
    {
      "entropy": 5.493776941299439,
      "epoch": 1.2934837580237308,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004835369067994971,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19356129318475723,
      "num_tokens": 28830375.0,
      "step": 16625
    },
    {
      "entropy": 5.523644351959229,
      "epoch": 1.2938727873954483,
      "grad_norm": 1.125,
      "learning_rate": 0.000483526498582266,
      "loss": 5.1439,
      "mean_token_accuracy": 0.18733605444431306,
      "num_tokens": 28839144.0,
      "step": 16630
    },
    {
      "entropy": 5.594898414611817,
      "epoch": 1.294261816767166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048351608720097715,
      "loss": 5.0949,
      "mean_token_accuracy": 0.18919123858213424,
      "num_tokens": 28847718.0,
      "step": 16635
    },
    {
      "entropy": 5.5137866020202635,
      "epoch": 1.2946508461388835,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048350567265578867,
      "loss": 5.0625,
      "mean_token_accuracy": 0.1982119008898735,
      "num_tokens": 28856528.0,
      "step": 16640
    },
    {
      "entropy": 5.595721483230591,
      "epoch": 1.295039875510601,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834952549468584,
      "loss": 5.1687,
      "mean_token_accuracy": 0.18531151562929155,
      "num_tokens": 28866328.0,
      "step": 16645
    },
    {
      "entropy": 5.538956260681152,
      "epoch": 1.2954289048823187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004834848340743446,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18923990726470946,
      "num_tokens": 28874977.0,
      "step": 16650
    },
    {
      "entropy": 5.486423826217651,
      "epoch": 1.2958179342540361,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004834744100384052,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20128463953733444,
      "num_tokens": 28883163.0,
      "step": 16655
    },
    {
      "entropy": 5.494342613220215,
      "epoch": 1.2962069636257538,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004834639828391984,
      "loss": 5.0384,
      "mean_token_accuracy": 0.19653960615396499,
      "num_tokens": 28891730.0,
      "step": 16660
    },
    {
      "entropy": 5.592054986953736,
      "epoch": 1.2965959929974713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048345355247688256,
      "loss": 5.2146,
      "mean_token_accuracy": 0.18622260987758638,
      "num_tokens": 28901065.0,
      "step": 16665
    },
    {
      "entropy": 5.531077241897583,
      "epoch": 1.296985022369189,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004834431189516158,
      "loss": 4.9931,
      "mean_token_accuracy": 0.19885778427124023,
      "num_tokens": 28909718.0,
      "step": 16670
    },
    {
      "entropy": 5.532374715805053,
      "epoch": 1.2973740517409065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004834326822635565,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19285767525434494,
      "num_tokens": 28918171.0,
      "step": 16675
    },
    {
      "entropy": 5.496357011795044,
      "epoch": 1.297763081112624,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048342224241286296,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20289627015590667,
      "num_tokens": 28926475.0,
      "step": 16680
    },
    {
      "entropy": 5.558602237701416,
      "epoch": 1.2981521104843416,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004834117993996937,
      "loss": 5.1093,
      "mean_token_accuracy": 0.1954582676291466,
      "num_tokens": 28935521.0,
      "step": 16685
    },
    {
      "entropy": 5.514357852935791,
      "epoch": 1.2985411398560591,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004834013532242071,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18770486861467361,
      "num_tokens": 28943905.0,
      "step": 16690
    },
    {
      "entropy": 5.479977941513061,
      "epoch": 1.2989301692277766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004833909038865616,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1846332222223282,
      "num_tokens": 28952375.0,
      "step": 16695
    },
    {
      "entropy": 5.515368223190308,
      "epoch": 1.2993191985994943,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004833804513869159,
      "loss": 5.0856,
      "mean_token_accuracy": 0.19784984439611436,
      "num_tokens": 28960640.0,
      "step": 16700
    },
    {
      "entropy": 5.525284051895142,
      "epoch": 1.2997082279712118,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004833699957254284,
      "loss": 5.0684,
      "mean_token_accuracy": 0.19260049164295195,
      "num_tokens": 28969322.0,
      "step": 16705
    },
    {
      "entropy": 5.613003730773926,
      "epoch": 1.3000972573429295,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000483359536902258,
      "loss": 5.1646,
      "mean_token_accuracy": 0.19030515998601913,
      "num_tokens": 28978471.0,
      "step": 16710
    },
    {
      "entropy": 5.500603294372558,
      "epoch": 1.300486286714647,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004833490749175632,
      "loss": 4.9782,
      "mean_token_accuracy": 0.1997521013021469,
      "num_tokens": 28986627.0,
      "step": 16715
    },
    {
      "entropy": 5.485197257995606,
      "epoch": 1.3008753160863646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048333860977150286,
      "loss": 5.0703,
      "mean_token_accuracy": 0.19697431474924088,
      "num_tokens": 28995647.0,
      "step": 16720
    },
    {
      "entropy": 5.6010294437408445,
      "epoch": 1.3012643454580821,
      "grad_norm": 1.125,
      "learning_rate": 0.00048332814146423566,
      "loss": 5.055,
      "mean_token_accuracy": 0.19352222979068756,
      "num_tokens": 29003591.0,
      "step": 16725
    },
    {
      "entropy": 5.499473285675049,
      "epoch": 1.3016533748297996,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004833176699959206,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20607816129922868,
      "num_tokens": 29012150.0,
      "step": 16730
    },
    {
      "entropy": 5.501286172866822,
      "epoch": 1.3020424042015173,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048330719536671633,
      "loss": 5.1095,
      "mean_token_accuracy": 0.19094256460666656,
      "num_tokens": 29020806.0,
      "step": 16735
    },
    {
      "entropy": 5.587902069091797,
      "epoch": 1.3024314335732348,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000483296717576782,
      "loss": 5.1694,
      "mean_token_accuracy": 0.18659210205078125,
      "num_tokens": 29030074.0,
      "step": 16740
    },
    {
      "entropy": 5.5728682518005375,
      "epoch": 1.3028204629449522,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004832862366262765,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1951389193534851,
      "num_tokens": 29038861.0,
      "step": 16745
    },
    {
      "entropy": 5.480522871017456,
      "epoch": 1.30320949231667,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048327575251535886,
      "loss": 4.9338,
      "mean_token_accuracy": 0.20013366341590882,
      "num_tokens": 29046260.0,
      "step": 16750
    },
    {
      "entropy": 5.456730604171753,
      "epoch": 1.3035985216883874,
      "grad_norm": 1.125,
      "learning_rate": 0.0004832652652441883,
      "loss": 5.0021,
      "mean_token_accuracy": 0.19843949377536774,
      "num_tokens": 29054722.0,
      "step": 16755
    },
    {
      "entropy": 5.527504110336304,
      "epoch": 1.303987551060105,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048325477481292375,
      "loss": 5.0964,
      "mean_token_accuracy": 0.1871419832110405,
      "num_tokens": 29063700.0,
      "step": 16760
    },
    {
      "entropy": 5.522088956832886,
      "epoch": 1.3043765804318226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004832442812217244,
      "loss": 4.9979,
      "mean_token_accuracy": 0.19686779230833054,
      "num_tokens": 29072254.0,
      "step": 16765
    },
    {
      "entropy": 5.524970865249633,
      "epoch": 1.3047656098035403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004832337844707496,
      "loss": 5.0952,
      "mean_token_accuracy": 0.19259236603975297,
      "num_tokens": 29081999.0,
      "step": 16770
    },
    {
      "entropy": 5.503440237045288,
      "epoch": 1.3051546391752578,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048322328456015855,
      "loss": 5.0112,
      "mean_token_accuracy": 0.19622527211904525,
      "num_tokens": 29090493.0,
      "step": 16775
    },
    {
      "entropy": 5.571767854690552,
      "epoch": 1.3055436685469752,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048321278149011053,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19958551824092866,
      "num_tokens": 29099172.0,
      "step": 16780
    },
    {
      "entropy": 5.569439268112182,
      "epoch": 1.305932697918693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048320227526076504,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19150739014148713,
      "num_tokens": 29108234.0,
      "step": 16785
    },
    {
      "entropy": 5.576681232452392,
      "epoch": 1.3063217272904104,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004831917658722814,
      "loss": 5.1133,
      "mean_token_accuracy": 0.19355372190475464,
      "num_tokens": 29117307.0,
      "step": 16790
    },
    {
      "entropy": 5.555779218673706,
      "epoch": 1.3067107566621279,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048318125332481903,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19167981594800948,
      "num_tokens": 29125552.0,
      "step": 16795
    },
    {
      "entropy": 5.531487274169922,
      "epoch": 1.3070997860338456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048317073761853764,
      "loss": 4.9819,
      "mean_token_accuracy": 0.20302668809890748,
      "num_tokens": 29134559.0,
      "step": 16800
    },
    {
      "entropy": 5.574425792694091,
      "epoch": 1.307488815405563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004831602187535965,
      "loss": 5.1536,
      "mean_token_accuracy": 0.18791132271289826,
      "num_tokens": 29143121.0,
      "step": 16805
    },
    {
      "entropy": 5.612791109085083,
      "epoch": 1.3078778447772808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004831496967301554,
      "loss": 5.165,
      "mean_token_accuracy": 0.18933375626802446,
      "num_tokens": 29153012.0,
      "step": 16810
    },
    {
      "entropy": 5.61829981803894,
      "epoch": 1.3082668741489982,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048313917154837386,
      "loss": 5.0861,
      "mean_token_accuracy": 0.18930458575487136,
      "num_tokens": 29161451.0,
      "step": 16815
    },
    {
      "entropy": 5.496663665771484,
      "epoch": 1.308655903520716,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004831286432084118,
      "loss": 5.0673,
      "mean_token_accuracy": 0.1879603996872902,
      "num_tokens": 29170691.0,
      "step": 16820
    },
    {
      "entropy": 5.569409036636353,
      "epoch": 1.3090449328924334,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004831181117104289,
      "loss": 5.0409,
      "mean_token_accuracy": 0.19415445178747176,
      "num_tokens": 29179111.0,
      "step": 16825
    },
    {
      "entropy": 5.630125999450684,
      "epoch": 1.3094339622641509,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048310757705458476,
      "loss": 5.1631,
      "mean_token_accuracy": 0.18859304636716842,
      "num_tokens": 29187444.0,
      "step": 16830
    },
    {
      "entropy": 5.504996013641358,
      "epoch": 1.3098229916358686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048309703924103944,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2017731711268425,
      "num_tokens": 29196830.0,
      "step": 16835
    },
    {
      "entropy": 5.459016180038452,
      "epoch": 1.310212021007586,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048308649826995283,
      "loss": 5.0153,
      "mean_token_accuracy": 0.19524496495723725,
      "num_tokens": 29205587.0,
      "step": 16840
    },
    {
      "entropy": 5.538132858276367,
      "epoch": 1.3106010503793035,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048307595414148475,
      "loss": 5.1125,
      "mean_token_accuracy": 0.1906437784433365,
      "num_tokens": 29213961.0,
      "step": 16845
    },
    {
      "entropy": 5.5479536056518555,
      "epoch": 1.3109900797510212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004830654068557952,
      "loss": 5.1051,
      "mean_token_accuracy": 0.18761806339025497,
      "num_tokens": 29222744.0,
      "step": 16850
    },
    {
      "entropy": 5.5231012344360355,
      "epoch": 1.3113791091227387,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004830548564130444,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20277958512306213,
      "num_tokens": 29231378.0,
      "step": 16855
    },
    {
      "entropy": 5.511657762527466,
      "epoch": 1.3117681384944564,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048304430281339216,
      "loss": 5.0414,
      "mean_token_accuracy": 0.19683454036712647,
      "num_tokens": 29238909.0,
      "step": 16860
    },
    {
      "entropy": 5.457151556015015,
      "epoch": 1.3121571678661739,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004830337460569989,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2008628159761429,
      "num_tokens": 29247132.0,
      "step": 16865
    },
    {
      "entropy": 5.6079387187957765,
      "epoch": 1.3125461972378916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004830231861440246,
      "loss": 5.112,
      "mean_token_accuracy": 0.19339798837900163,
      "num_tokens": 29255322.0,
      "step": 16870
    },
    {
      "entropy": 5.635947132110596,
      "epoch": 1.312935226609609,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004830126230746295,
      "loss": 5.1624,
      "mean_token_accuracy": 0.18725260496139526,
      "num_tokens": 29265000.0,
      "step": 16875
    },
    {
      "entropy": 5.538443183898925,
      "epoch": 1.3133242559813265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048300205684897405,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19440464824438095,
      "num_tokens": 29273139.0,
      "step": 16880
    },
    {
      "entropy": 5.567435216903687,
      "epoch": 1.3137132853530442,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004829914874672184,
      "loss": 5.1128,
      "mean_token_accuracy": 0.1865583539009094,
      "num_tokens": 29281734.0,
      "step": 16885
    },
    {
      "entropy": 5.5537354946136475,
      "epoch": 1.3141023147247617,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048298091492952297,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18811049163341523,
      "num_tokens": 29290296.0,
      "step": 16890
    },
    {
      "entropy": 5.614425373077393,
      "epoch": 1.3144913440964792,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004829703392360482,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19014693051576614,
      "num_tokens": 29298164.0,
      "step": 16895
    },
    {
      "entropy": 5.506893873214722,
      "epoch": 1.3148803734681969,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048295976038695455,
      "loss": 4.9782,
      "mean_token_accuracy": 0.19886327683925628,
      "num_tokens": 29307420.0,
      "step": 16900
    },
    {
      "entropy": 5.579666471481323,
      "epoch": 1.3152694028399143,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004829491783824025,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19567200541496277,
      "num_tokens": 29316041.0,
      "step": 16905
    },
    {
      "entropy": 5.520770645141601,
      "epoch": 1.315658432211632,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048293859322255276,
      "loss": 5.058,
      "mean_token_accuracy": 0.19597916156053544,
      "num_tokens": 29323750.0,
      "step": 16910
    },
    {
      "entropy": 5.471966791152954,
      "epoch": 1.3160474615833495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004829280049075657,
      "loss": 4.9623,
      "mean_token_accuracy": 0.203752002120018,
      "num_tokens": 29333108.0,
      "step": 16915
    },
    {
      "entropy": 5.585264730453491,
      "epoch": 1.3164364909550672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048291741343760216,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19052098244428634,
      "num_tokens": 29341976.0,
      "step": 16920
    },
    {
      "entropy": 5.592809057235717,
      "epoch": 1.3168255203267847,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004829068188128229,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19758483469486238,
      "num_tokens": 29350517.0,
      "step": 16925
    },
    {
      "entropy": 5.584476661682129,
      "epoch": 1.3172145496985022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004828962210333885,
      "loss": 5.1044,
      "mean_token_accuracy": 0.18689894825220107,
      "num_tokens": 29359410.0,
      "step": 16930
    },
    {
      "entropy": 5.537783575057984,
      "epoch": 1.3176035790702199,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004828856200994598,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2034487783908844,
      "num_tokens": 29367164.0,
      "step": 16935
    },
    {
      "entropy": 5.539615249633789,
      "epoch": 1.3179926084419373,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004828750160111978,
      "loss": 5.1279,
      "mean_token_accuracy": 0.19479843229055405,
      "num_tokens": 29375700.0,
      "step": 16940
    },
    {
      "entropy": 5.482031726837159,
      "epoch": 1.3183816378136548,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048286440876876325,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20310419648885727,
      "num_tokens": 29383953.0,
      "step": 16945
    },
    {
      "entropy": 5.452842140197754,
      "epoch": 1.3187706671853725,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048285379837231714,
      "loss": 4.9269,
      "mean_token_accuracy": 0.20126338303089142,
      "num_tokens": 29391995.0,
      "step": 16950
    },
    {
      "entropy": 5.524240970611572,
      "epoch": 1.31915969655709,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004828431848220205,
      "loss": 5.0205,
      "mean_token_accuracy": 0.19970037788152695,
      "num_tokens": 29400358.0,
      "step": 16955
    },
    {
      "entropy": 5.528271722793579,
      "epoch": 1.3195487259288077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004828325681180343,
      "loss": 5.1426,
      "mean_token_accuracy": 0.1882537841796875,
      "num_tokens": 29408598.0,
      "step": 16960
    },
    {
      "entropy": 5.609574794769287,
      "epoch": 1.3199377553005252,
      "grad_norm": 1.125,
      "learning_rate": 0.0004828219482605197,
      "loss": 5.1898,
      "mean_token_accuracy": 0.18205541670322417,
      "num_tokens": 29416623.0,
      "step": 16965
    },
    {
      "entropy": 5.557714891433716,
      "epoch": 1.3203267846722428,
      "grad_norm": 1.125,
      "learning_rate": 0.00048281132524963786,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20188557207584382,
      "num_tokens": 29424593.0,
      "step": 16970
    },
    {
      "entropy": 5.5288519859313965,
      "epoch": 1.3207158140439603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004828006990855499,
      "loss": 5.0791,
      "mean_token_accuracy": 0.19430007934570312,
      "num_tokens": 29433405.0,
      "step": 16975
    },
    {
      "entropy": 5.528368854522705,
      "epoch": 1.3211048434156778,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048279006976841704,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1956000655889511,
      "num_tokens": 29441714.0,
      "step": 16980
    },
    {
      "entropy": 5.546284246444702,
      "epoch": 1.3214938727873955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004827794372984007,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20161654353141784,
      "num_tokens": 29450594.0,
      "step": 16985
    },
    {
      "entropy": 5.506742858886719,
      "epoch": 1.321882902159113,
      "grad_norm": 1.078125,
      "learning_rate": 0.000482768801675662,
      "loss": 5.0154,
      "mean_token_accuracy": 0.19701900482177734,
      "num_tokens": 29459208.0,
      "step": 16990
    },
    {
      "entropy": 5.556023168563843,
      "epoch": 1.3222719315308304,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004827581629003625,
      "loss": 5.1031,
      "mean_token_accuracy": 0.18960306495428086,
      "num_tokens": 29467766.0,
      "step": 16995
    },
    {
      "entropy": 5.583490800857544,
      "epoch": 1.3226609609025481,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048274752097266356,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19262328445911409,
      "num_tokens": 29476732.0,
      "step": 17000
    },
    {
      "entropy": 5.605873012542725,
      "epoch": 1.3230499902742658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004827368758927266,
      "loss": 5.1497,
      "mean_token_accuracy": 0.18411410748958587,
      "num_tokens": 29485199.0,
      "step": 17005
    },
    {
      "entropy": 5.537188386917114,
      "epoch": 1.3234390196459833,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004827262276607132,
      "loss": 5.0599,
      "mean_token_accuracy": 0.19417423605918885,
      "num_tokens": 29493039.0,
      "step": 17010
    },
    {
      "entropy": 5.56510009765625,
      "epoch": 1.3238280490177008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000482715576276785,
      "loss": 5.0585,
      "mean_token_accuracy": 0.19691295474767684,
      "num_tokens": 29501655.0,
      "step": 17015
    },
    {
      "entropy": 5.613816547393799,
      "epoch": 1.3242170783894185,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827049217411035,
      "loss": 5.111,
      "mean_token_accuracy": 0.1867526188492775,
      "num_tokens": 29510261.0,
      "step": 17020
    },
    {
      "entropy": 5.519961786270142,
      "epoch": 1.324606107761136,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048269426405383043,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2027512952685356,
      "num_tokens": 29518382.0,
      "step": 17025
    },
    {
      "entropy": 5.512136936187744,
      "epoch": 1.3249951371328534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048268360321512745,
      "loss": 5.057,
      "mean_token_accuracy": 0.190300789475441,
      "num_tokens": 29526814.0,
      "step": 17030
    },
    {
      "entropy": 5.599355983734131,
      "epoch": 1.3253841665045711,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004826729392251564,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19459279924631118,
      "num_tokens": 29535169.0,
      "step": 17035
    },
    {
      "entropy": 5.626461219787598,
      "epoch": 1.3257731958762886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000482662272084079,
      "loss": 5.1579,
      "mean_token_accuracy": 0.19368110597133636,
      "num_tokens": 29543368.0,
      "step": 17040
    },
    {
      "entropy": 5.597419738769531,
      "epoch": 1.326162225248006,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004826516017920571,
      "loss": 5.118,
      "mean_token_accuracy": 0.18922984153032302,
      "num_tokens": 29553042.0,
      "step": 17045
    },
    {
      "entropy": 5.6120974063873295,
      "epoch": 1.3265512546197238,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004826409283492528,
      "loss": 5.1149,
      "mean_token_accuracy": 0.19020629227161406,
      "num_tokens": 29561787.0,
      "step": 17050
    },
    {
      "entropy": 5.544434881210327,
      "epoch": 1.3269402839914415,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048263025175582783,
      "loss": 5.1137,
      "mean_token_accuracy": 0.1880372181534767,
      "num_tokens": 29570243.0,
      "step": 17055
    },
    {
      "entropy": 5.544651651382447,
      "epoch": 1.327329313363159,
      "grad_norm": 1.125,
      "learning_rate": 0.0004826195720119442,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1981926068663597,
      "num_tokens": 29578560.0,
      "step": 17060
    },
    {
      "entropy": 5.598849821090698,
      "epoch": 1.3277183427348764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004826088891177641,
      "loss": 5.1197,
      "mean_token_accuracy": 0.1871596395969391,
      "num_tokens": 29587441.0,
      "step": 17065
    },
    {
      "entropy": 5.5302722454071045,
      "epoch": 1.3281073721065941,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048259820307344954,
      "loss": 5.0239,
      "mean_token_accuracy": 0.19634164571762086,
      "num_tokens": 29596372.0,
      "step": 17070
    },
    {
      "entropy": 5.546082544326782,
      "epoch": 1.3284964014783116,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048258751387916267,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19748930633068085,
      "num_tokens": 29605263.0,
      "step": 17075
    },
    {
      "entropy": 5.537404870986938,
      "epoch": 1.328885430850029,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048257682153506564,
      "loss": 5.1308,
      "mean_token_accuracy": 0.193301922082901,
      "num_tokens": 29614397.0,
      "step": 17080
    },
    {
      "entropy": 5.630432844161987,
      "epoch": 1.3292744602217468,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004825661260413208,
      "loss": 5.1837,
      "mean_token_accuracy": 0.18694810271263124,
      "num_tokens": 29622996.0,
      "step": 17085
    },
    {
      "entropy": 5.539497661590576,
      "epoch": 1.3296634895934643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048255542739809035,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2008394867181778,
      "num_tokens": 29630960.0,
      "step": 17090
    },
    {
      "entropy": 5.59538164138794,
      "epoch": 1.330052518965182,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048254472560553665,
      "loss": 5.1535,
      "mean_token_accuracy": 0.18763921707868575,
      "num_tokens": 29639520.0,
      "step": 17095
    },
    {
      "entropy": 5.516068029403686,
      "epoch": 1.3304415483368994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048253402066382207,
      "loss": 5.0172,
      "mean_token_accuracy": 0.19877759665250777,
      "num_tokens": 29648542.0,
      "step": 17100
    },
    {
      "entropy": 5.539023351669312,
      "epoch": 1.3308305777086171,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004825233125731091,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18925854861736296,
      "num_tokens": 29657604.0,
      "step": 17105
    },
    {
      "entropy": 5.609941720962524,
      "epoch": 1.3312196070803346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048251260133356014,
      "loss": 5.1401,
      "mean_token_accuracy": 0.19245071709156036,
      "num_tokens": 29665811.0,
      "step": 17110
    },
    {
      "entropy": 5.503051328659057,
      "epoch": 1.331608636452052,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048250188694533774,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18603211492300034,
      "num_tokens": 29674851.0,
      "step": 17115
    },
    {
      "entropy": 5.535083389282226,
      "epoch": 1.3319976658237698,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004824911694086044,
      "loss": 5.0187,
      "mean_token_accuracy": 0.19593706130981445,
      "num_tokens": 29683478.0,
      "step": 17120
    },
    {
      "entropy": 5.499434614181519,
      "epoch": 1.3323866951954872,
      "grad_norm": 1.046875,
      "learning_rate": 0.000482480448723523,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19217196404933928,
      "num_tokens": 29691765.0,
      "step": 17125
    },
    {
      "entropy": 5.57268009185791,
      "epoch": 1.3327757245672047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048246972489025586,
      "loss": 5.0339,
      "mean_token_accuracy": 0.19644560813903808,
      "num_tokens": 29700271.0,
      "step": 17130
    },
    {
      "entropy": 5.59065203666687,
      "epoch": 1.3331647539389224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048245899790896597,
      "loss": 5.1107,
      "mean_token_accuracy": 0.19127732813358306,
      "num_tokens": 29709004.0,
      "step": 17135
    },
    {
      "entropy": 5.555433130264282,
      "epoch": 1.33355378331064,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004824482677798159,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19528018236160277,
      "num_tokens": 29717479.0,
      "step": 17140
    },
    {
      "entropy": 5.543634223937988,
      "epoch": 1.3339428126823576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048243753450296863,
      "loss": 5.0705,
      "mean_token_accuracy": 0.19502720683813096,
      "num_tokens": 29725859.0,
      "step": 17145
    },
    {
      "entropy": 5.494951343536377,
      "epoch": 1.334331842054075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048242679807858693,
      "loss": 5.0405,
      "mean_token_accuracy": 0.1928912043571472,
      "num_tokens": 29734201.0,
      "step": 17150
    },
    {
      "entropy": 5.561113882064819,
      "epoch": 1.3347208714257928,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048241605850683365,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1934991106390953,
      "num_tokens": 29742552.0,
      "step": 17155
    },
    {
      "entropy": 5.488109445571899,
      "epoch": 1.3351099007975102,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048240531578787197,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19432728886604309,
      "num_tokens": 29751297.0,
      "step": 17160
    },
    {
      "entropy": 5.496108865737915,
      "epoch": 1.3354989301692277,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004823945699218647,
      "loss": 5.113,
      "mean_token_accuracy": 0.19043429642915727,
      "num_tokens": 29759262.0,
      "step": 17165
    },
    {
      "entropy": 5.567288780212403,
      "epoch": 1.3358879595409454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004823838209089749,
      "loss": 5.104,
      "mean_token_accuracy": 0.18988242894411086,
      "num_tokens": 29767428.0,
      "step": 17170
    },
    {
      "entropy": 5.628966188430786,
      "epoch": 1.3362769889126629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048237306874936565,
      "loss": 5.0888,
      "mean_token_accuracy": 0.19549930542707444,
      "num_tokens": 29776252.0,
      "step": 17175
    },
    {
      "entropy": 5.540105962753296,
      "epoch": 1.3366660182843804,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004823623134432001,
      "loss": 5.0378,
      "mean_token_accuracy": 0.18961155712604522,
      "num_tokens": 29784520.0,
      "step": 17180
    },
    {
      "entropy": 5.551808071136475,
      "epoch": 1.337055047656098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048235155499064166,
      "loss": 5.165,
      "mean_token_accuracy": 0.18573916852474212,
      "num_tokens": 29793452.0,
      "step": 17185
    },
    {
      "entropy": 5.497555875778199,
      "epoch": 1.3374440770278155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048234079339185335,
      "loss": 5.0021,
      "mean_token_accuracy": 0.19956104457378387,
      "num_tokens": 29802278.0,
      "step": 17190
    },
    {
      "entropy": 5.6366273880004885,
      "epoch": 1.3378331063995332,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048233002864699856,
      "loss": 5.1621,
      "mean_token_accuracy": 0.1898314505815506,
      "num_tokens": 29811302.0,
      "step": 17195
    },
    {
      "entropy": 5.480880784988403,
      "epoch": 1.3382221357712507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004823192607562405,
      "loss": 4.9221,
      "mean_token_accuracy": 0.20390674471855164,
      "num_tokens": 29819555.0,
      "step": 17200
    },
    {
      "entropy": 5.515711927413941,
      "epoch": 1.3386111651429684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048230848971974265,
      "loss": 5.1968,
      "mean_token_accuracy": 0.18707700520753862,
      "num_tokens": 29828929.0,
      "step": 17205
    },
    {
      "entropy": 5.635153341293335,
      "epoch": 1.3390001945146859,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004822977155376684,
      "loss": 5.184,
      "mean_token_accuracy": 0.1915266200900078,
      "num_tokens": 29837703.0,
      "step": 17210
    },
    {
      "entropy": 5.5148398876190186,
      "epoch": 1.3393892238864034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004822869382101814,
      "loss": 4.9623,
      "mean_token_accuracy": 0.198391392827034,
      "num_tokens": 29846053.0,
      "step": 17215
    },
    {
      "entropy": 5.455366611480713,
      "epoch": 1.339778253258121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004822761577374449,
      "loss": 4.9895,
      "mean_token_accuracy": 0.204795441031456,
      "num_tokens": 29854970.0,
      "step": 17220
    },
    {
      "entropy": 5.545156049728393,
      "epoch": 1.3401672826298385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004822653741196227,
      "loss": 5.0912,
      "mean_token_accuracy": 0.19040034115314483,
      "num_tokens": 29863835.0,
      "step": 17225
    },
    {
      "entropy": 5.541007852554321,
      "epoch": 1.340556312001556,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004822545873568783,
      "loss": 5.0564,
      "mean_token_accuracy": 0.1968723341822624,
      "num_tokens": 29872149.0,
      "step": 17230
    },
    {
      "entropy": 5.551089239120484,
      "epoch": 1.3409453413732737,
      "grad_norm": 1.125,
      "learning_rate": 0.00048224379744937545,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19499320685863494,
      "num_tokens": 29880503.0,
      "step": 17235
    },
    {
      "entropy": 5.506267499923706,
      "epoch": 1.3413343707449912,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048223300439727783,
      "loss": 5.1295,
      "mean_token_accuracy": 0.18950352370738982,
      "num_tokens": 29889090.0,
      "step": 17240
    },
    {
      "entropy": 5.558257341384888,
      "epoch": 1.3417234001167089,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048222220820074926,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19792360812425613,
      "num_tokens": 29897681.0,
      "step": 17245
    },
    {
      "entropy": 5.57754545211792,
      "epoch": 1.3421124294884264,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048221140885995346,
      "loss": 5.1354,
      "mean_token_accuracy": 0.18972157388925553,
      "num_tokens": 29905415.0,
      "step": 17250
    },
    {
      "entropy": 5.565930652618408,
      "epoch": 1.342501458860144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004822006063750543,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19440682530403136,
      "num_tokens": 29914876.0,
      "step": 17255
    },
    {
      "entropy": 5.576021575927735,
      "epoch": 1.3428904882318615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048218980074621575,
      "loss": 5.1366,
      "mean_token_accuracy": 0.18949638605117797,
      "num_tokens": 29923965.0,
      "step": 17260
    },
    {
      "entropy": 5.564385652542114,
      "epoch": 1.343279517603579,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048217899197360183,
      "loss": 5.0582,
      "mean_token_accuracy": 0.19049679934978486,
      "num_tokens": 29932618.0,
      "step": 17265
    },
    {
      "entropy": 5.616783475875854,
      "epoch": 1.3436685469752967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004821681800573764,
      "loss": 5.1547,
      "mean_token_accuracy": 0.19399770945310593,
      "num_tokens": 29941314.0,
      "step": 17270
    },
    {
      "entropy": 5.609663391113282,
      "epoch": 1.3440575763470142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004821573649977036,
      "loss": 5.1169,
      "mean_token_accuracy": 0.18683778196573259,
      "num_tokens": 29949808.0,
      "step": 17275
    },
    {
      "entropy": 5.537212228775024,
      "epoch": 1.3444466057187316,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048214654679474753,
      "loss": 5.0833,
      "mean_token_accuracy": 0.1958432376384735,
      "num_tokens": 29959225.0,
      "step": 17280
    },
    {
      "entropy": 5.546075868606567,
      "epoch": 1.3448356350904493,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048213572544867224,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1891117811203003,
      "num_tokens": 29968668.0,
      "step": 17285
    },
    {
      "entropy": 5.535118865966797,
      "epoch": 1.3452246644621668,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048212490095964213,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18940555453300476,
      "num_tokens": 29977502.0,
      "step": 17290
    },
    {
      "entropy": 5.488551950454712,
      "epoch": 1.3456136938338845,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004821140733278212,
      "loss": 4.9988,
      "mean_token_accuracy": 0.19297264218330384,
      "num_tokens": 29986444.0,
      "step": 17295
    },
    {
      "entropy": 5.625052833557129,
      "epoch": 1.346002723205602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004821032425533739,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1863955855369568,
      "num_tokens": 29995071.0,
      "step": 17300
    },
    {
      "entropy": 5.583033704757691,
      "epoch": 1.3463917525773197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004820924086364646,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18945569545030594,
      "num_tokens": 30003071.0,
      "step": 17305
    },
    {
      "entropy": 5.5127723693847654,
      "epoch": 1.3467807819490372,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048208157157725756,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19626446962356567,
      "num_tokens": 30011899.0,
      "step": 17310
    },
    {
      "entropy": 5.529442024230957,
      "epoch": 1.3471698113207546,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048207073137591726,
      "loss": 4.992,
      "mean_token_accuracy": 0.20535602569580078,
      "num_tokens": 30020454.0,
      "step": 17315
    },
    {
      "entropy": 5.515955877304077,
      "epoch": 1.3475588406924723,
      "grad_norm": 1.125,
      "learning_rate": 0.00048205988803260824,
      "loss": 5.053,
      "mean_token_accuracy": 0.19165019541978837,
      "num_tokens": 30028095.0,
      "step": 17320
    },
    {
      "entropy": 5.53240270614624,
      "epoch": 1.3479478700641898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048204904154749496,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1942714512348175,
      "num_tokens": 30037051.0,
      "step": 17325
    },
    {
      "entropy": 5.593975305557251,
      "epoch": 1.3483368994359073,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048203819192074206,
      "loss": 5.061,
      "mean_token_accuracy": 0.19956636130809785,
      "num_tokens": 30045513.0,
      "step": 17330
    },
    {
      "entropy": 5.596064901351928,
      "epoch": 1.348725928807625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048202733915251407,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1856134831905365,
      "num_tokens": 30054132.0,
      "step": 17335
    },
    {
      "entropy": 5.548170518875122,
      "epoch": 1.3491149581793425,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048201648324297573,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19005759209394454,
      "num_tokens": 30062850.0,
      "step": 17340
    },
    {
      "entropy": 5.498459196090698,
      "epoch": 1.3495039875510602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048200562419229185,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1962674990296364,
      "num_tokens": 30072186.0,
      "step": 17345
    },
    {
      "entropy": 5.557011508941651,
      "epoch": 1.3498930169227776,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000481994762000627,
      "loss": 5.052,
      "mean_token_accuracy": 0.1971317023038864,
      "num_tokens": 30080931.0,
      "step": 17350
    },
    {
      "entropy": 5.5418037414550785,
      "epoch": 1.3502820462944953,
      "grad_norm": 1.125,
      "learning_rate": 0.00048198389666814616,
      "loss": 5.0152,
      "mean_token_accuracy": 0.19672925770282745,
      "num_tokens": 30089607.0,
      "step": 17355
    },
    {
      "entropy": 5.583463191986084,
      "epoch": 1.3506710756662128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048197302819501416,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18593638986349106,
      "num_tokens": 30098893.0,
      "step": 17360
    },
    {
      "entropy": 5.629031991958618,
      "epoch": 1.3510601050379303,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048196215658139586,
      "loss": 5.1011,
      "mean_token_accuracy": 0.19326601624488832,
      "num_tokens": 30107304.0,
      "step": 17365
    },
    {
      "entropy": 5.646771192550659,
      "epoch": 1.351449134409648,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004819512818274562,
      "loss": 5.1607,
      "mean_token_accuracy": 0.19044991135597228,
      "num_tokens": 30116196.0,
      "step": 17370
    },
    {
      "entropy": 5.5932622909545895,
      "epoch": 1.3518381637813655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004819404039333603,
      "loss": 5.0905,
      "mean_token_accuracy": 0.18964900821447372,
      "num_tokens": 30124691.0,
      "step": 17375
    },
    {
      "entropy": 5.584123229980468,
      "epoch": 1.352227193153083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004819295228992731,
      "loss": 5.0706,
      "mean_token_accuracy": 0.189251109957695,
      "num_tokens": 30133824.0,
      "step": 17380
    },
    {
      "entropy": 5.555778312683105,
      "epoch": 1.3526162225248006,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048191863872535984,
      "loss": 5.0254,
      "mean_token_accuracy": 0.19817333221435546,
      "num_tokens": 30142327.0,
      "step": 17385
    },
    {
      "entropy": 5.572787523269653,
      "epoch": 1.3530052518965183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004819077514117855,
      "loss": 5.0624,
      "mean_token_accuracy": 0.19490636140108109,
      "num_tokens": 30150774.0,
      "step": 17390
    },
    {
      "entropy": 5.510891914367676,
      "epoch": 1.3533942812682358,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048189686095871545,
      "loss": 4.9694,
      "mean_token_accuracy": 0.19359280169010162,
      "num_tokens": 30158593.0,
      "step": 17395
    },
    {
      "entropy": 5.5235076427459715,
      "epoch": 1.3537833106399533,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048188596736631475,
      "loss": 5.0428,
      "mean_token_accuracy": 0.19750061333179475,
      "num_tokens": 30166556.0,
      "step": 17400
    },
    {
      "entropy": 5.499352025985718,
      "epoch": 1.354172340011671,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048187507063474885,
      "loss": 5.0143,
      "mean_token_accuracy": 0.19385875314474105,
      "num_tokens": 30175082.0,
      "step": 17405
    },
    {
      "entropy": 5.531732988357544,
      "epoch": 1.3545613693833884,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048186417076418294,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20032892972230912,
      "num_tokens": 30184024.0,
      "step": 17410
    },
    {
      "entropy": 5.597299003601075,
      "epoch": 1.354950398755106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048185326775478256,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19109995812177658,
      "num_tokens": 30192083.0,
      "step": 17415
    },
    {
      "entropy": 5.5226678371429445,
      "epoch": 1.3553394281268236,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048184236160671306,
      "loss": 4.9534,
      "mean_token_accuracy": 0.20559277534484863,
      "num_tokens": 30199731.0,
      "step": 17420
    },
    {
      "entropy": 5.465492868423462,
      "epoch": 1.355728457498541,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048183145232013995,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1919456973671913,
      "num_tokens": 30207444.0,
      "step": 17425
    },
    {
      "entropy": 5.577214622497559,
      "epoch": 1.3561174868702586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048182053989522883,
      "loss": 5.1771,
      "mean_token_accuracy": 0.18968395590782167,
      "num_tokens": 30215716.0,
      "step": 17430
    },
    {
      "entropy": 5.468042850494385,
      "epoch": 1.3565065162419763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048180962433214515,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2024299219250679,
      "num_tokens": 30223910.0,
      "step": 17435
    },
    {
      "entropy": 5.5320042133331295,
      "epoch": 1.356895545613694,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048179870563105453,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1996099129319191,
      "num_tokens": 30233234.0,
      "step": 17440
    },
    {
      "entropy": 5.501065111160278,
      "epoch": 1.3572845749854114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004817877837921228,
      "loss": 5.0286,
      "mean_token_accuracy": 0.19719744175672532,
      "num_tokens": 30242264.0,
      "step": 17445
    },
    {
      "entropy": 5.572187805175782,
      "epoch": 1.357673604357129,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004817768588155155,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19251794666051864,
      "num_tokens": 30250606.0,
      "step": 17450
    },
    {
      "entropy": 5.611683988571167,
      "epoch": 1.3580626337288466,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048176593070139865,
      "loss": 5.1933,
      "mean_token_accuracy": 0.18453987091779708,
      "num_tokens": 30259197.0,
      "step": 17455
    },
    {
      "entropy": 5.484611940383911,
      "epoch": 1.358451663100564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048175499944993767,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2062402233481407,
      "num_tokens": 30267896.0,
      "step": 17460
    },
    {
      "entropy": 5.57444429397583,
      "epoch": 1.3588406924722816,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048174406506129884,
      "loss": 5.108,
      "mean_token_accuracy": 0.1988934725522995,
      "num_tokens": 30277053.0,
      "step": 17465
    },
    {
      "entropy": 5.617016077041626,
      "epoch": 1.3592297218439993,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048173312753564787,
      "loss": 5.117,
      "mean_token_accuracy": 0.19550347775220872,
      "num_tokens": 30286075.0,
      "step": 17470
    },
    {
      "entropy": 5.537977838516236,
      "epoch": 1.3596187512157167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004817221868731506,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19449779391288757,
      "num_tokens": 30295329.0,
      "step": 17475
    },
    {
      "entropy": 5.576938629150391,
      "epoch": 1.3600077805874344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048171124307397334,
      "loss": 5.1429,
      "mean_token_accuracy": 0.19321171939373016,
      "num_tokens": 30303685.0,
      "step": 17480
    },
    {
      "entropy": 5.496486568450928,
      "epoch": 1.360396809959152,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004817002961382819,
      "loss": 5.0076,
      "mean_token_accuracy": 0.1937185913324356,
      "num_tokens": 30312885.0,
      "step": 17485
    },
    {
      "entropy": 5.465707874298095,
      "epoch": 1.3607858393308696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048168934606624255,
      "loss": 4.9154,
      "mean_token_accuracy": 0.20480719655752183,
      "num_tokens": 30321368.0,
      "step": 17490
    },
    {
      "entropy": 5.55322790145874,
      "epoch": 1.361174868702587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048167839285802116,
      "loss": 4.9919,
      "mean_token_accuracy": 0.196986024081707,
      "num_tokens": 30330252.0,
      "step": 17495
    },
    {
      "entropy": 5.539842748641968,
      "epoch": 1.3615638980743046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004816674365137843,
      "loss": 5.0809,
      "mean_token_accuracy": 0.19607275277376174,
      "num_tokens": 30340095.0,
      "step": 17500
    },
    {
      "entropy": 5.606916284561157,
      "epoch": 1.3619529274460223,
      "grad_norm": 1.09375,
      "learning_rate": 0.000481656477033698,
      "loss": 5.1293,
      "mean_token_accuracy": 0.1895449161529541,
      "num_tokens": 30348303.0,
      "step": 17505
    },
    {
      "entropy": 5.53062334060669,
      "epoch": 1.3623419568177397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004816455144179286,
      "loss": 5.0417,
      "mean_token_accuracy": 0.19066710770130157,
      "num_tokens": 30356309.0,
      "step": 17510
    },
    {
      "entropy": 5.622492408752441,
      "epoch": 1.3627309861894572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004816345486666424,
      "loss": 5.1551,
      "mean_token_accuracy": 0.18634538352489471,
      "num_tokens": 30365128.0,
      "step": 17515
    },
    {
      "entropy": 5.540809059143067,
      "epoch": 1.363120015561175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004816235797800058,
      "loss": 5.1362,
      "mean_token_accuracy": 0.19195785969495774,
      "num_tokens": 30373656.0,
      "step": 17520
    },
    {
      "entropy": 5.525598764419556,
      "epoch": 1.3635090449328924,
      "grad_norm": 1.125,
      "learning_rate": 0.0004816126077581852,
      "loss": 5.1055,
      "mean_token_accuracy": 0.19161802679300308,
      "num_tokens": 30382612.0,
      "step": 17525
    },
    {
      "entropy": 5.506517553329468,
      "epoch": 1.36389807430461,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048160163260134724,
      "loss": 5.1361,
      "mean_token_accuracy": 0.19014995843172072,
      "num_tokens": 30392060.0,
      "step": 17530
    },
    {
      "entropy": 5.554969024658203,
      "epoch": 1.3642871036763276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004815906543096583,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20168478637933732,
      "num_tokens": 30400751.0,
      "step": 17535
    },
    {
      "entropy": 5.562521743774414,
      "epoch": 1.3646761330480452,
      "grad_norm": 1.046875,
      "learning_rate": 0.000481579672883285,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19417935460805893,
      "num_tokens": 30409584.0,
      "step": 17540
    },
    {
      "entropy": 5.510841989517212,
      "epoch": 1.3650651624197627,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048156868832239396,
      "loss": 5.0793,
      "mean_token_accuracy": 0.19371015280485154,
      "num_tokens": 30418433.0,
      "step": 17545
    },
    {
      "entropy": 5.571598148345947,
      "epoch": 1.3654541917914802,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004815577006271519,
      "loss": 5.0437,
      "mean_token_accuracy": 0.19983194172382354,
      "num_tokens": 30426943.0,
      "step": 17550
    },
    {
      "entropy": 5.563754415512085,
      "epoch": 1.365843221163198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048154670979772555,
      "loss": 5.0905,
      "mean_token_accuracy": 0.18966715782880783,
      "num_tokens": 30435763.0,
      "step": 17555
    },
    {
      "entropy": 5.455461835861206,
      "epoch": 1.3662322505349154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004815357158342815,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20787291526794432,
      "num_tokens": 30444009.0,
      "step": 17560
    },
    {
      "entropy": 5.547334003448486,
      "epoch": 1.3666212799066328,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004815247187369868,
      "loss": 5.0176,
      "mean_token_accuracy": 0.19565975666046143,
      "num_tokens": 30452652.0,
      "step": 17565
    },
    {
      "entropy": 5.539991283416748,
      "epoch": 1.3670103092783505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048151371850600814,
      "loss": 5.0452,
      "mean_token_accuracy": 0.18872333765029908,
      "num_tokens": 30462129.0,
      "step": 17570
    },
    {
      "entropy": 5.515231370925903,
      "epoch": 1.367399338650068,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048150271514151255,
      "loss": 5.0848,
      "mean_token_accuracy": 0.19493391811847688,
      "num_tokens": 30471264.0,
      "step": 17575
    },
    {
      "entropy": 5.461421155929566,
      "epoch": 1.3677883680217857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048149170864366693,
      "loss": 5.0022,
      "mean_token_accuracy": 0.19715942144393922,
      "num_tokens": 30479524.0,
      "step": 17580
    },
    {
      "entropy": 5.6382089138031,
      "epoch": 1.3681773973935032,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048148069901263836,
      "loss": 5.0658,
      "mean_token_accuracy": 0.1952446863055229,
      "num_tokens": 30487925.0,
      "step": 17585
    },
    {
      "entropy": 5.558881568908691,
      "epoch": 1.368566426765221,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048146968624859375,
      "loss": 5.0844,
      "mean_token_accuracy": 0.19245128631591796,
      "num_tokens": 30496543.0,
      "step": 17590
    },
    {
      "entropy": 5.481025171279907,
      "epoch": 1.3689554561369384,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004814586703517003,
      "loss": 5.0107,
      "mean_token_accuracy": 0.1927877724170685,
      "num_tokens": 30504780.0,
      "step": 17595
    },
    {
      "entropy": 5.50847806930542,
      "epoch": 1.3693444855086558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814476513221251,
      "loss": 5.0751,
      "mean_token_accuracy": 0.18846897631883622,
      "num_tokens": 30513355.0,
      "step": 17600
    },
    {
      "entropy": 5.4651679515838625,
      "epoch": 1.3697335148803735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048143662916003545,
      "loss": 4.984,
      "mean_token_accuracy": 0.2010539010167122,
      "num_tokens": 30522445.0,
      "step": 17605
    },
    {
      "entropy": 5.514300918579101,
      "epoch": 1.370122544252091,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814256038655985,
      "loss": 5.0273,
      "mean_token_accuracy": 0.19752272665500642,
      "num_tokens": 30530506.0,
      "step": 17610
    },
    {
      "entropy": 5.613162565231323,
      "epoch": 1.3705115736238085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004814145754389816,
      "loss": 5.134,
      "mean_token_accuracy": 0.18942491710186005,
      "num_tokens": 30539032.0,
      "step": 17615
    },
    {
      "entropy": 5.5770402431488035,
      "epoch": 1.3709006029955262,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048140354388035204,
      "loss": 5.0146,
      "mean_token_accuracy": 0.19519609957933426,
      "num_tokens": 30547052.0,
      "step": 17620
    },
    {
      "entropy": 5.516412734985352,
      "epoch": 1.3712896323672437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004813925091898772,
      "loss": 5.1268,
      "mean_token_accuracy": 0.19294073879718782,
      "num_tokens": 30556111.0,
      "step": 17625
    },
    {
      "entropy": 5.5713520526885985,
      "epoch": 1.3716786617389614,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004813814713677246,
      "loss": 5.1106,
      "mean_token_accuracy": 0.190730719268322,
      "num_tokens": 30565021.0,
      "step": 17630
    },
    {
      "entropy": 5.608752107620239,
      "epoch": 1.3720676911106788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004813704304140616,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19582546651363372,
      "num_tokens": 30573411.0,
      "step": 17635
    },
    {
      "entropy": 5.617388725280762,
      "epoch": 1.3724567204823965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004813593863290559,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19124250262975692,
      "num_tokens": 30582929.0,
      "step": 17640
    },
    {
      "entropy": 5.6182581901550295,
      "epoch": 1.372845749854114,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004813483391128748,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19197334051132203,
      "num_tokens": 30592087.0,
      "step": 17645
    },
    {
      "entropy": 5.612645769119263,
      "epoch": 1.3732347792258315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004813372887656862,
      "loss": 5.0965,
      "mean_token_accuracy": 0.1877231165766716,
      "num_tokens": 30600725.0,
      "step": 17650
    },
    {
      "entropy": 5.573830318450928,
      "epoch": 1.3736238085975492,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004813262352876576,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19162333011627197,
      "num_tokens": 30608345.0,
      "step": 17655
    },
    {
      "entropy": 5.530574798583984,
      "epoch": 1.3740128379692667,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048131517867895684,
      "loss": 5.0583,
      "mean_token_accuracy": 0.19639256298542024,
      "num_tokens": 30616951.0,
      "step": 17660
    },
    {
      "entropy": 5.518606233596802,
      "epoch": 1.3744018673409841,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048130411893975155,
      "loss": 5.0423,
      "mean_token_accuracy": 0.19841305911540985,
      "num_tokens": 30625840.0,
      "step": 17665
    },
    {
      "entropy": 5.546737337112427,
      "epoch": 1.3747908967127018,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004812930560702097,
      "loss": 4.9992,
      "mean_token_accuracy": 0.19794746190309526,
      "num_tokens": 30634473.0,
      "step": 17670
    },
    {
      "entropy": 5.505931329727173,
      "epoch": 1.3751799260844193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048128199007049907,
      "loss": 5.0287,
      "mean_token_accuracy": 0.1972987785935402,
      "num_tokens": 30643324.0,
      "step": 17675
    },
    {
      "entropy": 5.539654970169067,
      "epoch": 1.375568955456137,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048127092094078756,
      "loss": 5.0991,
      "mean_token_accuracy": 0.19085582494735717,
      "num_tokens": 30651681.0,
      "step": 17680
    },
    {
      "entropy": 5.5691431045532225,
      "epoch": 1.3759579848278545,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048125984868124317,
      "loss": 4.9936,
      "mean_token_accuracy": 0.20606574416160583,
      "num_tokens": 30660295.0,
      "step": 17685
    },
    {
      "entropy": 5.5411464214324955,
      "epoch": 1.3763470141995722,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004812487732920338,
      "loss": 4.9964,
      "mean_token_accuracy": 0.209442900121212,
      "num_tokens": 30668553.0,
      "step": 17690
    },
    {
      "entropy": 5.567028331756592,
      "epoch": 1.3767360435712896,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004812376947733277,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19114660173654557,
      "num_tokens": 30677378.0,
      "step": 17695
    },
    {
      "entropy": 5.546908235549926,
      "epoch": 1.3771250729430071,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048122661312529263,
      "loss": 5.0894,
      "mean_token_accuracy": 0.19561887830495833,
      "num_tokens": 30686662.0,
      "step": 17700
    },
    {
      "entropy": 5.512117528915406,
      "epoch": 1.3775141023147248,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004812155283480971,
      "loss": 4.989,
      "mean_token_accuracy": 0.2007271483540535,
      "num_tokens": 30695246.0,
      "step": 17705
    },
    {
      "entropy": 5.493115186691284,
      "epoch": 1.3779031316864423,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004812044404419091,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20013216882944107,
      "num_tokens": 30705275.0,
      "step": 17710
    },
    {
      "entropy": 5.582388925552368,
      "epoch": 1.3782921610581598,
      "grad_norm": 1.109375,
      "learning_rate": 0.000481193349406897,
      "loss": 5.113,
      "mean_token_accuracy": 0.18434123396873475,
      "num_tokens": 30713955.0,
      "step": 17715
    },
    {
      "entropy": 5.559508752822876,
      "epoch": 1.3786811904298775,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000481182255243229,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1919082671403885,
      "num_tokens": 30723213.0,
      "step": 17720
    },
    {
      "entropy": 5.534653377532959,
      "epoch": 1.379070219801595,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048117115795107335,
      "loss": 5.0072,
      "mean_token_accuracy": 0.19736116677522658,
      "num_tokens": 30731643.0,
      "step": 17725
    },
    {
      "entropy": 5.5371589183807375,
      "epoch": 1.3794592491733126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004811600575305987,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19474658817052842,
      "num_tokens": 30740779.0,
      "step": 17730
    },
    {
      "entropy": 5.594646978378296,
      "epoch": 1.3798482785450301,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004811489539819731,
      "loss": 5.1253,
      "mean_token_accuracy": 0.18977709710597992,
      "num_tokens": 30749685.0,
      "step": 17735
    },
    {
      "entropy": 5.532616472244262,
      "epoch": 1.3802373079167478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048113784730536544,
      "loss": 5.0166,
      "mean_token_accuracy": 0.19858538508415222,
      "num_tokens": 30758193.0,
      "step": 17740
    },
    {
      "entropy": 5.583976173400879,
      "epoch": 1.3806263372884653,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048112673750094396,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19237576574087142,
      "num_tokens": 30766870.0,
      "step": 17745
    },
    {
      "entropy": 5.543646621704101,
      "epoch": 1.3810153666601828,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004811156245688773,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19869203567504884,
      "num_tokens": 30776547.0,
      "step": 17750
    },
    {
      "entropy": 5.516203737258911,
      "epoch": 1.3814043960319005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004811045085093342,
      "loss": 4.9896,
      "mean_token_accuracy": 0.19963317960500718,
      "num_tokens": 30785579.0,
      "step": 17755
    },
    {
      "entropy": 5.5404270648956295,
      "epoch": 1.381793425403618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004810933893224831,
      "loss": 5.1591,
      "mean_token_accuracy": 0.18658055663108825,
      "num_tokens": 30794075.0,
      "step": 17760
    },
    {
      "entropy": 5.591392564773559,
      "epoch": 1.3821824547753354,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048108226700849287,
      "loss": 5.0522,
      "mean_token_accuracy": 0.19804604351520538,
      "num_tokens": 30802773.0,
      "step": 17765
    },
    {
      "entropy": 5.544975757598877,
      "epoch": 1.382571484147053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810711415675323,
      "loss": 5.0532,
      "mean_token_accuracy": 0.19035611748695375,
      "num_tokens": 30811588.0,
      "step": 17770
    },
    {
      "entropy": 5.554931879043579,
      "epoch": 1.3829605135187706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004810600129997702,
      "loss": 5.0553,
      "mean_token_accuracy": 0.19628656804561614,
      "num_tokens": 30820194.0,
      "step": 17775
    },
    {
      "entropy": 5.605623197555542,
      "epoch": 1.3833495428904883,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048104888130537534,
      "loss": 5.1394,
      "mean_token_accuracy": 0.18814265131950378,
      "num_tokens": 30829145.0,
      "step": 17780
    },
    {
      "entropy": 5.555330991744995,
      "epoch": 1.3837385722622058,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048103774648451664,
      "loss": 5.099,
      "mean_token_accuracy": 0.19463342875242234,
      "num_tokens": 30838158.0,
      "step": 17785
    },
    {
      "entropy": 5.524973964691162,
      "epoch": 1.3841276016339235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048102660853736314,
      "loss": 5.0469,
      "mean_token_accuracy": 0.19589153081178665,
      "num_tokens": 30847134.0,
      "step": 17790
    },
    {
      "entropy": 5.5677228450775145,
      "epoch": 1.384516631005641,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004810154674640837,
      "loss": 5.1235,
      "mean_token_accuracy": 0.18678848445415497,
      "num_tokens": 30856815.0,
      "step": 17795
    },
    {
      "entropy": 5.590299987792969,
      "epoch": 1.3849056603773584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048100432326484744,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19653042256832123,
      "num_tokens": 30865485.0,
      "step": 17800
    },
    {
      "entropy": 5.545382833480835,
      "epoch": 1.385294689749076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004809931759398235,
      "loss": 5.0518,
      "mean_token_accuracy": 0.19777002334594726,
      "num_tokens": 30873624.0,
      "step": 17805
    },
    {
      "entropy": 5.488943862915039,
      "epoch": 1.3856837191207936,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048098202548918094,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2004162162542343,
      "num_tokens": 30882059.0,
      "step": 17810
    },
    {
      "entropy": 5.566474342346192,
      "epoch": 1.386072748492511,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048097087191308914,
      "loss": 5.1408,
      "mean_token_accuracy": 0.18473455756902696,
      "num_tokens": 30891408.0,
      "step": 17815
    },
    {
      "entropy": 5.645927000045776,
      "epoch": 1.3864617778642288,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004809597152117171,
      "loss": 5.1008,
      "mean_token_accuracy": 0.192196324467659,
      "num_tokens": 30899832.0,
      "step": 17820
    },
    {
      "entropy": 5.553818655014038,
      "epoch": 1.3868508072359464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809485553852342,
      "loss": 5.0047,
      "mean_token_accuracy": 0.1986765444278717,
      "num_tokens": 30908393.0,
      "step": 17825
    },
    {
      "entropy": 5.532617235183716,
      "epoch": 1.387239836607664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004809373924338098,
      "loss": 5.047,
      "mean_token_accuracy": 0.19803311824798583,
      "num_tokens": 30917347.0,
      "step": 17830
    },
    {
      "entropy": 5.489804983139038,
      "epoch": 1.3876288659793814,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048092622635761316,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2038021832704544,
      "num_tokens": 30926386.0,
      "step": 17835
    },
    {
      "entropy": 5.5997583866119385,
      "epoch": 1.388017895351099,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048091505715681395,
      "loss": 5.2278,
      "mean_token_accuracy": 0.18122268915176393,
      "num_tokens": 30935047.0,
      "step": 17840
    },
    {
      "entropy": 5.6150205612182615,
      "epoch": 1.3884069247228166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809038848315814,
      "loss": 5.0968,
      "mean_token_accuracy": 0.1925975725054741,
      "num_tokens": 30943551.0,
      "step": 17845
    },
    {
      "entropy": 5.5871387958526615,
      "epoch": 1.388795954094534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004808927093820851,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20019707530736924,
      "num_tokens": 30950929.0,
      "step": 17850
    },
    {
      "entropy": 5.59451994895935,
      "epoch": 1.3891849834662517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004808815308084947,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1828531175851822,
      "num_tokens": 30959678.0,
      "step": 17855
    },
    {
      "entropy": 5.534245538711548,
      "epoch": 1.3895740128379692,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004808703491109798,
      "loss": 5.0392,
      "mean_token_accuracy": 0.19509243220090866,
      "num_tokens": 30968499.0,
      "step": 17860
    },
    {
      "entropy": 5.576900959014893,
      "epoch": 1.3899630422096867,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808591642897099,
      "loss": 5.1028,
      "mean_token_accuracy": 0.18686798214912415,
      "num_tokens": 30977146.0,
      "step": 17865
    },
    {
      "entropy": 5.48150897026062,
      "epoch": 1.3903520715814044,
      "grad_norm": 1.09375,
      "learning_rate": 0.000480847976344855,
      "loss": 4.8963,
      "mean_token_accuracy": 0.20184454321861267,
      "num_tokens": 30985285.0,
      "step": 17870
    },
    {
      "entropy": 5.502690601348877,
      "epoch": 1.390741100953122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048083678527658465,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20485545098781585,
      "num_tokens": 30993648.0,
      "step": 17875
    },
    {
      "entropy": 5.532558250427246,
      "epoch": 1.3911301303248396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004808255910850687,
      "loss": 4.9632,
      "mean_token_accuracy": 0.20521898567676544,
      "num_tokens": 31002007.0,
      "step": 17880
    },
    {
      "entropy": 5.535276746749878,
      "epoch": 1.391519159696557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004808143937704769,
      "loss": 5.0483,
      "mean_token_accuracy": 0.19593747556209565,
      "num_tokens": 31010107.0,
      "step": 17885
    },
    {
      "entropy": 5.536030960083008,
      "epoch": 1.3919081890682747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048080319333297937,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20344551503658295,
      "num_tokens": 31018754.0,
      "step": 17890
    },
    {
      "entropy": 5.574505996704102,
      "epoch": 1.3922972184399922,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048079198977274597,
      "loss": 5.1426,
      "mean_token_accuracy": 0.18835900574922562,
      "num_tokens": 31027929.0,
      "step": 17895
    },
    {
      "entropy": 5.535457563400269,
      "epoch": 1.3926862478117097,
      "grad_norm": 1.125,
      "learning_rate": 0.00048078078308994666,
      "loss": 5.0969,
      "mean_token_accuracy": 0.18722502738237382,
      "num_tokens": 31037023.0,
      "step": 17900
    },
    {
      "entropy": 5.5400307178497314,
      "epoch": 1.3930752771834274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004807695732847515,
      "loss": 4.986,
      "mean_token_accuracy": 0.20118181705474852,
      "num_tokens": 31045483.0,
      "step": 17905
    },
    {
      "entropy": 5.520857477188111,
      "epoch": 1.3934643065551449,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048075836035733053,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19901844263076782,
      "num_tokens": 31053905.0,
      "step": 17910
    },
    {
      "entropy": 5.514943170547485,
      "epoch": 1.3938533359268626,
      "grad_norm": 1.09375,
      "learning_rate": 0.000480747144307854,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1974051758646965,
      "num_tokens": 31062288.0,
      "step": 17915
    },
    {
      "entropy": 5.5517021179199215,
      "epoch": 1.39424236529858,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048073592513649203,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19578115791082382,
      "num_tokens": 31070165.0,
      "step": 17920
    },
    {
      "entropy": 5.599383115768433,
      "epoch": 1.3946313946702977,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004807247028434148,
      "loss": 5.0557,
      "mean_token_accuracy": 0.19163752049207688,
      "num_tokens": 31078448.0,
      "step": 17925
    },
    {
      "entropy": 5.5500242710113525,
      "epoch": 1.3950204240420152,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004807134774287927,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19304633140563965,
      "num_tokens": 31087511.0,
      "step": 17930
    },
    {
      "entropy": 5.4819310188293455,
      "epoch": 1.3954094534137327,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048070224889279603,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1946722760796547,
      "num_tokens": 31096893.0,
      "step": 17935
    },
    {
      "entropy": 5.555552053451538,
      "epoch": 1.3957984827854504,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048069101723559505,
      "loss": 5.062,
      "mean_token_accuracy": 0.19403376579284667,
      "num_tokens": 31106111.0,
      "step": 17940
    },
    {
      "entropy": 5.494966077804565,
      "epoch": 1.3961875121571679,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004806797824573603,
      "loss": 4.917,
      "mean_token_accuracy": 0.2073436811566353,
      "num_tokens": 31114780.0,
      "step": 17945
    },
    {
      "entropy": 5.575967264175415,
      "epoch": 1.3965765415288853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004806685445582624,
      "loss": 5.1495,
      "mean_token_accuracy": 0.18139443695545196,
      "num_tokens": 31123644.0,
      "step": 17950
    },
    {
      "entropy": 5.491543436050415,
      "epoch": 1.396965570900603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048065730353847143,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1960713803768158,
      "num_tokens": 31133133.0,
      "step": 17955
    },
    {
      "entropy": 5.528134346008301,
      "epoch": 1.3973546002723205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048064605939815837,
      "loss": 5.0546,
      "mean_token_accuracy": 0.19731059819459915,
      "num_tokens": 31141545.0,
      "step": 17960
    },
    {
      "entropy": 5.553800916671753,
      "epoch": 1.3977436296440382,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004806348121374936,
      "loss": 5.0803,
      "mean_token_accuracy": 0.1949324280023575,
      "num_tokens": 31150689.0,
      "step": 17965
    },
    {
      "entropy": 5.62173752784729,
      "epoch": 1.3981326590157557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004806235617566479,
      "loss": 5.1603,
      "mean_token_accuracy": 0.18845051676034927,
      "num_tokens": 31159114.0,
      "step": 17970
    },
    {
      "entropy": 5.564534091949463,
      "epoch": 1.3985216883874734,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000480612308255792,
      "loss": 5.1594,
      "mean_token_accuracy": 0.19304109066724778,
      "num_tokens": 31167938.0,
      "step": 17975
    },
    {
      "entropy": 5.628028774261475,
      "epoch": 1.3989107177591908,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048060105163509647,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1874155655503273,
      "num_tokens": 31176653.0,
      "step": 17980
    },
    {
      "entropy": 5.574227619171142,
      "epoch": 1.3992997471309083,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004805897918947323,
      "loss": 5.0521,
      "mean_token_accuracy": 0.19768701642751693,
      "num_tokens": 31185247.0,
      "step": 17985
    },
    {
      "entropy": 5.566842842102051,
      "epoch": 1.399688776502626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004805785290348702,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19304127395153045,
      "num_tokens": 31194504.0,
      "step": 17990
    },
    {
      "entropy": 5.5727410316467285,
      "epoch": 1.4000778058743435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048056726305568124,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19094185531139374,
      "num_tokens": 31202668.0,
      "step": 17995
    },
    {
      "entropy": 5.58998384475708,
      "epoch": 1.400466835246061,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048055599395733617,
      "loss": 5.054,
      "mean_token_accuracy": 0.1926966652274132,
      "num_tokens": 31211165.0,
      "step": 18000
    },
    {
      "epoch": 1.400466835246061,
      "eval_entropy": 5.31877617115465,
      "eval_loss": 5.138509750366211,
      "eval_mean_token_accuracy": 0.20082474479718893,
      "eval_num_tokens": 31211165.0,
      "eval_runtime": 21.6844,
      "eval_samples_per_second": 1916.492,
      "eval_steps_per_second": 239.573,
      "step": 18000
    },
    {
      "entropy": 5.517045497894287,
      "epoch": 1.4008558646177787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048054472174000615,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19554646015167237,
      "num_tokens": 31220502.0,
      "step": 18005
    },
    {
      "entropy": 5.597953271865845,
      "epoch": 1.4012448939894961,
      "grad_norm": 1.125,
      "learning_rate": 0.00048053344640386207,
      "loss": 5.1631,
      "mean_token_accuracy": 0.18142991960048677,
      "num_tokens": 31228771.0,
      "step": 18010
    },
    {
      "entropy": 5.53455662727356,
      "epoch": 1.4016339233612138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048052216794907505,
      "loss": 5.0517,
      "mean_token_accuracy": 0.19655922502279283,
      "num_tokens": 31237098.0,
      "step": 18015
    },
    {
      "entropy": 5.555712747573852,
      "epoch": 1.4020229527329313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048051088637581624,
      "loss": 5.0849,
      "mean_token_accuracy": 0.18785118460655212,
      "num_tokens": 31245879.0,
      "step": 18020
    },
    {
      "entropy": 5.538217544555664,
      "epoch": 1.402411982104649,
      "grad_norm": 1.15625,
      "learning_rate": 0.000480499601684257,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2021091639995575,
      "num_tokens": 31254041.0,
      "step": 18025
    },
    {
      "entropy": 5.507911491394043,
      "epoch": 1.4028010114763665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004804883138745682,
      "loss": 5.0151,
      "mean_token_accuracy": 0.19280405789613725,
      "num_tokens": 31262600.0,
      "step": 18030
    },
    {
      "entropy": 5.507891845703125,
      "epoch": 1.403190040848084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004804770229469214,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20524874180555344,
      "num_tokens": 31270844.0,
      "step": 18035
    },
    {
      "entropy": 5.521011781692505,
      "epoch": 1.4035790702198017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004804657289014878,
      "loss": 5.0961,
      "mean_token_accuracy": 0.1872323974967003,
      "num_tokens": 31280104.0,
      "step": 18040
    },
    {
      "entropy": 5.5206633567810055,
      "epoch": 1.4039680995915191,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048045443173843884,
      "loss": 4.9476,
      "mean_token_accuracy": 0.19926963299512862,
      "num_tokens": 31288280.0,
      "step": 18045
    },
    {
      "entropy": 5.456657648086548,
      "epoch": 1.4043571289632366,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048044313145794587,
      "loss": 4.8983,
      "mean_token_accuracy": 0.20874076336622238,
      "num_tokens": 31296806.0,
      "step": 18050
    },
    {
      "entropy": 5.6134788513183596,
      "epoch": 1.4047461583349543,
      "grad_norm": 1.125,
      "learning_rate": 0.00048043182806018034,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1895008161664009,
      "num_tokens": 31305687.0,
      "step": 18055
    },
    {
      "entropy": 5.588626956939697,
      "epoch": 1.4051351877066718,
      "grad_norm": 1.125,
      "learning_rate": 0.00048042052154531384,
      "loss": 5.1634,
      "mean_token_accuracy": 0.19666950702667235,
      "num_tokens": 31314332.0,
      "step": 18060
    },
    {
      "entropy": 5.530027246475219,
      "epoch": 1.4055242170783895,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004804092119135179,
      "loss": 5.0739,
      "mean_token_accuracy": 0.2026395931839943,
      "num_tokens": 31322630.0,
      "step": 18065
    },
    {
      "entropy": 5.532639360427856,
      "epoch": 1.405913246450107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004803978991649641,
      "loss": 5.0197,
      "mean_token_accuracy": 0.19600387960672377,
      "num_tokens": 31330778.0,
      "step": 18070
    },
    {
      "entropy": 5.505466270446777,
      "epoch": 1.4063022758218247,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048038658329982404,
      "loss": 4.9998,
      "mean_token_accuracy": 0.19533978551626205,
      "num_tokens": 31339637.0,
      "step": 18075
    },
    {
      "entropy": 5.589492273330689,
      "epoch": 1.4066913051935421,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004803752643182695,
      "loss": 5.1331,
      "mean_token_accuracy": 0.19073974341154099,
      "num_tokens": 31348255.0,
      "step": 18080
    },
    {
      "entropy": 5.494571590423584,
      "epoch": 1.4070803345652596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004803639422204723,
      "loss": 5.0105,
      "mean_token_accuracy": 0.19242308139801026,
      "num_tokens": 31357027.0,
      "step": 18085
    },
    {
      "entropy": 5.502149343490601,
      "epoch": 1.4074693639369773,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004803526170066041,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20116140991449355,
      "num_tokens": 31365475.0,
      "step": 18090
    },
    {
      "entropy": 5.561756181716919,
      "epoch": 1.4078583933086948,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048034128867683674,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18987676799297332,
      "num_tokens": 31374459.0,
      "step": 18095
    },
    {
      "entropy": 5.613378047943115,
      "epoch": 1.4082474226804123,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803299572313422,
      "loss": 5.0308,
      "mean_token_accuracy": 0.19178077727556228,
      "num_tokens": 31382720.0,
      "step": 18100
    },
    {
      "entropy": 5.4810240268707275,
      "epoch": 1.40863645205213,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004803186226702924,
      "loss": 5.0129,
      "mean_token_accuracy": 0.19117873758077622,
      "num_tokens": 31390543.0,
      "step": 18105
    },
    {
      "entropy": 5.560120725631714,
      "epoch": 1.4090254814238474,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004803072849938592,
      "loss": 5.0776,
      "mean_token_accuracy": 0.19085932970046998,
      "num_tokens": 31399428.0,
      "step": 18110
    },
    {
      "entropy": 5.5510985374450685,
      "epoch": 1.4094145107955651,
      "grad_norm": 1.125,
      "learning_rate": 0.0004802959442022149,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20323875695466995,
      "num_tokens": 31408287.0,
      "step": 18115
    },
    {
      "entropy": 5.580893611907959,
      "epoch": 1.4098035401672826,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048028460029553126,
      "loss": 5.1082,
      "mean_token_accuracy": 0.19309157580137254,
      "num_tokens": 31417340.0,
      "step": 18120
    },
    {
      "entropy": 5.556183671951294,
      "epoch": 1.4101925695390003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048027325327398065,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19392092972993852,
      "num_tokens": 31426363.0,
      "step": 18125
    },
    {
      "entropy": 5.627390718460083,
      "epoch": 1.4105815989107178,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004802619031377351,
      "loss": 5.1541,
      "mean_token_accuracy": 0.19699808657169343,
      "num_tokens": 31436107.0,
      "step": 18130
    },
    {
      "entropy": 5.562397909164429,
      "epoch": 1.4109706282824352,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048025054988696684,
      "loss": 5.0525,
      "mean_token_accuracy": 0.18827087581157684,
      "num_tokens": 31444899.0,
      "step": 18135
    },
    {
      "entropy": 5.576064205169677,
      "epoch": 1.411359657654153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048023919352184827,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20217641592025756,
      "num_tokens": 31453130.0,
      "step": 18140
    },
    {
      "entropy": 5.52488489151001,
      "epoch": 1.4117486870258704,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048022783404255156,
      "loss": 5.0272,
      "mean_token_accuracy": 0.19490601420402526,
      "num_tokens": 31461951.0,
      "step": 18145
    },
    {
      "entropy": 5.578972625732422,
      "epoch": 1.412137716397588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004802164714492491,
      "loss": 5.0978,
      "mean_token_accuracy": 0.18536048084497453,
      "num_tokens": 31470487.0,
      "step": 18150
    },
    {
      "entropy": 5.574371242523194,
      "epoch": 1.4125267457693056,
      "grad_norm": 1.125,
      "learning_rate": 0.00048020510574211335,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19663629531860352,
      "num_tokens": 31479067.0,
      "step": 18155
    },
    {
      "entropy": 5.545616245269775,
      "epoch": 1.412915775141023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048019373692131674,
      "loss": 5.0231,
      "mean_token_accuracy": 0.19367250949144363,
      "num_tokens": 31487993.0,
      "step": 18160
    },
    {
      "entropy": 5.531673336029053,
      "epoch": 1.4133048045127408,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048018236498703176,
      "loss": 5.0319,
      "mean_token_accuracy": 0.19408976882696152,
      "num_tokens": 31496436.0,
      "step": 18165
    },
    {
      "entropy": 5.559005546569824,
      "epoch": 1.4136938338844582,
      "grad_norm": 1.125,
      "learning_rate": 0.00048017098993943097,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2013227418065071,
      "num_tokens": 31504834.0,
      "step": 18170
    },
    {
      "entropy": 5.6224287986755375,
      "epoch": 1.414082863256176,
      "grad_norm": 1.109375,
      "learning_rate": 0.000480159611778687,
      "loss": 5.1492,
      "mean_token_accuracy": 0.18386515825986863,
      "num_tokens": 31513532.0,
      "step": 18175
    },
    {
      "entropy": 5.51333646774292,
      "epoch": 1.4144718926278934,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048014823050497243,
      "loss": 5.0225,
      "mean_token_accuracy": 0.19553097188472748,
      "num_tokens": 31522291.0,
      "step": 18180
    },
    {
      "entropy": 5.5416069507598875,
      "epoch": 1.4148609219996109,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048013684611846,
      "loss": 5.0431,
      "mean_token_accuracy": 0.1976754680275917,
      "num_tokens": 31530164.0,
      "step": 18185
    },
    {
      "entropy": 5.58240327835083,
      "epoch": 1.4152499513713286,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048012545861932245,
      "loss": 5.0756,
      "mean_token_accuracy": 0.19078683108091354,
      "num_tokens": 31538690.0,
      "step": 18190
    },
    {
      "entropy": 5.478669261932373,
      "epoch": 1.415638980743046,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004801140680077325,
      "loss": 5.0051,
      "mean_token_accuracy": 0.20108330845832825,
      "num_tokens": 31547109.0,
      "step": 18195
    },
    {
      "entropy": 5.574164867401123,
      "epoch": 1.4160280101147635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004801026742838631,
      "loss": 5.1092,
      "mean_token_accuracy": 0.1911427214741707,
      "num_tokens": 31555524.0,
      "step": 18200
    },
    {
      "entropy": 5.543385171890259,
      "epoch": 1.4164170394864812,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800912774478871,
      "loss": 5.0942,
      "mean_token_accuracy": 0.1923657551407814,
      "num_tokens": 31564093.0,
      "step": 18205
    },
    {
      "entropy": 5.51996078491211,
      "epoch": 1.416806068858199,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004800798774999773,
      "loss": 4.9959,
      "mean_token_accuracy": 0.19956326633691787,
      "num_tokens": 31572680.0,
      "step": 18210
    },
    {
      "entropy": 5.5452179431915285,
      "epoch": 1.4171950982299164,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048006847444030686,
      "loss": 5.055,
      "mean_token_accuracy": 0.19761183261871337,
      "num_tokens": 31581537.0,
      "step": 18215
    },
    {
      "entropy": 5.600072002410888,
      "epoch": 1.4175841276016339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004800570682690487,
      "loss": 5.0139,
      "mean_token_accuracy": 0.19170113950967788,
      "num_tokens": 31590151.0,
      "step": 18220
    },
    {
      "entropy": 5.548355340957642,
      "epoch": 1.4179731569733516,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004800456589863759,
      "loss": 4.997,
      "mean_token_accuracy": 0.20166392922401427,
      "num_tokens": 31598600.0,
      "step": 18225
    },
    {
      "entropy": 5.650712776184082,
      "epoch": 1.418362186345069,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004800342465924616,
      "loss": 5.1174,
      "mean_token_accuracy": 0.1892600730061531,
      "num_tokens": 31606357.0,
      "step": 18230
    },
    {
      "entropy": 5.593307065963745,
      "epoch": 1.4187512157167865,
      "grad_norm": 1.046875,
      "learning_rate": 0.000480022831087479,
      "loss": 5.1105,
      "mean_token_accuracy": 0.18833684623241426,
      "num_tokens": 31615794.0,
      "step": 18235
    },
    {
      "entropy": 5.583683109283447,
      "epoch": 1.4191402450885042,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004800114124716012,
      "loss": 5.0555,
      "mean_token_accuracy": 0.19954750090837478,
      "num_tokens": 31624000.0,
      "step": 18240
    },
    {
      "entropy": 5.565824222564697,
      "epoch": 1.4195292744602217,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004799999907450015,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19955568462610246,
      "num_tokens": 31632762.0,
      "step": 18245
    },
    {
      "entropy": 5.553789377212524,
      "epoch": 1.4199183038319392,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004799885659078533,
      "loss": 5.0851,
      "mean_token_accuracy": 0.18759685754776,
      "num_tokens": 31641218.0,
      "step": 18250
    },
    {
      "entropy": 5.565247344970703,
      "epoch": 1.4203073332036569,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004799771379603299,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19457255601882933,
      "num_tokens": 31650765.0,
      "step": 18255
    },
    {
      "entropy": 5.540380191802979,
      "epoch": 1.4206963625753746,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004799657069026046,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1982538804411888,
      "num_tokens": 31659139.0,
      "step": 18260
    },
    {
      "entropy": 5.564652585983277,
      "epoch": 1.421085391947092,
      "grad_norm": 1.125,
      "learning_rate": 0.00047995427273485097,
      "loss": 5.0906,
      "mean_token_accuracy": 0.19555256962776185,
      "num_tokens": 31667083.0,
      "step": 18265
    },
    {
      "entropy": 5.54394907951355,
      "epoch": 1.4214744213188095,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047994283545724247,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20335510075092317,
      "num_tokens": 31676409.0,
      "step": 18270
    },
    {
      "entropy": 5.5554248809814455,
      "epoch": 1.4218634506905272,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047993139506995273,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19831628799438478,
      "num_tokens": 31684507.0,
      "step": 18275
    },
    {
      "entropy": 5.544463729858398,
      "epoch": 1.4222524800622447,
      "grad_norm": 1.125,
      "learning_rate": 0.0004799199515731551,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20106517970561982,
      "num_tokens": 31693091.0,
      "step": 18280
    },
    {
      "entropy": 5.487980556488037,
      "epoch": 1.4226415094339622,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047990850496702346,
      "loss": 5.0227,
      "mean_token_accuracy": 0.2041850432753563,
      "num_tokens": 31701486.0,
      "step": 18285
    },
    {
      "entropy": 5.583898591995239,
      "epoch": 1.4230305388056799,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004798970552517314,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1941058397293091,
      "num_tokens": 31710107.0,
      "step": 18290
    },
    {
      "entropy": 5.650792407989502,
      "epoch": 1.4234195681773973,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047988560242745264,
      "loss": 5.1157,
      "mean_token_accuracy": 0.19557671546936034,
      "num_tokens": 31718378.0,
      "step": 18295
    },
    {
      "entropy": 5.580296659469605,
      "epoch": 1.423808597549115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000479874146494361,
      "loss": 5.0715,
      "mean_token_accuracy": 0.19503856003284453,
      "num_tokens": 31726956.0,
      "step": 18300
    },
    {
      "entropy": 5.625977849960327,
      "epoch": 1.4241976269208325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004798626874526302,
      "loss": 5.1388,
      "mean_token_accuracy": 0.19569528847932816,
      "num_tokens": 31735551.0,
      "step": 18305
    },
    {
      "entropy": 5.652529239654541,
      "epoch": 1.4245866562925502,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047985122530243426,
      "loss": 5.1974,
      "mean_token_accuracy": 0.17976532131433487,
      "num_tokens": 31744852.0,
      "step": 18310
    },
    {
      "entropy": 5.617678308486939,
      "epoch": 1.4249756856642677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000479839760043947,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19166087210178376,
      "num_tokens": 31754204.0,
      "step": 18315
    },
    {
      "entropy": 5.610816764831543,
      "epoch": 1.4253647150359852,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047982829167734245,
      "loss": 5.0713,
      "mean_token_accuracy": 0.1953285202383995,
      "num_tokens": 31762478.0,
      "step": 18320
    },
    {
      "entropy": 5.559537601470947,
      "epoch": 1.4257537444077029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047981682020279456,
      "loss": 5.055,
      "mean_token_accuracy": 0.1979062408208847,
      "num_tokens": 31770607.0,
      "step": 18325
    },
    {
      "entropy": 5.512546968460083,
      "epoch": 1.4261427737794203,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004798053456204775,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20497969835996627,
      "num_tokens": 31779441.0,
      "step": 18330
    },
    {
      "entropy": 5.553103923797607,
      "epoch": 1.4265318031511378,
      "grad_norm": 1.125,
      "learning_rate": 0.0004797938679305651,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1929819479584694,
      "num_tokens": 31788686.0,
      "step": 18335
    },
    {
      "entropy": 5.605119752883911,
      "epoch": 1.4269208325228555,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047978238713323193,
      "loss": 5.0253,
      "mean_token_accuracy": 0.1964220330119133,
      "num_tokens": 31796433.0,
      "step": 18340
    },
    {
      "entropy": 5.548340940475464,
      "epoch": 1.427309861894573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047977090322865183,
      "loss": 5.0027,
      "mean_token_accuracy": 0.19998253732919694,
      "num_tokens": 31804700.0,
      "step": 18345
    },
    {
      "entropy": 5.479428672790528,
      "epoch": 1.4276988912662907,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004797594162169993,
      "loss": 5.0308,
      "mean_token_accuracy": 0.19349924027919768,
      "num_tokens": 31813441.0,
      "step": 18350
    },
    {
      "entropy": 5.5475023746490475,
      "epoch": 1.4280879206380082,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797479260984485,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18988244384527206,
      "num_tokens": 31823553.0,
      "step": 18355
    },
    {
      "entropy": 5.585196876525879,
      "epoch": 1.4284769500097259,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047973643287317386,
      "loss": 5.0642,
      "mean_token_accuracy": 0.19464348405599594,
      "num_tokens": 31832093.0,
      "step": 18360
    },
    {
      "entropy": 5.5267777919769285,
      "epoch": 1.4288659793814433,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047972493654134963,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20176289826631547,
      "num_tokens": 31840282.0,
      "step": 18365
    },
    {
      "entropy": 5.593541669845581,
      "epoch": 1.4292550087531608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047971343710315043,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19293663799762725,
      "num_tokens": 31849230.0,
      "step": 18370
    },
    {
      "entropy": 5.595777177810669,
      "epoch": 1.4296440381248785,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004797019345587506,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19061076194047927,
      "num_tokens": 31857965.0,
      "step": 18375
    },
    {
      "entropy": 5.595154047012329,
      "epoch": 1.430033067496596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004796904289083248,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1981726586818695,
      "num_tokens": 31866188.0,
      "step": 18380
    },
    {
      "entropy": 5.550575160980225,
      "epoch": 1.4304220968683135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004796789201520474,
      "loss": 5.1041,
      "mean_token_accuracy": 0.1944649651646614,
      "num_tokens": 31875024.0,
      "step": 18385
    },
    {
      "entropy": 5.4641077518463135,
      "epoch": 1.4308111262400312,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047966740829009333,
      "loss": 5.0328,
      "mean_token_accuracy": 0.19584709405899048,
      "num_tokens": 31883563.0,
      "step": 18390
    },
    {
      "entropy": 5.555887603759766,
      "epoch": 1.4312001556117486,
      "grad_norm": 1.171875,
      "learning_rate": 0.000479655893322637,
      "loss": 5.0894,
      "mean_token_accuracy": 0.18810806274414063,
      "num_tokens": 31891435.0,
      "step": 18395
    },
    {
      "entropy": 5.605862617492676,
      "epoch": 1.4315891849834663,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004796443752498532,
      "loss": 5.066,
      "mean_token_accuracy": 0.19410179704427719,
      "num_tokens": 31899335.0,
      "step": 18400
    },
    {
      "entropy": 5.577321243286133,
      "epoch": 1.4319782143551838,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004796328540719169,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1899257257580757,
      "num_tokens": 31908101.0,
      "step": 18405
    },
    {
      "entropy": 5.554736852645874,
      "epoch": 1.4323672437269015,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004796213297890026,
      "loss": 5.0933,
      "mean_token_accuracy": 0.196576389670372,
      "num_tokens": 31916664.0,
      "step": 18410
    },
    {
      "entropy": 5.563580131530761,
      "epoch": 1.432756273098619,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004796098024012853,
      "loss": 5.0161,
      "mean_token_accuracy": 0.2002794161438942,
      "num_tokens": 31924785.0,
      "step": 18415
    },
    {
      "entropy": 5.573451709747315,
      "epoch": 1.4331453024703364,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047959827190894,
      "loss": 5.1971,
      "mean_token_accuracy": 0.18792109340429305,
      "num_tokens": 31933117.0,
      "step": 18420
    },
    {
      "entropy": 5.66439003944397,
      "epoch": 1.4335343318420541,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047958673831214157,
      "loss": 5.1479,
      "mean_token_accuracy": 0.18793282210826873,
      "num_tokens": 31941217.0,
      "step": 18425
    },
    {
      "entropy": 5.5384317398071286,
      "epoch": 1.4339233612137716,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047957520161106496,
      "loss": 4.981,
      "mean_token_accuracy": 0.20418466329574586,
      "num_tokens": 31949749.0,
      "step": 18430
    },
    {
      "entropy": 5.553807973861694,
      "epoch": 1.434312390585489,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004795636618058853,
      "loss": 5.0618,
      "mean_token_accuracy": 0.1991420269012451,
      "num_tokens": 31957631.0,
      "step": 18435
    },
    {
      "entropy": 5.606360530853271,
      "epoch": 1.4347014199572068,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004795521188967777,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19741732478141785,
      "num_tokens": 31966453.0,
      "step": 18440
    },
    {
      "entropy": 5.547063541412354,
      "epoch": 1.4350904493289243,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004795405728839172,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19225009232759477,
      "num_tokens": 31975487.0,
      "step": 18445
    },
    {
      "entropy": 5.5155597686767575,
      "epoch": 1.435479478700642,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004795290237674792,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19762931168079376,
      "num_tokens": 31984505.0,
      "step": 18450
    },
    {
      "entropy": 5.499327850341797,
      "epoch": 1.4358685080723594,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047951747154763866,
      "loss": 4.984,
      "mean_token_accuracy": 0.19115490168333055,
      "num_tokens": 31993555.0,
      "step": 18455
    },
    {
      "entropy": 5.601089572906494,
      "epoch": 1.4362575374440771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047950591622457106,
      "loss": 5.1543,
      "mean_token_accuracy": 0.18501971513032914,
      "num_tokens": 32002950.0,
      "step": 18460
    },
    {
      "entropy": 5.5887593746185305,
      "epoch": 1.4366465668157946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047949435779845174,
      "loss": 5.0973,
      "mean_token_accuracy": 0.19135268926620483,
      "num_tokens": 32011023.0,
      "step": 18465
    },
    {
      "entropy": 5.527636098861694,
      "epoch": 1.437035596187512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047948279626945596,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20359327048063278,
      "num_tokens": 32019498.0,
      "step": 18470
    },
    {
      "entropy": 5.461010265350342,
      "epoch": 1.4374246255592298,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047947123163775924,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20667457729578018,
      "num_tokens": 32028211.0,
      "step": 18475
    },
    {
      "entropy": 5.583381652832031,
      "epoch": 1.4378136549309473,
      "grad_norm": 1.1875,
      "learning_rate": 0.000479459663903537,
      "loss": 5.1022,
      "mean_token_accuracy": 0.19352666288614273,
      "num_tokens": 32036378.0,
      "step": 18480
    },
    {
      "entropy": 5.567234659194947,
      "epoch": 1.4382026843026647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004794480930669649,
      "loss": 5.0267,
      "mean_token_accuracy": 0.19825142472982407,
      "num_tokens": 32045492.0,
      "step": 18485
    },
    {
      "entropy": 5.653302383422852,
      "epoch": 1.4385917136743824,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004794365191282183,
      "loss": 5.215,
      "mean_token_accuracy": 0.18451925367116928,
      "num_tokens": 32054297.0,
      "step": 18490
    },
    {
      "entropy": 5.574009704589844,
      "epoch": 1.4389807430461,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047942494208747297,
      "loss": 4.9892,
      "mean_token_accuracy": 0.19539348036050797,
      "num_tokens": 32062602.0,
      "step": 18495
    },
    {
      "entropy": 5.518990516662598,
      "epoch": 1.4393697724178176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004794133619449045,
      "loss": 5.0882,
      "mean_token_accuracy": 0.18784102648496628,
      "num_tokens": 32071752.0,
      "step": 18500
    },
    {
      "entropy": 5.554832220077515,
      "epoch": 1.439758801789535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794017787006886,
      "loss": 5.0078,
      "mean_token_accuracy": 0.19152483344078064,
      "num_tokens": 32080519.0,
      "step": 18505
    },
    {
      "entropy": 5.6091086864471436,
      "epoch": 1.4401478311612528,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000479390192355001,
      "loss": 5.1461,
      "mean_token_accuracy": 0.18657961040735244,
      "num_tokens": 32089489.0,
      "step": 18510
    },
    {
      "entropy": 5.6602137088775635,
      "epoch": 1.4405368605329703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004793786029080176,
      "loss": 5.1068,
      "mean_token_accuracy": 0.1916726365685463,
      "num_tokens": 32098199.0,
      "step": 18515
    },
    {
      "entropy": 5.583471298217773,
      "epoch": 1.4409258899046877,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793670103599143,
      "loss": 5.022,
      "mean_token_accuracy": 0.19564177542924882,
      "num_tokens": 32107340.0,
      "step": 18520
    },
    {
      "entropy": 5.6067112445831295,
      "epoch": 1.4413149192764054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047935541471086677,
      "loss": 5.0732,
      "mean_token_accuracy": 0.19028545618057252,
      "num_tokens": 32115697.0,
      "step": 18525
    },
    {
      "entropy": 5.545618915557862,
      "epoch": 1.441703948648123,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793438159610511,
      "loss": 5.0347,
      "mean_token_accuracy": 0.1952441543340683,
      "num_tokens": 32124801.0,
      "step": 18530
    },
    {
      "entropy": 5.583477449417114,
      "epoch": 1.4420929780198404,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047933221411064334,
      "loss": 5.0231,
      "mean_token_accuracy": 0.20084090381860734,
      "num_tokens": 32132728.0,
      "step": 18535
    },
    {
      "entropy": 5.522182941436768,
      "epoch": 1.442482007391558,
      "grad_norm": 1.125,
      "learning_rate": 0.0004793206091598194,
      "loss": 5.1676,
      "mean_token_accuracy": 0.18840880692005157,
      "num_tokens": 32141325.0,
      "step": 18540
    },
    {
      "entropy": 5.529790925979614,
      "epoch": 1.4428710367632755,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004793090011087554,
      "loss": 4.9449,
      "mean_token_accuracy": 0.20645099878311157,
      "num_tokens": 32149352.0,
      "step": 18545
    },
    {
      "entropy": 5.514100646972656,
      "epoch": 1.4432600661349932,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047929738995762755,
      "loss": 5.0425,
      "mean_token_accuracy": 0.19252483546733856,
      "num_tokens": 32158010.0,
      "step": 18550
    },
    {
      "entropy": 5.54469633102417,
      "epoch": 1.4436490955067107,
      "grad_norm": 1.125,
      "learning_rate": 0.000479285775706612,
      "loss": 5.0958,
      "mean_token_accuracy": 0.19382050782442092,
      "num_tokens": 32166204.0,
      "step": 18555
    },
    {
      "entropy": 5.65263147354126,
      "epoch": 1.4440381248784284,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047927415835588496,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1996418908238411,
      "num_tokens": 32175005.0,
      "step": 18560
    },
    {
      "entropy": 5.611942291259766,
      "epoch": 1.444427154250146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047926253790562265,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20101397186517717,
      "num_tokens": 32183529.0,
      "step": 18565
    },
    {
      "entropy": 5.585473251342774,
      "epoch": 1.4448161836218634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047925091435600147,
      "loss": 5.1397,
      "mean_token_accuracy": 0.1907166585326195,
      "num_tokens": 32192164.0,
      "step": 18570
    },
    {
      "entropy": 5.592973327636718,
      "epoch": 1.445205212993581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047923928770719776,
      "loss": 5.1469,
      "mean_token_accuracy": 0.1953514188528061,
      "num_tokens": 32201852.0,
      "step": 18575
    },
    {
      "entropy": 5.514042663574219,
      "epoch": 1.4455942423652985,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047922765795938797,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19660549312829972,
      "num_tokens": 32210501.0,
      "step": 18580
    },
    {
      "entropy": 5.47713270187378,
      "epoch": 1.445983271737016,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792160251127485,
      "loss": 4.9376,
      "mean_token_accuracy": 0.20381859242916106,
      "num_tokens": 32219365.0,
      "step": 18585
    },
    {
      "entropy": 5.547822523117065,
      "epoch": 1.4463723011087337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047920438916745586,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18996737897396088,
      "num_tokens": 32228599.0,
      "step": 18590
    },
    {
      "entropy": 5.578655481338501,
      "epoch": 1.4467613304804512,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004791927501236866,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1950696513056755,
      "num_tokens": 32237544.0,
      "step": 18595
    },
    {
      "entropy": 5.55748176574707,
      "epoch": 1.4471503598521689,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047918110798161754,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20111645758152008,
      "num_tokens": 32245953.0,
      "step": 18600
    },
    {
      "entropy": 5.562480354309082,
      "epoch": 1.4475393892238864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047916946274142503,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2085634157061577,
      "num_tokens": 32254050.0,
      "step": 18605
    },
    {
      "entropy": 5.518645620346069,
      "epoch": 1.447928418595604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047915781440328593,
      "loss": 5.0151,
      "mean_token_accuracy": 0.19770674854516984,
      "num_tokens": 32262732.0,
      "step": 18610
    },
    {
      "entropy": 5.501862096786499,
      "epoch": 1.4483174479673215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004791461629673769,
      "loss": 4.9903,
      "mean_token_accuracy": 0.1996803566813469,
      "num_tokens": 32271026.0,
      "step": 18615
    },
    {
      "entropy": 5.55328483581543,
      "epoch": 1.448706477339039,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004791345084338748,
      "loss": 5.1333,
      "mean_token_accuracy": 0.18838211745023728,
      "num_tokens": 32280500.0,
      "step": 18620
    },
    {
      "entropy": 5.597522735595703,
      "epoch": 1.4490955067107567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004791228508029565,
      "loss": 5.0555,
      "mean_token_accuracy": 0.1952091410756111,
      "num_tokens": 32289652.0,
      "step": 18625
    },
    {
      "entropy": 5.559080457687378,
      "epoch": 1.4494845360824742,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047911119007479873,
      "loss": 5.0331,
      "mean_token_accuracy": 0.1981078267097473,
      "num_tokens": 32297938.0,
      "step": 18630
    },
    {
      "entropy": 5.531543207168579,
      "epoch": 1.4498735654541917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047909952624957866,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21116172671318054,
      "num_tokens": 32306771.0,
      "step": 18635
    },
    {
      "entropy": 5.55816798210144,
      "epoch": 1.4502625948259094,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479087859327473,
      "loss": 5.0901,
      "mean_token_accuracy": 0.1919173091650009,
      "num_tokens": 32315086.0,
      "step": 18640
    },
    {
      "entropy": 5.582638883590699,
      "epoch": 1.450651624197627,
      "grad_norm": 1.078125,
      "learning_rate": 0.000479076189308659,
      "loss": 5.0951,
      "mean_token_accuracy": 0.19438012987375258,
      "num_tokens": 32323763.0,
      "step": 18645
    },
    {
      "entropy": 5.568017435073853,
      "epoch": 1.4510406535693445,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047906451619331364,
      "loss": 5.0924,
      "mean_token_accuracy": 0.18953520357608794,
      "num_tokens": 32332742.0,
      "step": 18650
    },
    {
      "entropy": 5.544163656234741,
      "epoch": 1.451429682941062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047905283998161404,
      "loss": 5.1485,
      "mean_token_accuracy": 0.18568762838840486,
      "num_tokens": 32341598.0,
      "step": 18655
    },
    {
      "entropy": 5.5970807552337645,
      "epoch": 1.4518187123127797,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047904116067373743,
      "loss": 5.0973,
      "mean_token_accuracy": 0.19539392441511155,
      "num_tokens": 32350468.0,
      "step": 18660
    },
    {
      "entropy": 5.565386629104614,
      "epoch": 1.4522077416844972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004790294782698609,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19747879803180696,
      "num_tokens": 32359569.0,
      "step": 18665
    },
    {
      "entropy": 5.561293315887451,
      "epoch": 1.4525967710562147,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004790177927701618,
      "loss": 5.0712,
      "mean_token_accuracy": 0.19361127465963363,
      "num_tokens": 32368655.0,
      "step": 18670
    },
    {
      "entropy": 5.553066968917847,
      "epoch": 1.4529858004279324,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004790061041748174,
      "loss": 5.0351,
      "mean_token_accuracy": 0.1976610541343689,
      "num_tokens": 32377132.0,
      "step": 18675
    },
    {
      "entropy": 5.52764139175415,
      "epoch": 1.4533748297996498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004789944124840051,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19478783756494522,
      "num_tokens": 32385605.0,
      "step": 18680
    },
    {
      "entropy": 5.521400785446167,
      "epoch": 1.4537638591713673,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004789827176979022,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20212470591068268,
      "num_tokens": 32394195.0,
      "step": 18685
    },
    {
      "entropy": 5.598244857788086,
      "epoch": 1.454152888543085,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004789710198166864,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19377706348896026,
      "num_tokens": 32403278.0,
      "step": 18690
    },
    {
      "entropy": 5.592105197906494,
      "epoch": 1.4545419179148027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047895931884053497,
      "loss": 5.0736,
      "mean_token_accuracy": 0.19358156770467758,
      "num_tokens": 32411388.0,
      "step": 18695
    },
    {
      "entropy": 5.5809712409973145,
      "epoch": 1.4549309472865202,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047894761476962547,
      "loss": 5.0025,
      "mean_token_accuracy": 0.19814395010471345,
      "num_tokens": 32420860.0,
      "step": 18700
    },
    {
      "entropy": 5.546928548812867,
      "epoch": 1.4553199766582376,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047893590760413545,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20104021281003953,
      "num_tokens": 32428914.0,
      "step": 18705
    },
    {
      "entropy": 5.511069917678833,
      "epoch": 1.4557090060299553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047892419734424276,
      "loss": 5.0004,
      "mean_token_accuracy": 0.19954125583171844,
      "num_tokens": 32436704.0,
      "step": 18710
    },
    {
      "entropy": 5.522685098648071,
      "epoch": 1.4560980354016728,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047891248399012495,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1983110025525093,
      "num_tokens": 32444963.0,
      "step": 18715
    },
    {
      "entropy": 5.472441720962524,
      "epoch": 1.4564870647733903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789007675419597,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2016046166419983,
      "num_tokens": 32454148.0,
      "step": 18720
    },
    {
      "entropy": 5.486963081359863,
      "epoch": 1.456876094145108,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047888904799992486,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20393495857715607,
      "num_tokens": 32463084.0,
      "step": 18725
    },
    {
      "entropy": 5.511074876785278,
      "epoch": 1.4572651235168255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047887732536419826,
      "loss": 5.0189,
      "mean_token_accuracy": 0.19480252861976624,
      "num_tokens": 32471689.0,
      "step": 18730
    },
    {
      "entropy": 5.567481803894043,
      "epoch": 1.4576541528885432,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004788655996349577,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19184796810150145,
      "num_tokens": 32480243.0,
      "step": 18735
    },
    {
      "entropy": 5.492781686782837,
      "epoch": 1.4580431822602606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047885387081238125,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20447610318660736,
      "num_tokens": 32488782.0,
      "step": 18740
    },
    {
      "entropy": 5.530826997756958,
      "epoch": 1.4584322116319783,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004788421388966467,
      "loss": 5.0596,
      "mean_token_accuracy": 0.19475236684083938,
      "num_tokens": 32497692.0,
      "step": 18745
    },
    {
      "entropy": 5.5739185333251955,
      "epoch": 1.4588212410036958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047883040388793207,
      "loss": 5.094,
      "mean_token_accuracy": 0.19232002198696135,
      "num_tokens": 32507046.0,
      "step": 18750
    },
    {
      "entropy": 5.682011318206787,
      "epoch": 1.4592102703754133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047881866578641563,
      "loss": 5.1712,
      "mean_token_accuracy": 0.1911939039826393,
      "num_tokens": 32516602.0,
      "step": 18755
    },
    {
      "entropy": 5.600255727767944,
      "epoch": 1.459599299747131,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004788069245922753,
      "loss": 5.1044,
      "mean_token_accuracy": 0.1946917861700058,
      "num_tokens": 32524779.0,
      "step": 18760
    },
    {
      "entropy": 5.524578857421875,
      "epoch": 1.4599883291188485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004787951803056893,
      "loss": 4.9937,
      "mean_token_accuracy": 0.1964445099234581,
      "num_tokens": 32533207.0,
      "step": 18765
    },
    {
      "entropy": 5.5487542152404785,
      "epoch": 1.460377358490566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004787834329268358,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20234522074460984,
      "num_tokens": 32542447.0,
      "step": 18770
    },
    {
      "entropy": 5.509467840194702,
      "epoch": 1.4607663878622836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000478771682455893,
      "loss": 4.9865,
      "mean_token_accuracy": 0.19997165501117706,
      "num_tokens": 32551888.0,
      "step": 18775
    },
    {
      "entropy": 5.55902214050293,
      "epoch": 1.461155417234001,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004787599288930393,
      "loss": 5.0319,
      "mean_token_accuracy": 0.19427230209112167,
      "num_tokens": 32559639.0,
      "step": 18780
    },
    {
      "entropy": 5.567389869689942,
      "epoch": 1.4615444466057188,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000478748172238453,
      "loss": 5.0578,
      "mean_token_accuracy": 0.19651602506637572,
      "num_tokens": 32567518.0,
      "step": 18785
    },
    {
      "entropy": 5.61395001411438,
      "epoch": 1.4619334759774363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004787364124923125,
      "loss": 5.09,
      "mean_token_accuracy": 0.20352420657873155,
      "num_tokens": 32576040.0,
      "step": 18790
    },
    {
      "entropy": 5.572337007522583,
      "epoch": 1.462322505349154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047872464965479625,
      "loss": 5.075,
      "mean_token_accuracy": 0.19988814145326614,
      "num_tokens": 32584978.0,
      "step": 18795
    },
    {
      "entropy": 5.594401788711548,
      "epoch": 1.4627115347208715,
      "grad_norm": 1.125,
      "learning_rate": 0.0004787128837260826,
      "loss": 5.121,
      "mean_token_accuracy": 0.190612356364727,
      "num_tokens": 32593700.0,
      "step": 18800
    },
    {
      "entropy": 5.575336694717407,
      "epoch": 1.463100564092589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004787011147063503,
      "loss": 5.1079,
      "mean_token_accuracy": 0.18457038551568986,
      "num_tokens": 32602509.0,
      "step": 18805
    },
    {
      "entropy": 5.61750636100769,
      "epoch": 1.4634895934643066,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004786893425957777,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19278372526168824,
      "num_tokens": 32611740.0,
      "step": 18810
    },
    {
      "entropy": 5.54319748878479,
      "epoch": 1.463878622836024,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004786775673945436,
      "loss": 5.0269,
      "mean_token_accuracy": 0.209192718565464,
      "num_tokens": 32621285.0,
      "step": 18815
    },
    {
      "entropy": 5.5690044403076175,
      "epoch": 1.4642676522077416,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047866578910282656,
      "loss": 5.0156,
      "mean_token_accuracy": 0.19824934154748916,
      "num_tokens": 32629610.0,
      "step": 18820
    },
    {
      "entropy": 5.53894419670105,
      "epoch": 1.4646566815794593,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047865400772080535,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2051255449652672,
      "num_tokens": 32637907.0,
      "step": 18825
    },
    {
      "entropy": 5.488312768936157,
      "epoch": 1.4650457109511767,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047864222324865875,
      "loss": 5.0358,
      "mean_token_accuracy": 0.19553977251052856,
      "num_tokens": 32646783.0,
      "step": 18830
    },
    {
      "entropy": 5.512218618392945,
      "epoch": 1.4654347403228944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004786304356865655,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20518088340759277,
      "num_tokens": 32655538.0,
      "step": 18835
    },
    {
      "entropy": 5.499580574035645,
      "epoch": 1.465823769694612,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047861864503470457,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21253649592399598,
      "num_tokens": 32663683.0,
      "step": 18840
    },
    {
      "entropy": 5.554932737350464,
      "epoch": 1.4662127990663296,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004786068512932547,
      "loss": 5.0646,
      "mean_token_accuracy": 0.19895505160093307,
      "num_tokens": 32671961.0,
      "step": 18845
    },
    {
      "entropy": 5.62546739578247,
      "epoch": 1.466601828438047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000478595054462395,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19595526307821273,
      "num_tokens": 32680672.0,
      "step": 18850
    },
    {
      "entropy": 5.6233861446380615,
      "epoch": 1.4669908578097646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047858325454230437,
      "loss": 5.153,
      "mean_token_accuracy": 0.18422453254461288,
      "num_tokens": 32690481.0,
      "step": 18855
    },
    {
      "entropy": 5.556326580047608,
      "epoch": 1.4673798871814823,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004785714515331619,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1886402979493141,
      "num_tokens": 32698956.0,
      "step": 18860
    },
    {
      "entropy": 5.609915733337402,
      "epoch": 1.4677689165531997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047855964543514666,
      "loss": 5.1192,
      "mean_token_accuracy": 0.18813333958387374,
      "num_tokens": 32707018.0,
      "step": 18865
    },
    {
      "entropy": 5.562416028976441,
      "epoch": 1.4681579459249172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047854783624843786,
      "loss": 5.042,
      "mean_token_accuracy": 0.18667346686124803,
      "num_tokens": 32717278.0,
      "step": 18870
    },
    {
      "entropy": 5.550019550323486,
      "epoch": 1.468546975296635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047853602397321453,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19212803840637208,
      "num_tokens": 32725778.0,
      "step": 18875
    },
    {
      "entropy": 5.541028261184692,
      "epoch": 1.4689360046683524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047852420860965605,
      "loss": 5.0651,
      "mean_token_accuracy": 0.19158881306648254,
      "num_tokens": 32734457.0,
      "step": 18880
    },
    {
      "entropy": 5.532605409622192,
      "epoch": 1.46932503404007,
      "grad_norm": 1.125,
      "learning_rate": 0.00047851239015794166,
      "loss": 4.9764,
      "mean_token_accuracy": 0.19497962445020675,
      "num_tokens": 32742583.0,
      "step": 18885
    },
    {
      "entropy": 5.57095947265625,
      "epoch": 1.4697140634117876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047850056861825066,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19372896403074263,
      "num_tokens": 32751273.0,
      "step": 18890
    },
    {
      "entropy": 5.522993087768555,
      "epoch": 1.4701030927835053,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047848874399076245,
      "loss": 4.9468,
      "mean_token_accuracy": 0.19745635986328125,
      "num_tokens": 32759428.0,
      "step": 18895
    },
    {
      "entropy": 5.539993762969971,
      "epoch": 1.4704921221552227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004784769162756563,
      "loss": 5.0546,
      "mean_token_accuracy": 0.20103904008865356,
      "num_tokens": 32768032.0,
      "step": 18900
    },
    {
      "entropy": 5.605464696884155,
      "epoch": 1.4708811515269402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478465085473112,
      "loss": 5.124,
      "mean_token_accuracy": 0.19227155297994614,
      "num_tokens": 32777485.0,
      "step": 18905
    },
    {
      "entropy": 5.599140644073486,
      "epoch": 1.471270180898658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004784532515833088,
      "loss": 5.213,
      "mean_token_accuracy": 0.17923856526613235,
      "num_tokens": 32785794.0,
      "step": 18910
    },
    {
      "entropy": 5.54658784866333,
      "epoch": 1.4716592102703754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047844141460642636,
      "loss": 5.0358,
      "mean_token_accuracy": 0.19192324727773666,
      "num_tokens": 32794396.0,
      "step": 18915
    },
    {
      "entropy": 5.531925058364868,
      "epoch": 1.4720482396420929,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047842957454264425,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19278449118137359,
      "num_tokens": 32802912.0,
      "step": 18920
    },
    {
      "entropy": 5.48994722366333,
      "epoch": 1.4724372690138106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047841773139214213,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20175158828496934,
      "num_tokens": 32811683.0,
      "step": 18925
    },
    {
      "entropy": 5.5172168731689455,
      "epoch": 1.472826298385528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004784058851550997,
      "loss": 5.0274,
      "mean_token_accuracy": 0.19538899511098862,
      "num_tokens": 32821035.0,
      "step": 18930
    },
    {
      "entropy": 5.557153701782227,
      "epoch": 1.4732153277572457,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004783940358316967,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19770071357488633,
      "num_tokens": 32829678.0,
      "step": 18935
    },
    {
      "entropy": 5.543934488296509,
      "epoch": 1.4736043571289632,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047838218342211296,
      "loss": 5.117,
      "mean_token_accuracy": 0.1937295228242874,
      "num_tokens": 32838048.0,
      "step": 18940
    },
    {
      "entropy": 5.4855937480926515,
      "epoch": 1.473993386500681,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047837032792652837,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20186340659856797,
      "num_tokens": 32846380.0,
      "step": 18945
    },
    {
      "entropy": 5.6162858486175535,
      "epoch": 1.4743824158723984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783584693451226,
      "loss": 5.132,
      "mean_token_accuracy": 0.18568824827671052,
      "num_tokens": 32855318.0,
      "step": 18950
    },
    {
      "entropy": 5.654341459274292,
      "epoch": 1.4747714452441159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004783466076780759,
      "loss": 5.1094,
      "mean_token_accuracy": 0.18727703541517257,
      "num_tokens": 32864494.0,
      "step": 18955
    },
    {
      "entropy": 5.577995157241821,
      "epoch": 1.4751604746158336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004783347429255679,
      "loss": 4.9787,
      "mean_token_accuracy": 0.19697563350200653,
      "num_tokens": 32872770.0,
      "step": 18960
    },
    {
      "entropy": 5.515985774993896,
      "epoch": 1.475549503987551,
      "grad_norm": 1.1875,
      "learning_rate": 0.000478322875087779,
      "loss": 5.0694,
      "mean_token_accuracy": 0.19254512935876847,
      "num_tokens": 32881249.0,
      "step": 18965
    },
    {
      "entropy": 5.576406192779541,
      "epoch": 1.4759385333592685,
      "grad_norm": 1.125,
      "learning_rate": 0.00047831100416488906,
      "loss": 5.1155,
      "mean_token_accuracy": 0.19401678144931794,
      "num_tokens": 32890197.0,
      "step": 18970
    },
    {
      "entropy": 5.527444696426391,
      "epoch": 1.4763275627309862,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047829913015707816,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20043394416570665,
      "num_tokens": 32898424.0,
      "step": 18975
    },
    {
      "entropy": 5.536509990692139,
      "epoch": 1.4767165921027037,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047828725306452657,
      "loss": 5.0216,
      "mean_token_accuracy": 0.1982140436768532,
      "num_tokens": 32906527.0,
      "step": 18980
    },
    {
      "entropy": 5.52204737663269,
      "epoch": 1.4771056214744214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047827537288741453,
      "loss": 5.0231,
      "mean_token_accuracy": 0.19963836371898652,
      "num_tokens": 32914954.0,
      "step": 18985
    },
    {
      "entropy": 5.532572460174561,
      "epoch": 1.4774946508461388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004782634896259222,
      "loss": 5.0581,
      "mean_token_accuracy": 0.19437048137187957,
      "num_tokens": 32923562.0,
      "step": 18990
    },
    {
      "entropy": 5.546833848953247,
      "epoch": 1.4778836802178565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047825160328023,
      "loss": 5.0221,
      "mean_token_accuracy": 0.2039424732327461,
      "num_tokens": 32931735.0,
      "step": 18995
    },
    {
      "entropy": 5.585835981369018,
      "epoch": 1.478272709589574,
      "grad_norm": 1.125,
      "learning_rate": 0.0004782397138505181,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1972127839922905,
      "num_tokens": 32940282.0,
      "step": 19000
    },
    {
      "entropy": 5.526833629608154,
      "epoch": 1.4786617389612915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047822782133696715,
      "loss": 5.0178,
      "mean_token_accuracy": 0.18914930522441864,
      "num_tokens": 32948354.0,
      "step": 19005
    },
    {
      "entropy": 5.554025602340698,
      "epoch": 1.4790507683330092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004782159257397574,
      "loss": 5.0473,
      "mean_token_accuracy": 0.1947380468249321,
      "num_tokens": 32957489.0,
      "step": 19010
    },
    {
      "entropy": 5.676969861984253,
      "epoch": 1.4794397977047267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047820402705906953,
      "loss": 5.2671,
      "mean_token_accuracy": 0.17776482850313186,
      "num_tokens": 32967739.0,
      "step": 19015
    },
    {
      "entropy": 5.619980192184448,
      "epoch": 1.4798288270764441,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047819212529508394,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19449526816606522,
      "num_tokens": 32976131.0,
      "step": 19020
    },
    {
      "entropy": 5.613455104827881,
      "epoch": 1.4802178564481618,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004781802204479812,
      "loss": 5.119,
      "mean_token_accuracy": 0.19957145750522615,
      "num_tokens": 32984678.0,
      "step": 19025
    },
    {
      "entropy": 5.530535078048706,
      "epoch": 1.4806068858198795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004781683125179421,
      "loss": 4.9376,
      "mean_token_accuracy": 0.20044415444135666,
      "num_tokens": 32993117.0,
      "step": 19030
    },
    {
      "entropy": 5.5346996784210205,
      "epoch": 1.480995915191597,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004781564015051472,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2034488156437874,
      "num_tokens": 33002104.0,
      "step": 19035
    },
    {
      "entropy": 5.5476133823394775,
      "epoch": 1.4813849445633145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004781444874097772,
      "loss": 4.9966,
      "mean_token_accuracy": 0.19342254102230072,
      "num_tokens": 33010515.0,
      "step": 19040
    },
    {
      "entropy": 5.609853649139405,
      "epoch": 1.4817739739350322,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047813257023201307,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19935715645551683,
      "num_tokens": 33019448.0,
      "step": 19045
    },
    {
      "entropy": 5.603008127212524,
      "epoch": 1.4821630033067497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004781206499720354,
      "loss": 5.0694,
      "mean_token_accuracy": 0.19271350353956224,
      "num_tokens": 33027490.0,
      "step": 19050
    },
    {
      "entropy": 5.525425100326538,
      "epoch": 1.4825520326784671,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047810872663002523,
      "loss": 5.1014,
      "mean_token_accuracy": 0.19199770241975783,
      "num_tokens": 33036428.0,
      "step": 19055
    },
    {
      "entropy": 5.543790483474732,
      "epoch": 1.4829410620501848,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047809680020616333,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1909507155418396,
      "num_tokens": 33045366.0,
      "step": 19060
    },
    {
      "entropy": 5.60338864326477,
      "epoch": 1.4833300914219023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047808487070063083,
      "loss": 5.005,
      "mean_token_accuracy": 0.20054623782634734,
      "num_tokens": 33054280.0,
      "step": 19065
    },
    {
      "entropy": 5.527318048477173,
      "epoch": 1.4837191207936198,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004780729381136086,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19764809012413026,
      "num_tokens": 33062935.0,
      "step": 19070
    },
    {
      "entropy": 5.542918300628662,
      "epoch": 1.4841081501653375,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004780610024452778,
      "loss": 5.0178,
      "mean_token_accuracy": 0.20716737806797028,
      "num_tokens": 33071795.0,
      "step": 19075
    },
    {
      "entropy": 5.543208837509155,
      "epoch": 1.4844971795370552,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047804906369581954,
      "loss": 5.0316,
      "mean_token_accuracy": 0.19775291234254838,
      "num_tokens": 33080280.0,
      "step": 19080
    },
    {
      "entropy": 5.533397769927978,
      "epoch": 1.4848862089087727,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004780371218654148,
      "loss": 5.0203,
      "mean_token_accuracy": 0.2004026934504509,
      "num_tokens": 33089125.0,
      "step": 19085
    },
    {
      "entropy": 5.469947862625122,
      "epoch": 1.4852752382804901,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047802517695424496,
      "loss": 4.974,
      "mean_token_accuracy": 0.20168286710977554,
      "num_tokens": 33097632.0,
      "step": 19090
    },
    {
      "entropy": 5.632231855392456,
      "epoch": 1.4856642676522078,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004780132289624913,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19432080686092376,
      "num_tokens": 33106318.0,
      "step": 19095
    },
    {
      "entropy": 5.558058166503907,
      "epoch": 1.4860532970239253,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004780012778903349,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20190412253141404,
      "num_tokens": 33115268.0,
      "step": 19100
    },
    {
      "entropy": 5.500111627578735,
      "epoch": 1.4864423263956428,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047798932373795724,
      "loss": 5.0698,
      "mean_token_accuracy": 0.19048128724098207,
      "num_tokens": 33123968.0,
      "step": 19105
    },
    {
      "entropy": 5.627884340286255,
      "epoch": 1.4868313557673605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047797736650553977,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20432619899511337,
      "num_tokens": 33132775.0,
      "step": 19110
    },
    {
      "entropy": 5.551082754135132,
      "epoch": 1.487220385139078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047796540619326366,
      "loss": 4.9859,
      "mean_token_accuracy": 0.1998939499258995,
      "num_tokens": 33141319.0,
      "step": 19115
    },
    {
      "entropy": 5.514258337020874,
      "epoch": 1.4876094145107956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004779534428013107,
      "loss": 5.0423,
      "mean_token_accuracy": 0.19480806738138198,
      "num_tokens": 33150175.0,
      "step": 19120
    },
    {
      "entropy": 5.49319634437561,
      "epoch": 1.4879984438825131,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047794147632986223,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19951022714376448,
      "num_tokens": 33159961.0,
      "step": 19125
    },
    {
      "entropy": 5.55402398109436,
      "epoch": 1.4883874732542308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004779295067790999,
      "loss": 5.0472,
      "mean_token_accuracy": 0.19604973047971724,
      "num_tokens": 33168672.0,
      "step": 19130
    },
    {
      "entropy": 5.574004983901977,
      "epoch": 1.4887765026259483,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004779175341492053,
      "loss": 5.0091,
      "mean_token_accuracy": 0.19269197881221772,
      "num_tokens": 33177229.0,
      "step": 19135
    },
    {
      "entropy": 5.58456563949585,
      "epoch": 1.4891655319976658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004779055584403601,
      "loss": 5.0828,
      "mean_token_accuracy": 0.18931804448366166,
      "num_tokens": 33185673.0,
      "step": 19140
    },
    {
      "entropy": 5.542715454101563,
      "epoch": 1.4895545613693835,
      "grad_norm": 1.09375,
      "learning_rate": 0.000477893579652746,
      "loss": 4.9757,
      "mean_token_accuracy": 0.20158514082431794,
      "num_tokens": 33194860.0,
      "step": 19145
    },
    {
      "entropy": 5.587748908996582,
      "epoch": 1.489943590741101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047788159778654475,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19882684648036958,
      "num_tokens": 33203205.0,
      "step": 19150
    },
    {
      "entropy": 5.54861536026001,
      "epoch": 1.4903326201128184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047786961284193813,
      "loss": 5.0886,
      "mean_token_accuracy": 0.19064337462186814,
      "num_tokens": 33211359.0,
      "step": 19155
    },
    {
      "entropy": 5.617289638519287,
      "epoch": 1.4907216494845361,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004778576248191081,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20279954075813295,
      "num_tokens": 33220176.0,
      "step": 19160
    },
    {
      "entropy": 5.479634475708008,
      "epoch": 1.4911106788562536,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004778456337182365,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2093699350953102,
      "num_tokens": 33228254.0,
      "step": 19165
    },
    {
      "entropy": 5.474156475067138,
      "epoch": 1.4914997082279713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004778336395395052,
      "loss": 4.956,
      "mean_token_accuracy": 0.20219864696264267,
      "num_tokens": 33236411.0,
      "step": 19170
    },
    {
      "entropy": 5.525721693038941,
      "epoch": 1.4918887375996888,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047782164228309636,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19211238920688628,
      "num_tokens": 33244222.0,
      "step": 19175
    },
    {
      "entropy": 5.567611265182495,
      "epoch": 1.4922777669714065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047780964194919193,
      "loss": 5.0757,
      "mean_token_accuracy": 0.19291103333234788,
      "num_tokens": 33253609.0,
      "step": 19180
    },
    {
      "entropy": 5.602247667312622,
      "epoch": 1.492666796343124,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047779763853797387,
      "loss": 5.0371,
      "mean_token_accuracy": 0.19836950898170472,
      "num_tokens": 33262002.0,
      "step": 19185
    },
    {
      "entropy": 5.561488008499145,
      "epoch": 1.4930558257148414,
      "grad_norm": 1.125,
      "learning_rate": 0.0004777856320496245,
      "loss": 5.0692,
      "mean_token_accuracy": 0.19719235748052596,
      "num_tokens": 33271185.0,
      "step": 19190
    },
    {
      "entropy": 5.534680128097534,
      "epoch": 1.493444855086559,
      "grad_norm": 1.046875,
      "learning_rate": 0.000477773622484326,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19225019216537476,
      "num_tokens": 33280205.0,
      "step": 19195
    },
    {
      "entropy": 5.617606544494629,
      "epoch": 1.4938338844582766,
      "grad_norm": 1.25,
      "learning_rate": 0.0004777616098422604,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19693068712949752,
      "num_tokens": 33288846.0,
      "step": 19200
    },
    {
      "entropy": 5.5117240905761715,
      "epoch": 1.494222913829994,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047774959412361014,
      "loss": 5.0209,
      "mean_token_accuracy": 0.19191986918449402,
      "num_tokens": 33297057.0,
      "step": 19205
    },
    {
      "entropy": 5.50296196937561,
      "epoch": 1.4946119432017118,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004777375753285575,
      "loss": 4.9524,
      "mean_token_accuracy": 0.20566382110118867,
      "num_tokens": 33305091.0,
      "step": 19210
    },
    {
      "entropy": 5.616590070724487,
      "epoch": 1.4950009725734292,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047772555345728486,
      "loss": 5.1389,
      "mean_token_accuracy": 0.20009548217058182,
      "num_tokens": 33313383.0,
      "step": 19215
    },
    {
      "entropy": 5.551234436035156,
      "epoch": 1.495390001945147,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004777135285099746,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20243726521730424,
      "num_tokens": 33322521.0,
      "step": 19220
    },
    {
      "entropy": 5.602776145935058,
      "epoch": 1.4957790313168644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004777015004868092,
      "loss": 5.2338,
      "mean_token_accuracy": 0.18354346603155136,
      "num_tokens": 33331474.0,
      "step": 19225
    },
    {
      "entropy": 5.6186521530151365,
      "epoch": 1.496168060688582,
      "grad_norm": 1.125,
      "learning_rate": 0.0004776894693879712,
      "loss": 5.0888,
      "mean_token_accuracy": 0.1937849074602127,
      "num_tokens": 33341354.0,
      "step": 19230
    },
    {
      "entropy": 5.526188516616822,
      "epoch": 1.4965570900602996,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004776774352136431,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20695056319236754,
      "num_tokens": 33349256.0,
      "step": 19235
    },
    {
      "entropy": 5.594116353988648,
      "epoch": 1.496946119432017,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047766539796400756,
      "loss": 5.1141,
      "mean_token_accuracy": 0.19094278663396835,
      "num_tokens": 33358543.0,
      "step": 19240
    },
    {
      "entropy": 5.5897684574127195,
      "epoch": 1.4973351488037348,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004776533576392471,
      "loss": 5.1171,
      "mean_token_accuracy": 0.19825131297111512,
      "num_tokens": 33366695.0,
      "step": 19245
    },
    {
      "entropy": 5.598502349853516,
      "epoch": 1.4977241781754522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004776413142395445,
      "loss": 5.1075,
      "mean_token_accuracy": 0.19641636312007904,
      "num_tokens": 33375596.0,
      "step": 19250
    },
    {
      "entropy": 5.592920112609863,
      "epoch": 1.4981132075471697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047762926776508244,
      "loss": 5.0582,
      "mean_token_accuracy": 0.19774041920900345,
      "num_tokens": 33385304.0,
      "step": 19255
    },
    {
      "entropy": 5.490079021453857,
      "epoch": 1.4985022369188874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047761721821604387,
      "loss": 5.0022,
      "mean_token_accuracy": 0.19547391384840013,
      "num_tokens": 33394110.0,
      "step": 19260
    },
    {
      "entropy": 5.545273017883301,
      "epoch": 1.4988912662906049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004776051655926115,
      "loss": 5.0182,
      "mean_token_accuracy": 0.19738116413354873,
      "num_tokens": 33403430.0,
      "step": 19265
    },
    {
      "entropy": 5.535129356384277,
      "epoch": 1.4992802956623226,
      "grad_norm": 1.125,
      "learning_rate": 0.00047759310989496813,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20044422298669815,
      "num_tokens": 33411853.0,
      "step": 19270
    },
    {
      "entropy": 5.550904893875122,
      "epoch": 1.49966932503404,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004775810511232969,
      "loss": 5.0601,
      "mean_token_accuracy": 0.1886581525206566,
      "num_tokens": 33420579.0,
      "step": 19275
    },
    {
      "entropy": 5.528168821334839,
      "epoch": 1.5000583544057577,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047756898927778056,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2010014221072197,
      "num_tokens": 33429290.0,
      "step": 19280
    },
    {
      "entropy": 5.637648963928223,
      "epoch": 1.5004473837774752,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047755692435860227,
      "loss": 5.1133,
      "mean_token_accuracy": 0.1969171106815338,
      "num_tokens": 33437962.0,
      "step": 19285
    },
    {
      "entropy": 5.618480014801025,
      "epoch": 1.5008364131491927,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477544856365945,
      "loss": 5.119,
      "mean_token_accuracy": 0.19115120321512222,
      "num_tokens": 33446755.0,
      "step": 19290
    },
    {
      "entropy": 5.574110698699951,
      "epoch": 1.5012254425209104,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047753278529999205,
      "loss": 5.0515,
      "mean_token_accuracy": 0.19534226953983308,
      "num_tokens": 33455223.0,
      "step": 19295
    },
    {
      "entropy": 5.584450054168701,
      "epoch": 1.5016144718926279,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047752071116092637,
      "loss": 5.0471,
      "mean_token_accuracy": 0.19488006830215454,
      "num_tokens": 33463089.0,
      "step": 19300
    },
    {
      "entropy": 5.625916004180908,
      "epoch": 1.5020035012643453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004775086339489312,
      "loss": 5.0369,
      "mean_token_accuracy": 0.1957094356417656,
      "num_tokens": 33470519.0,
      "step": 19305
    },
    {
      "entropy": 5.5653684616088865,
      "epoch": 1.502392530636063,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004774965536641899,
      "loss": 5.081,
      "mean_token_accuracy": 0.190675051510334,
      "num_tokens": 33479256.0,
      "step": 19310
    },
    {
      "entropy": 5.551696348190307,
      "epoch": 1.5027815600077807,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047748447030688575,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2024025246500969,
      "num_tokens": 33487601.0,
      "step": 19315
    },
    {
      "entropy": 5.588086700439453,
      "epoch": 1.503170589379498,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047747238387720194,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19902506917715074,
      "num_tokens": 33497124.0,
      "step": 19320
    },
    {
      "entropy": 5.572511911392212,
      "epoch": 1.5035596187512157,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477460294375322,
      "loss": 5.076,
      "mean_token_accuracy": 0.18755635917186736,
      "num_tokens": 33506022.0,
      "step": 19325
    },
    {
      "entropy": 5.500413179397583,
      "epoch": 1.5039486481229334,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047744820180142935,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1977880299091339,
      "num_tokens": 33514109.0,
      "step": 19330
    },
    {
      "entropy": 5.47395544052124,
      "epoch": 1.5043376774946509,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047743610615570746,
      "loss": 5.0256,
      "mean_token_accuracy": 0.193344946205616,
      "num_tokens": 33523097.0,
      "step": 19335
    },
    {
      "entropy": 5.624269008636475,
      "epoch": 1.5047267068663683,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047742400743833993,
      "loss": 5.1276,
      "mean_token_accuracy": 0.18903702795505523,
      "num_tokens": 33531149.0,
      "step": 19340
    },
    {
      "entropy": 5.57778697013855,
      "epoch": 1.505115736238086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004774119056495102,
      "loss": 5.0216,
      "mean_token_accuracy": 0.19675463140010835,
      "num_tokens": 33539974.0,
      "step": 19345
    },
    {
      "entropy": 5.554604578018188,
      "epoch": 1.5055047656098035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047739980078940206,
      "loss": 4.9862,
      "mean_token_accuracy": 0.193359899520874,
      "num_tokens": 33548670.0,
      "step": 19350
    },
    {
      "entropy": 5.600244569778442,
      "epoch": 1.505893794981521,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047738769285819894,
      "loss": 5.0968,
      "mean_token_accuracy": 0.19069682955741882,
      "num_tokens": 33556904.0,
      "step": 19355
    },
    {
      "entropy": 5.59004430770874,
      "epoch": 1.5062828243532387,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004773755818560849,
      "loss": 5.1403,
      "mean_token_accuracy": 0.1927887663245201,
      "num_tokens": 33565583.0,
      "step": 19360
    },
    {
      "entropy": 5.5753296375274655,
      "epoch": 1.5066718537249564,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004773634677832434,
      "loss": 5.035,
      "mean_token_accuracy": 0.19719114303588867,
      "num_tokens": 33573644.0,
      "step": 19365
    },
    {
      "entropy": 5.636075401306153,
      "epoch": 1.5070608830966739,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004773513506398584,
      "loss": 5.2082,
      "mean_token_accuracy": 0.18971898406744003,
      "num_tokens": 33582098.0,
      "step": 19370
    },
    {
      "entropy": 5.604250431060791,
      "epoch": 1.5074499124683913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004773392304261137,
      "loss": 5.1552,
      "mean_token_accuracy": 0.1846052423119545,
      "num_tokens": 33591299.0,
      "step": 19375
    },
    {
      "entropy": 5.591132307052613,
      "epoch": 1.507838941840109,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004773271071421933,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20295005589723586,
      "num_tokens": 33599442.0,
      "step": 19380
    },
    {
      "entropy": 5.60074520111084,
      "epoch": 1.5082279712118265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047731498078828105,
      "loss": 5.1556,
      "mean_token_accuracy": 0.2008211374282837,
      "num_tokens": 33608445.0,
      "step": 19385
    },
    {
      "entropy": 5.578325176239014,
      "epoch": 1.508617000583544,
      "grad_norm": 1.125,
      "learning_rate": 0.000477302851364561,
      "loss": 5.1311,
      "mean_token_accuracy": 0.19358229488134385,
      "num_tokens": 33617025.0,
      "step": 19390
    },
    {
      "entropy": 5.5581684589385985,
      "epoch": 1.5090060299552617,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047729071887121717,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20439456701278685,
      "num_tokens": 33625671.0,
      "step": 19395
    },
    {
      "entropy": 5.633315658569336,
      "epoch": 1.5093950593269791,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004772785833084337,
      "loss": 5.1269,
      "mean_token_accuracy": 0.18746764957904816,
      "num_tokens": 33634672.0,
      "step": 19400
    },
    {
      "entropy": 5.607293319702149,
      "epoch": 1.5097840886986966,
      "grad_norm": 1.125,
      "learning_rate": 0.0004772664446763946,
      "loss": 5.1339,
      "mean_token_accuracy": 0.1864413097500801,
      "num_tokens": 33643810.0,
      "step": 19405
    },
    {
      "entropy": 5.548423528671265,
      "epoch": 1.5101731180704143,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004772543029752842,
      "loss": 4.8954,
      "mean_token_accuracy": 0.20861364156007767,
      "num_tokens": 33652638.0,
      "step": 19410
    },
    {
      "entropy": 5.5491053581237795,
      "epoch": 1.510562147442132,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047724215820528666,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1972549542784691,
      "num_tokens": 33660705.0,
      "step": 19415
    },
    {
      "entropy": 5.609587335586548,
      "epoch": 1.5109511768138495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004772300103665863,
      "loss": 5.1282,
      "mean_token_accuracy": 0.19567057192325593,
      "num_tokens": 33669446.0,
      "step": 19420
    },
    {
      "entropy": 5.517504692077637,
      "epoch": 1.511340206185567,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047721785945936733,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20746591240167617,
      "num_tokens": 33678505.0,
      "step": 19425
    },
    {
      "entropy": 5.554972410202026,
      "epoch": 1.5117292355572847,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004772057054838143,
      "loss": 5.0854,
      "mean_token_accuracy": 0.19702490717172622,
      "num_tokens": 33687213.0,
      "step": 19430
    },
    {
      "entropy": 5.541841506958008,
      "epoch": 1.5121182649290021,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047719354844011146,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20011643767356874,
      "num_tokens": 33696040.0,
      "step": 19435
    },
    {
      "entropy": 5.610608100891113,
      "epoch": 1.5125072943007196,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004771813883284434,
      "loss": 5.1297,
      "mean_token_accuracy": 0.18853561878204345,
      "num_tokens": 33705210.0,
      "step": 19440
    },
    {
      "entropy": 5.502421903610229,
      "epoch": 1.5128963236724373,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047716922514899455,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20888576358556749,
      "num_tokens": 33714255.0,
      "step": 19445
    },
    {
      "entropy": 5.5027073383331295,
      "epoch": 1.5132853530441548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047715705890194953,
      "loss": 5.003,
      "mean_token_accuracy": 0.20668907910585405,
      "num_tokens": 33723368.0,
      "step": 19450
    },
    {
      "entropy": 5.584422397613525,
      "epoch": 1.5136743824158723,
      "grad_norm": 1.0,
      "learning_rate": 0.00047714488958749285,
      "loss": 5.031,
      "mean_token_accuracy": 0.19946435540914537,
      "num_tokens": 33732743.0,
      "step": 19455
    },
    {
      "entropy": 5.59551477432251,
      "epoch": 1.51406341178759,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047713271720580924,
      "loss": 5.104,
      "mean_token_accuracy": 0.1872835174202919,
      "num_tokens": 33741744.0,
      "step": 19460
    },
    {
      "entropy": 5.548620271682739,
      "epoch": 1.5144524411593077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004771205417570834,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20384987890720369,
      "num_tokens": 33749646.0,
      "step": 19465
    },
    {
      "entropy": 5.554851531982422,
      "epoch": 1.5148414705310251,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047710836324150004,
      "loss": 5.055,
      "mean_token_accuracy": 0.1971933737397194,
      "num_tokens": 33758031.0,
      "step": 19470
    },
    {
      "entropy": 5.52572865486145,
      "epoch": 1.5152304999027426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477096181659244,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19707555174827576,
      "num_tokens": 33766592.0,
      "step": 19475
    },
    {
      "entropy": 5.567147922515869,
      "epoch": 1.5156195292744603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004770839970105,
      "loss": 5.1863,
      "mean_token_accuracy": 0.18891493678092958,
      "num_tokens": 33775741.0,
      "step": 19480
    },
    {
      "entropy": 5.592911720275879,
      "epoch": 1.5160085586461778,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047707180929545295,
      "loss": 5.1022,
      "mean_token_accuracy": 0.193851974606514,
      "num_tokens": 33784551.0,
      "step": 19485
    },
    {
      "entropy": 5.598547410964966,
      "epoch": 1.5163975880178953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047705961851428786,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19360142797231675,
      "num_tokens": 33793051.0,
      "step": 19490
    },
    {
      "entropy": 5.584777116775513,
      "epoch": 1.516786617389613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047704742466718973,
      "loss": 5.0091,
      "mean_token_accuracy": 0.19565768986940385,
      "num_tokens": 33802016.0,
      "step": 19495
    },
    {
      "entropy": 5.538855981826782,
      "epoch": 1.5171756467613304,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047703522775434354,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1965561494231224,
      "num_tokens": 33810515.0,
      "step": 19500
    },
    {
      "entropy": 5.559140872955322,
      "epoch": 1.517564676133048,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047702302777593425,
      "loss": 5.0577,
      "mean_token_accuracy": 0.19679870903491975,
      "num_tokens": 33820295.0,
      "step": 19505
    },
    {
      "entropy": 5.6061439037323,
      "epoch": 1.5179537055047656,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047701082473214715,
      "loss": 5.0843,
      "mean_token_accuracy": 0.19542277455329896,
      "num_tokens": 33829124.0,
      "step": 19510
    },
    {
      "entropy": 5.552377939224243,
      "epoch": 1.5183427348764833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047699861862316725,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20136120021343232,
      "num_tokens": 33838575.0,
      "step": 19515
    },
    {
      "entropy": 5.579402780532837,
      "epoch": 1.5187317642482008,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004769864094491798,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19320601969957352,
      "num_tokens": 33847262.0,
      "step": 19520
    },
    {
      "entropy": 5.564948654174804,
      "epoch": 1.5191207936199183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004769741972103702,
      "loss": 5.1213,
      "mean_token_accuracy": 0.19218202531337739,
      "num_tokens": 33856056.0,
      "step": 19525
    },
    {
      "entropy": 5.5473710060119625,
      "epoch": 1.519509822991636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047696198190692353,
      "loss": 5.1129,
      "mean_token_accuracy": 0.19225240349769593,
      "num_tokens": 33865408.0,
      "step": 19530
    },
    {
      "entropy": 5.549135351181031,
      "epoch": 1.5198988523633534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004769497635390253,
      "loss": 5.0188,
      "mean_token_accuracy": 0.20721402913331985,
      "num_tokens": 33873927.0,
      "step": 19535
    },
    {
      "entropy": 5.5910484790802,
      "epoch": 1.520287881735071,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004769375421068608,
      "loss": 5.0683,
      "mean_token_accuracy": 0.18575547337532045,
      "num_tokens": 33881851.0,
      "step": 19540
    },
    {
      "entropy": 5.533737230300903,
      "epoch": 1.5206769111067886,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047692531761061555,
      "loss": 4.9862,
      "mean_token_accuracy": 0.203743115067482,
      "num_tokens": 33890462.0,
      "step": 19545
    },
    {
      "entropy": 5.49003791809082,
      "epoch": 1.521065940478506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000476913090050475,
      "loss": 5.0525,
      "mean_token_accuracy": 0.19478340893983842,
      "num_tokens": 33899357.0,
      "step": 19550
    },
    {
      "entropy": 5.526320314407348,
      "epoch": 1.5214549698502235,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047690085942662464,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20074675232172012,
      "num_tokens": 33907451.0,
      "step": 19555
    },
    {
      "entropy": 5.57744779586792,
      "epoch": 1.5218439992219412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047688862573925015,
      "loss": 5.0611,
      "mean_token_accuracy": 0.19656474888324738,
      "num_tokens": 33915808.0,
      "step": 19560
    },
    {
      "entropy": 5.615764617919922,
      "epoch": 1.522233028593659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047687638898853707,
      "loss": 5.1406,
      "mean_token_accuracy": 0.19714242815971375,
      "num_tokens": 33924317.0,
      "step": 19565
    },
    {
      "entropy": 5.577002143859863,
      "epoch": 1.5226220579653764,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004768641491746711,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1939859628677368,
      "num_tokens": 33932788.0,
      "step": 19570
    },
    {
      "entropy": 5.555511856079102,
      "epoch": 1.523011087337094,
      "grad_norm": 1.15625,
      "learning_rate": 0.000476851906297838,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1965989425778389,
      "num_tokens": 33941358.0,
      "step": 19575
    },
    {
      "entropy": 5.547682857513427,
      "epoch": 1.5234001167088116,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047683966035822346,
      "loss": 5.0142,
      "mean_token_accuracy": 0.1982572928071022,
      "num_tokens": 33950155.0,
      "step": 19580
    },
    {
      "entropy": 5.51738657951355,
      "epoch": 1.523789146080529,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047682741135601336,
      "loss": 5.0254,
      "mean_token_accuracy": 0.1976988285779953,
      "num_tokens": 33958559.0,
      "step": 19585
    },
    {
      "entropy": 5.578416204452514,
      "epoch": 1.5241781754522465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047681515929139356,
      "loss": 5.0944,
      "mean_token_accuracy": 0.19835715293884276,
      "num_tokens": 33967215.0,
      "step": 19590
    },
    {
      "entropy": 5.570069932937622,
      "epoch": 1.5245672048239642,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047680290416454995,
      "loss": 5.0777,
      "mean_token_accuracy": 0.19689611792564393,
      "num_tokens": 33976248.0,
      "step": 19595
    },
    {
      "entropy": 5.622149324417114,
      "epoch": 1.524956234195682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767906459756684,
      "loss": 5.1914,
      "mean_token_accuracy": 0.18724040687084198,
      "num_tokens": 33985189.0,
      "step": 19600
    },
    {
      "entropy": 5.553195905685425,
      "epoch": 1.5253452635673992,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047677838472493504,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19655317664146424,
      "num_tokens": 33993998.0,
      "step": 19605
    },
    {
      "entropy": 5.573758840560913,
      "epoch": 1.5257342929391169,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004767661204125358,
      "loss": 4.9991,
      "mean_token_accuracy": 0.2049141451716423,
      "num_tokens": 34002366.0,
      "step": 19610
    },
    {
      "entropy": 5.587067031860352,
      "epoch": 1.5261233223108346,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004767538530386569,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20680000633001328,
      "num_tokens": 34011404.0,
      "step": 19615
    },
    {
      "entropy": 5.498363018035889,
      "epoch": 1.526512351682552,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047674158260348437,
      "loss": 5.0027,
      "mean_token_accuracy": 0.19868876039981842,
      "num_tokens": 34020290.0,
      "step": 19620
    },
    {
      "entropy": 5.519731140136718,
      "epoch": 1.5269013810542695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047672930910720436,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1909203290939331,
      "num_tokens": 34029249.0,
      "step": 19625
    },
    {
      "entropy": 5.570719146728516,
      "epoch": 1.5272904104259872,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047671703255000326,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1951455295085907,
      "num_tokens": 34037640.0,
      "step": 19630
    },
    {
      "entropy": 5.509208345413208,
      "epoch": 1.5276794397977047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004767047529320673,
      "loss": 5.0184,
      "mean_token_accuracy": 0.19293158799409865,
      "num_tokens": 34045779.0,
      "step": 19635
    },
    {
      "entropy": 5.548420190811157,
      "epoch": 1.5280684691694222,
      "grad_norm": 1.0,
      "learning_rate": 0.00047669247025358257,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19421012103557586,
      "num_tokens": 34054888.0,
      "step": 19640
    },
    {
      "entropy": 5.580225849151612,
      "epoch": 1.5284574985411399,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047668018451473584,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2010911598801613,
      "num_tokens": 34063547.0,
      "step": 19645
    },
    {
      "entropy": 5.548436307907105,
      "epoch": 1.5288465279128576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004766678957157132,
      "loss": 4.9829,
      "mean_token_accuracy": 0.19771626442670823,
      "num_tokens": 34072306.0,
      "step": 19650
    },
    {
      "entropy": 5.590692520141602,
      "epoch": 1.5292355572845748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004766556038567013,
      "loss": 5.0347,
      "mean_token_accuracy": 0.19475244581699372,
      "num_tokens": 34080335.0,
      "step": 19655
    },
    {
      "entropy": 5.565773963928223,
      "epoch": 1.5296245866562925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004766433089378865,
      "loss": 4.9686,
      "mean_token_accuracy": 0.1977250635623932,
      "num_tokens": 34089083.0,
      "step": 19660
    },
    {
      "entropy": 5.5273607730865475,
      "epoch": 1.5300136160280102,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047663101095945544,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20454098880290986,
      "num_tokens": 34097586.0,
      "step": 19665
    },
    {
      "entropy": 5.534102535247802,
      "epoch": 1.5304026453997277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047661870992159476,
      "loss": 4.9985,
      "mean_token_accuracy": 0.1913001924753189,
      "num_tokens": 34106069.0,
      "step": 19670
    },
    {
      "entropy": 5.531040906906128,
      "epoch": 1.5307916747714452,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047660640582449106,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2074694126844406,
      "num_tokens": 34114804.0,
      "step": 19675
    },
    {
      "entropy": 5.542651510238647,
      "epoch": 1.5311807041431629,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047659409866833104,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20633669793605805,
      "num_tokens": 34123605.0,
      "step": 19680
    },
    {
      "entropy": 5.6436842918396,
      "epoch": 1.5315697335148803,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004765817884533014,
      "loss": 5.1347,
      "mean_token_accuracy": 0.18847558051347732,
      "num_tokens": 34132628.0,
      "step": 19685
    },
    {
      "entropy": 5.595562076568603,
      "epoch": 1.5319587628865978,
      "grad_norm": 1.03125,
      "learning_rate": 0.000476569475179589,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19493800401687622,
      "num_tokens": 34141546.0,
      "step": 19690
    },
    {
      "entropy": 5.517737102508545,
      "epoch": 1.5323477922583155,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047655715884738074,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20552916526794435,
      "num_tokens": 34150126.0,
      "step": 19695
    },
    {
      "entropy": 5.546139478683472,
      "epoch": 1.5327368216300332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004765448394568632,
      "loss": 5.0948,
      "mean_token_accuracy": 0.19115078896284105,
      "num_tokens": 34159598.0,
      "step": 19700
    },
    {
      "entropy": 5.55718674659729,
      "epoch": 1.5331258510017505,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004765325170082237,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19157752096652986,
      "num_tokens": 34168952.0,
      "step": 19705
    },
    {
      "entropy": 5.576296806335449,
      "epoch": 1.5335148803734682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000476520191501649,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19357323050498962,
      "num_tokens": 34178409.0,
      "step": 19710
    },
    {
      "entropy": 5.645197629928589,
      "epoch": 1.5339039097451859,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047650786293732607,
      "loss": 5.1164,
      "mean_token_accuracy": 0.18972488343715668,
      "num_tokens": 34187216.0,
      "step": 19715
    },
    {
      "entropy": 5.59799599647522,
      "epoch": 1.5342929391169033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004764955313154421,
      "loss": 5.062,
      "mean_token_accuracy": 0.20320827066898345,
      "num_tokens": 34196184.0,
      "step": 19720
    },
    {
      "entropy": 5.528744316101074,
      "epoch": 1.5346819684886208,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047648319663618415,
      "loss": 5.0026,
      "mean_token_accuracy": 0.19944600909948348,
      "num_tokens": 34205016.0,
      "step": 19725
    },
    {
      "entropy": 5.555101156234741,
      "epoch": 1.5350709978603385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047647085889973936,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1935126841068268,
      "num_tokens": 34213467.0,
      "step": 19730
    },
    {
      "entropy": 5.558152532577514,
      "epoch": 1.535460027232056,
      "grad_norm": 1.125,
      "learning_rate": 0.00047645851810629503,
      "loss": 5.0134,
      "mean_token_accuracy": 0.1996410682797432,
      "num_tokens": 34222009.0,
      "step": 19735
    },
    {
      "entropy": 5.585694360733032,
      "epoch": 1.5358490566037735,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047644617425603825,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20683494806289673,
      "num_tokens": 34230160.0,
      "step": 19740
    },
    {
      "entropy": 5.567558336257934,
      "epoch": 1.5362380859754912,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004764338273491565,
      "loss": 5.1049,
      "mean_token_accuracy": 0.18704537600278853,
      "num_tokens": 34237898.0,
      "step": 19745
    },
    {
      "entropy": 5.57094292640686,
      "epoch": 1.5366271153472089,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047642147738583695,
      "loss": 5.036,
      "mean_token_accuracy": 0.19600681960582733,
      "num_tokens": 34246612.0,
      "step": 19750
    },
    {
      "entropy": 5.632332611083984,
      "epoch": 1.5370161447189261,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004764091243662671,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1923511356115341,
      "num_tokens": 34254981.0,
      "step": 19755
    },
    {
      "entropy": 5.563598680496216,
      "epoch": 1.5374051740906438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047639676829063437,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2003023460507393,
      "num_tokens": 34263834.0,
      "step": 19760
    },
    {
      "entropy": 5.567883014678955,
      "epoch": 1.5377942034623615,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047638440915912623,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19422474652528762,
      "num_tokens": 34272451.0,
      "step": 19765
    },
    {
      "entropy": 5.567710447311401,
      "epoch": 1.538183232834079,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763720469719303,
      "loss": 5.1052,
      "mean_token_accuracy": 0.19952731877565383,
      "num_tokens": 34281363.0,
      "step": 19770
    },
    {
      "entropy": 5.57563362121582,
      "epoch": 1.5385722622057965,
      "grad_norm": 1.140625,
      "learning_rate": 0.000476359681729234,
      "loss": 5.0005,
      "mean_token_accuracy": 0.195310015976429,
      "num_tokens": 34290259.0,
      "step": 19775
    },
    {
      "entropy": 5.586501693725586,
      "epoch": 1.5389612915775142,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000476347313431225,
      "loss": 5.078,
      "mean_token_accuracy": 0.1924600660800934,
      "num_tokens": 34298847.0,
      "step": 19780
    },
    {
      "entropy": 5.558987236022949,
      "epoch": 1.5393503209492316,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047633494207809096,
      "loss": 5.0447,
      "mean_token_accuracy": 0.1966763436794281,
      "num_tokens": 34307937.0,
      "step": 19785
    },
    {
      "entropy": 5.557637453079224,
      "epoch": 1.539739350320949,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047632256767001977,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2039127379655838,
      "num_tokens": 34316850.0,
      "step": 19790
    },
    {
      "entropy": 5.560368061065674,
      "epoch": 1.5401283796926668,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000476310190207199,
      "loss": 5.0403,
      "mean_token_accuracy": 0.19783806651830674,
      "num_tokens": 34325247.0,
      "step": 19795
    },
    {
      "entropy": 5.533462238311768,
      "epoch": 1.5405174090643845,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047629780968981653,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2166338473558426,
      "num_tokens": 34332789.0,
      "step": 19800
    },
    {
      "entropy": 5.587894105911255,
      "epoch": 1.540906438436102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047628542611806007,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1953209549188614,
      "num_tokens": 34340972.0,
      "step": 19805
    },
    {
      "entropy": 5.573417234420776,
      "epoch": 1.5412954678078195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047627303949211773,
      "loss": 5.0337,
      "mean_token_accuracy": 0.1979909747838974,
      "num_tokens": 34349580.0,
      "step": 19810
    },
    {
      "entropy": 5.622489166259766,
      "epoch": 1.5416844971795371,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004762606498121774,
      "loss": 5.1555,
      "mean_token_accuracy": 0.1863750398159027,
      "num_tokens": 34358971.0,
      "step": 19815
    },
    {
      "entropy": 5.584664440155029,
      "epoch": 1.5420735265512546,
      "grad_norm": 1.1875,
      "learning_rate": 0.000476248257078427,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19373337179422379,
      "num_tokens": 34368000.0,
      "step": 19820
    },
    {
      "entropy": 5.591545629501343,
      "epoch": 1.542462555922972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762358612910547,
      "loss": 5.039,
      "mean_token_accuracy": 0.1960557371377945,
      "num_tokens": 34376745.0,
      "step": 19825
    },
    {
      "entropy": 5.652286100387573,
      "epoch": 1.5428515852946898,
      "grad_norm": 1.125,
      "learning_rate": 0.0004762234624502484,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1887320265173912,
      "num_tokens": 34384635.0,
      "step": 19830
    },
    {
      "entropy": 5.637888097763062,
      "epoch": 1.5432406146664073,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047621106055619644,
      "loss": 5.1197,
      "mean_token_accuracy": 0.1959066942334175,
      "num_tokens": 34393394.0,
      "step": 19835
    },
    {
      "entropy": 5.561007070541382,
      "epoch": 1.5436296440381247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047619865560908687,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19618954062461852,
      "num_tokens": 34401343.0,
      "step": 19840
    },
    {
      "entropy": 5.597097635269165,
      "epoch": 1.5440186734098424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004761862476091079,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2001650422811508,
      "num_tokens": 34410096.0,
      "step": 19845
    },
    {
      "entropy": 5.617214775085449,
      "epoch": 1.5444077027815601,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047617383655644785,
      "loss": 5.177,
      "mean_token_accuracy": 0.18933719992637635,
      "num_tokens": 34418828.0,
      "step": 19850
    },
    {
      "entropy": 5.651484107971191,
      "epoch": 1.5447967321532776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004761614224512951,
      "loss": 5.1213,
      "mean_token_accuracy": 0.19720311164855958,
      "num_tokens": 34428061.0,
      "step": 19855
    },
    {
      "entropy": 5.582531595230103,
      "epoch": 1.545185761524995,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004761490052938379,
      "loss": 5.0449,
      "mean_token_accuracy": 0.19460816085338592,
      "num_tokens": 34436866.0,
      "step": 19860
    },
    {
      "entropy": 5.5498675346374515,
      "epoch": 1.5455747908967128,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004761365850842646,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1988269343972206,
      "num_tokens": 34445493.0,
      "step": 19865
    },
    {
      "entropy": 5.5747761726379395,
      "epoch": 1.5459638202684303,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004761241618227639,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2036134883761406,
      "num_tokens": 34454090.0,
      "step": 19870
    },
    {
      "entropy": 5.60131688117981,
      "epoch": 1.5463528496401477,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047611173550952414,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1917569160461426,
      "num_tokens": 34462410.0,
      "step": 19875
    },
    {
      "entropy": 5.5870284080505375,
      "epoch": 1.5467418790118654,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047609930614473387,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1901540771126747,
      "num_tokens": 34471288.0,
      "step": 19880
    },
    {
      "entropy": 5.600883817672729,
      "epoch": 1.547130908383583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047608687372858175,
      "loss": 5.0759,
      "mean_token_accuracy": 0.18993970155715942,
      "num_tokens": 34480250.0,
      "step": 19885
    },
    {
      "entropy": 5.634124612808227,
      "epoch": 1.5475199377553004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047607443826125633,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20465150028467177,
      "num_tokens": 34489182.0,
      "step": 19890
    },
    {
      "entropy": 5.526502323150635,
      "epoch": 1.547908967127018,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047606199974294637,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20352284908294677,
      "num_tokens": 34497474.0,
      "step": 19895
    },
    {
      "entropy": 5.563107347488403,
      "epoch": 1.5482979964987358,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004760495581738406,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19231471866369249,
      "num_tokens": 34505882.0,
      "step": 19900
    },
    {
      "entropy": 5.501498413085938,
      "epoch": 1.5486870258704533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004760371135541278,
      "loss": 4.9596,
      "mean_token_accuracy": 0.20587319284677505,
      "num_tokens": 34514939.0,
      "step": 19905
    },
    {
      "entropy": 5.547081184387207,
      "epoch": 1.5490760552421707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004760246658839967,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20251590013504028,
      "num_tokens": 34523314.0,
      "step": 19910
    },
    {
      "entropy": 5.4852393627166744,
      "epoch": 1.5494650846138884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004760122151636364,
      "loss": 4.972,
      "mean_token_accuracy": 0.20333033055067062,
      "num_tokens": 34532354.0,
      "step": 19915
    },
    {
      "entropy": 5.571914577484131,
      "epoch": 1.549854113985606,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004759997613932356,
      "loss": 5.0736,
      "mean_token_accuracy": 0.18885176330804826,
      "num_tokens": 34541150.0,
      "step": 19920
    },
    {
      "entropy": 5.559107303619385,
      "epoch": 1.5502431433573234,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047598730457298335,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1940785378217697,
      "num_tokens": 34549859.0,
      "step": 19925
    },
    {
      "entropy": 5.658868312835693,
      "epoch": 1.550632172729041,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047597484470306866,
      "loss": 5.2006,
      "mean_token_accuracy": 0.18259195238351822,
      "num_tokens": 34558327.0,
      "step": 19930
    },
    {
      "entropy": 5.615940904617309,
      "epoch": 1.5510212021007586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004759623817836806,
      "loss": 5.1204,
      "mean_token_accuracy": 0.19350760132074357,
      "num_tokens": 34567275.0,
      "step": 19935
    },
    {
      "entropy": 5.594376468658448,
      "epoch": 1.551410231472476,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004759499158150082,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20364888310432433,
      "num_tokens": 34575967.0,
      "step": 19940
    },
    {
      "entropy": 5.566449451446533,
      "epoch": 1.5517992608441937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047593744679724076,
      "loss": 5.0053,
      "mean_token_accuracy": 0.19955439120531082,
      "num_tokens": 34583997.0,
      "step": 19945
    },
    {
      "entropy": 5.547320890426636,
      "epoch": 1.5521882902159114,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047592497473056733,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2116280660033226,
      "num_tokens": 34592468.0,
      "step": 19950
    },
    {
      "entropy": 5.554736089706421,
      "epoch": 1.552577319587629,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047591249961517724,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19396212995052337,
      "num_tokens": 34601113.0,
      "step": 19955
    },
    {
      "entropy": 5.5614235401153564,
      "epoch": 1.5529663489593464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004759000214512598,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1922504112124443,
      "num_tokens": 34611027.0,
      "step": 19960
    },
    {
      "entropy": 5.505017948150635,
      "epoch": 1.553355378331064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004758875402390042,
      "loss": 4.8549,
      "mean_token_accuracy": 0.21692269444465637,
      "num_tokens": 34619881.0,
      "step": 19965
    },
    {
      "entropy": 5.502843713760376,
      "epoch": 1.5537444077027815,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047587505597859996,
      "loss": 4.9148,
      "mean_token_accuracy": 0.20295417606830596,
      "num_tokens": 34628770.0,
      "step": 19970
    },
    {
      "entropy": 5.5637979984283445,
      "epoch": 1.554133437074499,
      "grad_norm": 1.125,
      "learning_rate": 0.00047586256867023646,
      "loss": 5.0589,
      "mean_token_accuracy": 0.19705702364444733,
      "num_tokens": 34637014.0,
      "step": 19975
    },
    {
      "entropy": 5.537198495864868,
      "epoch": 1.5545224664462167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004758500783141032,
      "loss": 5.0664,
      "mean_token_accuracy": 0.19632147401571273,
      "num_tokens": 34646103.0,
      "step": 19980
    },
    {
      "entropy": 5.57470121383667,
      "epoch": 1.5549114958179344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004758375849103897,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1966407999396324,
      "num_tokens": 34654390.0,
      "step": 19985
    },
    {
      "entropy": 5.593024253845215,
      "epoch": 1.5553005251896517,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004758250884592855,
      "loss": 5.1816,
      "mean_token_accuracy": 0.18332546353340148,
      "num_tokens": 34663076.0,
      "step": 19990
    },
    {
      "entropy": 5.612808704376221,
      "epoch": 1.5556895545613694,
      "grad_norm": 1.125,
      "learning_rate": 0.00047581258896098024,
      "loss": 5.167,
      "mean_token_accuracy": 0.18644006550312042,
      "num_tokens": 34671972.0,
      "step": 19995
    },
    {
      "entropy": 5.585372877120972,
      "epoch": 1.556078583933087,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004758000864156636,
      "loss": 5.0413,
      "mean_token_accuracy": 0.19634005874395372,
      "num_tokens": 34681426.0,
      "step": 20000
    },
    {
      "entropy": 5.4840006828308105,
      "epoch": 1.5564676133048045,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047578758082352527,
      "loss": 5.0045,
      "mean_token_accuracy": 0.1981190860271454,
      "num_tokens": 34690095.0,
      "step": 20005
    },
    {
      "entropy": 5.460370206832886,
      "epoch": 1.556856642676522,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047577507218475487,
      "loss": 4.953,
      "mean_token_accuracy": 0.2045361429452896,
      "num_tokens": 34698177.0,
      "step": 20010
    },
    {
      "entropy": 5.607020425796509,
      "epoch": 1.5572456720482397,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047576256049954236,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19621413499116896,
      "num_tokens": 34706739.0,
      "step": 20015
    },
    {
      "entropy": 5.5696169376373295,
      "epoch": 1.5576347014199572,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004757500457680776,
      "loss": 5.063,
      "mean_token_accuracy": 0.19199937880039214,
      "num_tokens": 34715531.0,
      "step": 20020
    },
    {
      "entropy": 5.572199583053589,
      "epoch": 1.5580237307916747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004757375279905504,
      "loss": 5.1033,
      "mean_token_accuracy": 0.1893068566918373,
      "num_tokens": 34724049.0,
      "step": 20025
    },
    {
      "entropy": 5.519538068771363,
      "epoch": 1.5584127601633924,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047572500716715075,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2084044635295868,
      "num_tokens": 34732612.0,
      "step": 20030
    },
    {
      "entropy": 5.5287535190582275,
      "epoch": 1.55880178953511,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047571248329806855,
      "loss": 5.018,
      "mean_token_accuracy": 0.20132597535848618,
      "num_tokens": 34741500.0,
      "step": 20035
    },
    {
      "entropy": 5.53885293006897,
      "epoch": 1.5591908189068273,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047569995638349383,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19901140481233598,
      "num_tokens": 34749685.0,
      "step": 20040
    },
    {
      "entropy": 5.6017231941223145,
      "epoch": 1.559579848278545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004756874264236168,
      "loss": 5.0128,
      "mean_token_accuracy": 0.1954830378293991,
      "num_tokens": 34758463.0,
      "step": 20045
    },
    {
      "entropy": 5.562734746932984,
      "epoch": 1.5599688776502627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047567489341862753,
      "loss": 5.0287,
      "mean_token_accuracy": 0.20074363350868224,
      "num_tokens": 34767327.0,
      "step": 20050
    },
    {
      "entropy": 5.517592859268189,
      "epoch": 1.5603579070219802,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047566235736871607,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20742545574903487,
      "num_tokens": 34775453.0,
      "step": 20055
    },
    {
      "entropy": 5.53900146484375,
      "epoch": 1.5607469363936977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047564981827407277,
      "loss": 5.0303,
      "mean_token_accuracy": 0.19763220846652985,
      "num_tokens": 34784093.0,
      "step": 20060
    },
    {
      "entropy": 5.543224334716797,
      "epoch": 1.5611359657654154,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047563727613488785,
      "loss": 5.106,
      "mean_token_accuracy": 0.1996412917971611,
      "num_tokens": 34793097.0,
      "step": 20065
    },
    {
      "entropy": 5.544621849060059,
      "epoch": 1.5615249951371328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047562473095135154,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20307688862085344,
      "num_tokens": 34801833.0,
      "step": 20070
    },
    {
      "entropy": 5.63671236038208,
      "epoch": 1.5619140245088503,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004756121827236544,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19739456176757814,
      "num_tokens": 34810816.0,
      "step": 20075
    },
    {
      "entropy": 5.62411413192749,
      "epoch": 1.562303053880568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004755996314519866,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1944834604859352,
      "num_tokens": 34819523.0,
      "step": 20080
    },
    {
      "entropy": 5.595847320556641,
      "epoch": 1.5626920832522857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004755870771365387,
      "loss": 5.0661,
      "mean_token_accuracy": 0.20002146810293198,
      "num_tokens": 34827422.0,
      "step": 20085
    },
    {
      "entropy": 5.557934379577636,
      "epoch": 1.563081112624003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047557451977750113,
      "loss": 5.156,
      "mean_token_accuracy": 0.1901546373963356,
      "num_tokens": 34836474.0,
      "step": 20090
    },
    {
      "entropy": 5.506678915023803,
      "epoch": 1.5634701419957207,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755619593750645,
      "loss": 4.9239,
      "mean_token_accuracy": 0.20390913337469102,
      "num_tokens": 34845078.0,
      "step": 20095
    },
    {
      "entropy": 5.604987668991089,
      "epoch": 1.5638591713674383,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004755493959294194,
      "loss": 5.0692,
      "mean_token_accuracy": 0.19761287569999694,
      "num_tokens": 34854530.0,
      "step": 20100
    },
    {
      "entropy": 5.595241975784302,
      "epoch": 1.5642482007391558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004755368294407564,
      "loss": 5.0949,
      "mean_token_accuracy": 0.19602259993553162,
      "num_tokens": 34863925.0,
      "step": 20105
    },
    {
      "entropy": 5.593959379196167,
      "epoch": 1.5646372301108733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004755242599092662,
      "loss": 5.0382,
      "mean_token_accuracy": 0.19763870388269425,
      "num_tokens": 34872528.0,
      "step": 20110
    },
    {
      "entropy": 5.609878587722778,
      "epoch": 1.565026259482591,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047551168733513956,
      "loss": 5.03,
      "mean_token_accuracy": 0.19871195554733276,
      "num_tokens": 34881664.0,
      "step": 20115
    },
    {
      "entropy": 5.534467363357544,
      "epoch": 1.5654152888543085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047549911171856717,
      "loss": 5.0175,
      "mean_token_accuracy": 0.19365352392196655,
      "num_tokens": 34891128.0,
      "step": 20120
    },
    {
      "entropy": 5.557845592498779,
      "epoch": 1.565804318226026,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004754865330597398,
      "loss": 5.0209,
      "mean_token_accuracy": 0.19392119497060775,
      "num_tokens": 34900083.0,
      "step": 20125
    },
    {
      "entropy": 5.575722932815552,
      "epoch": 1.5661933475977436,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047547395135884854,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20190232247114182,
      "num_tokens": 34909339.0,
      "step": 20130
    },
    {
      "entropy": 5.584120464324951,
      "epoch": 1.5665823769694613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004754613666160841,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2002462163567543,
      "num_tokens": 34917350.0,
      "step": 20135
    },
    {
      "entropy": 5.579108667373657,
      "epoch": 1.5669714063411786,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047544877883163753,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19806021302938462,
      "num_tokens": 34925233.0,
      "step": 20140
    },
    {
      "entropy": 5.5688190937042235,
      "epoch": 1.5673604357128963,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047543618800569975,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19202273041009904,
      "num_tokens": 34934102.0,
      "step": 20145
    },
    {
      "entropy": 5.640570211410522,
      "epoch": 1.567749465084614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047542359413846184,
      "loss": 5.2059,
      "mean_token_accuracy": 0.1855194464325905,
      "num_tokens": 34943325.0,
      "step": 20150
    },
    {
      "entropy": 5.583875846862793,
      "epoch": 1.5681384944563315,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004754109972301149,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19533329457044601,
      "num_tokens": 34951725.0,
      "step": 20155
    },
    {
      "entropy": 5.644171524047851,
      "epoch": 1.568527523828049,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047539839728085007,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1893131360411644,
      "num_tokens": 34960068.0,
      "step": 20160
    },
    {
      "entropy": 5.604251289367676,
      "epoch": 1.5689165531997666,
      "grad_norm": 1.125,
      "learning_rate": 0.0004753857942908585,
      "loss": 5.0117,
      "mean_token_accuracy": 0.20191127061843872,
      "num_tokens": 34968400.0,
      "step": 20165
    },
    {
      "entropy": 5.571971130371094,
      "epoch": 1.5693055825714841,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004753731882603315,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19202990680933,
      "num_tokens": 34976535.0,
      "step": 20170
    },
    {
      "entropy": 5.605853271484375,
      "epoch": 1.5696946119432016,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047536057918946026,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1929836764931679,
      "num_tokens": 34985652.0,
      "step": 20175
    },
    {
      "entropy": 5.580571174621582,
      "epoch": 1.5700836413149193,
      "grad_norm": 1.125,
      "learning_rate": 0.0004753479670784361,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19444266259670256,
      "num_tokens": 34994045.0,
      "step": 20180
    },
    {
      "entropy": 5.6530585289001465,
      "epoch": 1.570472670686637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004753353519274505,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2003965273499489,
      "num_tokens": 35003143.0,
      "step": 20185
    },
    {
      "entropy": 5.601285219192505,
      "epoch": 1.5708617000583545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004753227337366948,
      "loss": 5.0954,
      "mean_token_accuracy": 0.19261547923088074,
      "num_tokens": 35012330.0,
      "step": 20190
    },
    {
      "entropy": 5.573415613174438,
      "epoch": 1.571250729430072,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047531011250636046,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20100121647119523,
      "num_tokens": 35020753.0,
      "step": 20195
    },
    {
      "entropy": 5.567460346221924,
      "epoch": 1.5716397588017896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047529748823663896,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20820355117321016,
      "num_tokens": 35029574.0,
      "step": 20200
    },
    {
      "entropy": 5.642651557922363,
      "epoch": 1.572028788173507,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004752848609277219,
      "loss": 5.1604,
      "mean_token_accuracy": 0.18785342574119568,
      "num_tokens": 35038658.0,
      "step": 20205
    },
    {
      "entropy": 5.610845375061035,
      "epoch": 1.5724178175452246,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047527223057980086,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19252055734395981,
      "num_tokens": 35046958.0,
      "step": 20210
    },
    {
      "entropy": 5.667481374740601,
      "epoch": 1.5728068469169423,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047525959719306736,
      "loss": 5.118,
      "mean_token_accuracy": 0.1984435126185417,
      "num_tokens": 35054873.0,
      "step": 20215
    },
    {
      "entropy": 5.642356204986572,
      "epoch": 1.5731958762886598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004752469607677134,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1913568302989006,
      "num_tokens": 35064696.0,
      "step": 20220
    },
    {
      "entropy": 5.558316278457641,
      "epoch": 1.5735849056603772,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004752343213039305,
      "loss": 5.1148,
      "mean_token_accuracy": 0.1941127747297287,
      "num_tokens": 35074107.0,
      "step": 20225
    },
    {
      "entropy": 5.588084173202515,
      "epoch": 1.573973935032095,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004752216788019104,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2007480412721634,
      "num_tokens": 35082583.0,
      "step": 20230
    },
    {
      "entropy": 5.595478248596192,
      "epoch": 1.5743629644038126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004752090332618451,
      "loss": 5.0773,
      "mean_token_accuracy": 0.19921378940343856,
      "num_tokens": 35091225.0,
      "step": 20235
    },
    {
      "entropy": 5.583647346496582,
      "epoch": 1.57475199377553,
      "grad_norm": 1.125,
      "learning_rate": 0.0004751963846839263,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19982710033655166,
      "num_tokens": 35099274.0,
      "step": 20240
    },
    {
      "entropy": 5.559940195083618,
      "epoch": 1.5751410231472476,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047518373306834605,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19410531967878342,
      "num_tokens": 35108515.0,
      "step": 20245
    },
    {
      "entropy": 5.620254039764404,
      "epoch": 1.5755300525189653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047517107841529626,
      "loss": 5.1244,
      "mean_token_accuracy": 0.19127087593078612,
      "num_tokens": 35118141.0,
      "step": 20250
    },
    {
      "entropy": 5.6663134574890135,
      "epoch": 1.5759190818906827,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000475158420724969,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20318174809217454,
      "num_tokens": 35126731.0,
      "step": 20255
    },
    {
      "entropy": 5.578728818893433,
      "epoch": 1.5763081112624002,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047514575999755627,
      "loss": 5.0587,
      "mean_token_accuracy": 0.19607993960380554,
      "num_tokens": 35135530.0,
      "step": 20260
    },
    {
      "entropy": 5.514973211288452,
      "epoch": 1.576697140634118,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047513309623325024,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20772783905267717,
      "num_tokens": 35143690.0,
      "step": 20265
    },
    {
      "entropy": 5.554438352584839,
      "epoch": 1.5770861700058354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004751204294322429,
      "loss": 5.032,
      "mean_token_accuracy": 0.19991711974143983,
      "num_tokens": 35152456.0,
      "step": 20270
    },
    {
      "entropy": 5.569545936584473,
      "epoch": 1.5774751993775529,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047510775959472675,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1994073897600174,
      "num_tokens": 35160750.0,
      "step": 20275
    },
    {
      "entropy": 5.5645819187164305,
      "epoch": 1.5778642287492706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750950867208937,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19553184509277344,
      "num_tokens": 35169928.0,
      "step": 20280
    },
    {
      "entropy": 5.492950677871704,
      "epoch": 1.5782532581209883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004750824108109362,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19902653843164445,
      "num_tokens": 35178187.0,
      "step": 20285
    },
    {
      "entropy": 5.551076364517212,
      "epoch": 1.5786422874927057,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004750697318650466,
      "loss": 5.0093,
      "mean_token_accuracy": 0.19380830079317093,
      "num_tokens": 35186771.0,
      "step": 20290
    },
    {
      "entropy": 5.581680250167847,
      "epoch": 1.5790313168644232,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004750570498834173,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21059935092926024,
      "num_tokens": 35194687.0,
      "step": 20295
    },
    {
      "entropy": 5.547849559783936,
      "epoch": 1.579420346236141,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004750443648662407,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20247087627649307,
      "num_tokens": 35203529.0,
      "step": 20300
    },
    {
      "entropy": 5.561748743057251,
      "epoch": 1.5798093756078584,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047503167681370924,
      "loss": 5.1139,
      "mean_token_accuracy": 0.1913180723786354,
      "num_tokens": 35212016.0,
      "step": 20305
    },
    {
      "entropy": 5.47457857131958,
      "epoch": 1.5801984049795759,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004750189857260154,
      "loss": 4.9405,
      "mean_token_accuracy": 0.20975393205881118,
      "num_tokens": 35219782.0,
      "step": 20310
    },
    {
      "entropy": 5.553012371063232,
      "epoch": 1.5805874343512936,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004750062916033519,
      "loss": 5.0853,
      "mean_token_accuracy": 0.19093867093324662,
      "num_tokens": 35228051.0,
      "step": 20315
    },
    {
      "entropy": 5.600552368164062,
      "epoch": 1.580976463723011,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004749935944459113,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19938493520021439,
      "num_tokens": 35236089.0,
      "step": 20320
    },
    {
      "entropy": 5.6133567810058596,
      "epoch": 1.5813654930947285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004749808942538862,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18891471177339553,
      "num_tokens": 35245123.0,
      "step": 20325
    },
    {
      "entropy": 5.611199522018433,
      "epoch": 1.5817545224664462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004749681910274693,
      "loss": 5.0838,
      "mean_token_accuracy": 0.19618035107851028,
      "num_tokens": 35253697.0,
      "step": 20330
    },
    {
      "entropy": 5.632259035110474,
      "epoch": 1.582143551838164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047495548476685334,
      "loss": 5.0088,
      "mean_token_accuracy": 0.20494149774312972,
      "num_tokens": 35262359.0,
      "step": 20335
    },
    {
      "entropy": 5.558789253234863,
      "epoch": 1.5825325812098814,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047494277547223125,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20229747295379638,
      "num_tokens": 35271248.0,
      "step": 20340
    },
    {
      "entropy": 5.557511043548584,
      "epoch": 1.5829216105815989,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047493006314379573,
      "loss": 5.1515,
      "mean_token_accuracy": 0.19013188481330873,
      "num_tokens": 35279641.0,
      "step": 20345
    },
    {
      "entropy": 5.630762004852295,
      "epoch": 1.5833106399533166,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047491734778173976,
      "loss": 5.0751,
      "mean_token_accuracy": 0.18951729238033294,
      "num_tokens": 35288720.0,
      "step": 20350
    },
    {
      "entropy": 5.578945064544678,
      "epoch": 1.583699669325034,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047490462938625617,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20137051343917847,
      "num_tokens": 35296909.0,
      "step": 20355
    },
    {
      "entropy": 5.566259574890137,
      "epoch": 1.5840886986967515,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047489190795753806,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19862593710422516,
      "num_tokens": 35305764.0,
      "step": 20360
    },
    {
      "entropy": 5.543924856185913,
      "epoch": 1.5844777280684692,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004748791834957784,
      "loss": 5.0347,
      "mean_token_accuracy": 0.19862818568944932,
      "num_tokens": 35314228.0,
      "step": 20365
    },
    {
      "entropy": 5.547152280807495,
      "epoch": 1.5848667574401867,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047486645600117037,
      "loss": 4.9426,
      "mean_token_accuracy": 0.20487884432077408,
      "num_tokens": 35321730.0,
      "step": 20370
    },
    {
      "entropy": 5.63907823562622,
      "epoch": 1.5852557868119042,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004748537254739068,
      "loss": 5.1975,
      "mean_token_accuracy": 0.18654660135507584,
      "num_tokens": 35330895.0,
      "step": 20375
    },
    {
      "entropy": 5.520685911178589,
      "epoch": 1.5856448161836219,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004748409919141812,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2046720176935196,
      "num_tokens": 35339499.0,
      "step": 20380
    },
    {
      "entropy": 5.566541576385498,
      "epoch": 1.5860338455553395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004748282553221865,
      "loss": 5.0466,
      "mean_token_accuracy": 0.19844391494989394,
      "num_tokens": 35347709.0,
      "step": 20385
    },
    {
      "entropy": 5.53077187538147,
      "epoch": 1.586422874927057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004748155156981162,
      "loss": 4.9113,
      "mean_token_accuracy": 0.1977192670106888,
      "num_tokens": 35356203.0,
      "step": 20390
    },
    {
      "entropy": 5.5226222515106205,
      "epoch": 1.5868119042987745,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047480277304216346,
      "loss": 5.0668,
      "mean_token_accuracy": 0.19889849722385405,
      "num_tokens": 35365850.0,
      "step": 20395
    },
    {
      "entropy": 5.529595851898193,
      "epoch": 1.5872009336704922,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047479002735452164,
      "loss": 5.0011,
      "mean_token_accuracy": 0.1998235747218132,
      "num_tokens": 35374428.0,
      "step": 20400
    },
    {
      "entropy": 5.540267276763916,
      "epoch": 1.5875899630422097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047477727863538415,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20316480994224548,
      "num_tokens": 35383012.0,
      "step": 20405
    },
    {
      "entropy": 5.629328107833862,
      "epoch": 1.5879789924139271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047476452688494444,
      "loss": 5.0977,
      "mean_token_accuracy": 0.18049991577863694,
      "num_tokens": 35392462.0,
      "step": 20410
    },
    {
      "entropy": 5.627099800109863,
      "epoch": 1.5883680217856448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000474751772103396,
      "loss": 5.0667,
      "mean_token_accuracy": 0.19105274826288224,
      "num_tokens": 35400200.0,
      "step": 20415
    },
    {
      "entropy": 5.5756693363189695,
      "epoch": 1.5887570511573625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004747390142909323,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1954803615808487,
      "num_tokens": 35408272.0,
      "step": 20420
    },
    {
      "entropy": 5.589681243896484,
      "epoch": 1.5891460805290798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047472625344774713,
      "loss": 5.0771,
      "mean_token_accuracy": 0.19289946258068086,
      "num_tokens": 35417796.0,
      "step": 20425
    },
    {
      "entropy": 5.576386022567749,
      "epoch": 1.5895351099007975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047471348957403374,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1920172393321991,
      "num_tokens": 35426173.0,
      "step": 20430
    },
    {
      "entropy": 5.664482259750367,
      "epoch": 1.5899241392725152,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004747007226699862,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1942828893661499,
      "num_tokens": 35434972.0,
      "step": 20435
    },
    {
      "entropy": 5.671311235427856,
      "epoch": 1.5903131686442327,
      "grad_norm": 1.125,
      "learning_rate": 0.00047468795273579803,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19163405001163483,
      "num_tokens": 35442824.0,
      "step": 20440
    },
    {
      "entropy": 5.551054000854492,
      "epoch": 1.5907021980159501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004746751797716629,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19766245782375336,
      "num_tokens": 35451519.0,
      "step": 20445
    },
    {
      "entropy": 5.568986129760742,
      "epoch": 1.5910912273876678,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004746624037777748,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20214746594429017,
      "num_tokens": 35459473.0,
      "step": 20450
    },
    {
      "entropy": 5.5514037132263185,
      "epoch": 1.5914802567593853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047464962475432754,
      "loss": 5.0925,
      "mean_token_accuracy": 0.19424415230751038,
      "num_tokens": 35468648.0,
      "step": 20455
    },
    {
      "entropy": 5.5837212085723875,
      "epoch": 1.5918692861311028,
      "grad_norm": 1.25,
      "learning_rate": 0.000474636842701515,
      "loss": 4.9846,
      "mean_token_accuracy": 0.20390194654464722,
      "num_tokens": 35477626.0,
      "step": 20460
    },
    {
      "entropy": 5.46216778755188,
      "epoch": 1.5922583155028205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004746240576195311,
      "loss": 4.8704,
      "mean_token_accuracy": 0.2154863804578781,
      "num_tokens": 35486116.0,
      "step": 20465
    },
    {
      "entropy": 5.581904697418213,
      "epoch": 1.5926473448745382,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047461126950856987,
      "loss": 5.0767,
      "mean_token_accuracy": 0.19188373684883117,
      "num_tokens": 35495860.0,
      "step": 20470
    },
    {
      "entropy": 5.582309436798096,
      "epoch": 1.5930363742462554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004745984783688253,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2064759075641632,
      "num_tokens": 35504306.0,
      "step": 20475
    },
    {
      "entropy": 5.574326372146606,
      "epoch": 1.5934254036179731,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047458568420049153,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19911552220582962,
      "num_tokens": 35513585.0,
      "step": 20480
    },
    {
      "entropy": 5.532439565658569,
      "epoch": 1.5938144329896908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047457288700376274,
      "loss": 4.999,
      "mean_token_accuracy": 0.19503778219223022,
      "num_tokens": 35521929.0,
      "step": 20485
    },
    {
      "entropy": 5.575844430923462,
      "epoch": 1.5942034623614083,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047456008677883304,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20033130496740342,
      "num_tokens": 35530498.0,
      "step": 20490
    },
    {
      "entropy": 5.534327125549316,
      "epoch": 1.5945924917331258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004745472835258966,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2025996670126915,
      "num_tokens": 35538958.0,
      "step": 20495
    },
    {
      "entropy": 5.604906129837036,
      "epoch": 1.5949815211048435,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047453447724514773,
      "loss": 5.1136,
      "mean_token_accuracy": 0.18190063238143922,
      "num_tokens": 35548488.0,
      "step": 20500
    },
    {
      "entropy": 5.652164602279663,
      "epoch": 1.595370550476561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745216679367808,
      "loss": 5.0965,
      "mean_token_accuracy": 0.19454507678747177,
      "num_tokens": 35557374.0,
      "step": 20505
    },
    {
      "entropy": 5.5773053646087645,
      "epoch": 1.5957595798482784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745088556009901,
      "loss": 5.0517,
      "mean_token_accuracy": 0.19406992048025132,
      "num_tokens": 35565521.0,
      "step": 20510
    },
    {
      "entropy": 5.519552707672119,
      "epoch": 1.5961486092199961,
      "grad_norm": 1.125,
      "learning_rate": 0.0004744960402379701,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19322532862424852,
      "num_tokens": 35574833.0,
      "step": 20515
    },
    {
      "entropy": 5.550961446762085,
      "epoch": 1.5965376385917138,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047448322184791525,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19596573561429978,
      "num_tokens": 35583226.0,
      "step": 20520
    },
    {
      "entropy": 5.583236217498779,
      "epoch": 1.596926667963431,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047447040043101994,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2041420355439186,
      "num_tokens": 35591366.0,
      "step": 20525
    },
    {
      "entropy": 5.620168209075928,
      "epoch": 1.5973156973351488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047445757598747886,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1894078090786934,
      "num_tokens": 35599244.0,
      "step": 20530
    },
    {
      "entropy": 5.595584726333618,
      "epoch": 1.5977047267068665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744447485174864,
      "loss": 5.0731,
      "mean_token_accuracy": 0.1870184689760208,
      "num_tokens": 35608132.0,
      "step": 20535
    },
    {
      "entropy": 5.584169673919678,
      "epoch": 1.598093756078584,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047443191802123746,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19617429375648499,
      "num_tokens": 35616459.0,
      "step": 20540
    },
    {
      "entropy": 5.61034460067749,
      "epoch": 1.5984827854503014,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047441908449892664,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19541900753974914,
      "num_tokens": 35624802.0,
      "step": 20545
    },
    {
      "entropy": 5.600928783416748,
      "epoch": 1.5988718148220191,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047440624795074855,
      "loss": 5.0514,
      "mean_token_accuracy": 0.19177628606557845,
      "num_tokens": 35632787.0,
      "step": 20550
    },
    {
      "entropy": 5.547068548202515,
      "epoch": 1.5992608441937366,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047439340837689804,
      "loss": 5.0535,
      "mean_token_accuracy": 0.20551725924015046,
      "num_tokens": 35641848.0,
      "step": 20555
    },
    {
      "entropy": 5.56574239730835,
      "epoch": 1.599649873565454,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047438056577756986,
      "loss": 5.0251,
      "mean_token_accuracy": 0.2022228553891182,
      "num_tokens": 35650502.0,
      "step": 20560
    },
    {
      "entropy": 5.5688553810119625,
      "epoch": 1.6000389029371718,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047436772015295903,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20269134640693665,
      "num_tokens": 35658454.0,
      "step": 20565
    },
    {
      "entropy": 5.583420276641846,
      "epoch": 1.6004279323088895,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047435487150326036,
      "loss": 5.0964,
      "mean_token_accuracy": 0.19134222865104675,
      "num_tokens": 35667235.0,
      "step": 20570
    },
    {
      "entropy": 5.619916820526123,
      "epoch": 1.6008169616806067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004743420198286688,
      "loss": 5.0745,
      "mean_token_accuracy": 0.2064508840441704,
      "num_tokens": 35675671.0,
      "step": 20575
    },
    {
      "entropy": 5.632651948928833,
      "epoch": 1.6012059910523244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047432916512937936,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2008173644542694,
      "num_tokens": 35684949.0,
      "step": 20580
    },
    {
      "entropy": 5.519704437255859,
      "epoch": 1.6015950204240421,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004743163074055871,
      "loss": 4.9976,
      "mean_token_accuracy": 0.19276344329118728,
      "num_tokens": 35693241.0,
      "step": 20585
    },
    {
      "entropy": 5.553729057312012,
      "epoch": 1.6019840497957596,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004743034466574871,
      "loss": 5.077,
      "mean_token_accuracy": 0.1966675415635109,
      "num_tokens": 35702510.0,
      "step": 20590
    },
    {
      "entropy": 5.58604588508606,
      "epoch": 1.602373079167477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004742905828852746,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20207636058330536,
      "num_tokens": 35711779.0,
      "step": 20595
    },
    {
      "entropy": 5.607494735717774,
      "epoch": 1.6027621085391948,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004742777160891447,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1886927992105484,
      "num_tokens": 35719848.0,
      "step": 20600
    },
    {
      "entropy": 5.599776697158814,
      "epoch": 1.6031511379109122,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004742648462692926,
      "loss": 5.1963,
      "mean_token_accuracy": 0.18366478979587555,
      "num_tokens": 35728762.0,
      "step": 20605
    },
    {
      "entropy": 5.565124034881592,
      "epoch": 1.6035401672826297,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047425197342591363,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20134300887584686,
      "num_tokens": 35737811.0,
      "step": 20610
    },
    {
      "entropy": 5.677921915054322,
      "epoch": 1.6039291966543474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004742390975592031,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19549560099840163,
      "num_tokens": 35746470.0,
      "step": 20615
    },
    {
      "entropy": 5.574686050415039,
      "epoch": 1.604318226026065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047422621866935645,
      "loss": 5.097,
      "mean_token_accuracy": 0.19594002962112428,
      "num_tokens": 35756000.0,
      "step": 20620
    },
    {
      "entropy": 5.50658483505249,
      "epoch": 1.6047072553977826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000474213336756569,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20368831157684325,
      "num_tokens": 35764781.0,
      "step": 20625
    },
    {
      "entropy": 5.531512594223022,
      "epoch": 1.6050962847695,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004742004518210362,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19779135435819625,
      "num_tokens": 35772845.0,
      "step": 20630
    },
    {
      "entropy": 5.641934061050415,
      "epoch": 1.6054853141412178,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004741875638629536,
      "loss": 5.0794,
      "mean_token_accuracy": 0.19366559982299805,
      "num_tokens": 35781863.0,
      "step": 20635
    },
    {
      "entropy": 5.559184741973877,
      "epoch": 1.6058743435129352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004741746728825168,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20290449559688567,
      "num_tokens": 35791321.0,
      "step": 20640
    },
    {
      "entropy": 5.529626131057739,
      "epoch": 1.6062633728846527,
      "grad_norm": 1.125,
      "learning_rate": 0.0004741617788799213,
      "loss": 4.9243,
      "mean_token_accuracy": 0.20864196717739106,
      "num_tokens": 35799643.0,
      "step": 20645
    },
    {
      "entropy": 5.584804058074951,
      "epoch": 1.6066524022563704,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047414888185536285,
      "loss": 5.0846,
      "mean_token_accuracy": 0.19097966849803924,
      "num_tokens": 35808382.0,
      "step": 20650
    },
    {
      "entropy": 5.566513156890869,
      "epoch": 1.6070414316280879,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004741359818090371,
      "loss": 5.0207,
      "mean_token_accuracy": 0.1968013659119606,
      "num_tokens": 35817110.0,
      "step": 20655
    },
    {
      "entropy": 5.522380828857422,
      "epoch": 1.6074304609998054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047412307874113976,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21003714203834534,
      "num_tokens": 35824590.0,
      "step": 20660
    },
    {
      "entropy": 5.579947137832642,
      "epoch": 1.607819490371523,
      "grad_norm": 1.125,
      "learning_rate": 0.0004741101726518667,
      "loss": 5.0699,
      "mean_token_accuracy": 0.19348026365041732,
      "num_tokens": 35833391.0,
      "step": 20665
    },
    {
      "entropy": 5.62249755859375,
      "epoch": 1.6082085197432407,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004740972635414136,
      "loss": 5.1071,
      "mean_token_accuracy": 0.19125356674194335,
      "num_tokens": 35842586.0,
      "step": 20670
    },
    {
      "entropy": 5.572748422622681,
      "epoch": 1.6085975491149582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004740843514099765,
      "loss": 5.062,
      "mean_token_accuracy": 0.1981152266263962,
      "num_tokens": 35850973.0,
      "step": 20675
    },
    {
      "entropy": 5.593725633621216,
      "epoch": 1.6089865784866757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004740714362577512,
      "loss": 5.0847,
      "mean_token_accuracy": 0.1918594181537628,
      "num_tokens": 35859686.0,
      "step": 20680
    },
    {
      "entropy": 5.577735090255738,
      "epoch": 1.6093756078583934,
      "grad_norm": 1.125,
      "learning_rate": 0.00047405851808493364,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20335650593042373,
      "num_tokens": 35868161.0,
      "step": 20685
    },
    {
      "entropy": 5.5351881980896,
      "epoch": 1.6097646372301109,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047404559689172006,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1905168354511261,
      "num_tokens": 35876431.0,
      "step": 20690
    },
    {
      "entropy": 5.585420322418213,
      "epoch": 1.6101536666018283,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047403267267830617,
      "loss": 5.0375,
      "mean_token_accuracy": 0.19040782153606414,
      "num_tokens": 35884598.0,
      "step": 20695
    },
    {
      "entropy": 5.669983386993408,
      "epoch": 1.610542695973546,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047401974544488844,
      "loss": 5.1467,
      "mean_token_accuracy": 0.18747070133686067,
      "num_tokens": 35893395.0,
      "step": 20700
    },
    {
      "entropy": 5.646161699295044,
      "epoch": 1.6109317253452635,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004740068151916628,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18957708925008773,
      "num_tokens": 35902043.0,
      "step": 20705
    },
    {
      "entropy": 5.638747978210449,
      "epoch": 1.611320754716981,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004739938819188255,
      "loss": 5.1268,
      "mean_token_accuracy": 0.19207285791635514,
      "num_tokens": 35911514.0,
      "step": 20710
    },
    {
      "entropy": 5.643983602523804,
      "epoch": 1.6117097840886987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004739809456265727,
      "loss": 5.0784,
      "mean_token_accuracy": 0.19706311523914338,
      "num_tokens": 35919777.0,
      "step": 20715
    },
    {
      "entropy": 5.62511420249939,
      "epoch": 1.6120988134604164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004739680063151008,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20109447538852693,
      "num_tokens": 35929058.0,
      "step": 20720
    },
    {
      "entropy": 5.508691501617432,
      "epoch": 1.6124878428321339,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004739550639846061,
      "loss": 4.965,
      "mean_token_accuracy": 0.20061279684305192,
      "num_tokens": 35937855.0,
      "step": 20725
    },
    {
      "entropy": 5.5391318798065186,
      "epoch": 1.6128768722038513,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047394211863528496,
      "loss": 5.0956,
      "mean_token_accuracy": 0.19917786717414857,
      "num_tokens": 35946472.0,
      "step": 20730
    },
    {
      "entropy": 5.595883464813232,
      "epoch": 1.613265901575569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004739291702673338,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19751307368278503,
      "num_tokens": 35954876.0,
      "step": 20735
    },
    {
      "entropy": 5.652703428268433,
      "epoch": 1.6136549309472865,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004739162188809492,
      "loss": 5.1157,
      "mean_token_accuracy": 0.18689001202583314,
      "num_tokens": 35964525.0,
      "step": 20740
    },
    {
      "entropy": 5.570595741271973,
      "epoch": 1.614043960319004,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047390326447632747,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19348090291023254,
      "num_tokens": 35973172.0,
      "step": 20745
    },
    {
      "entropy": 5.643519687652588,
      "epoch": 1.6144329896907217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004738903070536654,
      "loss": 5.1505,
      "mean_token_accuracy": 0.19644696712493898,
      "num_tokens": 35982181.0,
      "step": 20750
    },
    {
      "entropy": 5.641771078109741,
      "epoch": 1.6148220190624392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004738773466131594,
      "loss": 5.1305,
      "mean_token_accuracy": 0.18445833325386046,
      "num_tokens": 35991511.0,
      "step": 20755
    },
    {
      "entropy": 5.634194564819336,
      "epoch": 1.6152110484341566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738643831550063,
      "loss": 5.106,
      "mean_token_accuracy": 0.1961176335811615,
      "num_tokens": 36000798.0,
      "step": 20760
    },
    {
      "entropy": 5.570387887954712,
      "epoch": 1.6156000778058743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004738514166794026,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20635077208280564,
      "num_tokens": 36009779.0,
      "step": 20765
    },
    {
      "entropy": 5.61924557685852,
      "epoch": 1.615989107177592,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047383844718654525,
      "loss": 5.06,
      "mean_token_accuracy": 0.1982255145907402,
      "num_tokens": 36017752.0,
      "step": 20770
    },
    {
      "entropy": 5.5329670906066895,
      "epoch": 1.6163781365493095,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004738254746766309,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20671554803848266,
      "num_tokens": 36026618.0,
      "step": 20775
    },
    {
      "entropy": 5.571297025680542,
      "epoch": 1.616767165921027,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004738124991498565,
      "loss": 5.0272,
      "mean_token_accuracy": 0.19699340015649797,
      "num_tokens": 36035651.0,
      "step": 20780
    },
    {
      "entropy": 5.541163969039917,
      "epoch": 1.6171561952927447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047379952060641866,
      "loss": 5.0274,
      "mean_token_accuracy": 0.19539180845022203,
      "num_tokens": 36044346.0,
      "step": 20785
    },
    {
      "entropy": 5.563560104370117,
      "epoch": 1.6175452246644622,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047378653904651475,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20308210998773574,
      "num_tokens": 36052895.0,
      "step": 20790
    },
    {
      "entropy": 5.615769290924073,
      "epoch": 1.6179342540361796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047377355447034136,
      "loss": 5.0307,
      "mean_token_accuracy": 0.1948792889714241,
      "num_tokens": 36061186.0,
      "step": 20795
    },
    {
      "entropy": 5.631786251068116,
      "epoch": 1.6183232834078973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737605668780958,
      "loss": 5.1442,
      "mean_token_accuracy": 0.1893509343266487,
      "num_tokens": 36069540.0,
      "step": 20800
    },
    {
      "entropy": 5.6162073612213135,
      "epoch": 1.618712312779615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047374757626997494,
      "loss": 5.0386,
      "mean_token_accuracy": 0.19843841344118118,
      "num_tokens": 36078241.0,
      "step": 20805
    },
    {
      "entropy": 5.5599769115447994,
      "epoch": 1.6191013421513323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737345826461759,
      "loss": 5.0837,
      "mean_token_accuracy": 0.19560853242874146,
      "num_tokens": 36087253.0,
      "step": 20810
    },
    {
      "entropy": 5.569096565246582,
      "epoch": 1.61949037152305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737215860068959,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2054625391960144,
      "num_tokens": 36095361.0,
      "step": 20815
    },
    {
      "entropy": 5.561181497573853,
      "epoch": 1.6198794008947677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047370858635233223,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19793290197849273,
      "num_tokens": 36104104.0,
      "step": 20820
    },
    {
      "entropy": 5.5188267707824705,
      "epoch": 1.6202684302664851,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047369558368268194,
      "loss": 4.928,
      "mean_token_accuracy": 0.20316271930933,
      "num_tokens": 36112636.0,
      "step": 20825
    },
    {
      "entropy": 5.566997385025024,
      "epoch": 1.6206574596382026,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004736825779981424,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2044930264353752,
      "num_tokens": 36121458.0,
      "step": 20830
    },
    {
      "entropy": 5.592270135879517,
      "epoch": 1.6210464890099203,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004736695692989111,
      "loss": 5.0514,
      "mean_token_accuracy": 0.1944114461541176,
      "num_tokens": 36129541.0,
      "step": 20835
    },
    {
      "entropy": 5.530717897415161,
      "epoch": 1.6214355183816378,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004736565575851852,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20032714903354645,
      "num_tokens": 36138118.0,
      "step": 20840
    },
    {
      "entropy": 5.646763706207276,
      "epoch": 1.6218245477533553,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004736435428571622,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19243379682302475,
      "num_tokens": 36147107.0,
      "step": 20845
    },
    {
      "entropy": 5.565678977966309,
      "epoch": 1.622213577125073,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004736305251150397,
      "loss": 5.0549,
      "mean_token_accuracy": 0.18922393918037414,
      "num_tokens": 36155353.0,
      "step": 20850
    },
    {
      "entropy": 5.610718870162964,
      "epoch": 1.6226026064967907,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004736175043590151,
      "loss": 5.129,
      "mean_token_accuracy": 0.19976239502429963,
      "num_tokens": 36163928.0,
      "step": 20855
    },
    {
      "entropy": 5.5744835376739506,
      "epoch": 1.622991635868508,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000473604480589286,
      "loss": 4.9268,
      "mean_token_accuracy": 0.20720416605472564,
      "num_tokens": 36172601.0,
      "step": 20860
    },
    {
      "entropy": 5.57391676902771,
      "epoch": 1.6233806652402256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047359145380605007,
      "loss": 5.0217,
      "mean_token_accuracy": 0.1961013361811638,
      "num_tokens": 36182098.0,
      "step": 20865
    },
    {
      "entropy": 5.546414852142334,
      "epoch": 1.6237696946119433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004735784240095049,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20005569905042647,
      "num_tokens": 36191022.0,
      "step": 20870
    },
    {
      "entropy": 5.538806295394897,
      "epoch": 1.6241587239836608,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047356539119984813,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20235928297042846,
      "num_tokens": 36200229.0,
      "step": 20875
    },
    {
      "entropy": 5.567064332962036,
      "epoch": 1.6245477533553783,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004735523553772777,
      "loss": 5.0873,
      "mean_token_accuracy": 0.196621435880661,
      "num_tokens": 36208660.0,
      "step": 20880
    },
    {
      "entropy": 5.539689397811889,
      "epoch": 1.624936782727096,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004735393165419912,
      "loss": 5.0244,
      "mean_token_accuracy": 0.1990366369485855,
      "num_tokens": 36217034.0,
      "step": 20885
    },
    {
      "entropy": 5.601562881469727,
      "epoch": 1.6253258120988134,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047352627469418666,
      "loss": 5.0403,
      "mean_token_accuracy": 0.19395889043807985,
      "num_tokens": 36225653.0,
      "step": 20890
    },
    {
      "entropy": 5.593519020080566,
      "epoch": 1.625714841470531,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004735132298340619,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2052724242210388,
      "num_tokens": 36234449.0,
      "step": 20895
    },
    {
      "entropy": 5.544824314117432,
      "epoch": 1.6261038708422486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004735001819618148,
      "loss": 4.9995,
      "mean_token_accuracy": 0.19994664937257767,
      "num_tokens": 36243565.0,
      "step": 20900
    },
    {
      "entropy": 5.598747253417969,
      "epoch": 1.6264929002139663,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004734871310776434,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20233528912067414,
      "num_tokens": 36252848.0,
      "step": 20905
    },
    {
      "entropy": 5.553025960922241,
      "epoch": 1.6268819295856836,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004734740771817457,
      "loss": 5.0318,
      "mean_token_accuracy": 0.19314651787281037,
      "num_tokens": 36262125.0,
      "step": 20910
    },
    {
      "entropy": 5.589990663528442,
      "epoch": 1.6272709589574013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004734610202743198,
      "loss": 5.1331,
      "mean_token_accuracy": 0.1954658016562462,
      "num_tokens": 36271466.0,
      "step": 20915
    },
    {
      "entropy": 5.549453210830689,
      "epoch": 1.627659988329119,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734479603555638,
      "loss": 4.96,
      "mean_token_accuracy": 0.20164926052093507,
      "num_tokens": 36280207.0,
      "step": 20920
    },
    {
      "entropy": 5.617452001571655,
      "epoch": 1.6280490177008364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047343489742567593,
      "loss": 5.1297,
      "mean_token_accuracy": 0.19514364004135132,
      "num_tokens": 36288343.0,
      "step": 20925
    },
    {
      "entropy": 5.6378966808319095,
      "epoch": 1.628438047072554,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047342183148485424,
      "loss": 5.0164,
      "mean_token_accuracy": 0.19918595850467682,
      "num_tokens": 36296679.0,
      "step": 20930
    },
    {
      "entropy": 5.5936079025268555,
      "epoch": 1.6288270764442716,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047340876253329706,
      "loss": 4.9874,
      "mean_token_accuracy": 0.19952846467494964,
      "num_tokens": 36305806.0,
      "step": 20935
    },
    {
      "entropy": 5.497388696670532,
      "epoch": 1.629216105815989,
      "grad_norm": 1.125,
      "learning_rate": 0.00047339569057120275,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20910928696393966,
      "num_tokens": 36314861.0,
      "step": 20940
    },
    {
      "entropy": 5.59879059791565,
      "epoch": 1.6296051351877066,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047338261559876966,
      "loss": 5.0315,
      "mean_token_accuracy": 0.1960493341088295,
      "num_tokens": 36323611.0,
      "step": 20945
    },
    {
      "entropy": 5.633615779876709,
      "epoch": 1.6299941645594243,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047336953761619604,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1924738734960556,
      "num_tokens": 36331785.0,
      "step": 20950
    },
    {
      "entropy": 5.483160448074341,
      "epoch": 1.630383193931142,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047335645662368046,
      "loss": 4.954,
      "mean_token_accuracy": 0.20160435438156127,
      "num_tokens": 36340202.0,
      "step": 20955
    },
    {
      "entropy": 5.570060062408447,
      "epoch": 1.6307722233028592,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004733433726214214,
      "loss": 5.0245,
      "mean_token_accuracy": 0.19779227375984193,
      "num_tokens": 36348227.0,
      "step": 20960
    },
    {
      "entropy": 5.600245332717895,
      "epoch": 1.631161252674577,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047333028560961733,
      "loss": 5.0981,
      "mean_token_accuracy": 0.19460875391960145,
      "num_tokens": 36357801.0,
      "step": 20965
    },
    {
      "entropy": 5.590369367599488,
      "epoch": 1.6315502820462946,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047331719558846687,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20001079142093658,
      "num_tokens": 36366147.0,
      "step": 20970
    },
    {
      "entropy": 5.631360483169556,
      "epoch": 1.631939311418012,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047330410255816863,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1919443652033806,
      "num_tokens": 36374637.0,
      "step": 20975
    },
    {
      "entropy": 5.5797699928283695,
      "epoch": 1.6323283407897295,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047329100651892123,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2002417951822281,
      "num_tokens": 36383911.0,
      "step": 20980
    },
    {
      "entropy": 5.637818908691406,
      "epoch": 1.6327173701614472,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004732779074709234,
      "loss": 5.0247,
      "mean_token_accuracy": 0.19771524220705033,
      "num_tokens": 36392557.0,
      "step": 20985
    },
    {
      "entropy": 5.557787990570068,
      "epoch": 1.6331063995331647,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000473264805414374,
      "loss": 4.9888,
      "mean_token_accuracy": 0.19780325442552565,
      "num_tokens": 36400845.0,
      "step": 20990
    },
    {
      "entropy": 5.521533870697022,
      "epoch": 1.6334954289048822,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047325170034947167,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20563892126083375,
      "num_tokens": 36409360.0,
      "step": 20995
    },
    {
      "entropy": 5.518873596191407,
      "epoch": 1.6338844582766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047323859227641537,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2110530987381935,
      "num_tokens": 36418638.0,
      "step": 21000
    },
    {
      "epoch": 1.6338844582766,
      "eval_entropy": 5.358531058718082,
      "eval_loss": 5.107507705688477,
      "eval_mean_token_accuracy": 0.20434006118682627,
      "eval_num_tokens": 36418638.0,
      "eval_runtime": 21.6334,
      "eval_samples_per_second": 1921.01,
      "eval_steps_per_second": 240.138,
      "step": 21000
    },
    {
      "entropy": 5.5983349800109865,
      "epoch": 1.6342734876483176,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047322548119540397,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19308139532804489,
      "num_tokens": 36427571.0,
      "step": 21005
    },
    {
      "entropy": 5.6030237674713135,
      "epoch": 1.634662517020035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047321236710663636,
      "loss": 5.0571,
      "mean_token_accuracy": 0.19659548997879028,
      "num_tokens": 36436694.0,
      "step": 21010
    },
    {
      "entropy": 5.554193592071533,
      "epoch": 1.6350515463917525,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047319925001031165,
      "loss": 4.9789,
      "mean_token_accuracy": 0.20792436599731445,
      "num_tokens": 36445318.0,
      "step": 21015
    },
    {
      "entropy": 5.5952190399169925,
      "epoch": 1.6354405757634702,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004731861299066287,
      "loss": 5.0573,
      "mean_token_accuracy": 0.19558667838573457,
      "num_tokens": 36454116.0,
      "step": 21020
    },
    {
      "entropy": 5.531690549850464,
      "epoch": 1.6358296051351877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004731730067957867,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21292848736047745,
      "num_tokens": 36462937.0,
      "step": 21025
    },
    {
      "entropy": 5.567939949035645,
      "epoch": 1.6362186345069052,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047315988067798476,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20617568790912627,
      "num_tokens": 36471650.0,
      "step": 21030
    },
    {
      "entropy": 5.619936609268189,
      "epoch": 1.6366076638786229,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000473146751553422,
      "loss": 5.2007,
      "mean_token_accuracy": 0.1893467053771019,
      "num_tokens": 36480810.0,
      "step": 21035
    },
    {
      "entropy": 5.586515045166015,
      "epoch": 1.6369966932503404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004731336194222978,
      "loss": 5.0017,
      "mean_token_accuracy": 0.1984693005681038,
      "num_tokens": 36489698.0,
      "step": 21040
    },
    {
      "entropy": 5.571409702301025,
      "epoch": 1.6373857226220578,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047312048428481114,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2075487717986107,
      "num_tokens": 36498868.0,
      "step": 21045
    },
    {
      "entropy": 5.487791252136231,
      "epoch": 1.6377747519937755,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047310734614116153,
      "loss": 4.8893,
      "mean_token_accuracy": 0.205226269364357,
      "num_tokens": 36507447.0,
      "step": 21050
    },
    {
      "entropy": 5.577247571945191,
      "epoch": 1.6381637813654932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004730942049915483,
      "loss": 5.1874,
      "mean_token_accuracy": 0.19133370071649553,
      "num_tokens": 36516299.0,
      "step": 21055
    },
    {
      "entropy": 5.631215286254883,
      "epoch": 1.6385528107372107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004730810608361707,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20634507685899733,
      "num_tokens": 36524426.0,
      "step": 21060
    },
    {
      "entropy": 5.679188013076782,
      "epoch": 1.6389418401089282,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047306791367522833,
      "loss": 5.1082,
      "mean_token_accuracy": 0.1926860511302948,
      "num_tokens": 36533566.0,
      "step": 21065
    },
    {
      "entropy": 5.637728071212768,
      "epoch": 1.6393308694806459,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004730547635089206,
      "loss": 5.1296,
      "mean_token_accuracy": 0.19646978080272676,
      "num_tokens": 36541728.0,
      "step": 21070
    },
    {
      "entropy": 5.610584163665772,
      "epoch": 1.6397198988523634,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047304161033744714,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20313474237918855,
      "num_tokens": 36549196.0,
      "step": 21075
    },
    {
      "entropy": 5.578542757034302,
      "epoch": 1.6401089282240808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004730284541610075,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20234542936086655,
      "num_tokens": 36557354.0,
      "step": 21080
    },
    {
      "entropy": 5.482127857208252,
      "epoch": 1.6404979575957985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004730152949798012,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20173519551754,
      "num_tokens": 36566657.0,
      "step": 21085
    },
    {
      "entropy": 5.587358236312866,
      "epoch": 1.640886986967516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047300213279402787,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20067018568515776,
      "num_tokens": 36574313.0,
      "step": 21090
    },
    {
      "entropy": 5.647541332244873,
      "epoch": 1.6412760163392335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047298896760388745,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20562151223421096,
      "num_tokens": 36582158.0,
      "step": 21095
    },
    {
      "entropy": 5.542970418930054,
      "epoch": 1.6416650457109512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004729757994095796,
      "loss": 4.9903,
      "mean_token_accuracy": 0.20446429997682572,
      "num_tokens": 36590283.0,
      "step": 21100
    },
    {
      "entropy": 5.58388032913208,
      "epoch": 1.6420540750826689,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047296262821130405,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19780693650245668,
      "num_tokens": 36599571.0,
      "step": 21105
    },
    {
      "entropy": 5.628580713272095,
      "epoch": 1.6424431044543863,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047294945400926065,
      "loss": 5.073,
      "mean_token_accuracy": 0.19892978221178054,
      "num_tokens": 36608177.0,
      "step": 21110
    },
    {
      "entropy": 5.587074708938599,
      "epoch": 1.6428321338261038,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004729362768036494,
      "loss": 5.0731,
      "mean_token_accuracy": 0.1984366476535797,
      "num_tokens": 36617305.0,
      "step": 21115
    },
    {
      "entropy": 5.569946670532227,
      "epoch": 1.6432211631978215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047292309659467024,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2070159614086151,
      "num_tokens": 36625878.0,
      "step": 21120
    },
    {
      "entropy": 5.652110242843628,
      "epoch": 1.643610192569539,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004729099133825231,
      "loss": 5.1062,
      "mean_token_accuracy": 0.1925281912088394,
      "num_tokens": 36635728.0,
      "step": 21125
    },
    {
      "entropy": 5.639533758163452,
      "epoch": 1.6439992219412565,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472896727167408,
      "loss": 5.1501,
      "mean_token_accuracy": 0.19341049790382386,
      "num_tokens": 36644571.0,
      "step": 21130
    },
    {
      "entropy": 5.579378366470337,
      "epoch": 1.6443882513129742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047288353794952505,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20377596318721772,
      "num_tokens": 36653582.0,
      "step": 21135
    },
    {
      "entropy": 5.579152631759643,
      "epoch": 1.6447772806846916,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004728703457290744,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20910850167274475,
      "num_tokens": 36662872.0,
      "step": 21140
    },
    {
      "entropy": 5.565367650985718,
      "epoch": 1.6451663100564091,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004728571505062562,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20111654400825502,
      "num_tokens": 36671952.0,
      "step": 21145
    },
    {
      "entropy": 5.67355489730835,
      "epoch": 1.6455553394281268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004728439522812707,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1883334219455719,
      "num_tokens": 36681262.0,
      "step": 21150
    },
    {
      "entropy": 5.592899942398072,
      "epoch": 1.6459443687998445,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047283075105431806,
      "loss": 4.9577,
      "mean_token_accuracy": 0.20872252136468888,
      "num_tokens": 36689094.0,
      "step": 21155
    },
    {
      "entropy": 5.517189407348633,
      "epoch": 1.646333398171562,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004728175468255987,
      "loss": 4.937,
      "mean_token_accuracy": 0.2046843022108078,
      "num_tokens": 36697043.0,
      "step": 21160
    },
    {
      "entropy": 5.588912153244019,
      "epoch": 1.6467224275432795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047280433959531287,
      "loss": 5.132,
      "mean_token_accuracy": 0.19324016124010085,
      "num_tokens": 36705561.0,
      "step": 21165
    },
    {
      "entropy": 5.629413270950318,
      "epoch": 1.6471114569149972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472791129363661,
      "loss": 5.124,
      "mean_token_accuracy": 0.18882680535316468,
      "num_tokens": 36714973.0,
      "step": 21170
    },
    {
      "entropy": 5.626101970672607,
      "epoch": 1.6475004862867146,
      "grad_norm": 1.125,
      "learning_rate": 0.0004727779161308436,
      "loss": 5.1051,
      "mean_token_accuracy": 0.19349679499864578,
      "num_tokens": 36723218.0,
      "step": 21175
    },
    {
      "entropy": 5.617767095565796,
      "epoch": 1.6478895156584321,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004727646998970612,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20967632234096528,
      "num_tokens": 36731883.0,
      "step": 21180
    },
    {
      "entropy": 5.591440582275391,
      "epoch": 1.6482785450301498,
      "grad_norm": 1.125,
      "learning_rate": 0.00047275148066251417,
      "loss": 5.0906,
      "mean_token_accuracy": 0.18939222246408463,
      "num_tokens": 36739984.0,
      "step": 21185
    },
    {
      "entropy": 5.511768817901611,
      "epoch": 1.6486675744018673,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004727382584274032,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2095464289188385,
      "num_tokens": 36748551.0,
      "step": 21190
    },
    {
      "entropy": 5.541029405593872,
      "epoch": 1.6490566037735848,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047272503319192887,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20105783194303511,
      "num_tokens": 36756982.0,
      "step": 21195
    },
    {
      "entropy": 5.585467529296875,
      "epoch": 1.6494456331453025,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004727118049562919,
      "loss": 4.9806,
      "mean_token_accuracy": 0.19576646089553834,
      "num_tokens": 36765772.0,
      "step": 21200
    },
    {
      "entropy": 5.560279178619385,
      "epoch": 1.6498346625170202,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000472698573720693,
      "loss": 4.9703,
      "mean_token_accuracy": 0.19757506400346755,
      "num_tokens": 36774718.0,
      "step": 21205
    },
    {
      "entropy": 5.582103443145752,
      "epoch": 1.6502236918887376,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004726853394853329,
      "loss": 5.023,
      "mean_token_accuracy": 0.20149630755186082,
      "num_tokens": 36783575.0,
      "step": 21210
    },
    {
      "entropy": 5.62049822807312,
      "epoch": 1.650612721260455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047267210225041234,
      "loss": 5.0936,
      "mean_token_accuracy": 0.19383912831544875,
      "num_tokens": 36792665.0,
      "step": 21215
    },
    {
      "entropy": 5.553858137130737,
      "epoch": 1.6510017506321728,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047265886201613234,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2102370247244835,
      "num_tokens": 36800987.0,
      "step": 21220
    },
    {
      "entropy": 5.570660161972046,
      "epoch": 1.6513907800038903,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047264561878269365,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20491121411323548,
      "num_tokens": 36809652.0,
      "step": 21225
    },
    {
      "entropy": 5.591576957702637,
      "epoch": 1.6517798093756078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004726323725502973,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19046802520751954,
      "num_tokens": 36818177.0,
      "step": 21230
    },
    {
      "entropy": 5.602564430236816,
      "epoch": 1.6521688387473255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004726191233191443,
      "loss": 5.0795,
      "mean_token_accuracy": 0.19109975248575212,
      "num_tokens": 36828494.0,
      "step": 21235
    },
    {
      "entropy": 5.65405592918396,
      "epoch": 1.6525578681190431,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004726058710894357,
      "loss": 5.1372,
      "mean_token_accuracy": 0.19121405929327012,
      "num_tokens": 36836956.0,
      "step": 21240
    },
    {
      "entropy": 5.5333339214324955,
      "epoch": 1.6529468974907604,
      "grad_norm": 1.125,
      "learning_rate": 0.0004725926158613724,
      "loss": 5.0219,
      "mean_token_accuracy": 0.202458393573761,
      "num_tokens": 36845973.0,
      "step": 21245
    },
    {
      "entropy": 5.518373632431031,
      "epoch": 1.653335926862478,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004725793576351557,
      "loss": 4.9636,
      "mean_token_accuracy": 0.20332099348306656,
      "num_tokens": 36855090.0,
      "step": 21250
    },
    {
      "entropy": 5.590155410766601,
      "epoch": 1.6537249562341958,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004725660964109867,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2013632610440254,
      "num_tokens": 36864373.0,
      "step": 21255
    },
    {
      "entropy": 5.683975505828857,
      "epoch": 1.6541139856059133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047255283218906673,
      "loss": 5.1092,
      "mean_token_accuracy": 0.19302762299776077,
      "num_tokens": 36872969.0,
      "step": 21260
    },
    {
      "entropy": 5.68102912902832,
      "epoch": 1.6545030149776307,
      "grad_norm": 1.078125,
      "learning_rate": 0.000472539564969597,
      "loss": 5.1071,
      "mean_token_accuracy": 0.2057632476091385,
      "num_tokens": 36880828.0,
      "step": 21265
    },
    {
      "entropy": 5.572316884994507,
      "epoch": 1.6548920443493484,
      "grad_norm": 1.125,
      "learning_rate": 0.00047252629475277863,
      "loss": 4.976,
      "mean_token_accuracy": 0.19926813691854478,
      "num_tokens": 36888766.0,
      "step": 21270
    },
    {
      "entropy": 5.529380512237549,
      "epoch": 1.655281073721066,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004725130215388132,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20514314919710158,
      "num_tokens": 36896623.0,
      "step": 21275
    },
    {
      "entropy": 5.6231659889221195,
      "epoch": 1.6556701030927834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004724997453279021,
      "loss": 5.0474,
      "mean_token_accuracy": 0.1994338497519493,
      "num_tokens": 36904753.0,
      "step": 21280
    },
    {
      "entropy": 5.575541687011719,
      "epoch": 1.656059132464501,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004724864661202467,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20546438843011855,
      "num_tokens": 36914260.0,
      "step": 21285
    },
    {
      "entropy": 5.549522113800049,
      "epoch": 1.6564481618362188,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047247318391604846,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21007707715034485,
      "num_tokens": 36923114.0,
      "step": 21290
    },
    {
      "entropy": 5.505823945999145,
      "epoch": 1.656837191207936,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047245989871550897,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20518750846385955,
      "num_tokens": 36932042.0,
      "step": 21295
    },
    {
      "entropy": 5.621746158599853,
      "epoch": 1.6572262205796537,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004724466105188299,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19626017659902573,
      "num_tokens": 36940522.0,
      "step": 21300
    },
    {
      "entropy": 5.620790195465088,
      "epoch": 1.6576152499513714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047243331932621263,
      "loss": 5.0149,
      "mean_token_accuracy": 0.19953967332839967,
      "num_tokens": 36948757.0,
      "step": 21305
    },
    {
      "entropy": 5.543782043457031,
      "epoch": 1.658004279323089,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004724200251378591,
      "loss": 4.9189,
      "mean_token_accuracy": 0.20497383773326874,
      "num_tokens": 36957876.0,
      "step": 21310
    },
    {
      "entropy": 5.564902210235596,
      "epoch": 1.6583933086948064,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004724067279539709,
      "loss": 5.0688,
      "mean_token_accuracy": 0.1940697655081749,
      "num_tokens": 36967109.0,
      "step": 21315
    },
    {
      "entropy": 5.601346731185913,
      "epoch": 1.658782338066524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047239342777474975,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20013927817344665,
      "num_tokens": 36975865.0,
      "step": 21320
    },
    {
      "entropy": 5.661682653427124,
      "epoch": 1.6591713674382416,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047238012460039765,
      "loss": 5.0832,
      "mean_token_accuracy": 0.1947967678308487,
      "num_tokens": 36983847.0,
      "step": 21325
    },
    {
      "entropy": 5.558710050582886,
      "epoch": 1.659560396809959,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004723668184311162,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20364061892032623,
      "num_tokens": 36991996.0,
      "step": 21330
    },
    {
      "entropy": 5.585368347167969,
      "epoch": 1.6599494261816767,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004723535092671074,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19752856940031052,
      "num_tokens": 37001075.0,
      "step": 21335
    },
    {
      "entropy": 5.654236507415772,
      "epoch": 1.6603384555533944,
      "grad_norm": 1.125,
      "learning_rate": 0.00047234019710857335,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19622689336538315,
      "num_tokens": 37009352.0,
      "step": 21340
    },
    {
      "entropy": 5.613842535018921,
      "epoch": 1.6607274849251117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004723268819557158,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2045325368642807,
      "num_tokens": 37017608.0,
      "step": 21345
    },
    {
      "entropy": 5.608464574813842,
      "epoch": 1.6611165142968294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047231356380873687,
      "loss": 5.041,
      "mean_token_accuracy": 0.19824079424142838,
      "num_tokens": 37026288.0,
      "step": 21350
    },
    {
      "entropy": 5.645211029052734,
      "epoch": 1.661505543668547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004723002426678388,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1914190247654915,
      "num_tokens": 37034515.0,
      "step": 21355
    },
    {
      "entropy": 5.484783029556274,
      "epoch": 1.6618945730402646,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004722869185332235,
      "loss": 4.9028,
      "mean_token_accuracy": 0.20960735976696016,
      "num_tokens": 37043048.0,
      "step": 21360
    },
    {
      "entropy": 5.567171382904053,
      "epoch": 1.662283602411982,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047227359140509324,
      "loss": 5.0095,
      "mean_token_accuracy": 0.19764861464500427,
      "num_tokens": 37051458.0,
      "step": 21365
    },
    {
      "entropy": 5.594884204864502,
      "epoch": 1.6626726317836997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004722602612836501,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20250048339366913,
      "num_tokens": 37059685.0,
      "step": 21370
    },
    {
      "entropy": 5.549195003509522,
      "epoch": 1.6630616611554172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047224692816909665,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19346206188201903,
      "num_tokens": 37068568.0,
      "step": 21375
    },
    {
      "entropy": 5.576695823669434,
      "epoch": 1.6634506905271347,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004722335920616349,
      "loss": 4.9091,
      "mean_token_accuracy": 0.20346153378486634,
      "num_tokens": 37077018.0,
      "step": 21380
    },
    {
      "entropy": 5.56000657081604,
      "epoch": 1.6638397198988524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047222025296146745,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2047078639268875,
      "num_tokens": 37085177.0,
      "step": 21385
    },
    {
      "entropy": 5.596906995773315,
      "epoch": 1.66422874927057,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047220691086879643,
      "loss": 5.0839,
      "mean_token_accuracy": 0.19844681322574614,
      "num_tokens": 37093788.0,
      "step": 21390
    },
    {
      "entropy": 5.650668573379517,
      "epoch": 1.6646177786422873,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004721935657838245,
      "loss": 5.0384,
      "mean_token_accuracy": 0.20089267194271088,
      "num_tokens": 37101766.0,
      "step": 21395
    },
    {
      "entropy": 5.5776196956634525,
      "epoch": 1.665006808014005,
      "grad_norm": 1.140625,
      "learning_rate": 0.000472180217706754,
      "loss": 5.0346,
      "mean_token_accuracy": 0.19905794262886048,
      "num_tokens": 37110259.0,
      "step": 21400
    },
    {
      "entropy": 5.558304786682129,
      "epoch": 1.6653958373857227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004721668666377876,
      "loss": 4.9982,
      "mean_token_accuracy": 0.19484549760818481,
      "num_tokens": 37118898.0,
      "step": 21405
    },
    {
      "entropy": 5.6726305961608885,
      "epoch": 1.6657848667574402,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047215351257712783,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1889210447669029,
      "num_tokens": 37128022.0,
      "step": 21410
    },
    {
      "entropy": 5.64467568397522,
      "epoch": 1.6661738961291577,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047214015552497735,
      "loss": 5.0984,
      "mean_token_accuracy": 0.19312608391046523,
      "num_tokens": 37136415.0,
      "step": 21415
    },
    {
      "entropy": 5.598124980926514,
      "epoch": 1.6665629255008754,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047212679548153867,
      "loss": 5.069,
      "mean_token_accuracy": 0.19460731297731398,
      "num_tokens": 37144548.0,
      "step": 21420
    },
    {
      "entropy": 5.525355625152588,
      "epoch": 1.6669519548725928,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047211343244701475,
      "loss": 4.9369,
      "mean_token_accuracy": 0.20387574136257172,
      "num_tokens": 37153425.0,
      "step": 21425
    },
    {
      "entropy": 5.57338433265686,
      "epoch": 1.6673409842443103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004721000664216081,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19300612956285476,
      "num_tokens": 37162618.0,
      "step": 21430
    },
    {
      "entropy": 5.594841718673706,
      "epoch": 1.667730013616028,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004720866974055219,
      "loss": 5.0371,
      "mean_token_accuracy": 0.19377586394548416,
      "num_tokens": 37171684.0,
      "step": 21435
    },
    {
      "entropy": 5.6655463695526125,
      "epoch": 1.6681190429877457,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004720733253989585,
      "loss": 5.1226,
      "mean_token_accuracy": 0.19757217764854432,
      "num_tokens": 37180305.0,
      "step": 21440
    },
    {
      "entropy": 5.603850269317627,
      "epoch": 1.6685080723594632,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047205995040212114,
      "loss": 4.998,
      "mean_token_accuracy": 0.21022711247205733,
      "num_tokens": 37189475.0,
      "step": 21445
    },
    {
      "entropy": 5.612037706375122,
      "epoch": 1.6688971017311807,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004720465724152127,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20796894133090973,
      "num_tokens": 37197607.0,
      "step": 21450
    },
    {
      "entropy": 5.597958898544311,
      "epoch": 1.6692861311028984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047203319143843613,
      "loss": 4.98,
      "mean_token_accuracy": 0.20287234634160994,
      "num_tokens": 37206767.0,
      "step": 21455
    },
    {
      "entropy": 5.681886768341064,
      "epoch": 1.6696751604746158,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004720198074719946,
      "loss": 5.1747,
      "mean_token_accuracy": 0.19210042506456376,
      "num_tokens": 37216009.0,
      "step": 21460
    },
    {
      "entropy": 5.6130759716033936,
      "epoch": 1.6700641898463333,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047200642051609094,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20040407180786132,
      "num_tokens": 37225670.0,
      "step": 21465
    },
    {
      "entropy": 5.6227034568786625,
      "epoch": 1.670453219218051,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004719930305709285,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2009032756090164,
      "num_tokens": 37233399.0,
      "step": 21470
    },
    {
      "entropy": 5.5497785091400145,
      "epoch": 1.6708422485897685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047197963763671033,
      "loss": 5.056,
      "mean_token_accuracy": 0.19735658466815947,
      "num_tokens": 37242592.0,
      "step": 21475
    },
    {
      "entropy": 5.620804262161255,
      "epoch": 1.671231277961486,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004719662417136397,
      "loss": 5.0926,
      "mean_token_accuracy": 0.19097328782081605,
      "num_tokens": 37250489.0,
      "step": 21480
    },
    {
      "entropy": 5.653549718856811,
      "epoch": 1.6716203073332037,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047195284280191987,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19715687334537507,
      "num_tokens": 37259223.0,
      "step": 21485
    },
    {
      "entropy": 5.627438974380493,
      "epoch": 1.6720093367049214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004719394409017541,
      "loss": 5.1338,
      "mean_token_accuracy": 0.1968739703297615,
      "num_tokens": 37269017.0,
      "step": 21490
    },
    {
      "entropy": 5.681085062026978,
      "epoch": 1.6723983660766388,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047192603601334585,
      "loss": 5.0123,
      "mean_token_accuracy": 0.19787476658821107,
      "num_tokens": 37277314.0,
      "step": 21495
    },
    {
      "entropy": 5.695496034622193,
      "epoch": 1.6727873954483563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047191262813689835,
      "loss": 5.1404,
      "mean_token_accuracy": 0.19638393223285674,
      "num_tokens": 37285878.0,
      "step": 21500
    },
    {
      "entropy": 5.590741777420044,
      "epoch": 1.673176424820074,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004718992172726152,
      "loss": 5.024,
      "mean_token_accuracy": 0.19646441489458083,
      "num_tokens": 37294993.0,
      "step": 21505
    },
    {
      "entropy": 5.606803512573242,
      "epoch": 1.6735654541917915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047188580342069983,
      "loss": 5.0832,
      "mean_token_accuracy": 0.19628289639949797,
      "num_tokens": 37303934.0,
      "step": 21510
    },
    {
      "entropy": 5.553181743621826,
      "epoch": 1.673954483563509,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004718723865813557,
      "loss": 5.0748,
      "mean_token_accuracy": 0.19753180593252181,
      "num_tokens": 37312082.0,
      "step": 21515
    },
    {
      "entropy": 5.5780189514160154,
      "epoch": 1.6743435129352267,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004718589667547865,
      "loss": 5.0778,
      "mean_token_accuracy": 0.19303230047225953,
      "num_tokens": 37320702.0,
      "step": 21520
    },
    {
      "entropy": 5.669962310791016,
      "epoch": 1.6747325423069441,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004718455439411958,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20167287439107895,
      "num_tokens": 37328929.0,
      "step": 21525
    },
    {
      "entropy": 5.59612889289856,
      "epoch": 1.6751215716786616,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004718321181407873,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2020670935511589,
      "num_tokens": 37337239.0,
      "step": 21530
    },
    {
      "entropy": 5.504689407348633,
      "epoch": 1.6755106010503793,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004718186893537647,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2045491650700569,
      "num_tokens": 37345541.0,
      "step": 21535
    },
    {
      "entropy": 5.604449939727783,
      "epoch": 1.675899630422097,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004718052575803318,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1910080134868622,
      "num_tokens": 37354095.0,
      "step": 21540
    },
    {
      "entropy": 5.690257596969604,
      "epoch": 1.6762886597938145,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004717918228206923,
      "loss": 5.1376,
      "mean_token_accuracy": 0.20061299204826355,
      "num_tokens": 37362822.0,
      "step": 21545
    },
    {
      "entropy": 5.735221862792969,
      "epoch": 1.676677689165532,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004717783850750501,
      "loss": 5.1844,
      "mean_token_accuracy": 0.19224259555339812,
      "num_tokens": 37371452.0,
      "step": 21550
    },
    {
      "entropy": 5.620316839218139,
      "epoch": 1.6770667185372496,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004717649443436091,
      "loss": 4.9657,
      "mean_token_accuracy": 0.19309251606464387,
      "num_tokens": 37379922.0,
      "step": 21555
    },
    {
      "entropy": 5.545872640609741,
      "epoch": 1.6774557479089671,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047175150062657336,
      "loss": 4.9808,
      "mean_token_accuracy": 0.19941299557685851,
      "num_tokens": 37388480.0,
      "step": 21560
    },
    {
      "entropy": 5.59928560256958,
      "epoch": 1.6778447772806846,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047173805392414664,
      "loss": 5.067,
      "mean_token_accuracy": 0.2005523055791855,
      "num_tokens": 37397621.0,
      "step": 21565
    },
    {
      "entropy": 5.607751655578613,
      "epoch": 1.6782338066524023,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004717246042365332,
      "loss": 5.031,
      "mean_token_accuracy": 0.19792662858963012,
      "num_tokens": 37405654.0,
      "step": 21570
    },
    {
      "entropy": 5.598772382736206,
      "epoch": 1.6786228360241198,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047171115156393695,
      "loss": 5.1142,
      "mean_token_accuracy": 0.191181543469429,
      "num_tokens": 37414775.0,
      "step": 21575
    },
    {
      "entropy": 5.589687967300415,
      "epoch": 1.6790118653958372,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047169769590656207,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2001945346593857,
      "num_tokens": 37422990.0,
      "step": 21580
    },
    {
      "entropy": 5.565493774414063,
      "epoch": 1.679400894767555,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047168423726461273,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1983790934085846,
      "num_tokens": 37431164.0,
      "step": 21585
    },
    {
      "entropy": 5.62465763092041,
      "epoch": 1.6797899241392726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004716707756382931,
      "loss": 5.0763,
      "mean_token_accuracy": 0.19347990900278092,
      "num_tokens": 37439958.0,
      "step": 21590
    },
    {
      "entropy": 5.6375532150268555,
      "epoch": 1.6801789535109901,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047165731102780757,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2031334638595581,
      "num_tokens": 37448381.0,
      "step": 21595
    },
    {
      "entropy": 5.627219009399414,
      "epoch": 1.6805679828827076,
      "grad_norm": 1.125,
      "learning_rate": 0.00047164384343336037,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19322516024112701,
      "num_tokens": 37456705.0,
      "step": 21600
    },
    {
      "entropy": 5.566686058044434,
      "epoch": 1.6809570122544253,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047163037285515583,
      "loss": 5.1222,
      "mean_token_accuracy": 0.1980556443333626,
      "num_tokens": 37466323.0,
      "step": 21605
    },
    {
      "entropy": 5.5563271045684814,
      "epoch": 1.6813460416261428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716168992933982,
      "loss": 5.017,
      "mean_token_accuracy": 0.19530599564313889,
      "num_tokens": 37475570.0,
      "step": 21610
    },
    {
      "entropy": 5.600547122955322,
      "epoch": 1.6817350709978602,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047160342274829234,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20107959508895873,
      "num_tokens": 37484724.0,
      "step": 21615
    },
    {
      "entropy": 5.688745212554932,
      "epoch": 1.682124100369578,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047158994322004223,
      "loss": 5.1899,
      "mean_token_accuracy": 0.18916524201631546,
      "num_tokens": 37493747.0,
      "step": 21620
    },
    {
      "entropy": 5.540155506134033,
      "epoch": 1.6825131297412956,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004715764607088527,
      "loss": 4.8598,
      "mean_token_accuracy": 0.21754775792360306,
      "num_tokens": 37501557.0,
      "step": 21625
    },
    {
      "entropy": 5.523472118377685,
      "epoch": 1.6829021591130129,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004715629752149283,
      "loss": 5.0345,
      "mean_token_accuracy": 0.19737348556518555,
      "num_tokens": 37510231.0,
      "step": 21630
    },
    {
      "entropy": 5.597281408309937,
      "epoch": 1.6832911884847306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047154948673847356,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20183728635311127,
      "num_tokens": 37518038.0,
      "step": 21635
    },
    {
      "entropy": 5.657252788543701,
      "epoch": 1.6836802178564483,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004715359952796933,
      "loss": 5.17,
      "mean_token_accuracy": 0.18862509727478027,
      "num_tokens": 37527519.0,
      "step": 21640
    },
    {
      "entropy": 5.640777444839477,
      "epoch": 1.6840692472281658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047152250083879203,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19194298684597016,
      "num_tokens": 37536291.0,
      "step": 21645
    },
    {
      "entropy": 5.619144582748413,
      "epoch": 1.6844582765998832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047150900341597464,
      "loss": 5.0259,
      "mean_token_accuracy": 0.20478426218032836,
      "num_tokens": 37544343.0,
      "step": 21650
    },
    {
      "entropy": 5.511697912216187,
      "epoch": 1.684847305971601,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004714955030114459,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2022662118077278,
      "num_tokens": 37552013.0,
      "step": 21655
    },
    {
      "entropy": 5.537133836746216,
      "epoch": 1.6852363353433184,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047148199962541065,
      "loss": 5.056,
      "mean_token_accuracy": 0.19037532657384873,
      "num_tokens": 37561044.0,
      "step": 21660
    },
    {
      "entropy": 5.61952805519104,
      "epoch": 1.6856253647150359,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004714684932580738,
      "loss": 5.083,
      "mean_token_accuracy": 0.19459929913282395,
      "num_tokens": 37569459.0,
      "step": 21665
    },
    {
      "entropy": 5.593795824050903,
      "epoch": 1.6860143940867536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004714549839096403,
      "loss": 4.9696,
      "mean_token_accuracy": 0.201733823120594,
      "num_tokens": 37578488.0,
      "step": 21670
    },
    {
      "entropy": 5.514795207977295,
      "epoch": 1.6864034234584713,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047144147158031507,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2076156571507454,
      "num_tokens": 37586761.0,
      "step": 21675
    },
    {
      "entropy": 5.595983552932739,
      "epoch": 1.6867924528301885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004714279562703032,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1959920972585678,
      "num_tokens": 37595267.0,
      "step": 21680
    },
    {
      "entropy": 5.585158729553223,
      "epoch": 1.6871814822019062,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004714144379798098,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2055351570248604,
      "num_tokens": 37603271.0,
      "step": 21685
    },
    {
      "entropy": 5.565847778320313,
      "epoch": 1.687570511573624,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047140091670903986,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19494160562753676,
      "num_tokens": 37611152.0,
      "step": 21690
    },
    {
      "entropy": 5.5498144149780275,
      "epoch": 1.6879595409453414,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047138739245819857,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20409328788518905,
      "num_tokens": 37619465.0,
      "step": 21695
    },
    {
      "entropy": 5.662761068344116,
      "epoch": 1.6883485703170589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047137386522749124,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1841244637966156,
      "num_tokens": 37628588.0,
      "step": 21700
    },
    {
      "entropy": 5.612898445129394,
      "epoch": 1.6887375996887766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004713603350171231,
      "loss": 5.0797,
      "mean_token_accuracy": 0.20128399431705474,
      "num_tokens": 37638062.0,
      "step": 21705
    },
    {
      "entropy": 5.667411613464355,
      "epoch": 1.689126629060494,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047134680182729926,
      "loss": 5.0181,
      "mean_token_accuracy": 0.19881925582885743,
      "num_tokens": 37646968.0,
      "step": 21710
    },
    {
      "entropy": 5.675249099731445,
      "epoch": 1.6895156584322115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004713332656582254,
      "loss": 5.0969,
      "mean_token_accuracy": 0.19076040685176848,
      "num_tokens": 37655893.0,
      "step": 21715
    },
    {
      "entropy": 5.6023155689239506,
      "epoch": 1.6899046878039292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004713197265101066,
      "loss": 5.012,
      "mean_token_accuracy": 0.20391858518123626,
      "num_tokens": 37664638.0,
      "step": 21720
    },
    {
      "entropy": 5.555707502365112,
      "epoch": 1.690293717175647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047130618438314856,
      "loss": 5.0214,
      "mean_token_accuracy": 0.19854926615953444,
      "num_tokens": 37673582.0,
      "step": 21725
    },
    {
      "entropy": 5.6125373363494875,
      "epoch": 1.6906827465473642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004712926392775565,
      "loss": 5.0863,
      "mean_token_accuracy": 0.1962234228849411,
      "num_tokens": 37682310.0,
      "step": 21730
    },
    {
      "entropy": 5.539216995239258,
      "epoch": 1.6910717759190819,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004712790911935362,
      "loss": 4.9873,
      "mean_token_accuracy": 0.20257582813501357,
      "num_tokens": 37690476.0,
      "step": 21735
    },
    {
      "entropy": 5.561210060119629,
      "epoch": 1.6914608052907996,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000471265540131293,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20281372219324112,
      "num_tokens": 37698965.0,
      "step": 21740
    },
    {
      "entropy": 5.6243896484375,
      "epoch": 1.691849834662517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047125198609103267,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20186861902475356,
      "num_tokens": 37707499.0,
      "step": 21745
    },
    {
      "entropy": 5.570200109481812,
      "epoch": 1.6922388640342345,
      "grad_norm": 1.125,
      "learning_rate": 0.0004712384290729607,
      "loss": 5.0375,
      "mean_token_accuracy": 0.19135928601026536,
      "num_tokens": 37716686.0,
      "step": 21750
    },
    {
      "entropy": 5.584670305252075,
      "epoch": 1.6926278934059522,
      "grad_norm": 1.109375,
      "learning_rate": 0.000471224869077283,
      "loss": 5.0814,
      "mean_token_accuracy": 0.19294968694448472,
      "num_tokens": 37725607.0,
      "step": 21755
    },
    {
      "entropy": 5.613400983810425,
      "epoch": 1.6930169227776697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047121130610420527,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20611340254545213,
      "num_tokens": 37734103.0,
      "step": 21760
    },
    {
      "entropy": 5.656982660293579,
      "epoch": 1.6934059521493872,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047119774015393323,
      "loss": 5.077,
      "mean_token_accuracy": 0.19648510366678237,
      "num_tokens": 37743078.0,
      "step": 21765
    },
    {
      "entropy": 5.56747407913208,
      "epoch": 1.6937949815211049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047118417122667285,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2093929097056389,
      "num_tokens": 37751550.0,
      "step": 21770
    },
    {
      "entropy": 5.628477954864502,
      "epoch": 1.6941840108928226,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004711705993226298,
      "loss": 5.0711,
      "mean_token_accuracy": 0.19530853778123855,
      "num_tokens": 37759940.0,
      "step": 21775
    },
    {
      "entropy": 5.641937875747681,
      "epoch": 1.6945730402645398,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711570244420102,
      "loss": 5.103,
      "mean_token_accuracy": 0.1924607828259468,
      "num_tokens": 37769191.0,
      "step": 21780
    },
    {
      "entropy": 5.511842393875122,
      "epoch": 1.6949620696362575,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004711434465850199,
      "loss": 4.8949,
      "mean_token_accuracy": 0.21043619960546495,
      "num_tokens": 37777612.0,
      "step": 21785
    },
    {
      "entropy": 5.500446844100952,
      "epoch": 1.6953510990079752,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004711298657518651,
      "loss": 5.0267,
      "mean_token_accuracy": 0.1994296759366989,
      "num_tokens": 37785963.0,
      "step": 21790
    },
    {
      "entropy": 5.539422512054443,
      "epoch": 1.6957401283796927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004711162819427518,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20120040327310562,
      "num_tokens": 37795546.0,
      "step": 21795
    },
    {
      "entropy": 5.651784181594849,
      "epoch": 1.6961291577514102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000471102695157886,
      "loss": 5.1738,
      "mean_token_accuracy": 0.189884053170681,
      "num_tokens": 37803955.0,
      "step": 21800
    },
    {
      "entropy": 5.588523626327515,
      "epoch": 1.6965181871231279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004710891053974739,
      "loss": 5.0459,
      "mean_token_accuracy": 0.20582387149333953,
      "num_tokens": 37812659.0,
      "step": 21805
    },
    {
      "entropy": 5.577352476119995,
      "epoch": 1.6969072164948453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004710755126617217,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20160599648952485,
      "num_tokens": 37821064.0,
      "step": 21810
    },
    {
      "entropy": 5.659938669204712,
      "epoch": 1.6972962458665628,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004710619169508358,
      "loss": 5.1884,
      "mean_token_accuracy": 0.18609914034605027,
      "num_tokens": 37830353.0,
      "step": 21815
    },
    {
      "entropy": 5.611560869216919,
      "epoch": 1.6976852752382805,
      "grad_norm": 1.125,
      "learning_rate": 0.0004710483182650223,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19963879883289337,
      "num_tokens": 37838927.0,
      "step": 21820
    },
    {
      "entropy": 5.545606517791748,
      "epoch": 1.6980743046099982,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047103471660448767,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2035693809390068,
      "num_tokens": 37847845.0,
      "step": 21825
    },
    {
      "entropy": 5.616822719573975,
      "epoch": 1.6984633339817157,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047102111196943813,
      "loss": 5.0393,
      "mean_token_accuracy": 0.19011109918355942,
      "num_tokens": 37856065.0,
      "step": 21830
    },
    {
      "entropy": 5.5641180038452145,
      "epoch": 1.6988523633534331,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047100750436008035,
      "loss": 5.0099,
      "mean_token_accuracy": 0.1980397328734398,
      "num_tokens": 37865262.0,
      "step": 21835
    },
    {
      "entropy": 5.539391803741455,
      "epoch": 1.6992413927251508,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047099389377662054,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20115655809640884,
      "num_tokens": 37873994.0,
      "step": 21840
    },
    {
      "entropy": 5.597001647949218,
      "epoch": 1.6996304220968683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004709802802192654,
      "loss": 5.0656,
      "mean_token_accuracy": 0.19346749633550644,
      "num_tokens": 37882377.0,
      "step": 21845
    },
    {
      "entropy": 5.574456024169922,
      "epoch": 1.7000194514685858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047096666368822153,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2068011835217476,
      "num_tokens": 37890793.0,
      "step": 21850
    },
    {
      "entropy": 5.584441995620727,
      "epoch": 1.7004084808403035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047095304418369536,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20109118521213531,
      "num_tokens": 37899464.0,
      "step": 21855
    },
    {
      "entropy": 5.552498531341553,
      "epoch": 1.700797510212021,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004709394217058936,
      "loss": 5.0618,
      "mean_token_accuracy": 0.195765483379364,
      "num_tokens": 37907959.0,
      "step": 21860
    },
    {
      "entropy": 5.630438375473022,
      "epoch": 1.7011865395837384,
      "grad_norm": 1.125,
      "learning_rate": 0.0004709257962550231,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19461337178945542,
      "num_tokens": 37916751.0,
      "step": 21865
    },
    {
      "entropy": 5.576208543777466,
      "epoch": 1.7015755689554561,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047091216783129035,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20160162448883057,
      "num_tokens": 37925538.0,
      "step": 21870
    },
    {
      "entropy": 5.577460813522339,
      "epoch": 1.7019645983271738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004708985364349024,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21036584973335265,
      "num_tokens": 37933770.0,
      "step": 21875
    },
    {
      "entropy": 5.606276893615723,
      "epoch": 1.7023536276988913,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047088490206606586,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1932048887014389,
      "num_tokens": 37943006.0,
      "step": 21880
    },
    {
      "entropy": 5.647562026977539,
      "epoch": 1.7027426570706088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047087126472498785,
      "loss": 5.1559,
      "mean_token_accuracy": 0.19053927958011627,
      "num_tokens": 37951775.0,
      "step": 21885
    },
    {
      "entropy": 5.630889272689819,
      "epoch": 1.7031316864423265,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000470857624411875,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19639912247657776,
      "num_tokens": 37960301.0,
      "step": 21890
    },
    {
      "entropy": 5.5995525360107425,
      "epoch": 1.703520715814044,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047084398112693456,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20101989358663558,
      "num_tokens": 37969338.0,
      "step": 21895
    },
    {
      "entropy": 5.624325752258301,
      "epoch": 1.7039097451857614,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004708303348703734,
      "loss": 5.1007,
      "mean_token_accuracy": 0.1889639377593994,
      "num_tokens": 37978114.0,
      "step": 21900
    },
    {
      "entropy": 5.680545282363892,
      "epoch": 1.7042987745574791,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004708166856423986,
      "loss": 5.115,
      "mean_token_accuracy": 0.19899093359708786,
      "num_tokens": 37986581.0,
      "step": 21905
    },
    {
      "entropy": 5.660883092880249,
      "epoch": 1.7046878039291966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047080303344321727,
      "loss": 5.1037,
      "mean_token_accuracy": 0.1942038953304291,
      "num_tokens": 37995021.0,
      "step": 21910
    },
    {
      "entropy": 5.527478551864624,
      "epoch": 1.705076833300914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047078937827303653,
      "loss": 4.9583,
      "mean_token_accuracy": 0.19997209757566453,
      "num_tokens": 38003240.0,
      "step": 21915
    },
    {
      "entropy": 5.58984146118164,
      "epoch": 1.7054658626726318,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004707757201320636,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1986893370747566,
      "num_tokens": 38011416.0,
      "step": 21920
    },
    {
      "entropy": 5.5585814952850345,
      "epoch": 1.7058548920443495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047076205902050577,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2004631295800209,
      "num_tokens": 38020112.0,
      "step": 21925
    },
    {
      "entropy": 5.6159271717071535,
      "epoch": 1.706243921416067,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047074839493857024,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19398657977581024,
      "num_tokens": 38028935.0,
      "step": 21930
    },
    {
      "entropy": 5.556846761703492,
      "epoch": 1.7066329507877844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004707347278864644,
      "loss": 5.0372,
      "mean_token_accuracy": 0.19305555522441864,
      "num_tokens": 38037618.0,
      "step": 21935
    },
    {
      "entropy": 5.520416879653931,
      "epoch": 1.7070219801595021,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047072105786439563,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2003830000758171,
      "num_tokens": 38046853.0,
      "step": 21940
    },
    {
      "entropy": 5.625412464141846,
      "epoch": 1.7074110095312196,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047070738487257137,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20008161962032317,
      "num_tokens": 38056220.0,
      "step": 21945
    },
    {
      "entropy": 5.66451587677002,
      "epoch": 1.707800038902937,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047069370891119895,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19449239671230317,
      "num_tokens": 38064664.0,
      "step": 21950
    },
    {
      "entropy": 5.6162878513336185,
      "epoch": 1.7081890682746548,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000470680029980486,
      "loss": 5.1353,
      "mean_token_accuracy": 0.18884495198726653,
      "num_tokens": 38072963.0,
      "step": 21955
    },
    {
      "entropy": 5.556479215621948,
      "epoch": 1.7085780976463723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047066634808064006,
      "loss": 5.036,
      "mean_token_accuracy": 0.20238221883773805,
      "num_tokens": 38082320.0,
      "step": 21960
    },
    {
      "entropy": 5.580231142044068,
      "epoch": 1.7089671270180897,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047065266321186873,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20718733817338944,
      "num_tokens": 38091188.0,
      "step": 21965
    },
    {
      "entropy": 5.565281105041504,
      "epoch": 1.7093561563898074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004706389753743797,
      "loss": 5.0556,
      "mean_token_accuracy": 0.19353399723768233,
      "num_tokens": 38099584.0,
      "step": 21970
    },
    {
      "entropy": 5.524546766281128,
      "epoch": 1.7097451857615251,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706252845683805,
      "loss": 4.9645,
      "mean_token_accuracy": 0.20108680576086044,
      "num_tokens": 38108437.0,
      "step": 21975
    },
    {
      "entropy": 5.4553478240966795,
      "epoch": 1.7101342151332426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047061159079407903,
      "loss": 4.9222,
      "mean_token_accuracy": 0.20326072871685028,
      "num_tokens": 38116822.0,
      "step": 21980
    },
    {
      "entropy": 5.604338073730469,
      "epoch": 1.71052324450496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000470597894051683,
      "loss": 5.1576,
      "mean_token_accuracy": 0.19169563949108123,
      "num_tokens": 38125800.0,
      "step": 21985
    },
    {
      "entropy": 5.577329444885254,
      "epoch": 1.7109122738766778,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705841943414003,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2043044835329056,
      "num_tokens": 38134181.0,
      "step": 21990
    },
    {
      "entropy": 5.586867475509644,
      "epoch": 1.7113013032483952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047057049166343876,
      "loss": 5.0472,
      "mean_token_accuracy": 0.1978023409843445,
      "num_tokens": 38142820.0,
      "step": 21995
    },
    {
      "entropy": 5.525862503051758,
      "epoch": 1.7116903326201127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047055678601800623,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20009258836507798,
      "num_tokens": 38150571.0,
      "step": 22000
    },
    {
      "entropy": 5.63078465461731,
      "epoch": 1.7120793619918304,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047054307740531076,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1940620943903923,
      "num_tokens": 38159308.0,
      "step": 22005
    },
    {
      "entropy": 5.577589082717895,
      "epoch": 1.712468391363548,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047052936582556035,
      "loss": 5.007,
      "mean_token_accuracy": 0.20213858932256698,
      "num_tokens": 38167377.0,
      "step": 22010
    },
    {
      "entropy": 5.539031171798706,
      "epoch": 1.7128574207352654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047051565127896307,
      "loss": 4.9472,
      "mean_token_accuracy": 0.19885500520467758,
      "num_tokens": 38175381.0,
      "step": 22015
    },
    {
      "entropy": 5.525089693069458,
      "epoch": 1.713246450106983,
      "grad_norm": 1.125,
      "learning_rate": 0.00047050193376572686,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2042197361588478,
      "num_tokens": 38183821.0,
      "step": 22020
    },
    {
      "entropy": 5.55885763168335,
      "epoch": 1.7136354794787008,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047048821328606,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19958675801753997,
      "num_tokens": 38192181.0,
      "step": 22025
    },
    {
      "entropy": 5.563112688064575,
      "epoch": 1.7140245088504182,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004704744898401708,
      "loss": 5.0534,
      "mean_token_accuracy": 0.19329515546560289,
      "num_tokens": 38200703.0,
      "step": 22030
    },
    {
      "entropy": 5.527504968643188,
      "epoch": 1.7144135382221357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004704607634282672,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2062011644244194,
      "num_tokens": 38209398.0,
      "step": 22035
    },
    {
      "entropy": 5.545182514190674,
      "epoch": 1.7148025675938534,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047044703405055765,
      "loss": 4.984,
      "mean_token_accuracy": 0.20379881113767623,
      "num_tokens": 38217648.0,
      "step": 22040
    },
    {
      "entropy": 5.5601002216339115,
      "epoch": 1.7151915969655709,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047043330170725046,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20455377250909806,
      "num_tokens": 38226113.0,
      "step": 22045
    },
    {
      "entropy": 5.5248548030853275,
      "epoch": 1.7155806263372884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047041956639855406,
      "loss": 5.0006,
      "mean_token_accuracy": 0.1983527734875679,
      "num_tokens": 38234567.0,
      "step": 22050
    },
    {
      "entropy": 5.621897554397583,
      "epoch": 1.715969655709006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704058281246766,
      "loss": 5.1779,
      "mean_token_accuracy": 0.19026656150817872,
      "num_tokens": 38243477.0,
      "step": 22055
    },
    {
      "entropy": 5.58515772819519,
      "epoch": 1.7163586850807238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004703920868858268,
      "loss": 5.04,
      "mean_token_accuracy": 0.2020384907722473,
      "num_tokens": 38252008.0,
      "step": 22060
    },
    {
      "entropy": 5.58301043510437,
      "epoch": 1.716747714452441,
      "grad_norm": 1.125,
      "learning_rate": 0.00047037834268221315,
      "loss": 5.0508,
      "mean_token_accuracy": 0.19466104507446289,
      "num_tokens": 38260586.0,
      "step": 22065
    },
    {
      "entropy": 5.567405462265015,
      "epoch": 1.7171367438241587,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004703645955140441,
      "loss": 5.1095,
      "mean_token_accuracy": 0.1923338919878006,
      "num_tokens": 38268970.0,
      "step": 22070
    },
    {
      "entropy": 5.621124935150147,
      "epoch": 1.7175257731958764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047035084538152815,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20104631632566453,
      "num_tokens": 38277633.0,
      "step": 22075
    },
    {
      "entropy": 5.555356025695801,
      "epoch": 1.7179148025675939,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004703370922848742,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20047130435705185,
      "num_tokens": 38286122.0,
      "step": 22080
    },
    {
      "entropy": 5.573940849304199,
      "epoch": 1.7183038319393114,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047032333622429074,
      "loss": 5.0725,
      "mean_token_accuracy": 0.1988450363278389,
      "num_tokens": 38294060.0,
      "step": 22085
    },
    {
      "entropy": 5.55256667137146,
      "epoch": 1.718692861311029,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047030957719998656,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20489340275526047,
      "num_tokens": 38303446.0,
      "step": 22090
    },
    {
      "entropy": 5.586187505722046,
      "epoch": 1.7190818906827465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004702958152121704,
      "loss": 4.9991,
      "mean_token_accuracy": 0.1966421827673912,
      "num_tokens": 38311995.0,
      "step": 22095
    },
    {
      "entropy": 5.4701214790344235,
      "epoch": 1.719470920054464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004702820502610511,
      "loss": 4.9408,
      "mean_token_accuracy": 0.19998539686203004,
      "num_tokens": 38320132.0,
      "step": 22100
    },
    {
      "entropy": 5.590481519699097,
      "epoch": 1.7198599494261817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004702682823468375,
      "loss": 5.0309,
      "mean_token_accuracy": 0.19736093729734422,
      "num_tokens": 38328852.0,
      "step": 22105
    },
    {
      "entropy": 5.6275592803955075,
      "epoch": 1.7202489787978994,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047025451146973844,
      "loss": 5.0483,
      "mean_token_accuracy": 0.19381460249423982,
      "num_tokens": 38338142.0,
      "step": 22110
    },
    {
      "entropy": 5.595482015609742,
      "epoch": 1.7206380081696167,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047024073762996305,
      "loss": 5.0402,
      "mean_token_accuracy": 0.19861653000116347,
      "num_tokens": 38346522.0,
      "step": 22115
    },
    {
      "entropy": 5.576451635360717,
      "epoch": 1.7210270375413343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004702269608277202,
      "loss": 5.054,
      "mean_token_accuracy": 0.20477451533079147,
      "num_tokens": 38355203.0,
      "step": 22120
    },
    {
      "entropy": 5.5563994407653805,
      "epoch": 1.721416066913052,
      "grad_norm": 1.125,
      "learning_rate": 0.0004702131810632189,
      "loss": 4.9947,
      "mean_token_accuracy": 0.19546187222003936,
      "num_tokens": 38363657.0,
      "step": 22125
    },
    {
      "entropy": 5.520972013473511,
      "epoch": 1.7218050962847695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047019939833666837,
      "loss": 5.0162,
      "mean_token_accuracy": 0.19898925572633744,
      "num_tokens": 38372893.0,
      "step": 22130
    },
    {
      "entropy": 5.528195381164551,
      "epoch": 1.722194125656487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004701856126482776,
      "loss": 4.994,
      "mean_token_accuracy": 0.2053616926074028,
      "num_tokens": 38381856.0,
      "step": 22135
    },
    {
      "entropy": 5.664301776885987,
      "epoch": 1.7225831550282047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004701718239982559,
      "loss": 5.0886,
      "mean_token_accuracy": 0.18955252170562745,
      "num_tokens": 38390616.0,
      "step": 22140
    },
    {
      "entropy": 5.684468460083008,
      "epoch": 1.7229721843999222,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047015803238681234,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1896022826433182,
      "num_tokens": 38399024.0,
      "step": 22145
    },
    {
      "entropy": 5.58506588935852,
      "epoch": 1.7233612137716396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004701442378141563,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20057838410139084,
      "num_tokens": 38407605.0,
      "step": 22150
    },
    {
      "entropy": 5.640168333053589,
      "epoch": 1.7237502431433573,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000470130440280497,
      "loss": 5.1008,
      "mean_token_accuracy": 0.19637431502342223,
      "num_tokens": 38417228.0,
      "step": 22155
    },
    {
      "entropy": 5.653040266036987,
      "epoch": 1.724139272515075,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004701166397860439,
      "loss": 5.1071,
      "mean_token_accuracy": 0.18920597583055496,
      "num_tokens": 38425704.0,
      "step": 22160
    },
    {
      "entropy": 5.616227054595948,
      "epoch": 1.7245283018867923,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004701028363310064,
      "loss": 5.0827,
      "mean_token_accuracy": 0.19233149737119676,
      "num_tokens": 38434143.0,
      "step": 22165
    },
    {
      "entropy": 5.637010192871093,
      "epoch": 1.72491733125851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047008902991559385,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2015240713953972,
      "num_tokens": 38443708.0,
      "step": 22170
    },
    {
      "entropy": 5.613686370849609,
      "epoch": 1.7253063606302277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004700752205400158,
      "loss": 5.0284,
      "mean_token_accuracy": 0.198075008392334,
      "num_tokens": 38452399.0,
      "step": 22175
    },
    {
      "entropy": 5.598971176147461,
      "epoch": 1.7256953900019452,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047006140820448174,
      "loss": 5.043,
      "mean_token_accuracy": 0.1971168801188469,
      "num_tokens": 38460411.0,
      "step": 22180
    },
    {
      "entropy": 5.600412034988404,
      "epoch": 1.7260844193736626,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047004759290920136,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19836550056934357,
      "num_tokens": 38468542.0,
      "step": 22185
    },
    {
      "entropy": 5.567818975448608,
      "epoch": 1.7264734487453803,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004700337746543841,
      "loss": 5.0138,
      "mean_token_accuracy": 0.1999722346663475,
      "num_tokens": 38477027.0,
      "step": 22190
    },
    {
      "entropy": 5.5693676471710205,
      "epoch": 1.7268624781170978,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004700199534402398,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19828347563743592,
      "num_tokens": 38485853.0,
      "step": 22195
    },
    {
      "entropy": 5.553265142440796,
      "epoch": 1.7272515074888153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004700061292669781,
      "loss": 5.051,
      "mean_token_accuracy": 0.20213208943605424,
      "num_tokens": 38494177.0,
      "step": 22200
    },
    {
      "entropy": 5.601232814788818,
      "epoch": 1.727640536860533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004699923021348088,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1949697732925415,
      "num_tokens": 38503313.0,
      "step": 22205
    },
    {
      "entropy": 5.66887731552124,
      "epoch": 1.7280295662322507,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046997847204394166,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18958215713500975,
      "num_tokens": 38511950.0,
      "step": 22210
    },
    {
      "entropy": 5.694172620773315,
      "epoch": 1.728418595603968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046996463899458653,
      "loss": 5.1832,
      "mean_token_accuracy": 0.18583601713180542,
      "num_tokens": 38521331.0,
      "step": 22215
    },
    {
      "entropy": 5.533796930313111,
      "epoch": 1.7288076249756856,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004699508029869533,
      "loss": 5.0162,
      "mean_token_accuracy": 0.19839344173669815,
      "num_tokens": 38529709.0,
      "step": 22220
    },
    {
      "entropy": 5.501419162750244,
      "epoch": 1.7291966543474033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046993696402125205,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20062572956085206,
      "num_tokens": 38538083.0,
      "step": 22225
    },
    {
      "entropy": 5.545303678512573,
      "epoch": 1.7295856837191208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004699231220976925,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2004787653684616,
      "num_tokens": 38547282.0,
      "step": 22230
    },
    {
      "entropy": 5.6089537143707275,
      "epoch": 1.7299747130908383,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004699092772164849,
      "loss": 5.038,
      "mean_token_accuracy": 0.1952192783355713,
      "num_tokens": 38555674.0,
      "step": 22235
    },
    {
      "entropy": 5.649447393417359,
      "epoch": 1.730363742462556,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004698954293778392,
      "loss": 5.0555,
      "mean_token_accuracy": 0.1972043439745903,
      "num_tokens": 38564066.0,
      "step": 22240
    },
    {
      "entropy": 5.60773777961731,
      "epoch": 1.7307527718342735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046988157858196555,
      "loss": 5.0867,
      "mean_token_accuracy": 0.19835591465234756,
      "num_tokens": 38573299.0,
      "step": 22245
    },
    {
      "entropy": 5.574977159500122,
      "epoch": 1.731141801205991,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046986772482907426,
      "loss": 5.0241,
      "mean_token_accuracy": 0.19898533076047897,
      "num_tokens": 38581565.0,
      "step": 22250
    },
    {
      "entropy": 5.59044041633606,
      "epoch": 1.7315308305777086,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004698538681193754,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1994437500834465,
      "num_tokens": 38590192.0,
      "step": 22255
    },
    {
      "entropy": 5.560927104949951,
      "epoch": 1.7319198599494263,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046984000845307907,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1957292824983597,
      "num_tokens": 38599670.0,
      "step": 22260
    },
    {
      "entropy": 5.576747894287109,
      "epoch": 1.7323088893211438,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046982614583039584,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20853608548641206,
      "num_tokens": 38608432.0,
      "step": 22265
    },
    {
      "entropy": 5.654170179367066,
      "epoch": 1.7326979186928613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004698122802515359,
      "loss": 5.032,
      "mean_token_accuracy": 0.20013358443975449,
      "num_tokens": 38617359.0,
      "step": 22270
    },
    {
      "entropy": 5.571622276306153,
      "epoch": 1.733086948064579,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004697984117167097,
      "loss": 4.9969,
      "mean_token_accuracy": 0.19949166476726532,
      "num_tokens": 38626220.0,
      "step": 22275
    },
    {
      "entropy": 5.652535915374756,
      "epoch": 1.7334759774362964,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046978454022612767,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20276451855897903,
      "num_tokens": 38635441.0,
      "step": 22280
    },
    {
      "entropy": 5.57275619506836,
      "epoch": 1.733865006808014,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046977066578000025,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2012002944946289,
      "num_tokens": 38643948.0,
      "step": 22285
    },
    {
      "entropy": 5.568630933761597,
      "epoch": 1.7342540361797316,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046975678837853806,
      "loss": 5.1181,
      "mean_token_accuracy": 0.19392989873886107,
      "num_tokens": 38652755.0,
      "step": 22290
    },
    {
      "entropy": 5.526766681671143,
      "epoch": 1.734643065551449,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046974290802195156,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21727294921875,
      "num_tokens": 38661020.0,
      "step": 22295
    },
    {
      "entropy": 5.678378868103027,
      "epoch": 1.7350320949231666,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697290247104513,
      "loss": 5.1318,
      "mean_token_accuracy": 0.19324184358119964,
      "num_tokens": 38669628.0,
      "step": 22300
    },
    {
      "entropy": 5.612584018707276,
      "epoch": 1.7354211242948843,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046971513844424814,
      "loss": 5.0817,
      "mean_token_accuracy": 0.19404597133398055,
      "num_tokens": 38678861.0,
      "step": 22305
    },
    {
      "entropy": 5.60569314956665,
      "epoch": 1.735810153666602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046970124922355265,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20246932357549668,
      "num_tokens": 38688088.0,
      "step": 22310
    },
    {
      "entropy": 5.587804698944092,
      "epoch": 1.7361991830383194,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004696873570485756,
      "loss": 5.0325,
      "mean_token_accuracy": 0.19735390245914458,
      "num_tokens": 38697226.0,
      "step": 22315
    },
    {
      "entropy": 5.656267547607422,
      "epoch": 1.736588212410037,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004696734619195278,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19367065131664277,
      "num_tokens": 38705710.0,
      "step": 22320
    },
    {
      "entropy": 5.561840295791626,
      "epoch": 1.7369772417817546,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046965956383662,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2156523659825325,
      "num_tokens": 38714230.0,
      "step": 22325
    },
    {
      "entropy": 5.487687873840332,
      "epoch": 1.737366271153472,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004696456628000632,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20019209235906602,
      "num_tokens": 38723125.0,
      "step": 22330
    },
    {
      "entropy": 5.581688165664673,
      "epoch": 1.7377553005251896,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046963175881006825,
      "loss": 5.0633,
      "mean_token_accuracy": 0.19233470112085344,
      "num_tokens": 38731436.0,
      "step": 22335
    },
    {
      "entropy": 5.666593027114868,
      "epoch": 1.7381443298969073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004696178518668462,
      "loss": 5.1458,
      "mean_token_accuracy": 0.2003820687532425,
      "num_tokens": 38740662.0,
      "step": 22340
    },
    {
      "entropy": 5.530299949645996,
      "epoch": 1.7385333592686247,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046960394197060804,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20850037932395935,
      "num_tokens": 38748931.0,
      "step": 22345
    },
    {
      "entropy": 5.584829330444336,
      "epoch": 1.7389223886403422,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046959002912156473,
      "loss": 4.9724,
      "mean_token_accuracy": 0.19814537912607194,
      "num_tokens": 38757884.0,
      "step": 22350
    },
    {
      "entropy": 5.662018775939941,
      "epoch": 1.73931141801206,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004695761133199275,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19319060295820237,
      "num_tokens": 38766616.0,
      "step": 22355
    },
    {
      "entropy": 5.570777225494385,
      "epoch": 1.7397004473837776,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695621945659074,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20946432203054427,
      "num_tokens": 38775470.0,
      "step": 22360
    },
    {
      "entropy": 5.523472929000855,
      "epoch": 1.740089476755495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004695482728597157,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20536706894636153,
      "num_tokens": 38784566.0,
      "step": 22365
    },
    {
      "entropy": 5.5687579154968265,
      "epoch": 1.7404785061272126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046953434820156363,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2003973737359047,
      "num_tokens": 38792843.0,
      "step": 22370
    },
    {
      "entropy": 5.562903642654419,
      "epoch": 1.7408675354989303,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004695204205916624,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20416863709688188,
      "num_tokens": 38801425.0,
      "step": 22375
    },
    {
      "entropy": 5.578148078918457,
      "epoch": 1.7412565648706477,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004695064900302235,
      "loss": 5.046,
      "mean_token_accuracy": 0.20185983031988144,
      "num_tokens": 38810263.0,
      "step": 22380
    },
    {
      "entropy": 5.624987459182739,
      "epoch": 1.7416455942423652,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004694925565174581,
      "loss": 5.0627,
      "mean_token_accuracy": 0.19516354352235793,
      "num_tokens": 38818675.0,
      "step": 22385
    },
    {
      "entropy": 5.596960639953613,
      "epoch": 1.742034623614083,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046947862005357777,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19634397178888321,
      "num_tokens": 38827540.0,
      "step": 22390
    },
    {
      "entropy": 5.567427253723144,
      "epoch": 1.7424236529858004,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004694646806387939,
      "loss": 4.8947,
      "mean_token_accuracy": 0.21020916998386383,
      "num_tokens": 38837420.0,
      "step": 22395
    },
    {
      "entropy": 5.5541100025177,
      "epoch": 1.7428126823575179,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004694507382733181,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2062547743320465,
      "num_tokens": 38845604.0,
      "step": 22400
    },
    {
      "entropy": 5.515440130233765,
      "epoch": 1.7432017117292355,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004694367929573618,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2048572689294815,
      "num_tokens": 38854382.0,
      "step": 22405
    },
    {
      "entropy": 5.560438013076782,
      "epoch": 1.7435907411009532,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004694228446911367,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20275169163942336,
      "num_tokens": 38862889.0,
      "step": 22410
    },
    {
      "entropy": 5.6122509956359865,
      "epoch": 1.7439797704726707,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004694088934748542,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19653707891702651,
      "num_tokens": 38871640.0,
      "step": 22415
    },
    {
      "entropy": 5.523698997497559,
      "epoch": 1.7443687998443882,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004693949393087263,
      "loss": 4.9962,
      "mean_token_accuracy": 0.199367518723011,
      "num_tokens": 38880247.0,
      "step": 22420
    },
    {
      "entropy": 5.606223487854004,
      "epoch": 1.744757829216106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004693809821929647,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2004812091588974,
      "num_tokens": 38888857.0,
      "step": 22425
    },
    {
      "entropy": 5.634707260131836,
      "epoch": 1.7451468585878234,
      "grad_norm": 1.296875,
      "learning_rate": 0.000469367022127781,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19517911970615387,
      "num_tokens": 38897429.0,
      "step": 22430
    },
    {
      "entropy": 5.685098886489868,
      "epoch": 1.7455358879595408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046935305911338703,
      "loss": 5.2118,
      "mean_token_accuracy": 0.18373043984174728,
      "num_tokens": 38906730.0,
      "step": 22435
    },
    {
      "entropy": 5.6158239364624025,
      "epoch": 1.7459249173312585,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004693390931499948,
      "loss": 5.1343,
      "mean_token_accuracy": 0.1906909316778183,
      "num_tokens": 38916020.0,
      "step": 22440
    },
    {
      "entropy": 5.671753358840943,
      "epoch": 1.7463139467029762,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004693251242378162,
      "loss": 5.1038,
      "mean_token_accuracy": 0.19531947374343872,
      "num_tokens": 38925525.0,
      "step": 22445
    },
    {
      "entropy": 5.53682508468628,
      "epoch": 1.7467029760746935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046931115237706304,
      "loss": 4.8989,
      "mean_token_accuracy": 0.20466281622648239,
      "num_tokens": 38934270.0,
      "step": 22450
    },
    {
      "entropy": 5.560032796859741,
      "epoch": 1.7470920054464112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004692971775679475,
      "loss": 5.0517,
      "mean_token_accuracy": 0.1946267530322075,
      "num_tokens": 38943121.0,
      "step": 22455
    },
    {
      "entropy": 5.59547438621521,
      "epoch": 1.7474810348181289,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046928319981068154,
      "loss": 5.0493,
      "mean_token_accuracy": 0.19409110993146897,
      "num_tokens": 38952517.0,
      "step": 22460
    },
    {
      "entropy": 5.6408257484436035,
      "epoch": 1.7478700641898464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004692692191054773,
      "loss": 5.1324,
      "mean_token_accuracy": 0.1910709708929062,
      "num_tokens": 38961124.0,
      "step": 22465
    },
    {
      "entropy": 5.613063097000122,
      "epoch": 1.7482590935615638,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004692552354525468,
      "loss": 5.0413,
      "mean_token_accuracy": 0.1987176090478897,
      "num_tokens": 38969634.0,
      "step": 22470
    },
    {
      "entropy": 5.591943550109863,
      "epoch": 1.7486481229332815,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004692412488521023,
      "loss": 5.0816,
      "mean_token_accuracy": 0.19891417920589446,
      "num_tokens": 38978488.0,
      "step": 22475
    },
    {
      "entropy": 5.553132152557373,
      "epoch": 1.749037152304999,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469227259304356,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2062200576066971,
      "num_tokens": 38987012.0,
      "step": 22480
    },
    {
      "entropy": 5.595240259170533,
      "epoch": 1.7494261816767165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046921326680952023,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20488011091947556,
      "num_tokens": 38995254.0,
      "step": 22485
    },
    {
      "entropy": 5.6533815383911135,
      "epoch": 1.7498152110484342,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004691992713678072,
      "loss": 5.1596,
      "mean_token_accuracy": 0.18789693862199783,
      "num_tokens": 39003837.0,
      "step": 22490
    },
    {
      "entropy": 5.6098731517791744,
      "epoch": 1.7502042404201519,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004691852729794293,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20639403462409972,
      "num_tokens": 39012079.0,
      "step": 22495
    },
    {
      "entropy": 5.58169379234314,
      "epoch": 1.7505932697918691,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004691712716445991,
      "loss": 4.9848,
      "mean_token_accuracy": 0.20101028233766555,
      "num_tokens": 39021296.0,
      "step": 22500
    },
    {
      "entropy": 5.509814786911011,
      "epoch": 1.7509822991635868,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004691572673635288,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20133056342601777,
      "num_tokens": 39029557.0,
      "step": 22505
    },
    {
      "entropy": 5.572682952880859,
      "epoch": 1.7513713285353045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000469143260136431,
      "loss": 5.0638,
      "mean_token_accuracy": 0.19646933376789094,
      "num_tokens": 39037983.0,
      "step": 22510
    },
    {
      "entropy": 5.629957103729248,
      "epoch": 1.751760357907022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004691292499635183,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1992638736963272,
      "num_tokens": 39046874.0,
      "step": 22515
    },
    {
      "entropy": 5.541609430313111,
      "epoch": 1.7521493872787395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004691152368450032,
      "loss": 4.9398,
      "mean_token_accuracy": 0.2066134050488472,
      "num_tokens": 39054848.0,
      "step": 22520
    },
    {
      "entropy": 5.581417608261108,
      "epoch": 1.7525384166504572,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046910122078109835,
      "loss": 5.1059,
      "mean_token_accuracy": 0.1915219321846962,
      "num_tokens": 39063946.0,
      "step": 22525
    },
    {
      "entropy": 5.659631681442261,
      "epoch": 1.7529274460221747,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004690872017720163,
      "loss": 5.0392,
      "mean_token_accuracy": 0.20044841468334199,
      "num_tokens": 39072738.0,
      "step": 22530
    },
    {
      "entropy": 5.5723717212677,
      "epoch": 1.7533164753938921,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046907317981797006,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2027512699365616,
      "num_tokens": 39082449.0,
      "step": 22535
    },
    {
      "entropy": 5.5275407314300535,
      "epoch": 1.7537055047656098,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046905915491917213,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20208936035633088,
      "num_tokens": 39090892.0,
      "step": 22540
    },
    {
      "entropy": 5.630428457260132,
      "epoch": 1.7540945341373275,
      "grad_norm": 1.125,
      "learning_rate": 0.0004690451270758354,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19849861562252044,
      "num_tokens": 39099489.0,
      "step": 22545
    },
    {
      "entropy": 5.556709480285645,
      "epoch": 1.7544835635090448,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046903109628817276,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1989297479391098,
      "num_tokens": 39108057.0,
      "step": 22550
    },
    {
      "entropy": 5.636240243911743,
      "epoch": 1.7548725928807625,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046901706255639703,
      "loss": 5.1479,
      "mean_token_accuracy": 0.191128708422184,
      "num_tokens": 39116960.0,
      "step": 22555
    },
    {
      "entropy": 5.573104906082153,
      "epoch": 1.7552616222524802,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004690030258807212,
      "loss": 5.0212,
      "mean_token_accuracy": 0.19483115077018737,
      "num_tokens": 39126546.0,
      "step": 22560
    },
    {
      "entropy": 5.656208658218384,
      "epoch": 1.7556506516241976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046898898626135827,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19431264847517013,
      "num_tokens": 39135390.0,
      "step": 22565
    },
    {
      "entropy": 5.584502029418945,
      "epoch": 1.7560396809959151,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004689749436985211,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20115449279546738,
      "num_tokens": 39143182.0,
      "step": 22570
    },
    {
      "entropy": 5.635671520233155,
      "epoch": 1.7564287103676328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046896089819242304,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20134179294109344,
      "num_tokens": 39151525.0,
      "step": 22575
    },
    {
      "entropy": 5.608244276046753,
      "epoch": 1.7568177397393503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004689468497432771,
      "loss": 5.0974,
      "mean_token_accuracy": 0.19605673998594284,
      "num_tokens": 39160893.0,
      "step": 22580
    },
    {
      "entropy": 5.561341190338135,
      "epoch": 1.7572067691110678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004689327983512963,
      "loss": 4.9539,
      "mean_token_accuracy": 0.20230158120393754,
      "num_tokens": 39169682.0,
      "step": 22585
    },
    {
      "entropy": 5.635120582580567,
      "epoch": 1.7575957984827855,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046891874401669404,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19502562582492827,
      "num_tokens": 39178879.0,
      "step": 22590
    },
    {
      "entropy": 5.564996576309204,
      "epoch": 1.7579848278545032,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004689046867396834,
      "loss": 4.9697,
      "mean_token_accuracy": 0.19909309297800065,
      "num_tokens": 39186611.0,
      "step": 22595
    },
    {
      "entropy": 5.628575706481934,
      "epoch": 1.7583738572262204,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004688906265204779,
      "loss": 5.1428,
      "mean_token_accuracy": 0.18845791667699813,
      "num_tokens": 39195120.0,
      "step": 22600
    },
    {
      "entropy": 5.619038724899292,
      "epoch": 1.7587628865979381,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004688765633592907,
      "loss": 5.0338,
      "mean_token_accuracy": 0.19604857712984086,
      "num_tokens": 39203263.0,
      "step": 22605
    },
    {
      "entropy": 5.618705940246582,
      "epoch": 1.7591519159696558,
      "grad_norm": 1.125,
      "learning_rate": 0.0004688624972563352,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20032002329826354,
      "num_tokens": 39211486.0,
      "step": 22610
    },
    {
      "entropy": 5.524595928192139,
      "epoch": 1.7595409453413733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004688484282118249,
      "loss": 4.9488,
      "mean_token_accuracy": 0.20813057571649551,
      "num_tokens": 39220289.0,
      "step": 22615
    },
    {
      "entropy": 5.4896568775177,
      "epoch": 1.7599299747130908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004688343562259732,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2027808040380478,
      "num_tokens": 39229260.0,
      "step": 22620
    },
    {
      "entropy": 5.544314289093018,
      "epoch": 1.7603190040848085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004688202812989937,
      "loss": 5.0238,
      "mean_token_accuracy": 0.196371690928936,
      "num_tokens": 39237249.0,
      "step": 22625
    },
    {
      "entropy": 5.5726415634155275,
      "epoch": 1.760708033456526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004688062034311,
      "loss": 5.003,
      "mean_token_accuracy": 0.2005815714597702,
      "num_tokens": 39245973.0,
      "step": 22630
    },
    {
      "entropy": 5.650463199615478,
      "epoch": 1.7610970628282434,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046879212262250546,
      "loss": 5.0747,
      "mean_token_accuracy": 0.19567357748746872,
      "num_tokens": 39254488.0,
      "step": 22635
    },
    {
      "entropy": 5.523935270309448,
      "epoch": 1.761486092199961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046877803887342406,
      "loss": 5.0076,
      "mean_token_accuracy": 0.19874085783958434,
      "num_tokens": 39263596.0,
      "step": 22640
    },
    {
      "entropy": 5.576996278762818,
      "epoch": 1.7618751215716788,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004687639521840693,
      "loss": 4.9992,
      "mean_token_accuracy": 0.201176418364048,
      "num_tokens": 39272687.0,
      "step": 22645
    },
    {
      "entropy": 5.576949119567871,
      "epoch": 1.7622641509433963,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046874986255465495,
      "loss": 4.9831,
      "mean_token_accuracy": 0.19851980060338975,
      "num_tokens": 39281128.0,
      "step": 22650
    },
    {
      "entropy": 5.633670949935913,
      "epoch": 1.7626531803151138,
      "grad_norm": 1.125,
      "learning_rate": 0.00046873576998539485,
      "loss": 5.1046,
      "mean_token_accuracy": 0.1974513500928879,
      "num_tokens": 39290239.0,
      "step": 22655
    },
    {
      "entropy": 5.623898839950561,
      "epoch": 1.7630422096868315,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004687216744765028,
      "loss": 5.0781,
      "mean_token_accuracy": 0.19917575418949127,
      "num_tokens": 39299156.0,
      "step": 22660
    },
    {
      "entropy": 5.614734506607055,
      "epoch": 1.763431239058549,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004687075760281927,
      "loss": 5.0164,
      "mean_token_accuracy": 0.19556455910205842,
      "num_tokens": 39308081.0,
      "step": 22665
    },
    {
      "entropy": 5.589909458160401,
      "epoch": 1.7638202684302664,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004686934746406784,
      "loss": 5.0769,
      "mean_token_accuracy": 0.19757792353630066,
      "num_tokens": 39316219.0,
      "step": 22670
    },
    {
      "entropy": 5.593949937820435,
      "epoch": 1.764209297801984,
      "grad_norm": 1.125,
      "learning_rate": 0.00046867937031417397,
      "loss": 5.0519,
      "mean_token_accuracy": 0.19304109066724778,
      "num_tokens": 39324885.0,
      "step": 22675
    },
    {
      "entropy": 5.539704608917236,
      "epoch": 1.7645983271737016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004686652630488933,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20368342995643615,
      "num_tokens": 39333214.0,
      "step": 22680
    },
    {
      "entropy": 5.574097919464111,
      "epoch": 1.764987356545419,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046865115284505063,
      "loss": 5.0583,
      "mean_token_accuracy": 0.1962554082274437,
      "num_tokens": 39342020.0,
      "step": 22685
    },
    {
      "entropy": 5.641817617416382,
      "epoch": 1.7653763859171367,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046863703970285986,
      "loss": 5.061,
      "mean_token_accuracy": 0.19739140272140504,
      "num_tokens": 39351979.0,
      "step": 22690
    },
    {
      "entropy": 5.659980201721192,
      "epoch": 1.7657654152888544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004686229236225352,
      "loss": 5.1027,
      "mean_token_accuracy": 0.19560504853725433,
      "num_tokens": 39360218.0,
      "step": 22695
    },
    {
      "entropy": 5.5705938816070555,
      "epoch": 1.766154444660572,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004686088046042909,
      "loss": 5.058,
      "mean_token_accuracy": 0.1932186856865883,
      "num_tokens": 39369434.0,
      "step": 22700
    },
    {
      "entropy": 5.603907680511474,
      "epoch": 1.7665434740322894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046859468264834114,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20390002578496932,
      "num_tokens": 39377592.0,
      "step": 22705
    },
    {
      "entropy": 5.637730550765991,
      "epoch": 1.766932503404007,
      "grad_norm": 1.0,
      "learning_rate": 0.00046858055775490017,
      "loss": 5.0499,
      "mean_token_accuracy": 0.19977066665887833,
      "num_tokens": 39386792.0,
      "step": 22710
    },
    {
      "entropy": 5.622049856185913,
      "epoch": 1.7673215327757246,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004685664299241825,
      "loss": 5.04,
      "mean_token_accuracy": 0.1985357165336609,
      "num_tokens": 39395199.0,
      "step": 22715
    },
    {
      "entropy": 5.627436590194702,
      "epoch": 1.767710562147442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004685522991564022,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19292464554309846,
      "num_tokens": 39404319.0,
      "step": 22720
    },
    {
      "entropy": 5.624041509628296,
      "epoch": 1.7680995915191597,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685381654517738,
      "loss": 5.1011,
      "mean_token_accuracy": 0.19326256364583969,
      "num_tokens": 39413567.0,
      "step": 22725
    },
    {
      "entropy": 5.591647815704346,
      "epoch": 1.7684886208908772,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004685240288105119,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20360135585069655,
      "num_tokens": 39422394.0,
      "step": 22730
    },
    {
      "entropy": 5.5141438961029055,
      "epoch": 1.7688776502625947,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046850988923283085,
      "loss": 4.9114,
      "mean_token_accuracy": 0.20991896837949753,
      "num_tokens": 39432081.0,
      "step": 22735
    },
    {
      "entropy": 5.541791486740112,
      "epoch": 1.7692666796343124,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046849574671894523,
      "loss": 4.9539,
      "mean_token_accuracy": 0.20038838684558868,
      "num_tokens": 39440477.0,
      "step": 22740
    },
    {
      "entropy": 5.593561267852783,
      "epoch": 1.76965570900603,
      "grad_norm": 1.125,
      "learning_rate": 0.00046848160126906956,
      "loss": 5.1371,
      "mean_token_accuracy": 0.19569425731897355,
      "num_tokens": 39449253.0,
      "step": 22745
    },
    {
      "entropy": 5.589076328277588,
      "epoch": 1.7700447383777476,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004684674528834186,
      "loss": 5.0174,
      "mean_token_accuracy": 0.19810190051794052,
      "num_tokens": 39457686.0,
      "step": 22750
    },
    {
      "entropy": 5.652772426605225,
      "epoch": 1.770433767749465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046845330156220703,
      "loss": 5.1608,
      "mean_token_accuracy": 0.19582921862602234,
      "num_tokens": 39466198.0,
      "step": 22755
    },
    {
      "entropy": 5.59023003578186,
      "epoch": 1.7708227971211827,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004684391473056495,
      "loss": 5.0355,
      "mean_token_accuracy": 0.19748512506484986,
      "num_tokens": 39475098.0,
      "step": 22760
    },
    {
      "entropy": 5.636324119567871,
      "epoch": 1.7712118264929002,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046842499011396076,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20220395624637605,
      "num_tokens": 39483383.0,
      "step": 22765
    },
    {
      "entropy": 5.538096189498901,
      "epoch": 1.7716008558646177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046841082998735575,
      "loss": 4.9733,
      "mean_token_accuracy": 0.20839557945728301,
      "num_tokens": 39492428.0,
      "step": 22770
    },
    {
      "entropy": 5.659812688827515,
      "epoch": 1.7719898852363354,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046839666692604916,
      "loss": 5.1327,
      "mean_token_accuracy": 0.19164248257875444,
      "num_tokens": 39501390.0,
      "step": 22775
    },
    {
      "entropy": 5.589624881744385,
      "epoch": 1.7723789146080529,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000468382500930256,
      "loss": 5.0458,
      "mean_token_accuracy": 0.19889655858278274,
      "num_tokens": 39509463.0,
      "step": 22780
    },
    {
      "entropy": 5.577083015441895,
      "epoch": 1.7727679439797703,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046836833200019116,
      "loss": 4.994,
      "mean_token_accuracy": 0.2068038836121559,
      "num_tokens": 39518275.0,
      "step": 22785
    },
    {
      "entropy": 5.511835384368896,
      "epoch": 1.773156973351488,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004683541601360697,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20563874989748002,
      "num_tokens": 39527349.0,
      "step": 22790
    },
    {
      "entropy": 5.560085248947144,
      "epoch": 1.7735460027232057,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046833998533810653,
      "loss": 5.007,
      "mean_token_accuracy": 0.20325085371732712,
      "num_tokens": 39535411.0,
      "step": 22795
    },
    {
      "entropy": 5.6095030307769775,
      "epoch": 1.7739350320949232,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683258076065169,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19274890124797822,
      "num_tokens": 39543927.0,
      "step": 22800
    },
    {
      "entropy": 5.5761925220489506,
      "epoch": 1.7743240614666407,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046831162694151586,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21392301917076112,
      "num_tokens": 39551782.0,
      "step": 22805
    },
    {
      "entropy": 5.58159122467041,
      "epoch": 1.7747130908383584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046829744334331855,
      "loss": 5.1773,
      "mean_token_accuracy": 0.18794430494308473,
      "num_tokens": 39561369.0,
      "step": 22810
    },
    {
      "entropy": 5.5563068866729735,
      "epoch": 1.7751021202100759,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046828325681214013,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19780379682779312,
      "num_tokens": 39569559.0,
      "step": 22815
    },
    {
      "entropy": 5.552534437179565,
      "epoch": 1.7754911495817933,
      "grad_norm": 1.171875,
      "learning_rate": 0.000468269067348196,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20059692114591599,
      "num_tokens": 39578663.0,
      "step": 22820
    },
    {
      "entropy": 5.618351697921753,
      "epoch": 1.775880178953511,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004682548749517014,
      "loss": 5.2523,
      "mean_token_accuracy": 0.18005671948194504,
      "num_tokens": 39588336.0,
      "step": 22825
    },
    {
      "entropy": 5.647833251953125,
      "epoch": 1.7762692083252287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046824067962287163,
      "loss": 4.9938,
      "mean_token_accuracy": 0.20423362851142884,
      "num_tokens": 39596666.0,
      "step": 22830
    },
    {
      "entropy": 5.597459745407105,
      "epoch": 1.776658237696946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004682264813619221,
      "loss": 5.0481,
      "mean_token_accuracy": 0.19661538898944855,
      "num_tokens": 39604916.0,
      "step": 22835
    },
    {
      "entropy": 5.617653560638428,
      "epoch": 1.7770472670686637,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046821228016906836,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19308310151100158,
      "num_tokens": 39613808.0,
      "step": 22840
    },
    {
      "entropy": 5.571138143539429,
      "epoch": 1.7774362964403814,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004681980760445257,
      "loss": 5.0378,
      "mean_token_accuracy": 0.19837915897369385,
      "num_tokens": 39622476.0,
      "step": 22845
    },
    {
      "entropy": 5.6759086608886715,
      "epoch": 1.7778253258120988,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004681838689885098,
      "loss": 5.0158,
      "mean_token_accuracy": 0.19964097142219545,
      "num_tokens": 39630626.0,
      "step": 22850
    },
    {
      "entropy": 5.560094547271729,
      "epoch": 1.7782143551838163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004681696590012362,
      "loss": 5.0125,
      "mean_token_accuracy": 0.19630029797554016,
      "num_tokens": 39640067.0,
      "step": 22855
    },
    {
      "entropy": 5.559118461608887,
      "epoch": 1.778603384555534,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046815544608292043,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2069356307387352,
      "num_tokens": 39649576.0,
      "step": 22860
    },
    {
      "entropy": 5.583879375457764,
      "epoch": 1.7789924139272515,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004681412302337782,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20486710965633392,
      "num_tokens": 39658286.0,
      "step": 22865
    },
    {
      "entropy": 5.607027578353882,
      "epoch": 1.779381443298969,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004681270114540252,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20192814320325853,
      "num_tokens": 39667147.0,
      "step": 22870
    },
    {
      "entropy": 5.504190874099732,
      "epoch": 1.7797704726706867,
      "grad_norm": 1.25,
      "learning_rate": 0.0004681127897438772,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2061097353696823,
      "num_tokens": 39676069.0,
      "step": 22875
    },
    {
      "entropy": 5.598902893066406,
      "epoch": 1.7801595020424044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046809856510355007,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2045866534113884,
      "num_tokens": 39684378.0,
      "step": 22880
    },
    {
      "entropy": 5.598447513580322,
      "epoch": 1.7805485314141216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004680843375332595,
      "loss": 5.1119,
      "mean_token_accuracy": 0.19898812770843505,
      "num_tokens": 39693808.0,
      "step": 22885
    },
    {
      "entropy": 5.609540128707886,
      "epoch": 1.7809375607858393,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046807010703322143,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20833540856838226,
      "num_tokens": 39702375.0,
      "step": 22890
    },
    {
      "entropy": 5.709212923049927,
      "epoch": 1.781326590157557,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004680558736036518,
      "loss": 5.0902,
      "mean_token_accuracy": 0.19484387934207917,
      "num_tokens": 39711052.0,
      "step": 22895
    },
    {
      "entropy": 5.462477111816407,
      "epoch": 1.7817156195292745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004680416372447666,
      "loss": 4.7844,
      "mean_token_accuracy": 0.21905156672000886,
      "num_tokens": 39718942.0,
      "step": 22900
    },
    {
      "entropy": 5.6332207202911375,
      "epoch": 1.782104648900992,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046802739795678177,
      "loss": 5.1211,
      "mean_token_accuracy": 0.19188966304063798,
      "num_tokens": 39728048.0,
      "step": 22905
    },
    {
      "entropy": 5.552681255340576,
      "epoch": 1.7824936782727097,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046801315573991346,
      "loss": 5.0024,
      "mean_token_accuracy": 0.2045807123184204,
      "num_tokens": 39736579.0,
      "step": 22910
    },
    {
      "entropy": 5.561276626586914,
      "epoch": 1.7828827076444271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046799891059437764,
      "loss": 5.0199,
      "mean_token_accuracy": 0.19995367676019668,
      "num_tokens": 39744963.0,
      "step": 22915
    },
    {
      "entropy": 5.525426387786865,
      "epoch": 1.7832717370161446,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046798466252039056,
      "loss": 5.022,
      "mean_token_accuracy": 0.2055957704782486,
      "num_tokens": 39752880.0,
      "step": 22920
    },
    {
      "entropy": 5.540273427963257,
      "epoch": 1.7836607663878623,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046797041151816845,
      "loss": 4.9185,
      "mean_token_accuracy": 0.20775382071733475,
      "num_tokens": 39761069.0,
      "step": 22925
    },
    {
      "entropy": 5.672127723693848,
      "epoch": 1.78404979575958,
      "grad_norm": 1.125,
      "learning_rate": 0.00046795615758792747,
      "loss": 5.1242,
      "mean_token_accuracy": 0.19637293517589569,
      "num_tokens": 39769780.0,
      "step": 22930
    },
    {
      "entropy": 5.597632694244385,
      "epoch": 1.7844388251312973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004679419007298839,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20212045162916184,
      "num_tokens": 39778539.0,
      "step": 22935
    },
    {
      "entropy": 5.5482048988342285,
      "epoch": 1.784827854503015,
      "grad_norm": 1.125,
      "learning_rate": 0.0004679276409442541,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19957920014858246,
      "num_tokens": 39787175.0,
      "step": 22940
    },
    {
      "entropy": 5.609476041793823,
      "epoch": 1.7852168838747327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679133782312544,
      "loss": 5.1294,
      "mean_token_accuracy": 0.1929687336087227,
      "num_tokens": 39795953.0,
      "step": 22945
    },
    {
      "entropy": 5.597094249725342,
      "epoch": 1.7856059132464501,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004678991125911013,
      "loss": 4.9866,
      "mean_token_accuracy": 0.19987092167139053,
      "num_tokens": 39804858.0,
      "step": 22950
    },
    {
      "entropy": 5.608408117294312,
      "epoch": 1.7859949426181676,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004678848440240111,
      "loss": 5.1099,
      "mean_token_accuracy": 0.18953849524259567,
      "num_tokens": 39813091.0,
      "step": 22955
    },
    {
      "entropy": 5.5701898574829105,
      "epoch": 1.7863839719898853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004678705725302005,
      "loss": 5.0865,
      "mean_token_accuracy": 0.19091576784849168,
      "num_tokens": 39821489.0,
      "step": 22960
    },
    {
      "entropy": 5.623002290725708,
      "epoch": 1.7867730013616028,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004678562981098859,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19759983718395233,
      "num_tokens": 39830236.0,
      "step": 22965
    },
    {
      "entropy": 5.558730363845825,
      "epoch": 1.7871620307333203,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046784202076328394,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20138382762670518,
      "num_tokens": 39839064.0,
      "step": 22970
    },
    {
      "entropy": 5.587292242050171,
      "epoch": 1.787551060105038,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046782774049061124,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19815467447042465,
      "num_tokens": 39848101.0,
      "step": 22975
    },
    {
      "entropy": 5.704869985580444,
      "epoch": 1.7879400894767556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004678134572920845,
      "loss": 5.214,
      "mean_token_accuracy": 0.1935403600335121,
      "num_tokens": 39857053.0,
      "step": 22980
    },
    {
      "entropy": 5.629877281188965,
      "epoch": 1.788329118848473,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004677991711679204,
      "loss": 5.1246,
      "mean_token_accuracy": 0.19663288593292236,
      "num_tokens": 39865278.0,
      "step": 22985
    },
    {
      "entropy": 5.589210939407349,
      "epoch": 1.7887181482201906,
      "grad_norm": 1.125,
      "learning_rate": 0.00046778488211833585,
      "loss": 5.0211,
      "mean_token_accuracy": 0.1959928184747696,
      "num_tokens": 39874296.0,
      "step": 22990
    },
    {
      "entropy": 5.509317874908447,
      "epoch": 1.7891071775919083,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004677705901435475,
      "loss": 4.9337,
      "mean_token_accuracy": 0.209151254594326,
      "num_tokens": 39882867.0,
      "step": 22995
    },
    {
      "entropy": 5.560981607437133,
      "epoch": 1.7894962069636258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004677562952437723,
      "loss": 5.0376,
      "mean_token_accuracy": 0.19653394669294358,
      "num_tokens": 39891782.0,
      "step": 23000
    },
    {
      "entropy": 5.602384281158447,
      "epoch": 1.7898852363353432,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046774199741922705,
      "loss": 5.087,
      "mean_token_accuracy": 0.20199912190437316,
      "num_tokens": 39899914.0,
      "step": 23005
    },
    {
      "entropy": 5.630718755722046,
      "epoch": 1.790274265707061,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046772769667012884,
      "loss": 5.0478,
      "mean_token_accuracy": 0.19501936584711074,
      "num_tokens": 39909097.0,
      "step": 23010
    },
    {
      "entropy": 5.576741313934326,
      "epoch": 1.7906632950787784,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004677133929966946,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20243656039237976,
      "num_tokens": 39917922.0,
      "step": 23015
    },
    {
      "entropy": 5.590720558166504,
      "epoch": 1.791052324450496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046769908639914134,
      "loss": 4.954,
      "mean_token_accuracy": 0.20219176411628723,
      "num_tokens": 39925675.0,
      "step": 23020
    },
    {
      "entropy": 5.617328739166259,
      "epoch": 1.7914413538222136,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004676847768776861,
      "loss": 5.0801,
      "mean_token_accuracy": 0.19463513940572738,
      "num_tokens": 39933693.0,
      "step": 23025
    },
    {
      "entropy": 5.614121198654175,
      "epoch": 1.7918303831939313,
      "grad_norm": 1.25,
      "learning_rate": 0.0004676704644325462,
      "loss": 5.0974,
      "mean_token_accuracy": 0.19787005186080933,
      "num_tokens": 39942137.0,
      "step": 23030
    },
    {
      "entropy": 5.638820219039917,
      "epoch": 1.7922194125656488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004676561490639385,
      "loss": 5.0483,
      "mean_token_accuracy": 0.19255973100662233,
      "num_tokens": 39951268.0,
      "step": 23035
    },
    {
      "entropy": 5.569800138473511,
      "epoch": 1.7926084419373662,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004676418307720805,
      "loss": 4.9252,
      "mean_token_accuracy": 0.20710771083831786,
      "num_tokens": 39959930.0,
      "step": 23040
    },
    {
      "entropy": 5.470084762573242,
      "epoch": 1.792997471309084,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004676275095571892,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21053647696971894,
      "num_tokens": 39967694.0,
      "step": 23045
    },
    {
      "entropy": 5.554301786422729,
      "epoch": 1.7933865006808014,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046761318541948215,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20855779498815535,
      "num_tokens": 39976462.0,
      "step": 23050
    },
    {
      "entropy": 5.558436870574951,
      "epoch": 1.7937755300525189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004675988583591766,
      "loss": 4.9366,
      "mean_token_accuracy": 0.20703832060098648,
      "num_tokens": 39985089.0,
      "step": 23055
    },
    {
      "entropy": 5.628482103347778,
      "epoch": 1.7941645594242366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046758452837648997,
      "loss": 5.1046,
      "mean_token_accuracy": 0.19303470999002456,
      "num_tokens": 39994635.0,
      "step": 23060
    },
    {
      "entropy": 5.672496795654297,
      "epoch": 1.794553588795954,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004675701954716395,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20060535222291948,
      "num_tokens": 40003474.0,
      "step": 23065
    },
    {
      "entropy": 5.623958158493042,
      "epoch": 1.7949426181676715,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046755585964484286,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2043328434228897,
      "num_tokens": 40012171.0,
      "step": 23070
    },
    {
      "entropy": 5.635309553146362,
      "epoch": 1.7953316475393892,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004675415208963177,
      "loss": 5.0737,
      "mean_token_accuracy": 0.19803963005542755,
      "num_tokens": 40021101.0,
      "step": 23075
    },
    {
      "entropy": 5.5936370372772215,
      "epoch": 1.795720676911107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046752717922628125,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20822370797395706,
      "num_tokens": 40029699.0,
      "step": 23080
    },
    {
      "entropy": 5.627665090560913,
      "epoch": 1.7961097062828244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004675128346349514,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19998870342969893,
      "num_tokens": 40038443.0,
      "step": 23085
    },
    {
      "entropy": 5.580521821975708,
      "epoch": 1.7964987356545419,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046749848712254554,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20565702766180038,
      "num_tokens": 40046398.0,
      "step": 23090
    },
    {
      "entropy": 5.57363657951355,
      "epoch": 1.7968877650262596,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046748413668928164,
      "loss": 5.0242,
      "mean_token_accuracy": 0.19836025685071945,
      "num_tokens": 40054487.0,
      "step": 23095
    },
    {
      "entropy": 5.546409177780151,
      "epoch": 1.797276794397977,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004674697833353774,
      "loss": 5.0161,
      "mean_token_accuracy": 0.1999297559261322,
      "num_tokens": 40063088.0,
      "step": 23100
    },
    {
      "entropy": 5.619241333007812,
      "epoch": 1.7976658237696945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046745542706105045,
      "loss": 5.0609,
      "mean_token_accuracy": 0.19449102580547334,
      "num_tokens": 40071321.0,
      "step": 23105
    },
    {
      "entropy": 5.570852136611938,
      "epoch": 1.7980548531414122,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046744106786651875,
      "loss": 4.9657,
      "mean_token_accuracy": 0.2001722827553749,
      "num_tokens": 40080329.0,
      "step": 23110
    },
    {
      "entropy": 5.611421251296997,
      "epoch": 1.7984438825131297,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004674267057520001,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20843944996595382,
      "num_tokens": 40088548.0,
      "step": 23115
    },
    {
      "entropy": 5.64282865524292,
      "epoch": 1.7988329118848472,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004674123407177125,
      "loss": 5.0917,
      "mean_token_accuracy": 0.19349610209465026,
      "num_tokens": 40096517.0,
      "step": 23120
    },
    {
      "entropy": 5.55327615737915,
      "epoch": 1.7992219412565649,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046739797276387394,
      "loss": 5.0315,
      "mean_token_accuracy": 0.20174278020858766,
      "num_tokens": 40104838.0,
      "step": 23125
    },
    {
      "entropy": 5.632558012008667,
      "epoch": 1.7996109706282826,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046738360189070235,
      "loss": 5.0388,
      "mean_token_accuracy": 0.20199909061193466,
      "num_tokens": 40113095.0,
      "step": 23130
    },
    {
      "entropy": 5.570906782150269,
      "epoch": 1.8,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004673692280984157,
      "loss": 4.9732,
      "mean_token_accuracy": 0.20213625431060792,
      "num_tokens": 40121436.0,
      "step": 23135
    },
    {
      "entropy": 5.504435586929321,
      "epoch": 1.8003890293717175,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004673548513872324,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2140723541378975,
      "num_tokens": 40130495.0,
      "step": 23140
    },
    {
      "entropy": 5.61940565109253,
      "epoch": 1.8007780587434352,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046734047175737026,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20216100364923478,
      "num_tokens": 40138456.0,
      "step": 23145
    },
    {
      "entropy": 5.553161573410034,
      "epoch": 1.8011670881151527,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046732608920904754,
      "loss": 4.903,
      "mean_token_accuracy": 0.21437480300664902,
      "num_tokens": 40146898.0,
      "step": 23150
    },
    {
      "entropy": 5.53667311668396,
      "epoch": 1.8015561174868702,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673117037424827,
      "loss": 5.0302,
      "mean_token_accuracy": 0.19319395571947098,
      "num_tokens": 40155461.0,
      "step": 23155
    },
    {
      "entropy": 5.587999534606934,
      "epoch": 1.8019451468585879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046729731535789375,
      "loss": 5.0201,
      "mean_token_accuracy": 0.196914441883564,
      "num_tokens": 40164111.0,
      "step": 23160
    },
    {
      "entropy": 5.645179700851441,
      "epoch": 1.8023341762303053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046728292405549913,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20021609514951705,
      "num_tokens": 40172306.0,
      "step": 23165
    },
    {
      "entropy": 5.590686893463134,
      "epoch": 1.8027232056020228,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004672685298355172,
      "loss": 5.0524,
      "mean_token_accuracy": 0.19998696446418762,
      "num_tokens": 40180720.0,
      "step": 23170
    },
    {
      "entropy": 5.51874475479126,
      "epoch": 1.8031122349737405,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004672541326981664,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2108146384358406,
      "num_tokens": 40189536.0,
      "step": 23175
    },
    {
      "entropy": 5.590334844589234,
      "epoch": 1.8035012643454582,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000467239732643665,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20336873084306717,
      "num_tokens": 40198510.0,
      "step": 23180
    },
    {
      "entropy": 5.6000707149505615,
      "epoch": 1.8038902937171757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046722532967223183,
      "loss": 5.0842,
      "mean_token_accuracy": 0.19700941443443298,
      "num_tokens": 40206959.0,
      "step": 23185
    },
    {
      "entropy": 5.623350000381469,
      "epoch": 1.8042793230888932,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004672109237840851,
      "loss": 5.0908,
      "mean_token_accuracy": 0.19866990000009538,
      "num_tokens": 40215661.0,
      "step": 23190
    },
    {
      "entropy": 5.597944211959839,
      "epoch": 1.8046683524606109,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046719651497944355,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19420279413461686,
      "num_tokens": 40223221.0,
      "step": 23195
    },
    {
      "entropy": 5.511111068725586,
      "epoch": 1.8050573818323283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671821032585259,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2047031491994858,
      "num_tokens": 40231846.0,
      "step": 23200
    },
    {
      "entropy": 5.5718762397766115,
      "epoch": 1.8054464112040458,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004671676886215506,
      "loss": 5.0265,
      "mean_token_accuracy": 0.19999951422214507,
      "num_tokens": 40240538.0,
      "step": 23205
    },
    {
      "entropy": 5.612865829467774,
      "epoch": 1.8058354405757635,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671532710687365,
      "loss": 5.1311,
      "mean_token_accuracy": 0.19003996700048448,
      "num_tokens": 40249260.0,
      "step": 23210
    },
    {
      "entropy": 5.568214082717896,
      "epoch": 1.806224469947481,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004671388506003024,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20145997256040574,
      "num_tokens": 40258458.0,
      "step": 23215
    },
    {
      "entropy": 5.585395479202271,
      "epoch": 1.8066134993191985,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004671244272164671,
      "loss": 5.0676,
      "mean_token_accuracy": 0.19343768507242204,
      "num_tokens": 40267304.0,
      "step": 23220
    },
    {
      "entropy": 5.575503826141357,
      "epoch": 1.8070025286909162,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046711000091744935,
      "loss": 5.0564,
      "mean_token_accuracy": 0.19747590124607087,
      "num_tokens": 40276873.0,
      "step": 23225
    },
    {
      "entropy": 5.554568338394165,
      "epoch": 1.8073915580626339,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004670955717034681,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2034117043018341,
      "num_tokens": 40285003.0,
      "step": 23230
    },
    {
      "entropy": 5.527369594573974,
      "epoch": 1.8077805874343513,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046708113957474233,
      "loss": 5.0258,
      "mean_token_accuracy": 0.203046777844429,
      "num_tokens": 40294462.0,
      "step": 23235
    },
    {
      "entropy": 5.604878377914429,
      "epoch": 1.8081696168060688,
      "grad_norm": 1.28125,
      "learning_rate": 0.000467066704531491,
      "loss": 5.0003,
      "mean_token_accuracy": 0.202876915037632,
      "num_tokens": 40303040.0,
      "step": 23240
    },
    {
      "entropy": 5.644914150238037,
      "epoch": 1.8085586461777865,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004670522665739332,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19444627612829207,
      "num_tokens": 40311490.0,
      "step": 23245
    },
    {
      "entropy": 5.661421155929565,
      "epoch": 1.808947675549504,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004670378257022878,
      "loss": 5.104,
      "mean_token_accuracy": 0.19724383503198623,
      "num_tokens": 40319725.0,
      "step": 23250
    },
    {
      "entropy": 5.611379337310791,
      "epoch": 1.8093367049212215,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046702338191677414,
      "loss": 5.0328,
      "mean_token_accuracy": 0.20321171283721923,
      "num_tokens": 40328019.0,
      "step": 23255
    },
    {
      "entropy": 5.6197587013244625,
      "epoch": 1.8097257342929391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004670089352176113,
      "loss": 5.013,
      "mean_token_accuracy": 0.20126061588525773,
      "num_tokens": 40336857.0,
      "step": 23260
    },
    {
      "entropy": 5.552067136764526,
      "epoch": 1.8101147636646568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046699448560501847,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1964714601635933,
      "num_tokens": 40345367.0,
      "step": 23265
    },
    {
      "entropy": 5.634008169174194,
      "epoch": 1.810503793036374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004669800330792149,
      "loss": 5.071,
      "mean_token_accuracy": 0.19731670916080474,
      "num_tokens": 40353899.0,
      "step": 23270
    },
    {
      "entropy": 5.634049606323242,
      "epoch": 1.8108928224080918,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046696557764041994,
      "loss": 5.1067,
      "mean_token_accuracy": 0.19399033039808272,
      "num_tokens": 40362045.0,
      "step": 23275
    },
    {
      "entropy": 5.57297568321228,
      "epoch": 1.8112818517798095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004669511192888528,
      "loss": 4.9418,
      "mean_token_accuracy": 0.20357016026973723,
      "num_tokens": 40370581.0,
      "step": 23280
    },
    {
      "entropy": 5.538308000564575,
      "epoch": 1.811670881151527,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046693665802473294,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21260672062635422,
      "num_tokens": 40379481.0,
      "step": 23285
    },
    {
      "entropy": 5.662737894058227,
      "epoch": 1.8120599105232444,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046692219384827986,
      "loss": 5.1325,
      "mean_token_accuracy": 0.18956195712089538,
      "num_tokens": 40387839.0,
      "step": 23290
    },
    {
      "entropy": 5.625577545166015,
      "epoch": 1.8124489398949621,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004669077267597129,
      "loss": 5.0074,
      "mean_token_accuracy": 0.1966658875346184,
      "num_tokens": 40396030.0,
      "step": 23295
    },
    {
      "entropy": 5.6442272663116455,
      "epoch": 1.8128379692666796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046689325675925174,
      "loss": 5.0927,
      "mean_token_accuracy": 0.19145426899194717,
      "num_tokens": 40404413.0,
      "step": 23300
    },
    {
      "entropy": 5.595897483825683,
      "epoch": 1.813226998638397,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046687878384711574,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2019767388701439,
      "num_tokens": 40413474.0,
      "step": 23305
    },
    {
      "entropy": 5.522207307815552,
      "epoch": 1.8136160280101148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046686430802352476,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20177856534719468,
      "num_tokens": 40422564.0,
      "step": 23310
    },
    {
      "entropy": 5.5636693954467775,
      "epoch": 1.8140050573818325,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004668498292886982,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20211093723773957,
      "num_tokens": 40430371.0,
      "step": 23315
    },
    {
      "entropy": 5.649656915664673,
      "epoch": 1.8143940867535497,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046683534764285577,
      "loss": 5.1286,
      "mean_token_accuracy": 0.18882469236850738,
      "num_tokens": 40439346.0,
      "step": 23320
    },
    {
      "entropy": 5.720078229904175,
      "epoch": 1.8147831161252674,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046682086308621725,
      "loss": 5.1482,
      "mean_token_accuracy": 0.19283961802721022,
      "num_tokens": 40449503.0,
      "step": 23325
    },
    {
      "entropy": 5.612382411956787,
      "epoch": 1.8151721454969851,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004668063756190026,
      "loss": 5.056,
      "mean_token_accuracy": 0.19769258052110672,
      "num_tokens": 40457372.0,
      "step": 23330
    },
    {
      "entropy": 5.6008195877075195,
      "epoch": 1.8155611748687026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046679188524143136,
      "loss": 5.0643,
      "mean_token_accuracy": 0.19532227516174316,
      "num_tokens": 40465731.0,
      "step": 23335
    },
    {
      "entropy": 5.637818384170532,
      "epoch": 1.81595020424042,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004667773919537235,
      "loss": 5.0968,
      "mean_token_accuracy": 0.1865835666656494,
      "num_tokens": 40473941.0,
      "step": 23340
    },
    {
      "entropy": 5.688023567199707,
      "epoch": 1.8163392336121378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046676289575609914,
      "loss": 5.1647,
      "mean_token_accuracy": 0.19164917916059493,
      "num_tokens": 40482370.0,
      "step": 23345
    },
    {
      "entropy": 5.567803525924683,
      "epoch": 1.8167282629838553,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046674839664877796,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20602133572101594,
      "num_tokens": 40489975.0,
      "step": 23350
    },
    {
      "entropy": 5.624909067153931,
      "epoch": 1.8171172923555727,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004667338946319801,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19751333743333815,
      "num_tokens": 40499294.0,
      "step": 23355
    },
    {
      "entropy": 5.648792171478272,
      "epoch": 1.8175063217272904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004667193897059255,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19113535135984422,
      "num_tokens": 40508087.0,
      "step": 23360
    },
    {
      "entropy": 5.674963140487671,
      "epoch": 1.8178953510990081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046670488187083436,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2037469670176506,
      "num_tokens": 40516054.0,
      "step": 23365
    },
    {
      "entropy": 5.590182161331176,
      "epoch": 1.8182843804707254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004666903711269267,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20195274651050568,
      "num_tokens": 40524640.0,
      "step": 23370
    },
    {
      "entropy": 5.525883722305298,
      "epoch": 1.818673409842443,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004666758574744228,
      "loss": 4.974,
      "mean_token_accuracy": 0.19963400512933732,
      "num_tokens": 40533255.0,
      "step": 23375
    },
    {
      "entropy": 5.552801179885864,
      "epoch": 1.8190624392141608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004666613409135429,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20008007884025575,
      "num_tokens": 40542137.0,
      "step": 23380
    },
    {
      "entropy": 5.623146343231201,
      "epoch": 1.8194514685858783,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004666468214445072,
      "loss": 5.0489,
      "mean_token_accuracy": 0.19393593072891235,
      "num_tokens": 40550577.0,
      "step": 23385
    },
    {
      "entropy": 5.57143201828003,
      "epoch": 1.8198404979575957,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046663229906753595,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2025732308626175,
      "num_tokens": 40559203.0,
      "step": 23390
    },
    {
      "entropy": 5.57795205116272,
      "epoch": 1.8202295273293134,
      "grad_norm": 1.125,
      "learning_rate": 0.00046661777378284965,
      "loss": 4.9417,
      "mean_token_accuracy": 0.20421261042356492,
      "num_tokens": 40568254.0,
      "step": 23395
    },
    {
      "entropy": 5.6695321559906,
      "epoch": 1.820618556701031,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004666032455906685,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19353146851062775,
      "num_tokens": 40576659.0,
      "step": 23400
    },
    {
      "entropy": 5.6467445373535154,
      "epoch": 1.8210075860727484,
      "grad_norm": 1.125,
      "learning_rate": 0.00046658871449121325,
      "loss": 5.0274,
      "mean_token_accuracy": 0.20692488700151443,
      "num_tokens": 40585544.0,
      "step": 23405
    },
    {
      "entropy": 5.588854885101318,
      "epoch": 1.821396615444466,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004665741804847041,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20389530211687087,
      "num_tokens": 40594763.0,
      "step": 23410
    },
    {
      "entropy": 5.560424137115478,
      "epoch": 1.8217856448161838,
      "grad_norm": 1.125,
      "learning_rate": 0.00046655964357136164,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20090416222810745,
      "num_tokens": 40604350.0,
      "step": 23415
    },
    {
      "entropy": 5.60378589630127,
      "epoch": 1.822174674187901,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046654510375140657,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1935052141547203,
      "num_tokens": 40612799.0,
      "step": 23420
    },
    {
      "entropy": 5.675509071350097,
      "epoch": 1.8225637035596187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004665305610250594,
      "loss": 5.1203,
      "mean_token_accuracy": 0.19847144633531572,
      "num_tokens": 40621390.0,
      "step": 23425
    },
    {
      "entropy": 5.672425317764282,
      "epoch": 1.8229527329313364,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004665160153925408,
      "loss": 5.0845,
      "mean_token_accuracy": 0.19245277792215348,
      "num_tokens": 40629953.0,
      "step": 23430
    },
    {
      "entropy": 5.61926908493042,
      "epoch": 1.823341762303054,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046650146685407154,
      "loss": 5.0405,
      "mean_token_accuracy": 0.19744528830051422,
      "num_tokens": 40638663.0,
      "step": 23435
    },
    {
      "entropy": 5.688778734207153,
      "epoch": 1.8237307916747714,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046648691540987226,
      "loss": 5.0896,
      "mean_token_accuracy": 0.19857471734285354,
      "num_tokens": 40647693.0,
      "step": 23440
    },
    {
      "entropy": 5.629212284088135,
      "epoch": 1.824119821046489,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046647236106016387,
      "loss": 5.1158,
      "mean_token_accuracy": 0.19051593095064162,
      "num_tokens": 40656300.0,
      "step": 23445
    },
    {
      "entropy": 5.619501829147339,
      "epoch": 1.8245088504182065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664578038051672,
      "loss": 4.977,
      "mean_token_accuracy": 0.20272729992866517,
      "num_tokens": 40664085.0,
      "step": 23450
    },
    {
      "entropy": 5.618737316131591,
      "epoch": 1.824897879789924,
      "grad_norm": 1.15625,
      "learning_rate": 0.000466443243645103,
      "loss": 5.0154,
      "mean_token_accuracy": 0.20505328774452208,
      "num_tokens": 40672510.0,
      "step": 23455
    },
    {
      "entropy": 5.687450313568116,
      "epoch": 1.8252869091616417,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004664286805801923,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2021819159388542,
      "num_tokens": 40681273.0,
      "step": 23460
    },
    {
      "entropy": 5.585610818862915,
      "epoch": 1.8256759385333594,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004664141146106562,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20094319730997084,
      "num_tokens": 40689336.0,
      "step": 23465
    },
    {
      "entropy": 5.628610324859619,
      "epoch": 1.8260649679050769,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046639954573671547,
      "loss": 4.9596,
      "mean_token_accuracy": 0.20481908768415452,
      "num_tokens": 40697557.0,
      "step": 23470
    },
    {
      "entropy": 5.612027740478515,
      "epoch": 1.8264539972767944,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046638497395859127,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20712563395500183,
      "num_tokens": 40706129.0,
      "step": 23475
    },
    {
      "entropy": 5.563867855072021,
      "epoch": 1.826843026648512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004663703992765047,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20928830951452254,
      "num_tokens": 40714837.0,
      "step": 23480
    },
    {
      "entropy": 5.567662477493286,
      "epoch": 1.8272320560202295,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046635582169067693,
      "loss": 4.9489,
      "mean_token_accuracy": 0.20544512271881105,
      "num_tokens": 40723300.0,
      "step": 23485
    },
    {
      "entropy": 5.574252080917359,
      "epoch": 1.827621085391947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046634124120132915,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20527570992708205,
      "num_tokens": 40732337.0,
      "step": 23490
    },
    {
      "entropy": 5.680656671524048,
      "epoch": 1.8280101147636647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004663266578086826,
      "loss": 5.0452,
      "mean_token_accuracy": 0.19928997308015822,
      "num_tokens": 40741130.0,
      "step": 23495
    },
    {
      "entropy": 5.619625473022461,
      "epoch": 1.8283991441353822,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004663120715129585,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2007047563791275,
      "num_tokens": 40749686.0,
      "step": 23500
    },
    {
      "entropy": 5.625330400466919,
      "epoch": 1.8287881735070997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004662974823143783,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1989281490445137,
      "num_tokens": 40758181.0,
      "step": 23505
    },
    {
      "entropy": 5.6340419292449955,
      "epoch": 1.8291772028788174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004662828902131632,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20022480487823485,
      "num_tokens": 40767688.0,
      "step": 23510
    },
    {
      "entropy": 5.626961421966553,
      "epoch": 1.829566232250535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004662682952095348,
      "loss": 5.0009,
      "mean_token_accuracy": 0.19939922094345092,
      "num_tokens": 40776242.0,
      "step": 23515
    },
    {
      "entropy": 5.568482112884522,
      "epoch": 1.8299552616222525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046625369730371436,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2056102216243744,
      "num_tokens": 40784806.0,
      "step": 23520
    },
    {
      "entropy": 5.570433950424194,
      "epoch": 1.83034429099397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004662390964959235,
      "loss": 5.01,
      "mean_token_accuracy": 0.20464001297950746,
      "num_tokens": 40793823.0,
      "step": 23525
    },
    {
      "entropy": 5.581672954559326,
      "epoch": 1.8307333203656877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046622449278638375,
      "loss": 4.9175,
      "mean_token_accuracy": 0.20923811346292495,
      "num_tokens": 40802243.0,
      "step": 23530
    },
    {
      "entropy": 5.601703643798828,
      "epoch": 1.8311223497374052,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004662098861753167,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19091252982616425,
      "num_tokens": 40810641.0,
      "step": 23535
    },
    {
      "entropy": 5.541178274154663,
      "epoch": 1.8315113791091227,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046619527666294394,
      "loss": 4.9164,
      "mean_token_accuracy": 0.20411177724599838,
      "num_tokens": 40818967.0,
      "step": 23540
    },
    {
      "entropy": 5.668802261352539,
      "epoch": 1.8319004084808403,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004661806642494872,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19938068985939025,
      "num_tokens": 40827603.0,
      "step": 23545
    },
    {
      "entropy": 5.52700023651123,
      "epoch": 1.8322894378525578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004661660489351681,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20315391570329666,
      "num_tokens": 40836314.0,
      "step": 23550
    },
    {
      "entropy": 5.621227979660034,
      "epoch": 1.8326784672242753,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004661514307202086,
      "loss": 5.1486,
      "mean_token_accuracy": 0.19572524577379227,
      "num_tokens": 40845559.0,
      "step": 23555
    },
    {
      "entropy": 5.6129528999328615,
      "epoch": 1.833067496595993,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046613680960483036,
      "loss": 5.0313,
      "mean_token_accuracy": 0.19726206958293915,
      "num_tokens": 40853616.0,
      "step": 23560
    },
    {
      "entropy": 5.598602199554444,
      "epoch": 1.8334565259677107,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004661221855892552,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20465305596590042,
      "num_tokens": 40862919.0,
      "step": 23565
    },
    {
      "entropy": 5.568418836593628,
      "epoch": 1.8338455553394282,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046610755867370506,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20334067940711975,
      "num_tokens": 40872301.0,
      "step": 23570
    },
    {
      "entropy": 5.632410335540771,
      "epoch": 1.8342345847111456,
      "grad_norm": 2.15625,
      "learning_rate": 0.000466092928858402,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21323545277118683,
      "num_tokens": 40881964.0,
      "step": 23575
    },
    {
      "entropy": 5.648987770080566,
      "epoch": 1.8346236140828633,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046607829614356787,
      "loss": 5.0348,
      "mean_token_accuracy": 0.19798095226287843,
      "num_tokens": 40890801.0,
      "step": 23580
    },
    {
      "entropy": 5.596552133560181,
      "epoch": 1.8350126434545808,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004660636605294247,
      "loss": 4.935,
      "mean_token_accuracy": 0.207064488530159,
      "num_tokens": 40899993.0,
      "step": 23585
    },
    {
      "entropy": 5.594598960876465,
      "epoch": 1.8354016728262983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004660490220161946,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2030389666557312,
      "num_tokens": 40908329.0,
      "step": 23590
    },
    {
      "entropy": 5.586009550094604,
      "epoch": 1.835790702198016,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046603438060409966,
      "loss": 5.0982,
      "mean_token_accuracy": 0.19534702152013778,
      "num_tokens": 40916269.0,
      "step": 23595
    },
    {
      "entropy": 5.527313375473023,
      "epoch": 1.8361797315697335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000466019736293362,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21106261610984803,
      "num_tokens": 40924962.0,
      "step": 23600
    },
    {
      "entropy": 5.516798877716065,
      "epoch": 1.836568760941451,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046600508908420396,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20647090822458267,
      "num_tokens": 40933682.0,
      "step": 23605
    },
    {
      "entropy": 5.687616491317749,
      "epoch": 1.8369577903131686,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046599043897684766,
      "loss": 5.1482,
      "mean_token_accuracy": 0.19332049489021302,
      "num_tokens": 40943718.0,
      "step": 23610
    },
    {
      "entropy": 5.617151355743408,
      "epoch": 1.8373468196848863,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004659757859715155,
      "loss": 5.0339,
      "mean_token_accuracy": 0.19302555173635483,
      "num_tokens": 40952480.0,
      "step": 23615
    },
    {
      "entropy": 5.629652118682861,
      "epoch": 1.8377358490566038,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046596113006842975,
      "loss": 5.095,
      "mean_token_accuracy": 0.19981483966112137,
      "num_tokens": 40961439.0,
      "step": 23620
    },
    {
      "entropy": 5.632921409606934,
      "epoch": 1.8381248784283213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004659464712678128,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2003880113363266,
      "num_tokens": 40969882.0,
      "step": 23625
    },
    {
      "entropy": 5.576944780349732,
      "epoch": 1.838513907800039,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659318095698871,
      "loss": 5.0086,
      "mean_token_accuracy": 0.19245371371507644,
      "num_tokens": 40978685.0,
      "step": 23630
    },
    {
      "entropy": 5.5869471549987795,
      "epoch": 1.8389029371717565,
      "grad_norm": 1.015625,
      "learning_rate": 0.000465917144974875,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2001183420419693,
      "num_tokens": 40987476.0,
      "step": 23635
    },
    {
      "entropy": 5.570794582366943,
      "epoch": 1.839291966543474,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004659024774829992,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20841701328754425,
      "num_tokens": 40995754.0,
      "step": 23640
    },
    {
      "entropy": 5.577577924728393,
      "epoch": 1.8396809959151916,
      "grad_norm": 1.078125,
      "learning_rate": 0.000465887807094482,
      "loss": 4.973,
      "mean_token_accuracy": 0.20319348722696304,
      "num_tokens": 41004178.0,
      "step": 23645
    },
    {
      "entropy": 5.585713720321655,
      "epoch": 1.8400700252869093,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046587313380954635,
      "loss": 5.066,
      "mean_token_accuracy": 0.19207748770713806,
      "num_tokens": 41013211.0,
      "step": 23650
    },
    {
      "entropy": 5.667567539215088,
      "epoch": 1.8404590546586266,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046585845762841453,
      "loss": 5.1578,
      "mean_token_accuracy": 0.19035876393318177,
      "num_tokens": 41021870.0,
      "step": 23655
    },
    {
      "entropy": 5.705098628997803,
      "epoch": 1.8408480840303443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004658437785513095,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19650598764419555,
      "num_tokens": 41031647.0,
      "step": 23660
    },
    {
      "entropy": 5.610191679000854,
      "epoch": 1.841237113402062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004658290965784539,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19747078716754912,
      "num_tokens": 41041296.0,
      "step": 23665
    },
    {
      "entropy": 5.656903505325317,
      "epoch": 1.8416261427737795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004658144117100704,
      "loss": 5.1736,
      "mean_token_accuracy": 0.18967496752738952,
      "num_tokens": 41050716.0,
      "step": 23670
    },
    {
      "entropy": 5.634930372238159,
      "epoch": 1.842015172145497,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046579972394638204,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19532993733882903,
      "num_tokens": 41058777.0,
      "step": 23675
    },
    {
      "entropy": 5.6296008110046385,
      "epoch": 1.8424042015172146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046578503328761146,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19758632332086562,
      "num_tokens": 41067779.0,
      "step": 23680
    },
    {
      "entropy": 5.593875408172607,
      "epoch": 1.842793230888932,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046577033973398173,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20478224605321885,
      "num_tokens": 41075349.0,
      "step": 23685
    },
    {
      "entropy": 5.603654861450195,
      "epoch": 1.8431822602606496,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004657556432857157,
      "loss": 5.0504,
      "mean_token_accuracy": 0.19663824290037155,
      "num_tokens": 41084340.0,
      "step": 23690
    },
    {
      "entropy": 5.569454383850098,
      "epoch": 1.8435712896323673,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004657409439430364,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20277228504419326,
      "num_tokens": 41093018.0,
      "step": 23695
    },
    {
      "entropy": 5.597228908538819,
      "epoch": 1.843960319004085,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004657262417061669,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20528214871883393,
      "num_tokens": 41101492.0,
      "step": 23700
    },
    {
      "entropy": 5.569350385665894,
      "epoch": 1.8443493483758022,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004657115365753302,
      "loss": 5.0094,
      "mean_token_accuracy": 0.19747061729431153,
      "num_tokens": 41110373.0,
      "step": 23705
    },
    {
      "entropy": 5.623968029022217,
      "epoch": 1.84473837774752,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046569682855074953,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19229544550180436,
      "num_tokens": 41119383.0,
      "step": 23710
    },
    {
      "entropy": 5.60439624786377,
      "epoch": 1.8451274071192376,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046568211763264796,
      "loss": 5.0063,
      "mean_token_accuracy": 0.19818878471851348,
      "num_tokens": 41127055.0,
      "step": 23715
    },
    {
      "entropy": 5.650721311569214,
      "epoch": 1.845516436490955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004656674038212488,
      "loss": 5.1171,
      "mean_token_accuracy": 0.1911420091986656,
      "num_tokens": 41136428.0,
      "step": 23720
    },
    {
      "entropy": 5.676035308837891,
      "epoch": 1.8459054658626726,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046565268711677525,
      "loss": 5.0348,
      "mean_token_accuracy": 0.19619891196489334,
      "num_tokens": 41145723.0,
      "step": 23725
    },
    {
      "entropy": 5.592716884613037,
      "epoch": 1.8462944952343903,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046563796751945065,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19297757297754287,
      "num_tokens": 41153998.0,
      "step": 23730
    },
    {
      "entropy": 5.521500158309936,
      "epoch": 1.8466835246061077,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046562324502949834,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20525648891925813,
      "num_tokens": 41162443.0,
      "step": 23735
    },
    {
      "entropy": 5.545340299606323,
      "epoch": 1.8470725539778252,
      "grad_norm": 1.125,
      "learning_rate": 0.0004656085196471416,
      "loss": 4.9173,
      "mean_token_accuracy": 0.20413196831941605,
      "num_tokens": 41171267.0,
      "step": 23740
    },
    {
      "entropy": 5.68914828300476,
      "epoch": 1.847461583349543,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046559379137260404,
      "loss": 5.0937,
      "mean_token_accuracy": 0.1971474274992943,
      "num_tokens": 41179944.0,
      "step": 23745
    },
    {
      "entropy": 5.643242263793946,
      "epoch": 1.8478506127212606,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000465579060206109,
      "loss": 5.0783,
      "mean_token_accuracy": 0.19258589148521424,
      "num_tokens": 41189038.0,
      "step": 23750
    },
    {
      "entropy": 5.706858205795288,
      "epoch": 1.8482396420929779,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046556432614788015,
      "loss": 5.1096,
      "mean_token_accuracy": 0.19299572706222534,
      "num_tokens": 41197166.0,
      "step": 23755
    },
    {
      "entropy": 5.607342386245728,
      "epoch": 1.8486286714646956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004655495891981409,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1990988716483116,
      "num_tokens": 41205708.0,
      "step": 23760
    },
    {
      "entropy": 5.673098468780518,
      "epoch": 1.8490177008364133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000465534849357115,
      "loss": 5.0947,
      "mean_token_accuracy": 0.19573987573385238,
      "num_tokens": 41214195.0,
      "step": 23765
    },
    {
      "entropy": 5.601473951339722,
      "epoch": 1.8494067302081307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046552010662502595,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1977492779493332,
      "num_tokens": 41223203.0,
      "step": 23770
    },
    {
      "entropy": 5.6286139488220215,
      "epoch": 1.8497957595798482,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004655053610020976,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20184979736804962,
      "num_tokens": 41231982.0,
      "step": 23775
    },
    {
      "entropy": 5.59403772354126,
      "epoch": 1.850184788951566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004654906124885535,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20151068717241288,
      "num_tokens": 41241355.0,
      "step": 23780
    },
    {
      "entropy": 5.711835527420044,
      "epoch": 1.8505738183232834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004654758610846176,
      "loss": 5.1699,
      "mean_token_accuracy": 0.18628573566675186,
      "num_tokens": 41250272.0,
      "step": 23785
    },
    {
      "entropy": 5.631198883056641,
      "epoch": 1.8509628476950009,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004654611067905137,
      "loss": 5.0412,
      "mean_token_accuracy": 0.19747087210416794,
      "num_tokens": 41258831.0,
      "step": 23790
    },
    {
      "entropy": 5.710073089599609,
      "epoch": 1.8513518770667186,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046544634960646563,
      "loss": 5.1284,
      "mean_token_accuracy": 0.19221454560756684,
      "num_tokens": 41267344.0,
      "step": 23795
    },
    {
      "entropy": 5.658873653411865,
      "epoch": 1.8517409064384363,
      "grad_norm": 1.125,
      "learning_rate": 0.0004654315895326974,
      "loss": 5.0635,
      "mean_token_accuracy": 0.1972479298710823,
      "num_tokens": 41276207.0,
      "step": 23800
    },
    {
      "entropy": 5.616982269287109,
      "epoch": 1.8521299358101535,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004654168265694328,
      "loss": 5.029,
      "mean_token_accuracy": 0.19278495907783508,
      "num_tokens": 41284286.0,
      "step": 23805
    },
    {
      "entropy": 5.600574016571045,
      "epoch": 1.8525189651818712,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000465402060716896,
      "loss": 5.049,
      "mean_token_accuracy": 0.20046378672122955,
      "num_tokens": 41292998.0,
      "step": 23810
    },
    {
      "entropy": 5.623632287979126,
      "epoch": 1.852907994553589,
      "grad_norm": 1.171875,
      "learning_rate": 0.000465387291975311,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20010250210762023,
      "num_tokens": 41301188.0,
      "step": 23815
    },
    {
      "entropy": 5.599987888336182,
      "epoch": 1.8532970239253064,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046537252034490185,
      "loss": 5.0315,
      "mean_token_accuracy": 0.19491710066795348,
      "num_tokens": 41310225.0,
      "step": 23820
    },
    {
      "entropy": 5.66416916847229,
      "epoch": 1.8536860532970239,
      "grad_norm": 1.125,
      "learning_rate": 0.00046535774582589275,
      "loss": 5.0299,
      "mean_token_accuracy": 0.19920175075531005,
      "num_tokens": 41319692.0,
      "step": 23825
    },
    {
      "entropy": 5.668254375457764,
      "epoch": 1.8540750826687415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046534296841850776,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20260425060987472,
      "num_tokens": 41327840.0,
      "step": 23830
    },
    {
      "entropy": 5.547959518432617,
      "epoch": 1.854464112040459,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046532818812297124,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1955566644668579,
      "num_tokens": 41337734.0,
      "step": 23835
    },
    {
      "entropy": 5.573527574539185,
      "epoch": 1.8548531414121765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004653134049395074,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20796550810337067,
      "num_tokens": 41346237.0,
      "step": 23840
    },
    {
      "entropy": 5.601950645446777,
      "epoch": 1.8552421707838942,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004652986188683406,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20794565975666046,
      "num_tokens": 41355500.0,
      "step": 23845
    },
    {
      "entropy": 5.616783714294433,
      "epoch": 1.855631200155612,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046528382990969504,
      "loss": 5.0416,
      "mean_token_accuracy": 0.1937785655260086,
      "num_tokens": 41364710.0,
      "step": 23850
    },
    {
      "entropy": 5.6291553497314455,
      "epoch": 1.8560202295273294,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004652690380637953,
      "loss": 5.022,
      "mean_token_accuracy": 0.20264870822429656,
      "num_tokens": 41372397.0,
      "step": 23855
    },
    {
      "entropy": 5.590443181991577,
      "epoch": 1.8564092588990468,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004652542433308657,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20663865357637407,
      "num_tokens": 41381079.0,
      "step": 23860
    },
    {
      "entropy": 5.653665494918823,
      "epoch": 1.8567982882707645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004652394457111309,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20100135505199432,
      "num_tokens": 41390453.0,
      "step": 23865
    },
    {
      "entropy": 5.694218111038208,
      "epoch": 1.857187317642482,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046522464520481517,
      "loss": 5.0883,
      "mean_token_accuracy": 0.18989461958408355,
      "num_tokens": 41398971.0,
      "step": 23870
    },
    {
      "entropy": 5.648893737792969,
      "epoch": 1.8575763470141995,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046520984181214333,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20288860946893691,
      "num_tokens": 41407459.0,
      "step": 23875
    },
    {
      "entropy": 5.566096591949463,
      "epoch": 1.8579653763859172,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004651950355333398,
      "loss": 4.9911,
      "mean_token_accuracy": 0.19916386306285858,
      "num_tokens": 41416465.0,
      "step": 23880
    },
    {
      "entropy": 5.621599817276001,
      "epoch": 1.8583544057576347,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004651802263686294,
      "loss": 5.1178,
      "mean_token_accuracy": 0.19879331439733505,
      "num_tokens": 41425362.0,
      "step": 23885
    },
    {
      "entropy": 5.588563394546509,
      "epoch": 1.8587434351293521,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004651654143182367,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2020531788468361,
      "num_tokens": 41433758.0,
      "step": 23890
    },
    {
      "entropy": 5.5273370265960695,
      "epoch": 1.8591324645010698,
      "grad_norm": 1.125,
      "learning_rate": 0.0004651505993823866,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20331908166408538,
      "num_tokens": 41442357.0,
      "step": 23895
    },
    {
      "entropy": 5.623252630233765,
      "epoch": 1.8595214938727875,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046513578156130383,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20437546372413634,
      "num_tokens": 41451257.0,
      "step": 23900
    },
    {
      "entropy": 5.624139404296875,
      "epoch": 1.859910523244505,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004651209608552131,
      "loss": 5.0145,
      "mean_token_accuracy": 0.19945389479398729,
      "num_tokens": 41460499.0,
      "step": 23905
    },
    {
      "entropy": 5.669059467315674,
      "epoch": 1.8602995526162225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046510613726433945,
      "loss": 5.0375,
      "mean_token_accuracy": 0.19725742936134338,
      "num_tokens": 41468365.0,
      "step": 23910
    },
    {
      "entropy": 5.695423650741577,
      "epoch": 1.8606885819879402,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004650913107889078,
      "loss": 5.0573,
      "mean_token_accuracy": 0.20205554515123367,
      "num_tokens": 41476538.0,
      "step": 23915
    },
    {
      "entropy": 5.642839622497559,
      "epoch": 1.8610776113596577,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000465076481429143,
      "loss": 5.0917,
      "mean_token_accuracy": 0.188010773062706,
      "num_tokens": 41486310.0,
      "step": 23920
    },
    {
      "entropy": 5.61181869506836,
      "epoch": 1.8614666407313751,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650616491852701,
      "loss": 5.0636,
      "mean_token_accuracy": 0.19182274341583253,
      "num_tokens": 41495015.0,
      "step": 23925
    },
    {
      "entropy": 5.651263618469239,
      "epoch": 1.8618556701030928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046504681405751426,
      "loss": 5.1085,
      "mean_token_accuracy": 0.1996583253145218,
      "num_tokens": 41504440.0,
      "step": 23930
    },
    {
      "entropy": 5.569432735443115,
      "epoch": 1.8622446994748103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046503197604610047,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21530550420284272,
      "num_tokens": 41513143.0,
      "step": 23935
    },
    {
      "entropy": 5.642820930480957,
      "epoch": 1.8626337288465278,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046501713515125393,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20182627141475679,
      "num_tokens": 41522164.0,
      "step": 23940
    },
    {
      "entropy": 5.605938148498535,
      "epoch": 1.8630227582182455,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004650022913731998,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2042882576584816,
      "num_tokens": 41529456.0,
      "step": 23945
    },
    {
      "entropy": 5.512390041351319,
      "epoch": 1.8634117875899632,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046498744471216335,
      "loss": 4.9696,
      "mean_token_accuracy": 0.202773018181324,
      "num_tokens": 41538536.0,
      "step": 23950
    },
    {
      "entropy": 5.6517863273620605,
      "epoch": 1.8638008169616807,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046497259516836974,
      "loss": 5.053,
      "mean_token_accuracy": 0.20335968136787413,
      "num_tokens": 41547514.0,
      "step": 23955
    },
    {
      "entropy": 5.656624460220337,
      "epoch": 1.8641898463333981,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046495774274204446,
      "loss": 5.0365,
      "mean_token_accuracy": 0.19922229051589965,
      "num_tokens": 41555845.0,
      "step": 23960
    },
    {
      "entropy": 5.639614677429199,
      "epoch": 1.8645788757051158,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004649428874334127,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2031002938747406,
      "num_tokens": 41565007.0,
      "step": 23965
    },
    {
      "entropy": 5.600994729995728,
      "epoch": 1.8649679050768333,
      "grad_norm": 1.125,
      "learning_rate": 0.0004649280292427,
      "loss": 4.992,
      "mean_token_accuracy": 0.19983874559402465,
      "num_tokens": 41574346.0,
      "step": 23970
    },
    {
      "entropy": 5.623596286773681,
      "epoch": 1.8653569344485508,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004649131681701318,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19635965526103974,
      "num_tokens": 41583945.0,
      "step": 23975
    },
    {
      "entropy": 5.567094850540161,
      "epoch": 1.8657459638202685,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046489830421593347,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20169115960597991,
      "num_tokens": 41591987.0,
      "step": 23980
    },
    {
      "entropy": 5.632329893112183,
      "epoch": 1.866134993191986,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004648834373803307,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1927157998085022,
      "num_tokens": 41600533.0,
      "step": 23985
    },
    {
      "entropy": 5.637715196609497,
      "epoch": 1.8665240225637034,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046486856766354893,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20404044538736343,
      "num_tokens": 41608567.0,
      "step": 23990
    },
    {
      "entropy": 5.6085264682769775,
      "epoch": 1.8669130519354211,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046485369506581387,
      "loss": 5.0178,
      "mean_token_accuracy": 0.19657496958971024,
      "num_tokens": 41616940.0,
      "step": 23995
    },
    {
      "entropy": 5.652847623825073,
      "epoch": 1.8673020813071388,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046483881958735124,
      "loss": 5.0764,
      "mean_token_accuracy": 0.19929840117692948,
      "num_tokens": 41625230.0,
      "step": 24000
    },
    {
      "epoch": 1.8673020813071388,
      "eval_entropy": 5.375226250649416,
      "eval_loss": 5.086186408996582,
      "eval_mean_token_accuracy": 0.20648816128237185,
      "eval_num_tokens": 41625230.0,
      "eval_runtime": 21.7518,
      "eval_samples_per_second": 1910.558,
      "eval_steps_per_second": 238.831,
      "step": 24000
    },
    {
      "entropy": 5.575991296768189,
      "epoch": 1.8676911106788563,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046482394122838663,
      "loss": 4.9779,
      "mean_token_accuracy": 0.19728079289197922,
      "num_tokens": 41634068.0,
      "step": 24005
    },
    {
      "entropy": 5.540875625610352,
      "epoch": 1.8680801400505738,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046480905998914587,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2106352776288986,
      "num_tokens": 41642512.0,
      "step": 24010
    },
    {
      "entropy": 5.5621195316314695,
      "epoch": 1.8684691694222915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004647941758698547,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2007884830236435,
      "num_tokens": 41651398.0,
      "step": 24015
    },
    {
      "entropy": 5.616655254364014,
      "epoch": 1.868858198794009,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000464779288870739,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20005956888198853,
      "num_tokens": 41661007.0,
      "step": 24020
    },
    {
      "entropy": 5.582627820968628,
      "epoch": 1.8692472281657264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046476439899202464,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20378190875053406,
      "num_tokens": 41669334.0,
      "step": 24025
    },
    {
      "entropy": 5.582896280288696,
      "epoch": 1.8696362575374441,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046474950623393756,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20612671971321106,
      "num_tokens": 41678061.0,
      "step": 24030
    },
    {
      "entropy": 5.62394642829895,
      "epoch": 1.8700252869091616,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004647346105967038,
      "loss": 5.0657,
      "mean_token_accuracy": 0.19432123750448227,
      "num_tokens": 41687073.0,
      "step": 24035
    },
    {
      "entropy": 5.611329221725464,
      "epoch": 1.870414316280879,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004647197120805493,
      "loss": 5.0081,
      "mean_token_accuracy": 0.199280446767807,
      "num_tokens": 41694952.0,
      "step": 24040
    },
    {
      "entropy": 5.605014133453369,
      "epoch": 1.8708033456525968,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046470481068570013,
      "loss": 5.0103,
      "mean_token_accuracy": 0.19463880956172944,
      "num_tokens": 41703621.0,
      "step": 24045
    },
    {
      "entropy": 5.5716217994689945,
      "epoch": 1.8711923750243145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004646899064123824,
      "loss": 4.988,
      "mean_token_accuracy": 0.20291805118322373,
      "num_tokens": 41712811.0,
      "step": 24050
    },
    {
      "entropy": 5.567281818389892,
      "epoch": 1.871581404396032,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004646749992608223,
      "loss": 4.973,
      "mean_token_accuracy": 0.19884432703256608,
      "num_tokens": 41721774.0,
      "step": 24055
    },
    {
      "entropy": 5.528051042556763,
      "epoch": 1.8719704337677494,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004646600892312459,
      "loss": 4.963,
      "mean_token_accuracy": 0.19842429906129838,
      "num_tokens": 41729928.0,
      "step": 24060
    },
    {
      "entropy": 5.520887613296509,
      "epoch": 1.872359463139467,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004646451763238796,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2041843444108963,
      "num_tokens": 41738946.0,
      "step": 24065
    },
    {
      "entropy": 5.6467499256134035,
      "epoch": 1.8727484925111846,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004646302605389496,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2024302214384079,
      "num_tokens": 41747184.0,
      "step": 24070
    },
    {
      "entropy": 5.548565769195557,
      "epoch": 1.873137521882902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004646153418766822,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20846169888973237,
      "num_tokens": 41756231.0,
      "step": 24075
    },
    {
      "entropy": 5.602645683288574,
      "epoch": 1.8735265512546198,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046460042033730377,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19349369555711746,
      "num_tokens": 41765870.0,
      "step": 24080
    },
    {
      "entropy": 5.5710245132446286,
      "epoch": 1.8739155806263375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004645854959210408,
      "loss": 4.9378,
      "mean_token_accuracy": 0.209228079020977,
      "num_tokens": 41774443.0,
      "step": 24085
    },
    {
      "entropy": 5.575962257385254,
      "epoch": 1.8743046099980547,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046457056862811956,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2094591960310936,
      "num_tokens": 41782925.0,
      "step": 24090
    },
    {
      "entropy": 5.604145479202271,
      "epoch": 1.8746936393697724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004645556384587668,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20371678620576858,
      "num_tokens": 41791293.0,
      "step": 24095
    },
    {
      "entropy": 5.598045349121094,
      "epoch": 1.87508266874149,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004645407054132089,
      "loss": 4.902,
      "mean_token_accuracy": 0.2176859363913536,
      "num_tokens": 41800122.0,
      "step": 24100
    },
    {
      "entropy": 5.640039300918579,
      "epoch": 1.8754716981132076,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004645257694916725,
      "loss": 5.069,
      "mean_token_accuracy": 0.20053231120109558,
      "num_tokens": 41808426.0,
      "step": 24105
    },
    {
      "entropy": 5.61688494682312,
      "epoch": 1.875860727484925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004645108306943842,
      "loss": 5.0046,
      "mean_token_accuracy": 0.1989349290728569,
      "num_tokens": 41817228.0,
      "step": 24110
    },
    {
      "entropy": 5.575835752487182,
      "epoch": 1.8762497568566427,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004644958890215707,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2041890040040016,
      "num_tokens": 41825711.0,
      "step": 24115
    },
    {
      "entropy": 5.675631189346314,
      "epoch": 1.8766387862283602,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004644809444734586,
      "loss": 5.0984,
      "mean_token_accuracy": 0.18770757913589478,
      "num_tokens": 41833976.0,
      "step": 24120
    },
    {
      "entropy": 5.611126375198364,
      "epoch": 1.8770278156000777,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046446599705027487,
      "loss": 5.0187,
      "mean_token_accuracy": 0.20637762397527695,
      "num_tokens": 41842180.0,
      "step": 24125
    },
    {
      "entropy": 5.583420848846435,
      "epoch": 1.8774168449717954,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046445104675224607,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20120240449905397,
      "num_tokens": 41850470.0,
      "step": 24130
    },
    {
      "entropy": 5.542023801803589,
      "epoch": 1.877805874343513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004644360935795993,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2058403104543686,
      "num_tokens": 41859036.0,
      "step": 24135
    },
    {
      "entropy": 5.68872857093811,
      "epoch": 1.8781949037152303,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046442113753256126,
      "loss": 5.0977,
      "mean_token_accuracy": 0.19448558390140533,
      "num_tokens": 41867918.0,
      "step": 24140
    },
    {
      "entropy": 5.619909763336182,
      "epoch": 1.878583933086948,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046440617861135905,
      "loss": 5.0744,
      "mean_token_accuracy": 0.19753012657165528,
      "num_tokens": 41877281.0,
      "step": 24145
    },
    {
      "entropy": 5.544361972808838,
      "epoch": 1.8789729624586657,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004643912168162194,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20197461545467377,
      "num_tokens": 41885991.0,
      "step": 24150
    },
    {
      "entropy": 5.638756275177002,
      "epoch": 1.8793619918303832,
      "grad_norm": 1.125,
      "learning_rate": 0.0004643762521473696,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1996283620595932,
      "num_tokens": 41894318.0,
      "step": 24155
    },
    {
      "entropy": 5.665662097930908,
      "epoch": 1.8797510212021007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004643612846050366,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20223896354436874,
      "num_tokens": 41902026.0,
      "step": 24160
    },
    {
      "entropy": 5.609949350357056,
      "epoch": 1.8801400505738184,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046434631418944744,
      "loss": 5.126,
      "mean_token_accuracy": 0.1919356033205986,
      "num_tokens": 41910210.0,
      "step": 24165
    },
    {
      "entropy": 5.678350877761841,
      "epoch": 1.8805290799455359,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004643313409008294,
      "loss": 5.007,
      "mean_token_accuracy": 0.20774077475070954,
      "num_tokens": 41917959.0,
      "step": 24170
    },
    {
      "entropy": 5.641670989990234,
      "epoch": 1.8809181093172533,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046431636473940956,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19739083647727967,
      "num_tokens": 41928079.0,
      "step": 24175
    },
    {
      "entropy": 5.648298740386963,
      "epoch": 1.881307138688971,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004643013857054152,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20122355967760086,
      "num_tokens": 41936681.0,
      "step": 24180
    },
    {
      "entropy": 5.6649078845977785,
      "epoch": 1.8816961680606887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046428640379907367,
      "loss": 5.0976,
      "mean_token_accuracy": 0.19612004309892656,
      "num_tokens": 41945013.0,
      "step": 24185
    },
    {
      "entropy": 5.569679069519043,
      "epoch": 1.882085197432406,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046427141902061225,
      "loss": 4.9987,
      "mean_token_accuracy": 0.19731379449367523,
      "num_tokens": 41953753.0,
      "step": 24190
    },
    {
      "entropy": 5.589600372314453,
      "epoch": 1.8824742268041237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004642564313702582,
      "loss": 5.0121,
      "mean_token_accuracy": 0.19985999912023544,
      "num_tokens": 41962755.0,
      "step": 24195
    },
    {
      "entropy": 5.557780790328979,
      "epoch": 1.8828632561758414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046424144084823916,
      "loss": 4.8907,
      "mean_token_accuracy": 0.2116768553853035,
      "num_tokens": 41970648.0,
      "step": 24200
    },
    {
      "entropy": 5.623630952835083,
      "epoch": 1.8832522855475589,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004642264474547824,
      "loss": 4.9874,
      "mean_token_accuracy": 0.20694733858108522,
      "num_tokens": 41978805.0,
      "step": 24205
    },
    {
      "entropy": 5.587182569503784,
      "epoch": 1.8836413149192763,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046421145119011556,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20310329049825668,
      "num_tokens": 41987076.0,
      "step": 24210
    },
    {
      "entropy": 5.660914564132691,
      "epoch": 1.884030344290994,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000464196452054466,
      "loss": 5.045,
      "mean_token_accuracy": 0.19886306822299957,
      "num_tokens": 41996316.0,
      "step": 24215
    },
    {
      "entropy": 5.670384454727173,
      "epoch": 1.8844193736627115,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004641814500480615,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20068604201078416,
      "num_tokens": 42005058.0,
      "step": 24220
    },
    {
      "entropy": 5.6745490550994875,
      "epoch": 1.884808403034429,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004641664451711296,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20423104614019394,
      "num_tokens": 42013355.0,
      "step": 24225
    },
    {
      "entropy": 5.663799095153808,
      "epoch": 1.8851974324061467,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046415143742389793,
      "loss": 5.1233,
      "mean_token_accuracy": 0.19698395282030107,
      "num_tokens": 42021346.0,
      "step": 24230
    },
    {
      "entropy": 5.599025392532349,
      "epoch": 1.8855864617778644,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004641364268065943,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19144562929868697,
      "num_tokens": 42030407.0,
      "step": 24235
    },
    {
      "entropy": 5.6245129108428955,
      "epoch": 1.8859754911495816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046412141331944644,
      "loss": 4.9939,
      "mean_token_accuracy": 0.1960299700498581,
      "num_tokens": 42038933.0,
      "step": 24240
    },
    {
      "entropy": 5.727528047561646,
      "epoch": 1.8863645205212993,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046410639696268206,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19513828605413436,
      "num_tokens": 42047395.0,
      "step": 24245
    },
    {
      "entropy": 5.635821294784546,
      "epoch": 1.886753549893017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004640913777365292,
      "loss": 5.1239,
      "mean_token_accuracy": 0.19163146167993544,
      "num_tokens": 42056141.0,
      "step": 24250
    },
    {
      "entropy": 5.666148948669433,
      "epoch": 1.8871425792647345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046407635564121565,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19990883469581605,
      "num_tokens": 42064059.0,
      "step": 24255
    },
    {
      "entropy": 5.700064468383789,
      "epoch": 1.887531608636452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046406133067696936,
      "loss": 5.1453,
      "mean_token_accuracy": 0.1920527085661888,
      "num_tokens": 42073350.0,
      "step": 24260
    },
    {
      "entropy": 5.653461551666259,
      "epoch": 1.8879206380081697,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004640463028440182,
      "loss": 5.0762,
      "mean_token_accuracy": 0.19621252864599228,
      "num_tokens": 42082033.0,
      "step": 24265
    },
    {
      "entropy": 5.594650316238403,
      "epoch": 1.8883096673798871,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004640312721425904,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20000164657831193,
      "num_tokens": 42090974.0,
      "step": 24270
    },
    {
      "entropy": 5.715928459167481,
      "epoch": 1.8886986967516046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004640162385729139,
      "loss": 5.0918,
      "mean_token_accuracy": 0.19750449508428575,
      "num_tokens": 42100605.0,
      "step": 24275
    },
    {
      "entropy": 5.645242691040039,
      "epoch": 1.8890877261233223,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004640012021352168,
      "loss": 4.9234,
      "mean_token_accuracy": 0.209676830470562,
      "num_tokens": 42108892.0,
      "step": 24280
    },
    {
      "entropy": 5.60062346458435,
      "epoch": 1.88947675549504,
      "grad_norm": 1.125,
      "learning_rate": 0.0004639861628297273,
      "loss": 5.028,
      "mean_token_accuracy": 0.19674924463033677,
      "num_tokens": 42117706.0,
      "step": 24285
    },
    {
      "entropy": 5.6008713722229,
      "epoch": 1.8898657848667575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046397112065667354,
      "loss": 5.1059,
      "mean_token_accuracy": 0.19407981634140015,
      "num_tokens": 42127194.0,
      "step": 24290
    },
    {
      "entropy": 5.576604509353638,
      "epoch": 1.890254814238475,
      "grad_norm": 1.125,
      "learning_rate": 0.00046395607561628387,
      "loss": 5.0109,
      "mean_token_accuracy": 0.1982605293393135,
      "num_tokens": 42135349.0,
      "step": 24295
    },
    {
      "entropy": 5.613498401641846,
      "epoch": 1.8906438436101927,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046394102770878643,
      "loss": 4.9845,
      "mean_token_accuracy": 0.19908467233181,
      "num_tokens": 42144173.0,
      "step": 24300
    },
    {
      "entropy": 5.540602731704712,
      "epoch": 1.8910328729819101,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046392597693440974,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2097703531384468,
      "num_tokens": 42153076.0,
      "step": 24305
    },
    {
      "entropy": 5.570808410644531,
      "epoch": 1.8914219023536276,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004639109232933819,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20130694210529326,
      "num_tokens": 42161821.0,
      "step": 24310
    },
    {
      "entropy": 5.55162034034729,
      "epoch": 1.8918109317253453,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004638958667859316,
      "loss": 4.9291,
      "mean_token_accuracy": 0.20706473290920258,
      "num_tokens": 42170892.0,
      "step": 24315
    },
    {
      "entropy": 5.6717630386352536,
      "epoch": 1.8921999610970628,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004638808074122872,
      "loss": 5.1098,
      "mean_token_accuracy": 0.18956306278705598,
      "num_tokens": 42180084.0,
      "step": 24320
    },
    {
      "entropy": 5.6500293731689455,
      "epoch": 1.8925889904687803,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004638657451726771,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20576262027025222,
      "num_tokens": 42188992.0,
      "step": 24325
    },
    {
      "entropy": 5.674073171615601,
      "epoch": 1.892978019840498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046385068006732995,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20046926587820052,
      "num_tokens": 42197833.0,
      "step": 24330
    },
    {
      "entropy": 5.6073997020721436,
      "epoch": 1.8933670492122157,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046383561209647437,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21329436004161834,
      "num_tokens": 42206599.0,
      "step": 24335
    },
    {
      "entropy": 5.535683059692383,
      "epoch": 1.8937560785839331,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046382054126033884,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20395016819238662,
      "num_tokens": 42214673.0,
      "step": 24340
    },
    {
      "entropy": 5.561068916320801,
      "epoch": 1.8941451079556506,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004638054675591523,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2017168954014778,
      "num_tokens": 42223848.0,
      "step": 24345
    },
    {
      "entropy": 5.628382444381714,
      "epoch": 1.8945341373273683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004637903909931432,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20288896709680557,
      "num_tokens": 42232548.0,
      "step": 24350
    },
    {
      "entropy": 5.660288667678833,
      "epoch": 1.8949231666990858,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004637753115625404,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2049438863992691,
      "num_tokens": 42240882.0,
      "step": 24355
    },
    {
      "entropy": 5.604740142822266,
      "epoch": 1.8953121960708033,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004637602292675726,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2000296786427498,
      "num_tokens": 42249487.0,
      "step": 24360
    },
    {
      "entropy": 5.607293033599854,
      "epoch": 1.895701225442521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004637451441084689,
      "loss": 5.0783,
      "mean_token_accuracy": 0.199790620803833,
      "num_tokens": 42258859.0,
      "step": 24365
    },
    {
      "entropy": 5.671412420272827,
      "epoch": 1.8960902548142384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004637300560854581,
      "loss": 5.1093,
      "mean_token_accuracy": 0.19099180400371552,
      "num_tokens": 42267492.0,
      "step": 24370
    },
    {
      "entropy": 5.604776525497437,
      "epoch": 1.896479284185956,
      "grad_norm": 1.140625,
      "learning_rate": 0.000463714965198769,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19726796597242355,
      "num_tokens": 42276046.0,
      "step": 24375
    },
    {
      "entropy": 5.629160737991333,
      "epoch": 1.8968683135576736,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004636998714486307,
      "loss": 5.0052,
      "mean_token_accuracy": 0.19947371631860733,
      "num_tokens": 42284992.0,
      "step": 24380
    },
    {
      "entropy": 5.611885356903076,
      "epoch": 1.8972573429293913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046368477483527224,
      "loss": 5.034,
      "mean_token_accuracy": 0.19951482117176056,
      "num_tokens": 42294803.0,
      "step": 24385
    },
    {
      "entropy": 5.661969327926636,
      "epoch": 1.8976463723011088,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004636696753589226,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1995529517531395,
      "num_tokens": 42303329.0,
      "step": 24390
    },
    {
      "entropy": 5.623769617080688,
      "epoch": 1.8980354016728263,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004636545730198109,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20047936737537383,
      "num_tokens": 42311745.0,
      "step": 24395
    },
    {
      "entropy": 5.6326172828674315,
      "epoch": 1.898424431044544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046363946781816643,
      "loss": 5.0758,
      "mean_token_accuracy": 0.19763122946023942,
      "num_tokens": 42320295.0,
      "step": 24400
    },
    {
      "entropy": 5.674953889846802,
      "epoch": 1.8988134604162614,
      "grad_norm": 1.25,
      "learning_rate": 0.00046362435975421816,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20674734562635422,
      "num_tokens": 42328703.0,
      "step": 24405
    },
    {
      "entropy": 5.597205543518067,
      "epoch": 1.899202489787979,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046360924882819554,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20198270082473754,
      "num_tokens": 42338338.0,
      "step": 24410
    },
    {
      "entropy": 5.693012475967407,
      "epoch": 1.8995915191596966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004635941350403277,
      "loss": 5.0654,
      "mean_token_accuracy": 0.19830695986747743,
      "num_tokens": 42347223.0,
      "step": 24415
    },
    {
      "entropy": 5.640581035614014,
      "epoch": 1.899980548531414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004635790183908442,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2077213779091835,
      "num_tokens": 42355827.0,
      "step": 24420
    },
    {
      "entropy": 5.623261165618897,
      "epoch": 1.9003695779031315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046356389887997415,
      "loss": 5.0494,
      "mean_token_accuracy": 0.19702430367469786,
      "num_tokens": 42363985.0,
      "step": 24425
    },
    {
      "entropy": 5.669466257095337,
      "epoch": 1.9007586072748492,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046354877650794707,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20064836144447326,
      "num_tokens": 42372280.0,
      "step": 24430
    },
    {
      "entropy": 5.654238843917847,
      "epoch": 1.901147636646567,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046353365127499235,
      "loss": 5.0991,
      "mean_token_accuracy": 0.18549738079309464,
      "num_tokens": 42381238.0,
      "step": 24435
    },
    {
      "entropy": 5.634921932220459,
      "epoch": 1.9015366660182844,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004635185231813396,
      "loss": 5.0581,
      "mean_token_accuracy": 0.1985452026128769,
      "num_tokens": 42389609.0,
      "step": 24440
    },
    {
      "entropy": 5.658938264846801,
      "epoch": 1.901925695390002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004635033922272183,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20129372775554658,
      "num_tokens": 42397502.0,
      "step": 24445
    },
    {
      "entropy": 5.687984323501587,
      "epoch": 1.9023147247617196,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046348825841285813,
      "loss": 5.084,
      "mean_token_accuracy": 0.1942368522286415,
      "num_tokens": 42405566.0,
      "step": 24450
    },
    {
      "entropy": 5.605263757705688,
      "epoch": 1.902703754133437,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004634731217384886,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19701357334852218,
      "num_tokens": 42414367.0,
      "step": 24455
    },
    {
      "entropy": 5.649328947067261,
      "epoch": 1.9030927835051545,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004634579822043394,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19335432797670365,
      "num_tokens": 42422889.0,
      "step": 24460
    },
    {
      "entropy": 5.672682046890259,
      "epoch": 1.9034818128768722,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046344283981064035,
      "loss": 5.0852,
      "mean_token_accuracy": 0.19481225311756134,
      "num_tokens": 42432069.0,
      "step": 24465
    },
    {
      "entropy": 5.622801160812378,
      "epoch": 1.90387084224859,
      "grad_norm": 1.125,
      "learning_rate": 0.00046342769455762114,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19998427778482436,
      "num_tokens": 42440509.0,
      "step": 24470
    },
    {
      "entropy": 5.598542499542236,
      "epoch": 1.9042598716203072,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004634125464455116,
      "loss": 5.0293,
      "mean_token_accuracy": 0.19778774976730346,
      "num_tokens": 42448889.0,
      "step": 24475
    },
    {
      "entropy": 5.525557136535644,
      "epoch": 1.9046489009920249,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046339739547454146,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21493522673845292,
      "num_tokens": 42457379.0,
      "step": 24480
    },
    {
      "entropy": 5.668351411819458,
      "epoch": 1.9050379303637426,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046338224164494074,
      "loss": 5.0915,
      "mean_token_accuracy": 0.19329685866832733,
      "num_tokens": 42466114.0,
      "step": 24485
    },
    {
      "entropy": 5.648586416244507,
      "epoch": 1.90542695973546,
      "grad_norm": 1.125,
      "learning_rate": 0.00046336708495693933,
      "loss": 5.0332,
      "mean_token_accuracy": 0.1925637423992157,
      "num_tokens": 42475734.0,
      "step": 24490
    },
    {
      "entropy": 5.554257822036743,
      "epoch": 1.9058159891071775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046335192541076725,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21054467111825942,
      "num_tokens": 42484007.0,
      "step": 24495
    },
    {
      "entropy": 5.51367220878601,
      "epoch": 1.9062050184788952,
      "grad_norm": 1.125,
      "learning_rate": 0.0004633367630066545,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20452889651060105,
      "num_tokens": 42493157.0,
      "step": 24500
    },
    {
      "entropy": 5.658606243133545,
      "epoch": 1.9065940478506127,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046332159774483116,
      "loss": 5.023,
      "mean_token_accuracy": 0.20212386250495912,
      "num_tokens": 42501389.0,
      "step": 24505
    },
    {
      "entropy": 5.618350076675415,
      "epoch": 1.9069830772223302,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004633064296255273,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20227981209754944,
      "num_tokens": 42509624.0,
      "step": 24510
    },
    {
      "entropy": 5.607503604888916,
      "epoch": 1.9073721065940479,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046329125864897307,
      "loss": 5.0619,
      "mean_token_accuracy": 0.19819847792387008,
      "num_tokens": 42518193.0,
      "step": 24515
    },
    {
      "entropy": 5.607752513885498,
      "epoch": 1.9077611359657656,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004632760848153987,
      "loss": 5.03,
      "mean_token_accuracy": 0.2017349973320961,
      "num_tokens": 42526641.0,
      "step": 24520
    },
    {
      "entropy": 5.611910963058472,
      "epoch": 1.9081501653374828,
      "grad_norm": 1.125,
      "learning_rate": 0.0004632609081250345,
      "loss": 5.1167,
      "mean_token_accuracy": 0.19755523204803466,
      "num_tokens": 42535263.0,
      "step": 24525
    },
    {
      "entropy": 5.710028553009034,
      "epoch": 1.9085391947092005,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046324572857811054,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2026426076889038,
      "num_tokens": 42544465.0,
      "step": 24530
    },
    {
      "entropy": 5.65536789894104,
      "epoch": 1.9089282240809182,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004632305461748573,
      "loss": 5.1013,
      "mean_token_accuracy": 0.19924804717302322,
      "num_tokens": 42553198.0,
      "step": 24535
    },
    {
      "entropy": 5.589033555984497,
      "epoch": 1.9093172534526357,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046321536091550517,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21182842403650284,
      "num_tokens": 42562014.0,
      "step": 24540
    },
    {
      "entropy": 5.62154746055603,
      "epoch": 1.9097062828243532,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004632001728002845,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2019475668668747,
      "num_tokens": 42570966.0,
      "step": 24545
    },
    {
      "entropy": 5.645296335220337,
      "epoch": 1.9100953121960709,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004631849818294257,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20051083862781524,
      "num_tokens": 42579939.0,
      "step": 24550
    },
    {
      "entropy": 5.600928401947021,
      "epoch": 1.9104843415677883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046316978800315934,
      "loss": 5.0037,
      "mean_token_accuracy": 0.1961117386817932,
      "num_tokens": 42588874.0,
      "step": 24555
    },
    {
      "entropy": 5.583487272262573,
      "epoch": 1.9108733709395058,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463154591321716,
      "loss": 4.987,
      "mean_token_accuracy": 0.20640987008810044,
      "num_tokens": 42597632.0,
      "step": 24560
    },
    {
      "entropy": 5.536374187469482,
      "epoch": 1.9112624003112235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046313939178532624,
      "loss": 4.8972,
      "mean_token_accuracy": 0.20970050543546676,
      "num_tokens": 42606240.0,
      "step": 24565
    },
    {
      "entropy": 5.6503900527954105,
      "epoch": 1.9116514296829412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004631241893942206,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20256012231111525,
      "num_tokens": 42614713.0,
      "step": 24570
    },
    {
      "entropy": 5.574986600875855,
      "epoch": 1.9120404590546585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046310898414862983,
      "loss": 4.9537,
      "mean_token_accuracy": 0.20438385903835296,
      "num_tokens": 42623123.0,
      "step": 24575
    },
    {
      "entropy": 5.627835702896118,
      "epoch": 1.9124294884263762,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004630937760487847,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19106125831604004,
      "num_tokens": 42632056.0,
      "step": 24580
    },
    {
      "entropy": 5.592078924179077,
      "epoch": 1.9128185177980939,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004630785650949158,
      "loss": 5.0172,
      "mean_token_accuracy": 0.19737221598625182,
      "num_tokens": 42640809.0,
      "step": 24585
    },
    {
      "entropy": 5.578227186203003,
      "epoch": 1.9132075471698113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004630633512872541,
      "loss": 4.9613,
      "mean_token_accuracy": 0.1996218129992485,
      "num_tokens": 42649378.0,
      "step": 24590
    },
    {
      "entropy": 5.618298625946045,
      "epoch": 1.9135965765415288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046304813462603035,
      "loss": 4.9946,
      "mean_token_accuracy": 0.20309585332870483,
      "num_tokens": 42658895.0,
      "step": 24595
    },
    {
      "entropy": 5.599772930145264,
      "epoch": 1.9139856059132465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004630329151114754,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20456584095954894,
      "num_tokens": 42666867.0,
      "step": 24600
    },
    {
      "entropy": 5.573085784912109,
      "epoch": 1.914374635284964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046301769274382034,
      "loss": 5.0896,
      "mean_token_accuracy": 0.18964345306158065,
      "num_tokens": 42676101.0,
      "step": 24605
    },
    {
      "entropy": 5.696787357330322,
      "epoch": 1.9147636646566815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004630024675232961,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1930241495370865,
      "num_tokens": 42685773.0,
      "step": 24610
    },
    {
      "entropy": 5.679801988601684,
      "epoch": 1.9151526940283992,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046298723945013354,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21010109633207322,
      "num_tokens": 42694640.0,
      "step": 24615
    },
    {
      "entropy": 5.659233236312867,
      "epoch": 1.9155417234001169,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004629720085245639,
      "loss": 5.0655,
      "mean_token_accuracy": 0.19480959922075272,
      "num_tokens": 42703883.0,
      "step": 24620
    },
    {
      "entropy": 5.59961838722229,
      "epoch": 1.9159307527718341,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004629567747468182,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20743289589881897,
      "num_tokens": 42712623.0,
      "step": 24625
    },
    {
      "entropy": 5.598846530914306,
      "epoch": 1.9163197821435518,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004629415381171276,
      "loss": 5.0353,
      "mean_token_accuracy": 0.2015077769756317,
      "num_tokens": 42721052.0,
      "step": 24630
    },
    {
      "entropy": 5.65079550743103,
      "epoch": 1.9167088115152695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046292629863572325,
      "loss": 5.0222,
      "mean_token_accuracy": 0.20417391061782836,
      "num_tokens": 42729866.0,
      "step": 24635
    },
    {
      "entropy": 5.679056167602539,
      "epoch": 1.917097840886987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004629110563028365,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19206285923719407,
      "num_tokens": 42740127.0,
      "step": 24640
    },
    {
      "entropy": 5.646302986145019,
      "epoch": 1.9174868702587045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004628958111186985,
      "loss": 5.0623,
      "mean_token_accuracy": 0.1969013601541519,
      "num_tokens": 42748427.0,
      "step": 24645
    },
    {
      "entropy": 5.6928942680358885,
      "epoch": 1.9178758996304222,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004628805630835407,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19276100099086763,
      "num_tokens": 42756864.0,
      "step": 24650
    },
    {
      "entropy": 5.66128454208374,
      "epoch": 1.9182649290021396,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004628653121975944,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1954493835568428,
      "num_tokens": 42765636.0,
      "step": 24655
    },
    {
      "entropy": 5.636631441116333,
      "epoch": 1.918653958373857,
      "grad_norm": 1.109375,
      "learning_rate": 0.000462850058461091,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20000417828559874,
      "num_tokens": 42774486.0,
      "step": 24660
    },
    {
      "entropy": 5.630409049987793,
      "epoch": 1.9190429877455748,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004628348018742619,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20318261831998824,
      "num_tokens": 42783281.0,
      "step": 24665
    },
    {
      "entropy": 5.593532705307007,
      "epoch": 1.9194320171172925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004628195424373387,
      "loss": 4.9994,
      "mean_token_accuracy": 0.20536266416311263,
      "num_tokens": 42791710.0,
      "step": 24670
    },
    {
      "entropy": 5.554297256469726,
      "epoch": 1.91982104648901,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046280428015055286,
      "loss": 5.0257,
      "mean_token_accuracy": 0.19954635798931122,
      "num_tokens": 42799670.0,
      "step": 24675
    },
    {
      "entropy": 5.599506759643555,
      "epoch": 1.9202100758607274,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046278901501413606,
      "loss": 5.077,
      "mean_token_accuracy": 0.19978390336036683,
      "num_tokens": 42808279.0,
      "step": 24680
    },
    {
      "entropy": 5.65127272605896,
      "epoch": 1.9205991052324451,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046277374702831983,
      "loss": 5.0154,
      "mean_token_accuracy": 0.19719365239143372,
      "num_tokens": 42816818.0,
      "step": 24685
    },
    {
      "entropy": 5.658857011795044,
      "epoch": 1.9209881346041626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046275847619333576,
      "loss": 5.058,
      "mean_token_accuracy": 0.19376415610313416,
      "num_tokens": 42826270.0,
      "step": 24690
    },
    {
      "entropy": 5.704193162918091,
      "epoch": 1.92137716397588,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046274320250941576,
      "loss": 5.1144,
      "mean_token_accuracy": 0.19637055695056915,
      "num_tokens": 42835396.0,
      "step": 24695
    },
    {
      "entropy": 5.6632579326629635,
      "epoch": 1.9217661933475978,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046272792597679146,
      "loss": 5.0483,
      "mean_token_accuracy": 0.19475603997707366,
      "num_tokens": 42844184.0,
      "step": 24700
    },
    {
      "entropy": 5.637825059890747,
      "epoch": 1.9221552227193153,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046271264659569474,
      "loss": 5.0561,
      "mean_token_accuracy": 0.19995171427726746,
      "num_tokens": 42852323.0,
      "step": 24705
    },
    {
      "entropy": 5.5610185146331785,
      "epoch": 1.9225442520910327,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004626973643663573,
      "loss": 4.8911,
      "mean_token_accuracy": 0.21082558184862138,
      "num_tokens": 42860843.0,
      "step": 24710
    },
    {
      "entropy": 5.6457319259643555,
      "epoch": 1.9229332814627504,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004626820792890112,
      "loss": 5.0098,
      "mean_token_accuracy": 0.19838268756866456,
      "num_tokens": 42869508.0,
      "step": 24715
    },
    {
      "entropy": 5.583624267578125,
      "epoch": 1.9233223108344681,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004626667913638882,
      "loss": 4.9511,
      "mean_token_accuracy": 0.20015075355768203,
      "num_tokens": 42878142.0,
      "step": 24720
    },
    {
      "entropy": 5.627728891372681,
      "epoch": 1.9237113402061856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004626515005912203,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20185617357492447,
      "num_tokens": 42886562.0,
      "step": 24725
    },
    {
      "entropy": 5.618095016479492,
      "epoch": 1.924100369577903,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004626362069712397,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2044566586613655,
      "num_tokens": 42895243.0,
      "step": 24730
    },
    {
      "entropy": 5.551434087753296,
      "epoch": 1.9244893989496208,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004626209105041782,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20748575329780578,
      "num_tokens": 42904129.0,
      "step": 24735
    },
    {
      "entropy": 5.486058139801026,
      "epoch": 1.9248784283213383,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462605611190268,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2155091032385826,
      "num_tokens": 42913053.0,
      "step": 24740
    },
    {
      "entropy": 5.582480573654175,
      "epoch": 1.9252674576930557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046259030902974133,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20138343870639802,
      "num_tokens": 42921925.0,
      "step": 24745
    },
    {
      "entropy": 5.618669700622559,
      "epoch": 1.9256564870647734,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004625750040228302,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20188248455524443,
      "num_tokens": 42931225.0,
      "step": 24750
    },
    {
      "entropy": 5.654020166397094,
      "epoch": 1.926045516436491,
      "grad_norm": 1.125,
      "learning_rate": 0.00046255969616976704,
      "loss": 5.069,
      "mean_token_accuracy": 0.19711742848157882,
      "num_tokens": 42940594.0,
      "step": 24755
    },
    {
      "entropy": 5.573828268051147,
      "epoch": 1.9264345458082084,
      "grad_norm": 1.1875,
      "learning_rate": 0.000462544385470784,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20198239982128144,
      "num_tokens": 42948883.0,
      "step": 24760
    },
    {
      "entropy": 5.606162881851196,
      "epoch": 1.926823575179926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004625290719261134,
      "loss": 5.0523,
      "mean_token_accuracy": 0.19790069460868837,
      "num_tokens": 42957315.0,
      "step": 24765
    },
    {
      "entropy": 5.620489501953125,
      "epoch": 1.9272126045516438,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004625137555359876,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20735580027103423,
      "num_tokens": 42966460.0,
      "step": 24770
    },
    {
      "entropy": 5.6660316467285154,
      "epoch": 1.9276016339233613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046249843630063905,
      "loss": 5.1326,
      "mean_token_accuracy": 0.18913318514823912,
      "num_tokens": 42975177.0,
      "step": 24775
    },
    {
      "entropy": 5.648400783538818,
      "epoch": 1.9279906632950787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004624831142203001,
      "loss": 5.0563,
      "mean_token_accuracy": 0.1988622546195984,
      "num_tokens": 42983456.0,
      "step": 24780
    },
    {
      "entropy": 5.6128623485565186,
      "epoch": 1.9283796926667964,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046246778929520346,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20673609524965286,
      "num_tokens": 42992398.0,
      "step": 24785
    },
    {
      "entropy": 5.549891948699951,
      "epoch": 1.928768722038514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004624524615255814,
      "loss": 4.828,
      "mean_token_accuracy": 0.21897414922714234,
      "num_tokens": 43000630.0,
      "step": 24790
    },
    {
      "entropy": 5.581707286834717,
      "epoch": 1.9291577514102314,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046243713091166655,
      "loss": 5.0911,
      "mean_token_accuracy": 0.19989578574895858,
      "num_tokens": 43009166.0,
      "step": 24795
    },
    {
      "entropy": 5.588793134689331,
      "epoch": 1.929546780781949,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004624217974536916,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20191770941019058,
      "num_tokens": 43017152.0,
      "step": 24800
    },
    {
      "entropy": 5.6290717124938965,
      "epoch": 1.9299358101536666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046240646115188925,
      "loss": 5.0167,
      "mean_token_accuracy": 0.20349715203046798,
      "num_tokens": 43026161.0,
      "step": 24805
    },
    {
      "entropy": 5.694434785842896,
      "epoch": 1.930324839525384,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004623911220064921,
      "loss": 5.0834,
      "mean_token_accuracy": 0.19844716936349868,
      "num_tokens": 43034257.0,
      "step": 24810
    },
    {
      "entropy": 5.555358982086181,
      "epoch": 1.9307138688971017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046237578001773297,
      "loss": 4.9361,
      "mean_token_accuracy": 0.20811618864536285,
      "num_tokens": 43043842.0,
      "step": 24815
    },
    {
      "entropy": 5.623932981491089,
      "epoch": 1.9311028982688194,
      "grad_norm": 1.125,
      "learning_rate": 0.00046236043518584454,
      "loss": 5.0534,
      "mean_token_accuracy": 0.205864380300045,
      "num_tokens": 43052479.0,
      "step": 24820
    },
    {
      "entropy": 5.652993249893188,
      "epoch": 1.931491927640537,
      "grad_norm": 1.125,
      "learning_rate": 0.0004623450875110597,
      "loss": 5.0616,
      "mean_token_accuracy": 0.19825852513313294,
      "num_tokens": 43061948.0,
      "step": 24825
    },
    {
      "entropy": 5.599682140350342,
      "epoch": 1.9318809570122544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046232973699361147,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19991534501314162,
      "num_tokens": 43070395.0,
      "step": 24830
    },
    {
      "entropy": 5.505357885360718,
      "epoch": 1.932269986383972,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004623143836337326,
      "loss": 4.9256,
      "mean_token_accuracy": 0.1985604152083397,
      "num_tokens": 43079209.0,
      "step": 24835
    },
    {
      "entropy": 5.612361907958984,
      "epoch": 1.9326590157556895,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046229902743165607,
      "loss": 5.1397,
      "mean_token_accuracy": 0.19236479550600052,
      "num_tokens": 43088275.0,
      "step": 24840
    },
    {
      "entropy": 5.656156206130982,
      "epoch": 1.933048045127407,
      "grad_norm": 1.125,
      "learning_rate": 0.0004622836683876149,
      "loss": 5.0424,
      "mean_token_accuracy": 0.19812537729740143,
      "num_tokens": 43097239.0,
      "step": 24845
    },
    {
      "entropy": 5.651016807556152,
      "epoch": 1.9334370744991247,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004622683065018422,
      "loss": 5.1077,
      "mean_token_accuracy": 0.18665484338998795,
      "num_tokens": 43106081.0,
      "step": 24850
    },
    {
      "entropy": 5.597801923751831,
      "epoch": 1.9338261038708422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046225294177457105,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19591110348701476,
      "num_tokens": 43116056.0,
      "step": 24855
    },
    {
      "entropy": 5.671476554870606,
      "epoch": 1.9342151332425597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046223757420603445,
      "loss": 4.9952,
      "mean_token_accuracy": 0.208236400783062,
      "num_tokens": 43124945.0,
      "step": 24860
    },
    {
      "entropy": 5.681667423248291,
      "epoch": 1.9346041626142774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004622222037964658,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20116529762744903,
      "num_tokens": 43134038.0,
      "step": 24865
    },
    {
      "entropy": 5.598711442947388,
      "epoch": 1.934993191985995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046220683054609815,
      "loss": 4.9636,
      "mean_token_accuracy": 0.20030059814453124,
      "num_tokens": 43142296.0,
      "step": 24870
    },
    {
      "entropy": 5.623720264434814,
      "epoch": 1.9353822213577125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046219145445516483,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20603508055210112,
      "num_tokens": 43151789.0,
      "step": 24875
    },
    {
      "entropy": 5.649555778503418,
      "epoch": 1.93577125072943,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046217607552389905,
      "loss": 5.1087,
      "mean_token_accuracy": 0.19947748631238937,
      "num_tokens": 43160876.0,
      "step": 24880
    },
    {
      "entropy": 5.654290628433228,
      "epoch": 1.9361602801011477,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046216069375253435,
      "loss": 5.0046,
      "mean_token_accuracy": 0.20651188790798186,
      "num_tokens": 43169377.0,
      "step": 24885
    },
    {
      "entropy": 5.593468856811524,
      "epoch": 1.9365493094728652,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000462145309141304,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20554946660995482,
      "num_tokens": 43177468.0,
      "step": 24890
    },
    {
      "entropy": 5.615224456787109,
      "epoch": 1.9369383388445827,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004621299216904414,
      "loss": 4.998,
      "mean_token_accuracy": 0.20471084266901016,
      "num_tokens": 43186350.0,
      "step": 24895
    },
    {
      "entropy": 5.551385688781738,
      "epoch": 1.9373273682163004,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046211453140018006,
      "loss": 4.9913,
      "mean_token_accuracy": 0.1963523030281067,
      "num_tokens": 43195124.0,
      "step": 24900
    },
    {
      "entropy": 5.591964149475098,
      "epoch": 1.937716397588018,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004620991382707537,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20457567572593688,
      "num_tokens": 43203617.0,
      "step": 24905
    },
    {
      "entropy": 5.529897975921631,
      "epoch": 1.9381054269597353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046208374230239556,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2080569237470627,
      "num_tokens": 43212790.0,
      "step": 24910
    },
    {
      "entropy": 5.630673265457153,
      "epoch": 1.938494456331453,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004620683434953394,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20583195686340333,
      "num_tokens": 43221185.0,
      "step": 24915
    },
    {
      "entropy": 5.6958580493927,
      "epoch": 1.9388834857031707,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046205294184981896,
      "loss": 5.1428,
      "mean_token_accuracy": 0.1950034096837044,
      "num_tokens": 43230524.0,
      "step": 24920
    },
    {
      "entropy": 5.632674741744995,
      "epoch": 1.9392725150748882,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046203753736606787,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2101844295859337,
      "num_tokens": 43239171.0,
      "step": 24925
    },
    {
      "entropy": 5.556626272201538,
      "epoch": 1.9396615444466057,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004620221300443197,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20043401271104813,
      "num_tokens": 43247553.0,
      "step": 24930
    },
    {
      "entropy": 5.534998607635498,
      "epoch": 1.9400505738183234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004620067198848086,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2006531447172165,
      "num_tokens": 43256127.0,
      "step": 24935
    },
    {
      "entropy": 5.611560201644897,
      "epoch": 1.9404396031900408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046199130688776805,
      "loss": 5.0549,
      "mean_token_accuracy": 0.19863448590040206,
      "num_tokens": 43264959.0,
      "step": 24940
    },
    {
      "entropy": 5.609650468826294,
      "epoch": 1.9408286325617583,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046197589105343204,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2046726554632187,
      "num_tokens": 43273465.0,
      "step": 24945
    },
    {
      "entropy": 5.642408752441407,
      "epoch": 1.941217661933476,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004619604723820345,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2026524692773819,
      "num_tokens": 43282027.0,
      "step": 24950
    },
    {
      "entropy": 5.635498428344727,
      "epoch": 1.9416066913051937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004619450508738094,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19632991552352905,
      "num_tokens": 43291202.0,
      "step": 24955
    },
    {
      "entropy": 5.678465080261231,
      "epoch": 1.941995720676911,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046192962652899056,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20204724222421647,
      "num_tokens": 43299594.0,
      "step": 24960
    },
    {
      "entropy": 5.506816720962524,
      "epoch": 1.9423847500486286,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046191419934781236,
      "loss": 4.8516,
      "mean_token_accuracy": 0.21347858160734176,
      "num_tokens": 43307775.0,
      "step": 24965
    },
    {
      "entropy": 5.594444942474365,
      "epoch": 1.9427737794203463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004618987693305086,
      "loss": 5.0052,
      "mean_token_accuracy": 0.19163881987333298,
      "num_tokens": 43316737.0,
      "step": 24970
    },
    {
      "entropy": 5.585969686508179,
      "epoch": 1.9431628087920638,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618833364773135,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21105848848819733,
      "num_tokens": 43325220.0,
      "step": 24975
    },
    {
      "entropy": 5.672573614120483,
      "epoch": 1.9435518381637813,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046186790078846127,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19802486896514893,
      "num_tokens": 43333957.0,
      "step": 24980
    },
    {
      "entropy": 5.6479820728302,
      "epoch": 1.943940867535499,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046185246226418603,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1979891300201416,
      "num_tokens": 43342668.0,
      "step": 24985
    },
    {
      "entropy": 5.573852252960205,
      "epoch": 1.9443298969072165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046183702090472206,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20102375745773315,
      "num_tokens": 43351805.0,
      "step": 24990
    },
    {
      "entropy": 5.583888339996338,
      "epoch": 1.944718926278934,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618215767103037,
      "loss": 4.9774,
      "mean_token_accuracy": 0.1992226868867874,
      "num_tokens": 43360184.0,
      "step": 24995
    },
    {
      "entropy": 5.58269853591919,
      "epoch": 1.9451079556506516,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004618061296811653,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19429562240839005,
      "num_tokens": 43368576.0,
      "step": 25000
    },
    {
      "entropy": 5.619055843353271,
      "epoch": 1.9454969850223693,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046179067981754124,
      "loss": 4.9834,
      "mean_token_accuracy": 0.1975250646471977,
      "num_tokens": 43377561.0,
      "step": 25005
    },
    {
      "entropy": 5.576317691802979,
      "epoch": 1.9458860143940866,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046177522711966584,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20405238270759582,
      "num_tokens": 43386276.0,
      "step": 25010
    },
    {
      "entropy": 5.57113561630249,
      "epoch": 1.9462750437658043,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046175977158777377,
      "loss": 4.9919,
      "mean_token_accuracy": 0.19586019963026047,
      "num_tokens": 43394753.0,
      "step": 25015
    },
    {
      "entropy": 5.554373025894165,
      "epoch": 1.946664073137522,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004617443132220993,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21379523128271102,
      "num_tokens": 43403462.0,
      "step": 25020
    },
    {
      "entropy": 5.654671144485474,
      "epoch": 1.9470531025092395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046172885202287717,
      "loss": 4.9907,
      "mean_token_accuracy": 0.19641466289758683,
      "num_tokens": 43412192.0,
      "step": 25025
    },
    {
      "entropy": 5.642926788330078,
      "epoch": 1.947442131880957,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046171338799034194,
      "loss": 5.0438,
      "mean_token_accuracy": 0.20044593662023544,
      "num_tokens": 43421273.0,
      "step": 25030
    },
    {
      "entropy": 5.694089794158936,
      "epoch": 1.9478311612526746,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004616979211247282,
      "loss": 5.1391,
      "mean_token_accuracy": 0.19465840756893157,
      "num_tokens": 43430257.0,
      "step": 25035
    },
    {
      "entropy": 5.630743789672851,
      "epoch": 1.9482201906243921,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046168245142627065,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19577836841344834,
      "num_tokens": 43438258.0,
      "step": 25040
    },
    {
      "entropy": 5.649375867843628,
      "epoch": 1.9486092199961096,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004616669788952041,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2058025762438774,
      "num_tokens": 43447221.0,
      "step": 25045
    },
    {
      "entropy": 5.599972152709961,
      "epoch": 1.9489982493678273,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046165150353176315,
      "loss": 4.9486,
      "mean_token_accuracy": 0.20481569468975067,
      "num_tokens": 43455909.0,
      "step": 25050
    },
    {
      "entropy": 5.6341465473175045,
      "epoch": 1.949387278739545,
      "grad_norm": 1.125,
      "learning_rate": 0.0004616360253361828,
      "loss": 5.1049,
      "mean_token_accuracy": 0.19357359260320664,
      "num_tokens": 43464262.0,
      "step": 25055
    },
    {
      "entropy": 5.647072744369507,
      "epoch": 1.9497763081112622,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046162054430869777,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20909491926431656,
      "num_tokens": 43472897.0,
      "step": 25060
    },
    {
      "entropy": 5.699408292770386,
      "epoch": 1.95016533748298,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000461605060449543,
      "loss": 5.1164,
      "mean_token_accuracy": 0.19576744139194488,
      "num_tokens": 43481714.0,
      "step": 25065
    },
    {
      "entropy": 5.520859909057617,
      "epoch": 1.9505543668546976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046158957375895353,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20298203229904174,
      "num_tokens": 43490032.0,
      "step": 25070
    },
    {
      "entropy": 5.580363702774048,
      "epoch": 1.950943396226415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004615740842371643,
      "loss": 5.0228,
      "mean_token_accuracy": 0.1948147863149643,
      "num_tokens": 43498466.0,
      "step": 25075
    },
    {
      "entropy": 5.639126348495483,
      "epoch": 1.9513324255981326,
      "grad_norm": 1.125,
      "learning_rate": 0.00046155859188441023,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2060020700097084,
      "num_tokens": 43506874.0,
      "step": 25080
    },
    {
      "entropy": 5.59894871711731,
      "epoch": 1.9517214549698503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004615430967009265,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20507294535636902,
      "num_tokens": 43515657.0,
      "step": 25085
    },
    {
      "entropy": 5.603897523880005,
      "epoch": 1.9521104843415678,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046152759868694815,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20148796439170838,
      "num_tokens": 43524934.0,
      "step": 25090
    },
    {
      "entropy": 5.669860076904297,
      "epoch": 1.9524995137132852,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004615120978427104,
      "loss": 5.0674,
      "mean_token_accuracy": 0.19333964139223098,
      "num_tokens": 43533988.0,
      "step": 25095
    },
    {
      "entropy": 5.677996635437012,
      "epoch": 1.952888543085003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004614965941684485,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2044929727911949,
      "num_tokens": 43542868.0,
      "step": 25100
    },
    {
      "entropy": 5.674543476104736,
      "epoch": 1.9532775724567206,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004614810876643975,
      "loss": 5.119,
      "mean_token_accuracy": 0.19272132813930512,
      "num_tokens": 43552577.0,
      "step": 25105
    },
    {
      "entropy": 5.6838987350463865,
      "epoch": 1.953666601828438,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046146557833079286,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1969879686832428,
      "num_tokens": 43562088.0,
      "step": 25110
    },
    {
      "entropy": 5.641274070739746,
      "epoch": 1.9540556312001556,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004614500661678698,
      "loss": 5.0629,
      "mean_token_accuracy": 0.19834199249744416,
      "num_tokens": 43570268.0,
      "step": 25115
    },
    {
      "entropy": 5.647603845596313,
      "epoch": 1.9544446605718733,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004614345511758639,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20309446901082992,
      "num_tokens": 43578751.0,
      "step": 25120
    },
    {
      "entropy": 5.6244651794433596,
      "epoch": 1.9548336899435907,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046141903335501034,
      "loss": 4.957,
      "mean_token_accuracy": 0.2006782680749893,
      "num_tokens": 43588229.0,
      "step": 25125
    },
    {
      "entropy": 5.557975769042969,
      "epoch": 1.9552227193153082,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004614035127055447,
      "loss": 4.961,
      "mean_token_accuracy": 0.20105046033859253,
      "num_tokens": 43596623.0,
      "step": 25130
    },
    {
      "entropy": 5.530763149261475,
      "epoch": 1.955611748687026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004613879892277024,
      "loss": 4.9027,
      "mean_token_accuracy": 0.20824336707592012,
      "num_tokens": 43605954.0,
      "step": 25135
    },
    {
      "entropy": 5.614673089981079,
      "epoch": 1.9560007780587434,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000461372462921719,
      "loss": 5.0139,
      "mean_token_accuracy": 0.1966230183839798,
      "num_tokens": 43613822.0,
      "step": 25140
    },
    {
      "entropy": 5.577919960021973,
      "epoch": 1.9563898074304609,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004613569337878302,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21031264662742616,
      "num_tokens": 43622602.0,
      "step": 25145
    },
    {
      "entropy": 5.645202732086181,
      "epoch": 1.9567788368021786,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004613414018262715,
      "loss": 5.0612,
      "mean_token_accuracy": 0.19623580276966096,
      "num_tokens": 43630820.0,
      "step": 25150
    },
    {
      "entropy": 5.595655250549316,
      "epoch": 1.9571678661738963,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004613258670372786,
      "loss": 5.0231,
      "mean_token_accuracy": 0.20416345000267028,
      "num_tokens": 43638571.0,
      "step": 25155
    },
    {
      "entropy": 5.586620426177978,
      "epoch": 1.9575568955456137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004613103294210873,
      "loss": 5.0578,
      "mean_token_accuracy": 0.19677623063325883,
      "num_tokens": 43648575.0,
      "step": 25160
    },
    {
      "entropy": 5.610661840438842,
      "epoch": 1.9579459249173312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046129478897793323,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20606261640787124,
      "num_tokens": 43656241.0,
      "step": 25165
    },
    {
      "entropy": 5.750542020797729,
      "epoch": 1.958334954289049,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046127924570805236,
      "loss": 5.144,
      "mean_token_accuracy": 0.19841453731060027,
      "num_tokens": 43665023.0,
      "step": 25170
    },
    {
      "entropy": 5.588088846206665,
      "epoch": 1.9587239836607664,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004612636996116803,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20586448013782502,
      "num_tokens": 43673455.0,
      "step": 25175
    },
    {
      "entropy": 5.635645914077759,
      "epoch": 1.9591130130324839,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004612481506890532,
      "loss": 5.0304,
      "mean_token_accuracy": 0.19816944748163223,
      "num_tokens": 43681679.0,
      "step": 25180
    },
    {
      "entropy": 5.681969785690308,
      "epoch": 1.9595020424042016,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046123259894040677,
      "loss": 5.0225,
      "mean_token_accuracy": 0.1994195356965065,
      "num_tokens": 43690062.0,
      "step": 25185
    },
    {
      "entropy": 5.6451390266418455,
      "epoch": 1.959891071775919,
      "grad_norm": 1.203125,
      "learning_rate": 0.000461217044365977,
      "loss": 5.0453,
      "mean_token_accuracy": 0.19377863854169847,
      "num_tokens": 43698301.0,
      "step": 25190
    },
    {
      "entropy": 5.645369052886963,
      "epoch": 1.9602801011476365,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004612014869660002,
      "loss": 5.0778,
      "mean_token_accuracy": 0.19743055999279022,
      "num_tokens": 43707221.0,
      "step": 25195
    },
    {
      "entropy": 5.622927141189575,
      "epoch": 1.9606691305193542,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046118592674071197,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19979306012392045,
      "num_tokens": 43716552.0,
      "step": 25200
    },
    {
      "entropy": 5.652859354019165,
      "epoch": 1.961058159891072,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004611703636903488,
      "loss": 4.9869,
      "mean_token_accuracy": 0.19607416540384293,
      "num_tokens": 43725296.0,
      "step": 25205
    },
    {
      "entropy": 5.711751651763916,
      "epoch": 1.9614471892627894,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004611547978151466,
      "loss": 5.1173,
      "mean_token_accuracy": 0.1953204721212387,
      "num_tokens": 43733398.0,
      "step": 25210
    },
    {
      "entropy": 5.559988403320313,
      "epoch": 1.9618362186345069,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046113922911534167,
      "loss": 4.987,
      "mean_token_accuracy": 0.20678746253252028,
      "num_tokens": 43741679.0,
      "step": 25215
    },
    {
      "entropy": 5.636204862594605,
      "epoch": 1.9622252480062246,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004611236575911702,
      "loss": 5.0414,
      "mean_token_accuracy": 0.19247227311134338,
      "num_tokens": 43750734.0,
      "step": 25220
    },
    {
      "entropy": 5.681608963012695,
      "epoch": 1.962614277377942,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046110808324286844,
      "loss": 5.0719,
      "mean_token_accuracy": 0.19714713394641875,
      "num_tokens": 43759885.0,
      "step": 25225
    },
    {
      "entropy": 5.6347908020019535,
      "epoch": 1.9630033067496595,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046109250607067275,
      "loss": 5.0051,
      "mean_token_accuracy": 0.19859375357627868,
      "num_tokens": 43769199.0,
      "step": 25230
    },
    {
      "entropy": 5.651576519012451,
      "epoch": 1.9633923361213772,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004610769260748196,
      "loss": 4.9831,
      "mean_token_accuracy": 0.20084545016288757,
      "num_tokens": 43777930.0,
      "step": 25235
    },
    {
      "entropy": 5.648206329345703,
      "epoch": 1.9637813654930947,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004610613432555451,
      "loss": 5.0683,
      "mean_token_accuracy": 0.19863332957029342,
      "num_tokens": 43786297.0,
      "step": 25240
    },
    {
      "entropy": 5.577006053924561,
      "epoch": 1.9641703948648122,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046104575761308597,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2059046134352684,
      "num_tokens": 43794157.0,
      "step": 25245
    },
    {
      "entropy": 5.628083324432373,
      "epoch": 1.9645594242365298,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004610301691476786,
      "loss": 5.0922,
      "mean_token_accuracy": 0.20064326226711274,
      "num_tokens": 43802926.0,
      "step": 25250
    },
    {
      "entropy": 5.618759679794311,
      "epoch": 1.9649484536082475,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004610145778595594,
      "loss": 4.9761,
      "mean_token_accuracy": 0.1999736100435257,
      "num_tokens": 43811446.0,
      "step": 25255
    },
    {
      "entropy": 5.708820438385009,
      "epoch": 1.965337482979965,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004609989837489651,
      "loss": 5.0395,
      "mean_token_accuracy": 0.19431572258472443,
      "num_tokens": 43819851.0,
      "step": 25260
    },
    {
      "entropy": 5.647942113876343,
      "epoch": 1.9657265123516825,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046098338681613234,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20012189149856568,
      "num_tokens": 43828763.0,
      "step": 25265
    },
    {
      "entropy": 5.594118595123291,
      "epoch": 1.9661155417234002,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004609677870612976,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20427397787570953,
      "num_tokens": 43837151.0,
      "step": 25270
    },
    {
      "entropy": 5.635504579544067,
      "epoch": 1.9665045710951177,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004609521844846978,
      "loss": 5.0268,
      "mean_token_accuracy": 0.19738633334636688,
      "num_tokens": 43845823.0,
      "step": 25275
    },
    {
      "entropy": 5.581147527694702,
      "epoch": 1.9668936004668351,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004609365790865695,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21498498618602752,
      "num_tokens": 43854551.0,
      "step": 25280
    },
    {
      "entropy": 5.669602489471435,
      "epoch": 1.9672826298385528,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046092097086714956,
      "loss": 5.0608,
      "mean_token_accuracy": 0.1942002832889557,
      "num_tokens": 43863839.0,
      "step": 25285
    },
    {
      "entropy": 5.607984638214111,
      "epoch": 1.9676716592102705,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046090535982667486,
      "loss": 5.092,
      "mean_token_accuracy": 0.19586304277181626,
      "num_tokens": 43872401.0,
      "step": 25290
    },
    {
      "entropy": 5.6260762214660645,
      "epoch": 1.9680606885819878,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046088974596538216,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19866797775030137,
      "num_tokens": 43881378.0,
      "step": 25295
    },
    {
      "entropy": 5.772875547409058,
      "epoch": 1.9684497179537055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608741292835085,
      "loss": 5.2449,
      "mean_token_accuracy": 0.1882839471101761,
      "num_tokens": 43891047.0,
      "step": 25300
    },
    {
      "entropy": 5.70738844871521,
      "epoch": 1.9688387473254232,
      "grad_norm": 1.25,
      "learning_rate": 0.0004608585097812907,
      "loss": 5.148,
      "mean_token_accuracy": 0.18614716529846193,
      "num_tokens": 43898844.0,
      "step": 25305
    },
    {
      "entropy": 5.642058801651001,
      "epoch": 1.9692277766971407,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004608428874589659,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2027621313929558,
      "num_tokens": 43907398.0,
      "step": 25310
    },
    {
      "entropy": 5.659902667999267,
      "epoch": 1.9696168060688581,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004608272623167711,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20097365826368332,
      "num_tokens": 43915184.0,
      "step": 25315
    },
    {
      "entropy": 5.651368856430054,
      "epoch": 1.9700058354405758,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046081163435494335,
      "loss": 4.9868,
      "mean_token_accuracy": 0.20818507075309753,
      "num_tokens": 43923496.0,
      "step": 25320
    },
    {
      "entropy": 5.6052039623260494,
      "epoch": 1.9703948648122933,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004607960035737198,
      "loss": 4.8938,
      "mean_token_accuracy": 0.2096571519970894,
      "num_tokens": 43931435.0,
      "step": 25325
    },
    {
      "entropy": 5.606964159011841,
      "epoch": 1.9707838941840108,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004607803699733376,
      "loss": 5.0992,
      "mean_token_accuracy": 0.20533965080976485,
      "num_tokens": 43939644.0,
      "step": 25330
    },
    {
      "entropy": 5.624517822265625,
      "epoch": 1.9711729235557285,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000460764733554034,
      "loss": 4.9985,
      "mean_token_accuracy": 0.204326094686985,
      "num_tokens": 43947611.0,
      "step": 25335
    },
    {
      "entropy": 5.59662275314331,
      "epoch": 1.9715619529274462,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046074909431604623,
      "loss": 5.0026,
      "mean_token_accuracy": 0.20859443098306657,
      "num_tokens": 43956876.0,
      "step": 25340
    },
    {
      "entropy": 5.556484842300415,
      "epoch": 1.9719509822991634,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004607334522596116,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20653315484523774,
      "num_tokens": 43965573.0,
      "step": 25345
    },
    {
      "entropy": 5.641865873336792,
      "epoch": 1.9723400116708811,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004607178073849675,
      "loss": 5.0999,
      "mean_token_accuracy": 0.19097919762134552,
      "num_tokens": 43974937.0,
      "step": 25350
    },
    {
      "entropy": 5.690134334564209,
      "epoch": 1.9727290410425988,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004607021596923512,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20773436725139618,
      "num_tokens": 43983629.0,
      "step": 25355
    },
    {
      "entropy": 5.625349760055542,
      "epoch": 1.9731180704143163,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004606865091820003,
      "loss": 5.0748,
      "mean_token_accuracy": 0.19699892699718474,
      "num_tokens": 43992655.0,
      "step": 25360
    },
    {
      "entropy": 5.591175127029419,
      "epoch": 1.9735070997860338,
      "grad_norm": 1.25,
      "learning_rate": 0.0004606708558541521,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20070958584547044,
      "num_tokens": 44001789.0,
      "step": 25365
    },
    {
      "entropy": 5.523272371292114,
      "epoch": 1.9738961291577515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004606551997090442,
      "loss": 4.8909,
      "mean_token_accuracy": 0.20621975958347322,
      "num_tokens": 44010940.0,
      "step": 25370
    },
    {
      "entropy": 5.655936861038208,
      "epoch": 1.974285158529469,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606395407469141,
      "loss": 5.1015,
      "mean_token_accuracy": 0.19859305322170256,
      "num_tokens": 44019207.0,
      "step": 25375
    },
    {
      "entropy": 5.615401029586792,
      "epoch": 1.9746741879011864,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004606238789679995,
      "loss": 5.0561,
      "mean_token_accuracy": 0.19314056485891343,
      "num_tokens": 44027529.0,
      "step": 25380
    },
    {
      "entropy": 5.595882844924927,
      "epoch": 1.9750632172729041,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004606082143725381,
      "loss": 4.98,
      "mean_token_accuracy": 0.20304424166679383,
      "num_tokens": 44036036.0,
      "step": 25385
    },
    {
      "entropy": 5.684293556213379,
      "epoch": 1.9754522466446218,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004605925469607673,
      "loss": 5.0972,
      "mean_token_accuracy": 0.19621259421110154,
      "num_tokens": 44044211.0,
      "step": 25390
    },
    {
      "entropy": 5.601698780059815,
      "epoch": 1.975841276016339,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046057687673292506,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20113127678632736,
      "num_tokens": 44052735.0,
      "step": 25395
    },
    {
      "entropy": 5.55897011756897,
      "epoch": 1.9762303053880568,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046056120368924907,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2016784057021141,
      "num_tokens": 44061629.0,
      "step": 25400
    },
    {
      "entropy": 5.632994365692139,
      "epoch": 1.9766193347597745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004605455278299772,
      "loss": 5.0662,
      "mean_token_accuracy": 0.2004440262913704,
      "num_tokens": 44070381.0,
      "step": 25405
    },
    {
      "entropy": 5.6185524463653564,
      "epoch": 1.977008364131492,
      "grad_norm": 1.125,
      "learning_rate": 0.0004605298491553472,
      "loss": 5.0488,
      "mean_token_accuracy": 0.20301635712385177,
      "num_tokens": 44078609.0,
      "step": 25410
    },
    {
      "entropy": 5.531156730651856,
      "epoch": 1.9773973935032094,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004605141676655971,
      "loss": 4.934,
      "mean_token_accuracy": 0.20573256015777588,
      "num_tokens": 44087151.0,
      "step": 25415
    },
    {
      "entropy": 5.585275793075562,
      "epoch": 1.9777864228749271,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046049848336096485,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20254035145044327,
      "num_tokens": 44095658.0,
      "step": 25420
    },
    {
      "entropy": 5.601599407196045,
      "epoch": 1.9781754522466446,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004604827962416883,
      "loss": 5.028,
      "mean_token_accuracy": 0.20641974061727525,
      "num_tokens": 44103789.0,
      "step": 25425
    },
    {
      "entropy": 5.627568531036377,
      "epoch": 1.978564481618362,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004604671063080055,
      "loss": 5.055,
      "mean_token_accuracy": 0.19585973471403123,
      "num_tokens": 44113082.0,
      "step": 25430
    },
    {
      "entropy": 5.656610202789307,
      "epoch": 1.9789535109900798,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004604514135601546,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20328574478626252,
      "num_tokens": 44121338.0,
      "step": 25435
    },
    {
      "entropy": 5.638218402862549,
      "epoch": 1.9793425403617975,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004604357179983736,
      "loss": 5.0411,
      "mean_token_accuracy": 0.20142284482717515,
      "num_tokens": 44130551.0,
      "step": 25440
    },
    {
      "entropy": 5.608001661300659,
      "epoch": 1.9797315697335147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004604200196229009,
      "loss": 5.0651,
      "mean_token_accuracy": 0.19927018880844116,
      "num_tokens": 44139138.0,
      "step": 25445
    },
    {
      "entropy": 5.578807973861695,
      "epoch": 1.9801205991052324,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004604043184339744,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20072684735059737,
      "num_tokens": 44148123.0,
      "step": 25450
    },
    {
      "entropy": 5.536021137237549,
      "epoch": 1.9805096284769501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004603886144318325,
      "loss": 4.866,
      "mean_token_accuracy": 0.2106361582875252,
      "num_tokens": 44156062.0,
      "step": 25455
    },
    {
      "entropy": 5.563628053665161,
      "epoch": 1.9808986578486676,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046037290761671346,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20518584847450255,
      "num_tokens": 44165049.0,
      "step": 25460
    },
    {
      "entropy": 5.6947657585144045,
      "epoch": 1.981287687220385,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004603571979888556,
      "loss": 5.1269,
      "mean_token_accuracy": 0.19949647337198256,
      "num_tokens": 44173774.0,
      "step": 25465
    },
    {
      "entropy": 5.665951538085937,
      "epoch": 1.9816767165921028,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004603414855484973,
      "loss": 5.0454,
      "mean_token_accuracy": 0.2025372326374054,
      "num_tokens": 44182933.0,
      "step": 25470
    },
    {
      "entropy": 5.683129930496216,
      "epoch": 1.9820657459638202,
      "grad_norm": 1.15625,
      "learning_rate": 0.000460325770295877,
      "loss": 5.055,
      "mean_token_accuracy": 0.20221837610006332,
      "num_tokens": 44190952.0,
      "step": 25475
    },
    {
      "entropy": 5.711680173873901,
      "epoch": 1.9824547753355377,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046031005223123297,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20273138880729674,
      "num_tokens": 44199625.0,
      "step": 25480
    },
    {
      "entropy": 5.616965675354004,
      "epoch": 1.9828438047072554,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000460294331354804,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1987301617860794,
      "num_tokens": 44207948.0,
      "step": 25485
    },
    {
      "entropy": 5.666461372375489,
      "epoch": 1.983232834078973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004602786076668283,
      "loss": 5.209,
      "mean_token_accuracy": 0.18254822492599487,
      "num_tokens": 44217120.0,
      "step": 25490
    },
    {
      "entropy": 5.726787948608399,
      "epoch": 1.9836218634506906,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046026288116754477,
      "loss": 5.1129,
      "mean_token_accuracy": 0.19329866915941238,
      "num_tokens": 44226099.0,
      "step": 25495
    },
    {
      "entropy": 5.69171290397644,
      "epoch": 1.984010892822408,
      "grad_norm": 1.125,
      "learning_rate": 0.0004602471518571919,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20420074760913848,
      "num_tokens": 44234256.0,
      "step": 25500
    },
    {
      "entropy": 5.613227462768554,
      "epoch": 1.9843999221941258,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004602314197360083,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2008064717054367,
      "num_tokens": 44242283.0,
      "step": 25505
    },
    {
      "entropy": 5.549334335327148,
      "epoch": 1.9847889515658432,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004602156848042328,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2041727513074875,
      "num_tokens": 44251057.0,
      "step": 25510
    },
    {
      "entropy": 5.644768190383911,
      "epoch": 1.9851779809375607,
      "grad_norm": 1.140625,
      "learning_rate": 0.000460199947062104,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1953382283449173,
      "num_tokens": 44259726.0,
      "step": 25515
    },
    {
      "entropy": 5.717181539535522,
      "epoch": 1.9855670103092784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046018420650986074,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1979443073272705,
      "num_tokens": 44268995.0,
      "step": 25520
    },
    {
      "entropy": 5.606896495819091,
      "epoch": 1.9859560396809959,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000460168463147742,
      "loss": 4.9552,
      "mean_token_accuracy": 0.20048964768648148,
      "num_tokens": 44277598.0,
      "step": 25525
    },
    {
      "entropy": 5.586391592025757,
      "epoch": 1.9863450690527134,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004601527169759865,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20657315701246262,
      "num_tokens": 44286233.0,
      "step": 25530
    },
    {
      "entropy": 5.571649360656738,
      "epoch": 1.986734098424431,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004601369679948333,
      "loss": 4.95,
      "mean_token_accuracy": 0.2089812859892845,
      "num_tokens": 44295078.0,
      "step": 25535
    },
    {
      "entropy": 5.561828470230102,
      "epoch": 1.9871231277961487,
      "grad_norm": 1.125,
      "learning_rate": 0.00046012121620452127,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20704407691955568,
      "num_tokens": 44304050.0,
      "step": 25540
    },
    {
      "entropy": 5.689087533950806,
      "epoch": 1.9875121571678662,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004601054616052893,
      "loss": 5.0889,
      "mean_token_accuracy": 0.19555749297142028,
      "num_tokens": 44313025.0,
      "step": 25545
    },
    {
      "entropy": 5.590252542495728,
      "epoch": 1.9879011865395837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046008970419737674,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20949381738901138,
      "num_tokens": 44322295.0,
      "step": 25550
    },
    {
      "entropy": 5.559688234329224,
      "epoch": 1.9882902159113014,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004600739439810225,
      "loss": 4.925,
      "mean_token_accuracy": 0.20875637233257294,
      "num_tokens": 44329865.0,
      "step": 25555
    },
    {
      "entropy": 5.625823545455932,
      "epoch": 1.9886792452830189,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046005818095646564,
      "loss": 5.0544,
      "mean_token_accuracy": 0.2057286873459816,
      "num_tokens": 44338124.0,
      "step": 25560
    },
    {
      "entropy": 5.546399831771851,
      "epoch": 1.9890682746547363,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046004241512394544,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20488054156303406,
      "num_tokens": 44346841.0,
      "step": 25565
    },
    {
      "entropy": 5.536183309555054,
      "epoch": 1.989457304026454,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004600266464837012,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20167540907859802,
      "num_tokens": 44355677.0,
      "step": 25570
    },
    {
      "entropy": 5.664252805709839,
      "epoch": 1.9898463333981715,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004600108750359721,
      "loss": 4.9826,
      "mean_token_accuracy": 0.19899121671915054,
      "num_tokens": 44363916.0,
      "step": 25575
    },
    {
      "entropy": 5.659473419189453,
      "epoch": 1.990235362769889,
      "grad_norm": 1.125,
      "learning_rate": 0.00045999510078099736,
      "loss": 5.0541,
      "mean_token_accuracy": 0.19745057970285415,
      "num_tokens": 44372770.0,
      "step": 25580
    },
    {
      "entropy": 5.70248064994812,
      "epoch": 1.9906243921416067,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045997932371901645,
      "loss": 5.0982,
      "mean_token_accuracy": 0.1969503492116928,
      "num_tokens": 44380724.0,
      "step": 25585
    },
    {
      "entropy": 5.582770252227784,
      "epoch": 1.9910134215133244,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004599635438502687,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20297805219888687,
      "num_tokens": 44389915.0,
      "step": 25590
    },
    {
      "entropy": 5.557068634033203,
      "epoch": 1.9914024508850419,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045994776117499363,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21086338460445403,
      "num_tokens": 44398297.0,
      "step": 25595
    },
    {
      "entropy": 5.649922037124634,
      "epoch": 1.9917914802567593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045993197569343063,
      "loss": 5.1657,
      "mean_token_accuracy": 0.19753341376781464,
      "num_tokens": 44407713.0,
      "step": 25600
    },
    {
      "entropy": 5.645587825775147,
      "epoch": 1.992180509628477,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045991618740581926,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2027626022696495,
      "num_tokens": 44416422.0,
      "step": 25605
    },
    {
      "entropy": 5.596484375,
      "epoch": 1.9925695390001945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004599003963123991,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1974819779396057,
      "num_tokens": 44425899.0,
      "step": 25610
    },
    {
      "entropy": 5.636276721954346,
      "epoch": 1.992958568371912,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045988460241340966,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2004519060254097,
      "num_tokens": 44434233.0,
      "step": 25615
    },
    {
      "entropy": 5.612360525131225,
      "epoch": 1.9933475977436297,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045986880570909075,
      "loss": 5.0615,
      "mean_token_accuracy": 0.19459983259439467,
      "num_tokens": 44443082.0,
      "step": 25620
    },
    {
      "entropy": 5.629563474655152,
      "epoch": 1.9937366271153472,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045985300619968186,
      "loss": 5.0304,
      "mean_token_accuracy": 0.20334535986185073,
      "num_tokens": 44451482.0,
      "step": 25625
    },
    {
      "entropy": 5.614025545120239,
      "epoch": 1.9941256564870646,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045983720388542286,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20137010216712953,
      "num_tokens": 44460264.0,
      "step": 25630
    },
    {
      "entropy": 5.530329751968384,
      "epoch": 1.9945146858587823,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004598213987665535,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21302837878465652,
      "num_tokens": 44468686.0,
      "step": 25635
    },
    {
      "entropy": 5.577920961380005,
      "epoch": 1.9949037152305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045980559084331354,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20338212251663207,
      "num_tokens": 44476835.0,
      "step": 25640
    },
    {
      "entropy": 5.576545095443725,
      "epoch": 1.9952927446022175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000459789780115943,
      "loss": 5.0343,
      "mean_token_accuracy": 0.20248454809188843,
      "num_tokens": 44485626.0,
      "step": 25645
    },
    {
      "entropy": 5.696612119674683,
      "epoch": 1.995681773973935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045977396658468156,
      "loss": 5.0878,
      "mean_token_accuracy": 0.19952531307935714,
      "num_tokens": 44494590.0,
      "step": 25650
    },
    {
      "entropy": 5.691667652130127,
      "epoch": 1.9960708033456527,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004597581502497693,
      "loss": 5.0952,
      "mean_token_accuracy": 0.20225452184677123,
      "num_tokens": 44503115.0,
      "step": 25655
    },
    {
      "entropy": 5.642867755889893,
      "epoch": 1.9964598327173702,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004597423311114462,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20246998518705367,
      "num_tokens": 44511416.0,
      "step": 25660
    },
    {
      "entropy": 5.594935321807862,
      "epoch": 1.9968488620890876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004597265091699522,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2004894495010376,
      "num_tokens": 44519978.0,
      "step": 25665
    },
    {
      "entropy": 5.538840579986572,
      "epoch": 1.9972378914608053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004597106844255276,
      "loss": 4.9514,
      "mean_token_accuracy": 0.19916902631521224,
      "num_tokens": 44528438.0,
      "step": 25670
    },
    {
      "entropy": 5.594379615783692,
      "epoch": 1.9976269208325228,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045969485687841227,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1987212598323822,
      "num_tokens": 44537126.0,
      "step": 25675
    },
    {
      "entropy": 5.575970268249511,
      "epoch": 1.9980159502042403,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045967902652884645,
      "loss": 4.9194,
      "mean_token_accuracy": 0.20549869984388353,
      "num_tokens": 44545874.0,
      "step": 25680
    },
    {
      "entropy": 5.587207365036011,
      "epoch": 1.998404979575958,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004596631933770704,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2029881000518799,
      "num_tokens": 44554515.0,
      "step": 25685
    },
    {
      "entropy": 5.64042444229126,
      "epoch": 1.9987940089476757,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045964735742332427,
      "loss": 5.01,
      "mean_token_accuracy": 0.20413050055503845,
      "num_tokens": 44562902.0,
      "step": 25690
    },
    {
      "entropy": 5.67670168876648,
      "epoch": 1.9991830383193931,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004596315186678484,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20368304550647737,
      "num_tokens": 44571482.0,
      "step": 25695
    },
    {
      "entropy": 5.6492283821105955,
      "epoch": 1.9995720676911106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045961567711088307,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2074311524629593,
      "num_tokens": 44580686.0,
      "step": 25700
    },
    {
      "entropy": 5.616321229934693,
      "epoch": 1.9999610970628283,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045959983275266873,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20075726956129075,
      "num_tokens": 44588317.0,
      "step": 25705
    },
    {
      "entropy": 5.6290525860256615,
      "epoch": 2.000311223497374,
      "grad_norm": 1.125,
      "learning_rate": 0.00045958398559344573,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1971252593729231,
      "num_tokens": 44596453.0,
      "step": 25710
    },
    {
      "entropy": 5.65076699256897,
      "epoch": 2.0007002528690916,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045956813563345454,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2115452006459236,
      "num_tokens": 44605444.0,
      "step": 25715
    },
    {
      "entropy": 5.680522727966308,
      "epoch": 2.0010892822408093,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004595522828729357,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20203300714492797,
      "num_tokens": 44614156.0,
      "step": 25720
    },
    {
      "entropy": 5.534897184371948,
      "epoch": 2.0014783116125265,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004595364273121296,
      "loss": 4.9117,
      "mean_token_accuracy": 0.21203378289937974,
      "num_tokens": 44622413.0,
      "step": 25725
    },
    {
      "entropy": 5.502642726898193,
      "epoch": 2.0018673409842442,
      "grad_norm": 1.125,
      "learning_rate": 0.0004595205689512771,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2178743526339531,
      "num_tokens": 44630940.0,
      "step": 25730
    },
    {
      "entropy": 5.640612983703614,
      "epoch": 2.002256370355962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045950470779061855,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20280055999755858,
      "num_tokens": 44640951.0,
      "step": 25735
    },
    {
      "entropy": 5.620240211486816,
      "epoch": 2.0026453997276796,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045948884383039475,
      "loss": 4.951,
      "mean_token_accuracy": 0.2117963433265686,
      "num_tokens": 44650054.0,
      "step": 25740
    },
    {
      "entropy": 5.628820610046387,
      "epoch": 2.003034429099397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045947297707084637,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2045560970902443,
      "num_tokens": 44658018.0,
      "step": 25745
    },
    {
      "entropy": 5.634444332122802,
      "epoch": 2.0034234584711146,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004594571075122142,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20544345825910568,
      "num_tokens": 44666310.0,
      "step": 25750
    },
    {
      "entropy": 5.577261018753052,
      "epoch": 2.0038124878428323,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045944123515473905,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20551505833864211,
      "num_tokens": 44675008.0,
      "step": 25755
    },
    {
      "entropy": 5.643285036087036,
      "epoch": 2.0042015172145495,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045942535999866175,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2027924671769142,
      "num_tokens": 44683504.0,
      "step": 25760
    },
    {
      "entropy": 5.667280483245849,
      "epoch": 2.0045905465862672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004594094820442231,
      "loss": 5.092,
      "mean_token_accuracy": 0.1969456344842911,
      "num_tokens": 44692079.0,
      "step": 25765
    },
    {
      "entropy": 5.637122344970703,
      "epoch": 2.004979575957985,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000459393601291664,
      "loss": 4.9966,
      "mean_token_accuracy": 0.20347248017787933,
      "num_tokens": 44700256.0,
      "step": 25770
    },
    {
      "entropy": 5.600408840179443,
      "epoch": 2.005368605329702,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045937771774122563,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20849878638982772,
      "num_tokens": 44708597.0,
      "step": 25775
    },
    {
      "entropy": 5.715204954147339,
      "epoch": 2.00575763470142,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004593618313931488,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20541465282440186,
      "num_tokens": 44717583.0,
      "step": 25780
    },
    {
      "entropy": 5.626265573501587,
      "epoch": 2.0061466640731376,
      "grad_norm": 1.125,
      "learning_rate": 0.00045934594224767463,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20146530419588088,
      "num_tokens": 44725661.0,
      "step": 25785
    },
    {
      "entropy": 5.509394884109497,
      "epoch": 2.0065356934448553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045933005030504424,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2073071628808975,
      "num_tokens": 44735045.0,
      "step": 25790
    },
    {
      "entropy": 5.624396085739136,
      "epoch": 2.0069247228165725,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045931415556549863,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21229027807712555,
      "num_tokens": 44743307.0,
      "step": 25795
    },
    {
      "entropy": 5.623622894287109,
      "epoch": 2.0073137521882902,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004592982580292792,
      "loss": 4.9956,
      "mean_token_accuracy": 0.20081831514835358,
      "num_tokens": 44751895.0,
      "step": 25800
    },
    {
      "entropy": 5.585944366455078,
      "epoch": 2.007702781560008,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045928235769662697,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2034072086215019,
      "num_tokens": 44760801.0,
      "step": 25805
    },
    {
      "entropy": 5.663759231567383,
      "epoch": 2.008091810931725,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045926645456778327,
      "loss": 5.089,
      "mean_token_accuracy": 0.19501139223575592,
      "num_tokens": 44769634.0,
      "step": 25810
    },
    {
      "entropy": 5.6328671932220455,
      "epoch": 2.008480840303443,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045925054864298946,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20681709051132202,
      "num_tokens": 44778856.0,
      "step": 25815
    },
    {
      "entropy": 5.622316932678222,
      "epoch": 2.0088698696751606,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045923463992248684,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21183178275823594,
      "num_tokens": 44787750.0,
      "step": 25820
    },
    {
      "entropy": 5.622903490066529,
      "epoch": 2.009258899046878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045921872840651675,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2046786367893219,
      "num_tokens": 44796021.0,
      "step": 25825
    },
    {
      "entropy": 5.6192786693573,
      "epoch": 2.0096479284185955,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045920281409532064,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2015443727374077,
      "num_tokens": 44804247.0,
      "step": 25830
    },
    {
      "entropy": 5.674878549575806,
      "epoch": 2.010036957790313,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004591868969891401,
      "loss": 5.0134,
      "mean_token_accuracy": 0.19631267935037613,
      "num_tokens": 44813591.0,
      "step": 25835
    },
    {
      "entropy": 5.62251648902893,
      "epoch": 2.010425987162031,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004591709770882165,
      "loss": 4.9446,
      "mean_token_accuracy": 0.20036744624376296,
      "num_tokens": 44821722.0,
      "step": 25840
    },
    {
      "entropy": 5.646756935119629,
      "epoch": 2.010815016533748,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004591550543927915,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20686585307121277,
      "num_tokens": 44830022.0,
      "step": 25845
    },
    {
      "entropy": 5.546005964279175,
      "epoch": 2.011204045905466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004591391289031067,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21528953313827515,
      "num_tokens": 44838804.0,
      "step": 25850
    },
    {
      "entropy": 5.571292352676392,
      "epoch": 2.0115930752771836,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004591232006194036,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2109516128897667,
      "num_tokens": 44847585.0,
      "step": 25855
    },
    {
      "entropy": 5.634805011749267,
      "epoch": 2.011982104648901,
      "grad_norm": 1.125,
      "learning_rate": 0.00045910726954192404,
      "loss": 5.0129,
      "mean_token_accuracy": 0.19733098149299622,
      "num_tokens": 44857107.0,
      "step": 25860
    },
    {
      "entropy": 5.624984169006348,
      "epoch": 2.0123711340206185,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004590913356709097,
      "loss": 4.974,
      "mean_token_accuracy": 0.20323953181505203,
      "num_tokens": 44866233.0,
      "step": 25865
    },
    {
      "entropy": 5.627806806564331,
      "epoch": 2.012760163392336,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045907539900660237,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19578027576208115,
      "num_tokens": 44874944.0,
      "step": 25870
    },
    {
      "entropy": 5.6209249019622805,
      "epoch": 2.013149192764054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004590594595492438,
      "loss": 4.8545,
      "mean_token_accuracy": 0.22061439752578735,
      "num_tokens": 44884118.0,
      "step": 25875
    },
    {
      "entropy": 5.5893186092376705,
      "epoch": 2.013538222135771,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045904351729907593,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2095783069729805,
      "num_tokens": 44893101.0,
      "step": 25880
    },
    {
      "entropy": 5.573078918457031,
      "epoch": 2.013927251507489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045902757225634066,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21072138845920563,
      "num_tokens": 44901188.0,
      "step": 25885
    },
    {
      "entropy": 5.577827882766724,
      "epoch": 2.0143162808792066,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004590116244212799,
      "loss": 4.8855,
      "mean_token_accuracy": 0.20304518640041352,
      "num_tokens": 44909667.0,
      "step": 25890
    },
    {
      "entropy": 5.609911394119263,
      "epoch": 2.014705310250924,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004589956737941355,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20266708731651306,
      "num_tokens": 44918112.0,
      "step": 25895
    },
    {
      "entropy": 5.585029077529907,
      "epoch": 2.0150943396226415,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004589797203751497,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2067651018500328,
      "num_tokens": 44927556.0,
      "step": 25900
    },
    {
      "entropy": 5.608692741394043,
      "epoch": 2.015483368994359,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004589637641645645,
      "loss": 4.9532,
      "mean_token_accuracy": 0.20604204386472702,
      "num_tokens": 44936789.0,
      "step": 25905
    },
    {
      "entropy": 5.603762435913086,
      "epoch": 2.0158723983660765,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045894780516262193,
      "loss": 4.9366,
      "mean_token_accuracy": 0.19604863971471786,
      "num_tokens": 44945515.0,
      "step": 25910
    },
    {
      "entropy": 5.578456878662109,
      "epoch": 2.016261427737794,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004589318433695642,
      "loss": 4.956,
      "mean_token_accuracy": 0.20684794187545777,
      "num_tokens": 44953789.0,
      "step": 25915
    },
    {
      "entropy": 5.59338903427124,
      "epoch": 2.016650457109512,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004589158787856336,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21449631303548813,
      "num_tokens": 44961874.0,
      "step": 25920
    },
    {
      "entropy": 5.6510570526123045,
      "epoch": 2.0170394864812295,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004588999114110721,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2030657410621643,
      "num_tokens": 44969955.0,
      "step": 25925
    },
    {
      "entropy": 5.685905981063843,
      "epoch": 2.017428515852947,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004588839412461223,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2103424459695816,
      "num_tokens": 44978514.0,
      "step": 25930
    },
    {
      "entropy": 5.577225112915039,
      "epoch": 2.0178175452246645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004588679682910264,
      "loss": 4.9287,
      "mean_token_accuracy": 0.20726794749498367,
      "num_tokens": 44986650.0,
      "step": 25935
    },
    {
      "entropy": 5.535041856765747,
      "epoch": 2.018206574596382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004588519925460266,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20004806369543077,
      "num_tokens": 44996175.0,
      "step": 25940
    },
    {
      "entropy": 5.625651454925537,
      "epoch": 2.0185956039680995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004588360140113655,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20398198664188386,
      "num_tokens": 45005557.0,
      "step": 25945
    },
    {
      "entropy": 5.625854253768921,
      "epoch": 2.018984633339817,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004588200326872855,
      "loss": 4.8912,
      "mean_token_accuracy": 0.19530292600393295,
      "num_tokens": 45013780.0,
      "step": 25950
    },
    {
      "entropy": 5.51806263923645,
      "epoch": 2.019373662711535,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004588040485740291,
      "loss": 4.9247,
      "mean_token_accuracy": 0.20801512598991395,
      "num_tokens": 45023255.0,
      "step": 25955
    },
    {
      "entropy": 5.647888612747193,
      "epoch": 2.019762692083252,
      "grad_norm": 1.125,
      "learning_rate": 0.0004587880616718387,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1914384514093399,
      "num_tokens": 45032901.0,
      "step": 25960
    },
    {
      "entropy": 5.715079116821289,
      "epoch": 2.02015172145497,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004587720719809572,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21030139923095703,
      "num_tokens": 45040943.0,
      "step": 25965
    },
    {
      "entropy": 5.714957761764526,
      "epoch": 2.0205407508266875,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004587560795016269,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19679745733737947,
      "num_tokens": 45049802.0,
      "step": 25970
    },
    {
      "entropy": 5.595868062973023,
      "epoch": 2.020929780198405,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004587400842340905,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21422239691019057,
      "num_tokens": 45058759.0,
      "step": 25975
    },
    {
      "entropy": 5.667203426361084,
      "epoch": 2.0213188095701224,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045872408617859083,
      "loss": 5.0434,
      "mean_token_accuracy": 0.19054094851016998,
      "num_tokens": 45067403.0,
      "step": 25980
    },
    {
      "entropy": 5.633752489089966,
      "epoch": 2.02170783894184,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045870808533537066,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20937184393405914,
      "num_tokens": 45076157.0,
      "step": 25985
    },
    {
      "entropy": 5.634929227828979,
      "epoch": 2.022096868313558,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045869208170467256,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20683026909828187,
      "num_tokens": 45084253.0,
      "step": 25990
    },
    {
      "entropy": 5.591089916229248,
      "epoch": 2.022485897685275,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004586760752867396,
      "loss": 4.9461,
      "mean_token_accuracy": 0.20789675116539003,
      "num_tokens": 45092609.0,
      "step": 25995
    },
    {
      "entropy": 5.573893022537232,
      "epoch": 2.022874927056993,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045866006608181456,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21681669503450393,
      "num_tokens": 45101122.0,
      "step": 26000
    },
    {
      "entropy": 5.705096435546875,
      "epoch": 2.0232639564287105,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004586440540901403,
      "loss": 5.0559,
      "mean_token_accuracy": 0.19834366142749787,
      "num_tokens": 45110279.0,
      "step": 26005
    },
    {
      "entropy": 5.661575365066528,
      "epoch": 2.0236529858004277,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045862803931195976,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20676905661821365,
      "num_tokens": 45119449.0,
      "step": 26010
    },
    {
      "entropy": 5.656465673446656,
      "epoch": 2.0240420151721454,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004586120217475161,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21528079062700273,
      "num_tokens": 45128336.0,
      "step": 26015
    },
    {
      "entropy": 5.599501466751098,
      "epoch": 2.024431044543863,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004585960013970522,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2026561364531517,
      "num_tokens": 45136577.0,
      "step": 26020
    },
    {
      "entropy": 5.595861148834229,
      "epoch": 2.024820073915581,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004585799782608112,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20043786019086837,
      "num_tokens": 45145058.0,
      "step": 26025
    },
    {
      "entropy": 5.604718017578125,
      "epoch": 2.025209103287298,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045856395233903624,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2063464865088463,
      "num_tokens": 45153791.0,
      "step": 26030
    },
    {
      "entropy": 5.582047748565674,
      "epoch": 2.025598132659016,
      "grad_norm": 1.25,
      "learning_rate": 0.0004585479236319705,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21511571258306503,
      "num_tokens": 45162228.0,
      "step": 26035
    },
    {
      "entropy": 5.581586074829102,
      "epoch": 2.0259871620307335,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045853189213985714,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21080451756715773,
      "num_tokens": 45170601.0,
      "step": 26040
    },
    {
      "entropy": 5.625035142898559,
      "epoch": 2.0263761914024507,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045851585786293943,
      "loss": 5.0527,
      "mean_token_accuracy": 0.1987909957766533,
      "num_tokens": 45179602.0,
      "step": 26045
    },
    {
      "entropy": 5.733458471298218,
      "epoch": 2.0267652207741684,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045849982080146067,
      "loss": 5.0742,
      "mean_token_accuracy": 0.19620684385299683,
      "num_tokens": 45188366.0,
      "step": 26050
    },
    {
      "entropy": 5.641735410690307,
      "epoch": 2.027154250145886,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004584837809556642,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2043984740972519,
      "num_tokens": 45196861.0,
      "step": 26055
    },
    {
      "entropy": 5.6144720077514645,
      "epoch": 2.0275432795176034,
      "grad_norm": 1.125,
      "learning_rate": 0.00045846773832579346,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20786049515008925,
      "num_tokens": 45205176.0,
      "step": 26060
    },
    {
      "entropy": 5.6016723155975345,
      "epoch": 2.027932308889321,
      "grad_norm": 1.25,
      "learning_rate": 0.00045845169291209176,
      "loss": 4.9695,
      "mean_token_accuracy": 0.19751398265361786,
      "num_tokens": 45213876.0,
      "step": 26065
    },
    {
      "entropy": 5.6475670337677,
      "epoch": 2.0283213382610388,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045843564471480263,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21228935569524765,
      "num_tokens": 45223035.0,
      "step": 26070
    },
    {
      "entropy": 5.584300184249878,
      "epoch": 2.0287103676327565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004584195937341695,
      "loss": 4.9347,
      "mean_token_accuracy": 0.20869850367307663,
      "num_tokens": 45232018.0,
      "step": 26075
    },
    {
      "entropy": 5.553749418258667,
      "epoch": 2.0290993970044737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045840353997043606,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21031275093555452,
      "num_tokens": 45240618.0,
      "step": 26080
    },
    {
      "entropy": 5.566223382949829,
      "epoch": 2.0294884263761914,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004583874834238458,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2119940012693405,
      "num_tokens": 45249537.0,
      "step": 26085
    },
    {
      "entropy": 5.5970603942871096,
      "epoch": 2.029877455747909,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004583714240946423,
      "loss": 4.9623,
      "mean_token_accuracy": 0.20785782039165496,
      "num_tokens": 45258114.0,
      "step": 26090
    },
    {
      "entropy": 5.540555715560913,
      "epoch": 2.0302664851196264,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045835536198306936,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21585020720958709,
      "num_tokens": 45266502.0,
      "step": 26095
    },
    {
      "entropy": 5.551183223724365,
      "epoch": 2.030655514491344,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045833929708937067,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2054360881447792,
      "num_tokens": 45275312.0,
      "step": 26100
    },
    {
      "entropy": 5.543352174758911,
      "epoch": 2.0310445438630618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583232294137899,
      "loss": 4.8888,
      "mean_token_accuracy": 0.20975821763277053,
      "num_tokens": 45283588.0,
      "step": 26105
    },
    {
      "entropy": 5.635727405548096,
      "epoch": 2.031433573234779,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004583071589565709,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20056421756744386,
      "num_tokens": 45292601.0,
      "step": 26110
    },
    {
      "entropy": 5.59704475402832,
      "epoch": 2.0318226026064967,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004582910857179576,
      "loss": 4.8714,
      "mean_token_accuracy": 0.21467458605766296,
      "num_tokens": 45300712.0,
      "step": 26115
    },
    {
      "entropy": 5.565234231948852,
      "epoch": 2.0322116319782144,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004582750096981938,
      "loss": 4.9165,
      "mean_token_accuracy": 0.20175865888595582,
      "num_tokens": 45308102.0,
      "step": 26120
    },
    {
      "entropy": 5.558914136886597,
      "epoch": 2.032600661349932,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004582589308975234,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21172445267438889,
      "num_tokens": 45316586.0,
      "step": 26125
    },
    {
      "entropy": 5.576681280136109,
      "epoch": 2.0329896907216494,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045824284931619044,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20456379354000093,
      "num_tokens": 45325926.0,
      "step": 26130
    },
    {
      "entropy": 5.586765909194947,
      "epoch": 2.033378720093367,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045822676495443893,
      "loss": 4.8962,
      "mean_token_accuracy": 0.20778916478157045,
      "num_tokens": 45334514.0,
      "step": 26135
    },
    {
      "entropy": 5.612622261047363,
      "epoch": 2.0337677494650848,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045821067781251284,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2005477160215378,
      "num_tokens": 45342933.0,
      "step": 26140
    },
    {
      "entropy": 5.595442295074463,
      "epoch": 2.034156778836802,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045819458789065633,
      "loss": 4.8961,
      "mean_token_accuracy": 0.204743829369545,
      "num_tokens": 45351747.0,
      "step": 26145
    },
    {
      "entropy": 5.6030536651611325,
      "epoch": 2.0345458082085197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004581784951891135,
      "loss": 4.9726,
      "mean_token_accuracy": 0.20552002340555192,
      "num_tokens": 45360220.0,
      "step": 26150
    },
    {
      "entropy": 5.546155309677124,
      "epoch": 2.0349348375802374,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004581623997081286,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20708168298006058,
      "num_tokens": 45368619.0,
      "step": 26155
    },
    {
      "entropy": 5.61648588180542,
      "epoch": 2.0353238669519547,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004581463014479459,
      "loss": 4.967,
      "mean_token_accuracy": 0.20813992321491243,
      "num_tokens": 45376528.0,
      "step": 26160
    },
    {
      "entropy": 5.616622877120972,
      "epoch": 2.0357128963236724,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004581302004088095,
      "loss": 4.9012,
      "mean_token_accuracy": 0.20849689692258835,
      "num_tokens": 45385391.0,
      "step": 26165
    },
    {
      "entropy": 5.54474835395813,
      "epoch": 2.03610192569539,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004581140965909638,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22134160548448562,
      "num_tokens": 45393029.0,
      "step": 26170
    },
    {
      "entropy": 5.612124729156494,
      "epoch": 2.0364909550671078,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004580979899946531,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2041140988469124,
      "num_tokens": 45401602.0,
      "step": 26175
    },
    {
      "entropy": 5.64312515258789,
      "epoch": 2.036879984438825,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004580818806201219,
      "loss": 5.0203,
      "mean_token_accuracy": 0.19860402941703797,
      "num_tokens": 45410589.0,
      "step": 26180
    },
    {
      "entropy": 5.72422194480896,
      "epoch": 2.0372690138105427,
      "grad_norm": 1.125,
      "learning_rate": 0.00045806576846761453,
      "loss": 5.0854,
      "mean_token_accuracy": 0.20496964901685716,
      "num_tokens": 45419932.0,
      "step": 26185
    },
    {
      "entropy": 5.652226829528809,
      "epoch": 2.0376580431822604,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045804965353737556,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20599112808704376,
      "num_tokens": 45428696.0,
      "step": 26190
    },
    {
      "entropy": 5.696899366378784,
      "epoch": 2.0380470725539777,
      "grad_norm": 1.25,
      "learning_rate": 0.0004580335358296494,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2029459908604622,
      "num_tokens": 45437236.0,
      "step": 26195
    },
    {
      "entropy": 5.692737579345703,
      "epoch": 2.0384361019256954,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045801741534468065,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20233915895223617,
      "num_tokens": 45445989.0,
      "step": 26200
    },
    {
      "entropy": 5.64029541015625,
      "epoch": 2.038825131297413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004580012920827139,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21099424511194229,
      "num_tokens": 45455100.0,
      "step": 26205
    },
    {
      "entropy": 5.606807899475098,
      "epoch": 2.0392141606691303,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004579851660439939,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2062552571296692,
      "num_tokens": 45463302.0,
      "step": 26210
    },
    {
      "entropy": 5.628108501434326,
      "epoch": 2.039603190040848,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045796903722876523,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20065639317035674,
      "num_tokens": 45472181.0,
      "step": 26215
    },
    {
      "entropy": 5.719865608215332,
      "epoch": 2.0399922194125657,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004579529056372726,
      "loss": 5.056,
      "mean_token_accuracy": 0.2014218434691429,
      "num_tokens": 45480878.0,
      "step": 26220
    },
    {
      "entropy": 5.697950172424316,
      "epoch": 2.0403812487842834,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004579367712697609,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19562004059553145,
      "num_tokens": 45489826.0,
      "step": 26225
    },
    {
      "entropy": 5.636931133270264,
      "epoch": 2.0407702781560006,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045792063412647475,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1990480378270149,
      "num_tokens": 45498467.0,
      "step": 26230
    },
    {
      "entropy": 5.584613609313965,
      "epoch": 2.0411593075277183,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045790449420765925,
      "loss": 4.8816,
      "mean_token_accuracy": 0.20481869727373123,
      "num_tokens": 45507221.0,
      "step": 26235
    },
    {
      "entropy": 5.663542079925537,
      "epoch": 2.041548336899436,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045788835151355914,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20287013351917266,
      "num_tokens": 45516291.0,
      "step": 26240
    },
    {
      "entropy": 5.607553195953369,
      "epoch": 2.0419373662711533,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045787220604441933,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21028766483068467,
      "num_tokens": 45524210.0,
      "step": 26245
    },
    {
      "entropy": 5.6377815246582035,
      "epoch": 2.042326395642871,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045785605780048497,
      "loss": 4.977,
      "mean_token_accuracy": 0.2052060216665268,
      "num_tokens": 45533113.0,
      "step": 26250
    },
    {
      "entropy": 5.573250675201416,
      "epoch": 2.0427154250145887,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045783990678200086,
      "loss": 4.8522,
      "mean_token_accuracy": 0.21622478365898132,
      "num_tokens": 45542139.0,
      "step": 26255
    },
    {
      "entropy": 5.680645942687988,
      "epoch": 2.043104454386306,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045782375298921227,
      "loss": 5.0089,
      "mean_token_accuracy": 0.19603093415498735,
      "num_tokens": 45550780.0,
      "step": 26260
    },
    {
      "entropy": 5.668607378005982,
      "epoch": 2.0434934837580236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004578075964223642,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20450927317142487,
      "num_tokens": 45559263.0,
      "step": 26265
    },
    {
      "entropy": 5.636191987991333,
      "epoch": 2.0438825131297413,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004577914370817018,
      "loss": 4.9857,
      "mean_token_accuracy": 0.21452201902866364,
      "num_tokens": 45567346.0,
      "step": 26270
    },
    {
      "entropy": 5.545056867599487,
      "epoch": 2.044271542501459,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045777527496747034,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20835826247930528,
      "num_tokens": 45575580.0,
      "step": 26275
    },
    {
      "entropy": 5.621274709701538,
      "epoch": 2.0446605718731763,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045775911007991493,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20811544209718705,
      "num_tokens": 45583939.0,
      "step": 26280
    },
    {
      "entropy": 5.758362817764282,
      "epoch": 2.045049601244894,
      "grad_norm": 1.125,
      "learning_rate": 0.00045774294241928093,
      "loss": 5.0629,
      "mean_token_accuracy": 0.1999020665884018,
      "num_tokens": 45592232.0,
      "step": 26285
    },
    {
      "entropy": 5.691959619522095,
      "epoch": 2.0454386306166117,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004577267719858137,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20723421573638917,
      "num_tokens": 45600775.0,
      "step": 26290
    },
    {
      "entropy": 5.59557614326477,
      "epoch": 2.045827659988329,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045771059877975853,
      "loss": 4.9438,
      "mean_token_accuracy": 0.20963931530714036,
      "num_tokens": 45610068.0,
      "step": 26295
    },
    {
      "entropy": 5.643482494354248,
      "epoch": 2.0462166893600466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004576944228013608,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2099675253033638,
      "num_tokens": 45618480.0,
      "step": 26300
    },
    {
      "entropy": 5.692978477478027,
      "epoch": 2.0466057187317643,
      "grad_norm": 1.1875,
      "learning_rate": 0.000457678244050866,
      "loss": 5.0633,
      "mean_token_accuracy": 0.19626158922910691,
      "num_tokens": 45627534.0,
      "step": 26305
    },
    {
      "entropy": 5.5980347156524655,
      "epoch": 2.046994748103482,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004576620625285196,
      "loss": 4.9837,
      "mean_token_accuracy": 0.2055788069963455,
      "num_tokens": 45636450.0,
      "step": 26310
    },
    {
      "entropy": 5.646435880661011,
      "epoch": 2.0473837774751993,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045764587823456717,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2107054904103279,
      "num_tokens": 45646000.0,
      "step": 26315
    },
    {
      "entropy": 5.708975839614868,
      "epoch": 2.047772806846917,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045762969116925424,
      "loss": 5.0134,
      "mean_token_accuracy": 0.19655341655015945,
      "num_tokens": 45655211.0,
      "step": 26320
    },
    {
      "entropy": 5.6566015720367435,
      "epoch": 2.0481618362186347,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045761350133282643,
      "loss": 5.065,
      "mean_token_accuracy": 0.2012076810002327,
      "num_tokens": 45664397.0,
      "step": 26325
    },
    {
      "entropy": 5.629683303833008,
      "epoch": 2.048550865590352,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045759730872552937,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20859864354133606,
      "num_tokens": 45672734.0,
      "step": 26330
    },
    {
      "entropy": 5.610647535324096,
      "epoch": 2.0489398949620696,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004575811133476088,
      "loss": 4.9792,
      "mean_token_accuracy": 0.19796843081712723,
      "num_tokens": 45681386.0,
      "step": 26335
    },
    {
      "entropy": 5.688955879211425,
      "epoch": 2.0493289243337873,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045756491519931047,
      "loss": 5.0273,
      "mean_token_accuracy": 0.19835065305233002,
      "num_tokens": 45690726.0,
      "step": 26340
    },
    {
      "entropy": 5.648142910003662,
      "epoch": 2.0497179537055046,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004575487142808801,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2047296866774559,
      "num_tokens": 45699438.0,
      "step": 26345
    },
    {
      "entropy": 5.646206378936768,
      "epoch": 2.0501069830772223,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004575325105925636,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20050771832466124,
      "num_tokens": 45708446.0,
      "step": 26350
    },
    {
      "entropy": 5.722958421707153,
      "epoch": 2.05049601244894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045751630413460665,
      "loss": 5.1405,
      "mean_token_accuracy": 0.1967083990573883,
      "num_tokens": 45717667.0,
      "step": 26355
    },
    {
      "entropy": 5.585675859451294,
      "epoch": 2.0508850418206577,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004575000949072554,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2119559094309807,
      "num_tokens": 45727066.0,
      "step": 26360
    },
    {
      "entropy": 5.663078689575196,
      "epoch": 2.051274071192375,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004574838829107557,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20177053660154343,
      "num_tokens": 45735091.0,
      "step": 26365
    },
    {
      "entropy": 5.610561656951904,
      "epoch": 2.0516631005640926,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004574676681453535,
      "loss": 4.905,
      "mean_token_accuracy": 0.21462469249963761,
      "num_tokens": 45743715.0,
      "step": 26370
    },
    {
      "entropy": 5.604492425918579,
      "epoch": 2.0520521299358103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045745145061129485,
      "loss": 4.9083,
      "mean_token_accuracy": 0.20644875466823578,
      "num_tokens": 45751906.0,
      "step": 26375
    },
    {
      "entropy": 5.589149808883667,
      "epoch": 2.0524411593075276,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045743523030882584,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20341716408729554,
      "num_tokens": 45760627.0,
      "step": 26380
    },
    {
      "entropy": 5.579925012588501,
      "epoch": 2.0528301886792453,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004574190072381926,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2025664135813713,
      "num_tokens": 45769249.0,
      "step": 26385
    },
    {
      "entropy": 5.540699195861817,
      "epoch": 2.053219218050963,
      "grad_norm": 1.25,
      "learning_rate": 0.0004574027813996413,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2093506082892418,
      "num_tokens": 45776914.0,
      "step": 26390
    },
    {
      "entropy": 5.6074995517730715,
      "epoch": 2.05360824742268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004573865527934181,
      "loss": 4.9506,
      "mean_token_accuracy": 0.20277193188667297,
      "num_tokens": 45785978.0,
      "step": 26395
    },
    {
      "entropy": 5.672599411010742,
      "epoch": 2.053997276794398,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004573703214197692,
      "loss": 5.035,
      "mean_token_accuracy": 0.20061203241348266,
      "num_tokens": 45794477.0,
      "step": 26400
    },
    {
      "entropy": 5.732485771179199,
      "epoch": 2.0543863061661156,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045735408727894095,
      "loss": 5.05,
      "mean_token_accuracy": 0.200111548602581,
      "num_tokens": 45803223.0,
      "step": 26405
    },
    {
      "entropy": 5.62385516166687,
      "epoch": 2.0547753355378333,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045733785037117977,
      "loss": 4.93,
      "mean_token_accuracy": 0.20855015516281128,
      "num_tokens": 45812357.0,
      "step": 26410
    },
    {
      "entropy": 5.629200458526611,
      "epoch": 2.0551643649095506,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004573216106967319,
      "loss": 4.9492,
      "mean_token_accuracy": 0.20995305180549623,
      "num_tokens": 45821443.0,
      "step": 26415
    },
    {
      "entropy": 5.696522998809814,
      "epoch": 2.0555533942812683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045730536825584365,
      "loss": 5.0333,
      "mean_token_accuracy": 0.19735227227211,
      "num_tokens": 45829730.0,
      "step": 26420
    },
    {
      "entropy": 5.646669578552246,
      "epoch": 2.055942423652986,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045728912304876176,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20214956104755402,
      "num_tokens": 45839221.0,
      "step": 26425
    },
    {
      "entropy": 5.616559982299805,
      "epoch": 2.056331453024703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004572728750757326,
      "loss": 4.969,
      "mean_token_accuracy": 0.20486962497234346,
      "num_tokens": 45848385.0,
      "step": 26430
    },
    {
      "entropy": 5.6246153831481935,
      "epoch": 2.056720482396421,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004572566243370025,
      "loss": 4.8406,
      "mean_token_accuracy": 0.21495334208011627,
      "num_tokens": 45856443.0,
      "step": 26435
    },
    {
      "entropy": 5.575373697280884,
      "epoch": 2.0571095117681386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004572403708328183,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2129769966006279,
      "num_tokens": 45864797.0,
      "step": 26440
    },
    {
      "entropy": 5.560120296478272,
      "epoch": 2.057498541139856,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004572241145634266,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2054147630929947,
      "num_tokens": 45873266.0,
      "step": 26445
    },
    {
      "entropy": 5.543383550643921,
      "epoch": 2.0578875705115736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000457207855529074,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2077618733048439,
      "num_tokens": 45881496.0,
      "step": 26450
    },
    {
      "entropy": 5.4915107727050785,
      "epoch": 2.0582765998832913,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045719159373000713,
      "loss": 4.8341,
      "mean_token_accuracy": 0.217554672062397,
      "num_tokens": 45889638.0,
      "step": 26455
    },
    {
      "entropy": 5.5725353240966795,
      "epoch": 2.058665629255009,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004571753291664729,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21190561801195146,
      "num_tokens": 45898137.0,
      "step": 26460
    },
    {
      "entropy": 5.649677515029907,
      "epoch": 2.059054658626726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004571590618387179,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2069894477725029,
      "num_tokens": 45906533.0,
      "step": 26465
    },
    {
      "entropy": 5.646552848815918,
      "epoch": 2.059443687998444,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004571427917469892,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2084255263209343,
      "num_tokens": 45915497.0,
      "step": 26470
    },
    {
      "entropy": 5.6400614261627195,
      "epoch": 2.0598327173701616,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045712651889153345,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20079196542501448,
      "num_tokens": 45924883.0,
      "step": 26475
    },
    {
      "entropy": 5.6329563617706295,
      "epoch": 2.060221746741879,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045711024327259764,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20458641052246093,
      "num_tokens": 45933038.0,
      "step": 26480
    },
    {
      "entropy": 5.641636180877685,
      "epoch": 2.0606107761135966,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045709396489042884,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20292631089687346,
      "num_tokens": 45941191.0,
      "step": 26485
    },
    {
      "entropy": 5.622003221511841,
      "epoch": 2.0609998054853143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045707768374527385,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20311365276575089,
      "num_tokens": 45949957.0,
      "step": 26490
    },
    {
      "entropy": 5.630477571487427,
      "epoch": 2.0613888348570315,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004570613998373799,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20093651860952377,
      "num_tokens": 45958104.0,
      "step": 26495
    },
    {
      "entropy": 5.637386274337769,
      "epoch": 2.061777864228749,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045704511316699395,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2030844047665596,
      "num_tokens": 45966651.0,
      "step": 26500
    },
    {
      "entropy": 5.671005725860596,
      "epoch": 2.062166893600467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045702882373436317,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2016158476471901,
      "num_tokens": 45975352.0,
      "step": 26505
    },
    {
      "entropy": 5.560907793045044,
      "epoch": 2.0625559229721846,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004570125315397347,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21164680868387223,
      "num_tokens": 45983844.0,
      "step": 26510
    },
    {
      "entropy": 5.684816455841064,
      "epoch": 2.062944952343902,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004569962365833558,
      "loss": 5.0177,
      "mean_token_accuracy": 0.1996267929673195,
      "num_tokens": 45992871.0,
      "step": 26515
    },
    {
      "entropy": 5.547699213027954,
      "epoch": 2.0633339817156195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045697993886547374,
      "loss": 4.8711,
      "mean_token_accuracy": 0.21368182003498076,
      "num_tokens": 46001984.0,
      "step": 26520
    },
    {
      "entropy": 5.599789476394653,
      "epoch": 2.0637230110873372,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045696363838633566,
      "loss": 4.9499,
      "mean_token_accuracy": 0.20727853626012802,
      "num_tokens": 46010075.0,
      "step": 26525
    },
    {
      "entropy": 5.6519876480102536,
      "epoch": 2.0641120404590545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045694733514618904,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20114164650440217,
      "num_tokens": 46019597.0,
      "step": 26530
    },
    {
      "entropy": 5.653260564804077,
      "epoch": 2.064501069830772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004569310291452812,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20401397049427034,
      "num_tokens": 46027973.0,
      "step": 26535
    },
    {
      "entropy": 5.612961626052856,
      "epoch": 2.06489009920249,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004569147203838595,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20062805265188216,
      "num_tokens": 46037675.0,
      "step": 26540
    },
    {
      "entropy": 5.702121210098267,
      "epoch": 2.065279128574207,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045689840886217157,
      "loss": 5.0487,
      "mean_token_accuracy": 0.19459159225225447,
      "num_tokens": 46046281.0,
      "step": 26545
    },
    {
      "entropy": 5.680599355697632,
      "epoch": 2.065668157945925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045688209458046475,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2129101499915123,
      "num_tokens": 46054920.0,
      "step": 26550
    },
    {
      "entropy": 5.556121110916138,
      "epoch": 2.0660571873176425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045686577753898663,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21250634640455246,
      "num_tokens": 46063023.0,
      "step": 26555
    },
    {
      "entropy": 5.591476535797119,
      "epoch": 2.0664462166893602,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045684945773798483,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2085109308362007,
      "num_tokens": 46071912.0,
      "step": 26560
    },
    {
      "entropy": 5.620500135421753,
      "epoch": 2.0668352460610775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000456833135177707,
      "loss": 4.9545,
      "mean_token_accuracy": 0.20158397555351257,
      "num_tokens": 46080703.0,
      "step": 26565
    },
    {
      "entropy": 5.63456883430481,
      "epoch": 2.067224275432795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004568168098584007,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2025320664048195,
      "num_tokens": 46088761.0,
      "step": 26570
    },
    {
      "entropy": 5.59713191986084,
      "epoch": 2.067613304804513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004568004817803137,
      "loss": 5.019,
      "mean_token_accuracy": 0.2018290713429451,
      "num_tokens": 46097509.0,
      "step": 26575
    },
    {
      "entropy": 5.612489461898804,
      "epoch": 2.06800233417623,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004567841509436937,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20884002447128297,
      "num_tokens": 46105836.0,
      "step": 26580
    },
    {
      "entropy": 5.535917901992798,
      "epoch": 2.068391363547948,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567678173487887,
      "loss": 4.923,
      "mean_token_accuracy": 0.20524317920207977,
      "num_tokens": 46114771.0,
      "step": 26585
    },
    {
      "entropy": 5.651635551452637,
      "epoch": 2.0687803929196655,
      "grad_norm": 1.125,
      "learning_rate": 0.0004567514809958462,
      "loss": 5.0806,
      "mean_token_accuracy": 0.19648711681365966,
      "num_tokens": 46122702.0,
      "step": 26590
    },
    {
      "entropy": 5.632438468933105,
      "epoch": 2.069169422291383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004567351418851144,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20312962532043458,
      "num_tokens": 46131134.0,
      "step": 26595
    },
    {
      "entropy": 5.658031606674195,
      "epoch": 2.0695584516631005,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045671880001684113,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20567927211523057,
      "num_tokens": 46140558.0,
      "step": 26600
    },
    {
      "entropy": 5.64303731918335,
      "epoch": 2.069947481034818,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045670245539127413,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2022929459810257,
      "num_tokens": 46148886.0,
      "step": 26605
    },
    {
      "entropy": 5.622505712509155,
      "epoch": 2.070336510406536,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045668610800866173,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20936911106109618,
      "num_tokens": 46158011.0,
      "step": 26610
    },
    {
      "entropy": 5.476855373382568,
      "epoch": 2.070725539778253,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045666975786925177,
      "loss": 4.7649,
      "mean_token_accuracy": 0.21348540335893632,
      "num_tokens": 46166510.0,
      "step": 26615
    },
    {
      "entropy": 5.5634912014007565,
      "epoch": 2.071114569149971,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004566534049732923,
      "loss": 4.8375,
      "mean_token_accuracy": 0.21540530622005463,
      "num_tokens": 46174857.0,
      "step": 26620
    },
    {
      "entropy": 5.510746145248413,
      "epoch": 2.0715035985216885,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045663704932103166,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2111956536769867,
      "num_tokens": 46183221.0,
      "step": 26625
    },
    {
      "entropy": 5.590742206573486,
      "epoch": 2.071892627893406,
      "grad_norm": 1.125,
      "learning_rate": 0.00045662069091271785,
      "loss": 5.0047,
      "mean_token_accuracy": 0.20554569810628892,
      "num_tokens": 46192941.0,
      "step": 26630
    },
    {
      "entropy": 5.6010912418365475,
      "epoch": 2.0722816572651235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045660432974859924,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2066770911216736,
      "num_tokens": 46201900.0,
      "step": 26635
    },
    {
      "entropy": 5.595024251937867,
      "epoch": 2.072670686636841,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045658796582892383,
      "loss": 4.8666,
      "mean_token_accuracy": 0.21004868894815446,
      "num_tokens": 46209952.0,
      "step": 26640
    },
    {
      "entropy": 5.651596975326538,
      "epoch": 2.073059716008559,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045657159915394013,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20071132481098175,
      "num_tokens": 46218434.0,
      "step": 26645
    },
    {
      "entropy": 5.626325225830078,
      "epoch": 2.073448745380276,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004565552297238964,
      "loss": 4.9646,
      "mean_token_accuracy": 0.20503720939159392,
      "num_tokens": 46226712.0,
      "step": 26650
    },
    {
      "entropy": 5.5914007186889645,
      "epoch": 2.073837774751994,
      "grad_norm": 1.125,
      "learning_rate": 0.000456538857539041,
      "loss": 4.9224,
      "mean_token_accuracy": 0.20591310858726503,
      "num_tokens": 46236006.0,
      "step": 26655
    },
    {
      "entropy": 5.616474294662476,
      "epoch": 2.0742268041237115,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045652248259962254,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2052401304244995,
      "num_tokens": 46244666.0,
      "step": 26660
    },
    {
      "entropy": 5.616417694091797,
      "epoch": 2.0746158334954288,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004565061049058892,
      "loss": 4.9743,
      "mean_token_accuracy": 0.20762315690517424,
      "num_tokens": 46252816.0,
      "step": 26665
    },
    {
      "entropy": 5.612368202209472,
      "epoch": 2.0750048628671465,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045648972445808963,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20806040316820146,
      "num_tokens": 46261912.0,
      "step": 26670
    },
    {
      "entropy": 5.544836044311523,
      "epoch": 2.075393892238864,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045647334125647235,
      "loss": 4.842,
      "mean_token_accuracy": 0.2208395704627037,
      "num_tokens": 46270702.0,
      "step": 26675
    },
    {
      "entropy": 5.585010814666748,
      "epoch": 2.0757829216105814,
      "grad_norm": 1.21875,
      "learning_rate": 0.000456456955301286,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2122930943965912,
      "num_tokens": 46279590.0,
      "step": 26680
    },
    {
      "entropy": 5.589020919799805,
      "epoch": 2.076171950982299,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004564405665927791,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20397457629442214,
      "num_tokens": 46287330.0,
      "step": 26685
    },
    {
      "entropy": 5.633216619491577,
      "epoch": 2.076560980354017,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045642417513120043,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20234276354312897,
      "num_tokens": 46297048.0,
      "step": 26690
    },
    {
      "entropy": 5.5670215606689455,
      "epoch": 2.076950009725734,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045640778091679876,
      "loss": 4.8905,
      "mean_token_accuracy": 0.2078715980052948,
      "num_tokens": 46305554.0,
      "step": 26695
    },
    {
      "entropy": 5.608142423629761,
      "epoch": 2.0773390390974518,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004563913839498226,
      "loss": 4.9968,
      "mean_token_accuracy": 0.1995532065629959,
      "num_tokens": 46315592.0,
      "step": 26700
    },
    {
      "entropy": 5.570044755935669,
      "epoch": 2.0777280684691695,
      "grad_norm": 1.1875,
      "learning_rate": 0.000456374984230521,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21533555686473846,
      "num_tokens": 46323686.0,
      "step": 26705
    },
    {
      "entropy": 5.6297022819519045,
      "epoch": 2.078117097840887,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004563585817591427,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2044871598482132,
      "num_tokens": 46332954.0,
      "step": 26710
    },
    {
      "entropy": 5.5705564498901365,
      "epoch": 2.0785061272126044,
      "grad_norm": 1.25,
      "learning_rate": 0.0004563421765359365,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21645950376987458,
      "num_tokens": 46341126.0,
      "step": 26715
    },
    {
      "entropy": 5.634419870376587,
      "epoch": 2.078895156584322,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045632576856115156,
      "loss": 5.0474,
      "mean_token_accuracy": 0.19495265930891037,
      "num_tokens": 46349981.0,
      "step": 26720
    },
    {
      "entropy": 5.646301364898681,
      "epoch": 2.07928418595604,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045630935783503657,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21006859391927718,
      "num_tokens": 46358238.0,
      "step": 26725
    },
    {
      "entropy": 5.649518871307373,
      "epoch": 2.079673215327757,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004562929443578407,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2064436599612236,
      "num_tokens": 46367283.0,
      "step": 26730
    },
    {
      "entropy": 5.58906307220459,
      "epoch": 2.0800622446994748,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004562765281298129,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2081178531050682,
      "num_tokens": 46375622.0,
      "step": 26735
    },
    {
      "entropy": 5.641617631912231,
      "epoch": 2.0804512740711925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004562601091512024,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21279183477163316,
      "num_tokens": 46384106.0,
      "step": 26740
    },
    {
      "entropy": 5.6219072341918945,
      "epoch": 2.08084030344291,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004562436874222582,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20291046500205995,
      "num_tokens": 46392731.0,
      "step": 26745
    },
    {
      "entropy": 5.602282905578614,
      "epoch": 2.0812293328146274,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045622726294322955,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1947156846523285,
      "num_tokens": 46402474.0,
      "step": 26750
    },
    {
      "entropy": 5.699810266494751,
      "epoch": 2.081618362186345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045621083571436563,
      "loss": 5.0486,
      "mean_token_accuracy": 0.19503010660409928,
      "num_tokens": 46411994.0,
      "step": 26755
    },
    {
      "entropy": 5.666330194473266,
      "epoch": 2.082007391558063,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004561944057359157,
      "loss": 4.9332,
      "mean_token_accuracy": 0.20660813450813292,
      "num_tokens": 46420926.0,
      "step": 26760
    },
    {
      "entropy": 5.609892272949219,
      "epoch": 2.08239642092978,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004561779730081291,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20208125412464142,
      "num_tokens": 46429432.0,
      "step": 26765
    },
    {
      "entropy": 5.512168979644775,
      "epoch": 2.0827854503014978,
      "grad_norm": 1.125,
      "learning_rate": 0.0004561615375312551,
      "loss": 4.8878,
      "mean_token_accuracy": 0.20943790972232817,
      "num_tokens": 46437882.0,
      "step": 26770
    },
    {
      "entropy": 5.683145666122437,
      "epoch": 2.0831744796732155,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045614509930554304,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2025373548269272,
      "num_tokens": 46446856.0,
      "step": 26775
    },
    {
      "entropy": 5.70679874420166,
      "epoch": 2.0835635090449327,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045612865833124253,
      "loss": 5.0373,
      "mean_token_accuracy": 0.19571487307548524,
      "num_tokens": 46456424.0,
      "step": 26780
    },
    {
      "entropy": 5.607332324981689,
      "epoch": 2.0839525384166504,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004561122146086029,
      "loss": 4.9493,
      "mean_token_accuracy": 0.20983870923519135,
      "num_tokens": 46465082.0,
      "step": 26785
    },
    {
      "entropy": 5.630939245223999,
      "epoch": 2.084341567788368,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004560957681378737,
      "loss": 5.0802,
      "mean_token_accuracy": 0.19562057703733443,
      "num_tokens": 46474502.0,
      "step": 26790
    },
    {
      "entropy": 5.642990779876709,
      "epoch": 2.084730597160086,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045607931891930445,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20131320655345916,
      "num_tokens": 46483582.0,
      "step": 26795
    },
    {
      "entropy": 5.626100778579712,
      "epoch": 2.085119626531803,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004560628669531447,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20313573479652405,
      "num_tokens": 46492193.0,
      "step": 26800
    },
    {
      "entropy": 5.681998872756958,
      "epoch": 2.0855086559035207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045604641223964416,
      "loss": 5.024,
      "mean_token_accuracy": 0.20447239577770232,
      "num_tokens": 46499771.0,
      "step": 26805
    },
    {
      "entropy": 5.646428012847901,
      "epoch": 2.0858976852752384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045602995477905247,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20567684173583983,
      "num_tokens": 46508792.0,
      "step": 26810
    },
    {
      "entropy": 5.636980104446411,
      "epoch": 2.0862867146469557,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004560134945716194,
      "loss": 4.989,
      "mean_token_accuracy": 0.2018295094370842,
      "num_tokens": 46516384.0,
      "step": 26815
    },
    {
      "entropy": 5.618140983581543,
      "epoch": 2.0866757440186734,
      "grad_norm": 1.125,
      "learning_rate": 0.00045599703161759464,
      "loss": 5.0216,
      "mean_token_accuracy": 0.19898437708616257,
      "num_tokens": 46525663.0,
      "step": 26820
    },
    {
      "entropy": 5.727896118164063,
      "epoch": 2.087064773390391,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045598056591722805,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2045079290866852,
      "num_tokens": 46534284.0,
      "step": 26825
    },
    {
      "entropy": 5.648638486862183,
      "epoch": 2.0874538027621083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045596409747076936,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21205185800790788,
      "num_tokens": 46542695.0,
      "step": 26830
    },
    {
      "entropy": 5.646912145614624,
      "epoch": 2.087842832133826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004559476262784686,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20022041201591492,
      "num_tokens": 46552098.0,
      "step": 26835
    },
    {
      "entropy": 5.621468210220337,
      "epoch": 2.0882318615055437,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004559311523405755,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21050500273704528,
      "num_tokens": 46560434.0,
      "step": 26840
    },
    {
      "entropy": 5.582331609725952,
      "epoch": 2.0886208908772614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004559146756573402,
      "loss": 4.9235,
      "mean_token_accuracy": 0.205746790766716,
      "num_tokens": 46569015.0,
      "step": 26845
    },
    {
      "entropy": 5.627817106246948,
      "epoch": 2.0890099202489787,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045589819622901274,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2004441276192665,
      "num_tokens": 46578220.0,
      "step": 26850
    },
    {
      "entropy": 5.63268551826477,
      "epoch": 2.0893989496206964,
      "grad_norm": 1.328125,
      "learning_rate": 0.000455881714055843,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21371953934431076,
      "num_tokens": 46587449.0,
      "step": 26855
    },
    {
      "entropy": 5.575710153579712,
      "epoch": 2.089787978992414,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045586522913808114,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2046692430973053,
      "num_tokens": 46596150.0,
      "step": 26860
    },
    {
      "entropy": 5.618714475631714,
      "epoch": 2.0901770083641313,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004558487414759773,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20941583663225175,
      "num_tokens": 46604382.0,
      "step": 26865
    },
    {
      "entropy": 5.6272753238677975,
      "epoch": 2.090566037735849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045583225106978175,
      "loss": 4.9987,
      "mean_token_accuracy": 0.20499735623598098,
      "num_tokens": 46613247.0,
      "step": 26870
    },
    {
      "entropy": 5.554845571517944,
      "epoch": 2.0909550671075667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004558157579197446,
      "loss": 4.9378,
      "mean_token_accuracy": 0.20878576934337617,
      "num_tokens": 46621537.0,
      "step": 26875
    },
    {
      "entropy": 5.635019111633301,
      "epoch": 2.091344096479284,
      "grad_norm": 1.25,
      "learning_rate": 0.00045579926202611603,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2066166251897812,
      "num_tokens": 46629434.0,
      "step": 26880
    },
    {
      "entropy": 5.632823944091797,
      "epoch": 2.0917331258510017,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004557827633891465,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20207550525665283,
      "num_tokens": 46637331.0,
      "step": 26885
    },
    {
      "entropy": 5.540047597885132,
      "epoch": 2.0921221552227194,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004557662620090863,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21149309277534484,
      "num_tokens": 46646176.0,
      "step": 26890
    },
    {
      "entropy": 5.5788984298706055,
      "epoch": 2.092511184594437,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045574975788618564,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21030667573213577,
      "num_tokens": 46653785.0,
      "step": 26895
    },
    {
      "entropy": 5.508225631713867,
      "epoch": 2.0929002139661543,
      "grad_norm": 1.125,
      "learning_rate": 0.0004557332510206953,
      "loss": 4.8816,
      "mean_token_accuracy": 0.21081152707338333,
      "num_tokens": 46662457.0,
      "step": 26900
    },
    {
      "entropy": 5.550087594985962,
      "epoch": 2.093289243337872,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004557167414128654,
      "loss": 4.9469,
      "mean_token_accuracy": 0.20769478529691696,
      "num_tokens": 46670894.0,
      "step": 26905
    },
    {
      "entropy": 5.657681179046631,
      "epoch": 2.0936782727095897,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004557002290629467,
      "loss": 4.9748,
      "mean_token_accuracy": 0.203441846370697,
      "num_tokens": 46679128.0,
      "step": 26910
    },
    {
      "entropy": 5.584599018096924,
      "epoch": 2.094067302081307,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045568371397118954,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2040469229221344,
      "num_tokens": 46688453.0,
      "step": 26915
    },
    {
      "entropy": 5.694149827957153,
      "epoch": 2.0944563314530247,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004556671961378446,
      "loss": 5.0916,
      "mean_token_accuracy": 0.19881702363491058,
      "num_tokens": 46697591.0,
      "step": 26920
    },
    {
      "entropy": 5.623454189300537,
      "epoch": 2.0948453608247424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045565067556316264,
      "loss": 4.8942,
      "mean_token_accuracy": 0.2112753227353096,
      "num_tokens": 46706308.0,
      "step": 26925
    },
    {
      "entropy": 5.610533094406128,
      "epoch": 2.0952343901964596,
      "grad_norm": 1.125,
      "learning_rate": 0.0004556341522473941,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2164320319890976,
      "num_tokens": 46714739.0,
      "step": 26930
    },
    {
      "entropy": 5.651600122451782,
      "epoch": 2.0956234195681773,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004556176261907899,
      "loss": 5.0348,
      "mean_token_accuracy": 0.20308492034673692,
      "num_tokens": 46723586.0,
      "step": 26935
    },
    {
      "entropy": 5.674388742446899,
      "epoch": 2.096012448939895,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004556010973936006,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20166873782873154,
      "num_tokens": 46732519.0,
      "step": 26940
    },
    {
      "entropy": 5.644387435913086,
      "epoch": 2.0964014783116127,
      "grad_norm": 1.125,
      "learning_rate": 0.0004555845658560772,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20382509678602218,
      "num_tokens": 46741002.0,
      "step": 26945
    },
    {
      "entropy": 5.585394954681396,
      "epoch": 2.09679050768333,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045556803157847036,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21128481328487397,
      "num_tokens": 46749466.0,
      "step": 26950
    },
    {
      "entropy": 5.631072950363159,
      "epoch": 2.0971795370550477,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045555149456103113,
      "loss": 4.876,
      "mean_token_accuracy": 0.20800571143627167,
      "num_tokens": 46758981.0,
      "step": 26955
    },
    {
      "entropy": 5.588720035552979,
      "epoch": 2.0975685664267654,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004555349548040102,
      "loss": 4.8884,
      "mean_token_accuracy": 0.20364405065774918,
      "num_tokens": 46767574.0,
      "step": 26960
    },
    {
      "entropy": 5.664254474639892,
      "epoch": 2.0979575957984826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004555184123076589,
      "loss": 4.974,
      "mean_token_accuracy": 0.21018045991659165,
      "num_tokens": 46775804.0,
      "step": 26965
    },
    {
      "entropy": 5.630028343200683,
      "epoch": 2.0983466251702003,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045550186707222785,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21176621615886687,
      "num_tokens": 46784195.0,
      "step": 26970
    },
    {
      "entropy": 5.613404607772827,
      "epoch": 2.098735654541918,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004554853190979683,
      "loss": 4.9424,
      "mean_token_accuracy": 0.20522339642047882,
      "num_tokens": 46793430.0,
      "step": 26975
    },
    {
      "entropy": 5.620491981506348,
      "epoch": 2.0991246839136353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045546876838513134,
      "loss": 4.946,
      "mean_token_accuracy": 0.20572441518306733,
      "num_tokens": 46802111.0,
      "step": 26980
    },
    {
      "entropy": 5.602687644958496,
      "epoch": 2.099513713285353,
      "grad_norm": 1.125,
      "learning_rate": 0.00045545221493396805,
      "loss": 4.9818,
      "mean_token_accuracy": 0.20563649982213975,
      "num_tokens": 46811301.0,
      "step": 26985
    },
    {
      "entropy": 5.620456218719482,
      "epoch": 2.0999027426570707,
      "grad_norm": 1.125,
      "learning_rate": 0.00045543565874472963,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20078134536743164,
      "num_tokens": 46819957.0,
      "step": 26990
    },
    {
      "entropy": 5.613553619384765,
      "epoch": 2.1002917720287884,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045541909981766725,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2068886324763298,
      "num_tokens": 46828493.0,
      "step": 26995
    },
    {
      "entropy": 5.676879501342773,
      "epoch": 2.1006808014005056,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004554025381530322,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20754093378782273,
      "num_tokens": 46837213.0,
      "step": 27000
    },
    {
      "epoch": 2.1006808014005056,
      "eval_entropy": 5.3621790708775015,
      "eval_loss": 5.075078010559082,
      "eval_mean_token_accuracy": 0.2092321919049634,
      "eval_num_tokens": 46837213.0,
      "eval_runtime": 21.6549,
      "eval_samples_per_second": 1919.102,
      "eval_steps_per_second": 239.899,
      "step": 27000
    },
    {
      "entropy": 5.676884889602661,
      "epoch": 2.1010698307722233,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004553859737510758,
      "loss": 5.0935,
      "mean_token_accuracy": 0.20030743926763533,
      "num_tokens": 46845439.0,
      "step": 27005
    },
    {
      "entropy": 5.616102695465088,
      "epoch": 2.101458860143941,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004553694066120493,
      "loss": 4.9857,
      "mean_token_accuracy": 0.19951392263174056,
      "num_tokens": 46853940.0,
      "step": 27010
    },
    {
      "entropy": 5.628016185760498,
      "epoch": 2.1018478895156583,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045535283673620426,
      "loss": 4.913,
      "mean_token_accuracy": 0.20974893271923065,
      "num_tokens": 46862031.0,
      "step": 27015
    },
    {
      "entropy": 5.658685255050659,
      "epoch": 2.102236918887376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045533626412379194,
      "loss": 5.0287,
      "mean_token_accuracy": 0.19469947516918182,
      "num_tokens": 46871290.0,
      "step": 27020
    },
    {
      "entropy": 5.625765323638916,
      "epoch": 2.1026259482590937,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004553196887750638,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2120487928390503,
      "num_tokens": 46880071.0,
      "step": 27025
    },
    {
      "entropy": 5.67730941772461,
      "epoch": 2.103014977630811,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004553031106902714,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2056495279073715,
      "num_tokens": 46888686.0,
      "step": 27030
    },
    {
      "entropy": 5.634192752838135,
      "epoch": 2.1034040070025286,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045528652986966627,
      "loss": 5.1111,
      "mean_token_accuracy": 0.19219001978635789,
      "num_tokens": 46897498.0,
      "step": 27035
    },
    {
      "entropy": 5.578342533111572,
      "epoch": 2.1037930363742463,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045526994631350006,
      "loss": 4.8808,
      "mean_token_accuracy": 0.21497553884983062,
      "num_tokens": 46905668.0,
      "step": 27040
    },
    {
      "entropy": 5.697851610183716,
      "epoch": 2.104182065745964,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004552533600220243,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20828325748443605,
      "num_tokens": 46913934.0,
      "step": 27045
    },
    {
      "entropy": 5.575703144073486,
      "epoch": 2.1045710951176813,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045523677099549073,
      "loss": 4.8646,
      "mean_token_accuracy": 0.21362457871437074,
      "num_tokens": 46922598.0,
      "step": 27050
    },
    {
      "entropy": 5.567189359664917,
      "epoch": 2.104960124489399,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000455220179234151,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20393268913030624,
      "num_tokens": 46931654.0,
      "step": 27055
    },
    {
      "entropy": 5.637973260879517,
      "epoch": 2.1053491538611167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004552035847382569,
      "loss": 4.9711,
      "mean_token_accuracy": 0.211558236181736,
      "num_tokens": 46940799.0,
      "step": 27060
    },
    {
      "entropy": 5.6942976951599125,
      "epoch": 2.105738183232834,
      "grad_norm": 1.25,
      "learning_rate": 0.00045518698750806024,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21091592013835908,
      "num_tokens": 46948646.0,
      "step": 27065
    },
    {
      "entropy": 5.607877779006958,
      "epoch": 2.1061272126045516,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045517038754381287,
      "loss": 4.9354,
      "mean_token_accuracy": 0.20234255194664003,
      "num_tokens": 46957631.0,
      "step": 27070
    },
    {
      "entropy": 5.593134546279908,
      "epoch": 2.1065162419762693,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551537848457666,
      "loss": 5.0024,
      "mean_token_accuracy": 0.19777540862560272,
      "num_tokens": 46965937.0,
      "step": 27075
    },
    {
      "entropy": 5.679655456542969,
      "epoch": 2.106905271347987,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551371794141734,
      "loss": 5.0397,
      "mean_token_accuracy": 0.19979262202978135,
      "num_tokens": 46974439.0,
      "step": 27080
    },
    {
      "entropy": 5.669306755065918,
      "epoch": 2.1072943007197042,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004551205712492852,
      "loss": 5.0403,
      "mean_token_accuracy": 0.20294493138790132,
      "num_tokens": 46982787.0,
      "step": 27085
    },
    {
      "entropy": 5.597129440307617,
      "epoch": 2.107683330091422,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004551039603513541,
      "loss": 4.9593,
      "mean_token_accuracy": 0.206158410012722,
      "num_tokens": 46990904.0,
      "step": 27090
    },
    {
      "entropy": 5.5969585418701175,
      "epoch": 2.1080723594631396,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045508734672063204,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2105418398976326,
      "num_tokens": 47000332.0,
      "step": 27095
    },
    {
      "entropy": 5.603137636184693,
      "epoch": 2.108461388834857,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004550707303573711,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20705873519182205,
      "num_tokens": 47008618.0,
      "step": 27100
    },
    {
      "entropy": 5.615599060058594,
      "epoch": 2.1088504182065746,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045505411126182347,
      "loss": 4.9517,
      "mean_token_accuracy": 0.20357646644115449,
      "num_tokens": 47017022.0,
      "step": 27105
    },
    {
      "entropy": 5.664198160171509,
      "epoch": 2.1092394475782923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550374894342412,
      "loss": 5.0491,
      "mean_token_accuracy": 0.19924866110086442,
      "num_tokens": 47025720.0,
      "step": 27110
    },
    {
      "entropy": 5.724512434005737,
      "epoch": 2.1096284769500095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004550208648748767,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20419651567935942,
      "num_tokens": 47034762.0,
      "step": 27115
    },
    {
      "entropy": 5.635872459411621,
      "epoch": 2.1100175063217272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000455004237583982,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21029621064662934,
      "num_tokens": 47043382.0,
      "step": 27120
    },
    {
      "entropy": 5.638763189315796,
      "epoch": 2.110406535693445,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004549876075618096,
      "loss": 5.014,
      "mean_token_accuracy": 0.2040812373161316,
      "num_tokens": 47052028.0,
      "step": 27125
    },
    {
      "entropy": 5.5754478931427,
      "epoch": 2.110795565065162,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004549709748086117,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21414296478033065,
      "num_tokens": 47060762.0,
      "step": 27130
    },
    {
      "entropy": 5.678235340118408,
      "epoch": 2.11118459443688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045495433932464063,
      "loss": 5.0874,
      "mean_token_accuracy": 0.19485669881105422,
      "num_tokens": 47068547.0,
      "step": 27135
    },
    {
      "entropy": 5.675933408737182,
      "epoch": 2.1115736238085976,
      "grad_norm": 1.203125,
      "learning_rate": 0.000454937701110149,
      "loss": 5.0141,
      "mean_token_accuracy": 0.20014180094003678,
      "num_tokens": 47077083.0,
      "step": 27140
    },
    {
      "entropy": 5.585717058181762,
      "epoch": 2.1119626531803153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045492106016538917,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21279843300580978,
      "num_tokens": 47085337.0,
      "step": 27145
    },
    {
      "entropy": 5.603982400894165,
      "epoch": 2.1123516825520325,
      "grad_norm": 1.25,
      "learning_rate": 0.00045490441649061354,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20592147409915923,
      "num_tokens": 47093931.0,
      "step": 27150
    },
    {
      "entropy": 5.599470376968384,
      "epoch": 2.1127407119237502,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004548877700860747,
      "loss": 4.898,
      "mean_token_accuracy": 0.2149114191532135,
      "num_tokens": 47102572.0,
      "step": 27155
    },
    {
      "entropy": 5.6060785293579105,
      "epoch": 2.113129741295468,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045487112095202544,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2082992136478424,
      "num_tokens": 47111015.0,
      "step": 27160
    },
    {
      "entropy": 5.6501974105834964,
      "epoch": 2.113518770667185,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004548544690887181,
      "loss": 4.9466,
      "mean_token_accuracy": 0.205793097615242,
      "num_tokens": 47120538.0,
      "step": 27165
    },
    {
      "entropy": 5.7110669136047365,
      "epoch": 2.113907800038903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004548378144964054,
      "loss": 4.9574,
      "mean_token_accuracy": 0.20797051042318343,
      "num_tokens": 47129414.0,
      "step": 27170
    },
    {
      "entropy": 5.563403987884522,
      "epoch": 2.1142968294106206,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045482115717534024,
      "loss": 4.9059,
      "mean_token_accuracy": 0.20198312997817994,
      "num_tokens": 47137720.0,
      "step": 27175
    },
    {
      "entropy": 5.681821537017822,
      "epoch": 2.1146858587823383,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004548044971257752,
      "loss": 5.0204,
      "mean_token_accuracy": 0.200797139108181,
      "num_tokens": 47145842.0,
      "step": 27180
    },
    {
      "entropy": 5.636719131469727,
      "epoch": 2.1150748881540555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045478783434796307,
      "loss": 4.9403,
      "mean_token_accuracy": 0.20943872034549713,
      "num_tokens": 47154101.0,
      "step": 27185
    },
    {
      "entropy": 5.557391595840454,
      "epoch": 2.1154639175257732,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045477116884215675,
      "loss": 4.8759,
      "mean_token_accuracy": 0.20752248167991638,
      "num_tokens": 47162497.0,
      "step": 27190
    },
    {
      "entropy": 5.59216570854187,
      "epoch": 2.115852946897491,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000454754500608609,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20809964686632157,
      "num_tokens": 47170825.0,
      "step": 27195
    },
    {
      "entropy": 5.6206718444824215,
      "epoch": 2.116241976269208,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004547378296475729,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20670367926359176,
      "num_tokens": 47178908.0,
      "step": 27200
    },
    {
      "entropy": 5.5832719802856445,
      "epoch": 2.116631005640926,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045472115595930124,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2054883599281311,
      "num_tokens": 47187939.0,
      "step": 27205
    },
    {
      "entropy": 5.635928297042847,
      "epoch": 2.1170200350126436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045470447954404716,
      "loss": 4.9644,
      "mean_token_accuracy": 0.20308625549077988,
      "num_tokens": 47197670.0,
      "step": 27210
    },
    {
      "entropy": 5.682624244689942,
      "epoch": 2.117409064384361,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045468780040206364,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2145197197794914,
      "num_tokens": 47206927.0,
      "step": 27215
    },
    {
      "entropy": 5.675573253631592,
      "epoch": 2.1177980937560785,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004546711185336037,
      "loss": 4.983,
      "mean_token_accuracy": 0.19844935685396195,
      "num_tokens": 47215645.0,
      "step": 27220
    },
    {
      "entropy": 5.620171451568604,
      "epoch": 2.1181871231277962,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045465443393892047,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21145660281181336,
      "num_tokens": 47225184.0,
      "step": 27225
    },
    {
      "entropy": 5.579754495620728,
      "epoch": 2.118576152499514,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004546377466182673,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20013754814863205,
      "num_tokens": 47234701.0,
      "step": 27230
    },
    {
      "entropy": 5.704707050323487,
      "epoch": 2.118965181871231,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004546210565718972,
      "loss": 4.976,
      "mean_token_accuracy": 0.20020591020584105,
      "num_tokens": 47243186.0,
      "step": 27235
    },
    {
      "entropy": 5.673367738723755,
      "epoch": 2.119354211242949,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045460436380006337,
      "loss": 4.9147,
      "mean_token_accuracy": 0.20959510356187822,
      "num_tokens": 47252320.0,
      "step": 27240
    },
    {
      "entropy": 5.630733489990234,
      "epoch": 2.1197432406146666,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004545876683030192,
      "loss": 4.939,
      "mean_token_accuracy": 0.21444210410118103,
      "num_tokens": 47261084.0,
      "step": 27245
    },
    {
      "entropy": 5.645307207107544,
      "epoch": 2.120132269986384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004545709700810181,
      "loss": 5.038,
      "mean_token_accuracy": 0.2060595542192459,
      "num_tokens": 47269819.0,
      "step": 27250
    },
    {
      "entropy": 5.694859218597412,
      "epoch": 2.1205212993581015,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004545542691343133,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20177240669727325,
      "num_tokens": 47278100.0,
      "step": 27255
    },
    {
      "entropy": 5.6194438457489015,
      "epoch": 2.120910328729819,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004545375654631582,
      "loss": 4.9556,
      "mean_token_accuracy": 0.20483158081769942,
      "num_tokens": 47286113.0,
      "step": 27260
    },
    {
      "entropy": 5.601614999771118,
      "epoch": 2.1212993581015365,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045452085906780645,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2053009897470474,
      "num_tokens": 47294445.0,
      "step": 27265
    },
    {
      "entropy": 5.58880581855774,
      "epoch": 2.121688387473254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045450414994851123,
      "loss": 4.9326,
      "mean_token_accuracy": 0.20819024592638016,
      "num_tokens": 47304785.0,
      "step": 27270
    },
    {
      "entropy": 5.598636245727539,
      "epoch": 2.122077416844972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045448743810552636,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2141011968255043,
      "num_tokens": 47313561.0,
      "step": 27275
    },
    {
      "entropy": 5.617367601394653,
      "epoch": 2.1224664462166896,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045447072353910537,
      "loss": 4.9545,
      "mean_token_accuracy": 0.20897779613733292,
      "num_tokens": 47321870.0,
      "step": 27280
    },
    {
      "entropy": 5.618071460723877,
      "epoch": 2.122855475588407,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004544540062495016,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19870543330907822,
      "num_tokens": 47331208.0,
      "step": 27285
    },
    {
      "entropy": 5.581287431716919,
      "epoch": 2.1232445049601245,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045443728623696905,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19968386143445968,
      "num_tokens": 47339680.0,
      "step": 27290
    },
    {
      "entropy": 5.715861701965332,
      "epoch": 2.123633534331842,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045442056350176123,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1966221109032631,
      "num_tokens": 47348912.0,
      "step": 27295
    },
    {
      "entropy": 5.759345293045044,
      "epoch": 2.1240225637035595,
      "grad_norm": 1.15625,
      "learning_rate": 0.000454403838044132,
      "loss": 5.0754,
      "mean_token_accuracy": 0.19589758068323135,
      "num_tokens": 47357990.0,
      "step": 27300
    },
    {
      "entropy": 5.67279896736145,
      "epoch": 2.124411593075277,
      "grad_norm": 1.265625,
      "learning_rate": 0.000454387109864335,
      "loss": 4.9502,
      "mean_token_accuracy": 0.20262227803468705,
      "num_tokens": 47366233.0,
      "step": 27305
    },
    {
      "entropy": 5.653172779083252,
      "epoch": 2.124800622446995,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045437037896262425,
      "loss": 5.0107,
      "mean_token_accuracy": 0.19377756267786025,
      "num_tokens": 47374916.0,
      "step": 27310
    },
    {
      "entropy": 5.645000791549682,
      "epoch": 2.125189651818712,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004543536453392534,
      "loss": 5.0147,
      "mean_token_accuracy": 0.19965385645627975,
      "num_tokens": 47383456.0,
      "step": 27315
    },
    {
      "entropy": 5.6308996200561525,
      "epoch": 2.12557868119043,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045433690899447643,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20442797988653183,
      "num_tokens": 47391633.0,
      "step": 27320
    },
    {
      "entropy": 5.521139812469483,
      "epoch": 2.1259677105621475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045432016992854734,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2119564265012741,
      "num_tokens": 47400155.0,
      "step": 27325
    },
    {
      "entropy": 5.587801933288574,
      "epoch": 2.126356739933865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004543034281417201,
      "loss": 4.9083,
      "mean_token_accuracy": 0.20598985403776168,
      "num_tokens": 47408884.0,
      "step": 27330
    },
    {
      "entropy": 5.652851152420044,
      "epoch": 2.1267457693055825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004542866836342488,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20189899802207947,
      "num_tokens": 47417981.0,
      "step": 27335
    },
    {
      "entropy": 5.643093061447144,
      "epoch": 2.1271347986773,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004542699364063873,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2064901441335678,
      "num_tokens": 47426718.0,
      "step": 27340
    },
    {
      "entropy": 5.708825302124024,
      "epoch": 2.127523828049018,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004542531864583899,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20488932132720947,
      "num_tokens": 47434831.0,
      "step": 27345
    },
    {
      "entropy": 5.650969696044922,
      "epoch": 2.127912857420735,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004542364337905108,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2045737162232399,
      "num_tokens": 47443638.0,
      "step": 27350
    },
    {
      "entropy": 5.610740423202515,
      "epoch": 2.128301886792453,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045421967840300403,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20466667264699936,
      "num_tokens": 47452622.0,
      "step": 27355
    },
    {
      "entropy": 5.625663948059082,
      "epoch": 2.1286909161641705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004542029202961239,
      "loss": 4.9119,
      "mean_token_accuracy": 0.20957971960306168,
      "num_tokens": 47461675.0,
      "step": 27360
    },
    {
      "entropy": 5.6109191417694095,
      "epoch": 2.1290799455358878,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045418615947012464,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1980179712176323,
      "num_tokens": 47470229.0,
      "step": 27365
    },
    {
      "entropy": 5.611227750778198,
      "epoch": 2.1294689749076054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004541693959252607,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20363456010818481,
      "num_tokens": 47479387.0,
      "step": 27370
    },
    {
      "entropy": 5.608771562576294,
      "epoch": 2.129858004279323,
      "grad_norm": 1.125,
      "learning_rate": 0.00045415262966178635,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21077269464731216,
      "num_tokens": 47488674.0,
      "step": 27375
    },
    {
      "entropy": 5.646748161315918,
      "epoch": 2.130247033651041,
      "grad_norm": 1.125,
      "learning_rate": 0.0004541358606799559,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20020621120929719,
      "num_tokens": 47496626.0,
      "step": 27380
    },
    {
      "entropy": 5.606435632705688,
      "epoch": 2.130636063022758,
      "grad_norm": 1.125,
      "learning_rate": 0.000454119088980024,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20396506637334824,
      "num_tokens": 47505735.0,
      "step": 27385
    },
    {
      "entropy": 5.676372766494751,
      "epoch": 2.131025092394476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004541023145622449,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20978566259145737,
      "num_tokens": 47514082.0,
      "step": 27390
    },
    {
      "entropy": 5.633269596099853,
      "epoch": 2.1314141217661935,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045408553742687336,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21442373245954513,
      "num_tokens": 47523383.0,
      "step": 27395
    },
    {
      "entropy": 5.607184934616089,
      "epoch": 2.1318031511379107,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004540687575741638,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21150895655155183,
      "num_tokens": 47532546.0,
      "step": 27400
    },
    {
      "entropy": 5.582590532302857,
      "epoch": 2.1321921805096284,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045405197500437077,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21065830439329147,
      "num_tokens": 47541238.0,
      "step": 27405
    },
    {
      "entropy": 5.61484317779541,
      "epoch": 2.132581209881346,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045403518971774915,
      "loss": 4.9457,
      "mean_token_accuracy": 0.19634849578142166,
      "num_tokens": 47549822.0,
      "step": 27410
    },
    {
      "entropy": 5.657980680465698,
      "epoch": 2.132970239253064,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045401840171455343,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20605187863111496,
      "num_tokens": 47558669.0,
      "step": 27415
    },
    {
      "entropy": 5.673924493789673,
      "epoch": 2.133359268624781,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004540016109950384,
      "loss": 5.0558,
      "mean_token_accuracy": 0.19639279395341874,
      "num_tokens": 47566894.0,
      "step": 27420
    },
    {
      "entropy": 5.6893524646759035,
      "epoch": 2.133748297996499,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004539848175594589,
      "loss": 5.012,
      "mean_token_accuracy": 0.20483334064483644,
      "num_tokens": 47575859.0,
      "step": 27425
    },
    {
      "entropy": 5.654035806655884,
      "epoch": 2.1341373273682165,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004539680214080695,
      "loss": 5.0606,
      "mean_token_accuracy": 0.19311740398406982,
      "num_tokens": 47584408.0,
      "step": 27430
    },
    {
      "entropy": 5.649634456634521,
      "epoch": 2.1345263567399337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045395122254112536,
      "loss": 4.928,
      "mean_token_accuracy": 0.20387642383575438,
      "num_tokens": 47593243.0,
      "step": 27435
    },
    {
      "entropy": 5.7537096500396725,
      "epoch": 2.1349153861116514,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004539344209588812,
      "loss": 5.0471,
      "mean_token_accuracy": 0.19927849918603896,
      "num_tokens": 47602343.0,
      "step": 27440
    },
    {
      "entropy": 5.698724699020386,
      "epoch": 2.135304415483369,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045391761666159204,
      "loss": 5.0859,
      "mean_token_accuracy": 0.197279192507267,
      "num_tokens": 47612617.0,
      "step": 27445
    },
    {
      "entropy": 5.698676013946534,
      "epoch": 2.1356934448550864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045390080964951294,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18966654539108277,
      "num_tokens": 47621764.0,
      "step": 27450
    },
    {
      "entropy": 5.652096700668335,
      "epoch": 2.136082474226804,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045388399992289864,
      "loss": 4.942,
      "mean_token_accuracy": 0.20775261521339417,
      "num_tokens": 47630910.0,
      "step": 27455
    },
    {
      "entropy": 5.697013235092163,
      "epoch": 2.136471503598522,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045386718748200436,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20340099781751633,
      "num_tokens": 47639794.0,
      "step": 27460
    },
    {
      "entropy": 5.653557300567627,
      "epoch": 2.136860532970239,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045385037232708534,
      "loss": 4.985,
      "mean_token_accuracy": 0.19947068840265275,
      "num_tokens": 47648326.0,
      "step": 27465
    },
    {
      "entropy": 5.603629064559937,
      "epoch": 2.1372495623419567,
      "grad_norm": 1.125,
      "learning_rate": 0.0004538335544583964,
      "loss": 4.9755,
      "mean_token_accuracy": 0.20340578854084015,
      "num_tokens": 47657974.0,
      "step": 27470
    },
    {
      "entropy": 5.61540994644165,
      "epoch": 2.1376385917136744,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045381673387619306,
      "loss": 4.8966,
      "mean_token_accuracy": 0.20648950934410096,
      "num_tokens": 47666521.0,
      "step": 27475
    },
    {
      "entropy": 5.6765398502349855,
      "epoch": 2.138027621085392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045379991058073035,
      "loss": 5.004,
      "mean_token_accuracy": 0.20573042929172516,
      "num_tokens": 47675723.0,
      "step": 27480
    },
    {
      "entropy": 5.674753618240357,
      "epoch": 2.1384166504571094,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004537830845722636,
      "loss": 4.9354,
      "mean_token_accuracy": 0.20032002478837968,
      "num_tokens": 47684594.0,
      "step": 27485
    },
    {
      "entropy": 5.680707550048828,
      "epoch": 2.138805679828827,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004537662558510481,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20203856378793716,
      "num_tokens": 47694145.0,
      "step": 27490
    },
    {
      "entropy": 5.683885765075684,
      "epoch": 2.1391947092005448,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045374942441733925,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20480004996061324,
      "num_tokens": 47702448.0,
      "step": 27495
    },
    {
      "entropy": 5.691501665115356,
      "epoch": 2.139583738572262,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004537325902713923,
      "loss": 5.0029,
      "mean_token_accuracy": 0.1962115943431854,
      "num_tokens": 47711624.0,
      "step": 27500
    },
    {
      "entropy": 5.676574325561523,
      "epoch": 2.1399727679439797,
      "grad_norm": 1.125,
      "learning_rate": 0.0004537157534134629,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20001849234104158,
      "num_tokens": 47720374.0,
      "step": 27505
    },
    {
      "entropy": 5.725689840316773,
      "epoch": 2.1403617973156974,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045369891384380623,
      "loss": 5.1007,
      "mean_token_accuracy": 0.19410979747772217,
      "num_tokens": 47729656.0,
      "step": 27510
    },
    {
      "entropy": 5.6624054431915285,
      "epoch": 2.140750826687415,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004536820715626781,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20551721900701522,
      "num_tokens": 47738357.0,
      "step": 27515
    },
    {
      "entropy": 5.674378967285156,
      "epoch": 2.1411398560591324,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045366522657033385,
      "loss": 5.0166,
      "mean_token_accuracy": 0.19888090789318086,
      "num_tokens": 47747110.0,
      "step": 27520
    },
    {
      "entropy": 5.659598875045776,
      "epoch": 2.14152888543085,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045364837886702924,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20608027428388595,
      "num_tokens": 47756571.0,
      "step": 27525
    },
    {
      "entropy": 5.695581436157227,
      "epoch": 2.1419179148025678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004536315284530198,
      "loss": 4.9467,
      "mean_token_accuracy": 0.20389514565467834,
      "num_tokens": 47765327.0,
      "step": 27530
    },
    {
      "entropy": 5.605331373214722,
      "epoch": 2.142306944174285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004536146753285612,
      "loss": 4.8994,
      "mean_token_accuracy": 0.20527543425559996,
      "num_tokens": 47774091.0,
      "step": 27535
    },
    {
      "entropy": 5.632266998291016,
      "epoch": 2.1426959735460027,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004535978194939093,
      "loss": 4.9923,
      "mean_token_accuracy": 0.20189673751592635,
      "num_tokens": 47781908.0,
      "step": 27540
    },
    {
      "entropy": 5.61867356300354,
      "epoch": 2.1430850029177204,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045358096094931966,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21373921930789946,
      "num_tokens": 47791845.0,
      "step": 27545
    },
    {
      "entropy": 5.565458154678344,
      "epoch": 2.1434740322894377,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004535640996950482,
      "loss": 4.889,
      "mean_token_accuracy": 0.20583549290895461,
      "num_tokens": 47800200.0,
      "step": 27550
    },
    {
      "entropy": 5.667959499359131,
      "epoch": 2.1438630616611554,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004535472357313507,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20027893036603928,
      "num_tokens": 47808459.0,
      "step": 27555
    },
    {
      "entropy": 5.67978801727295,
      "epoch": 2.144252091032873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045353036905848316,
      "loss": 4.9283,
      "mean_token_accuracy": 0.20958651155233382,
      "num_tokens": 47816495.0,
      "step": 27560
    },
    {
      "entropy": 5.687069845199585,
      "epoch": 2.1446411204045903,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004535134996767014,
      "loss": 4.9865,
      "mean_token_accuracy": 0.20812389701604844,
      "num_tokens": 47825452.0,
      "step": 27565
    },
    {
      "entropy": 5.651744222640991,
      "epoch": 2.145030149776308,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004534966275862614,
      "loss": 4.9522,
      "mean_token_accuracy": 0.2072868376970291,
      "num_tokens": 47834003.0,
      "step": 27570
    },
    {
      "entropy": 5.625974845886231,
      "epoch": 2.1454191791480257,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004534797527874192,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20492086857557296,
      "num_tokens": 47842499.0,
      "step": 27575
    },
    {
      "entropy": 5.603460788726807,
      "epoch": 2.1458082085197434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045346287528043083,
      "loss": 5.0123,
      "mean_token_accuracy": 0.20081672370433806,
      "num_tokens": 47851835.0,
      "step": 27580
    },
    {
      "entropy": 5.726094198226929,
      "epoch": 2.1461972378914607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045344599506555244,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20083667784929277,
      "num_tokens": 47861070.0,
      "step": 27585
    },
    {
      "entropy": 5.650775194168091,
      "epoch": 2.1465862672631784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045342911214304,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2102648690342903,
      "num_tokens": 47869852.0,
      "step": 27590
    },
    {
      "entropy": 5.643073511123657,
      "epoch": 2.146975296634896,
      "grad_norm": 1.25,
      "learning_rate": 0.0004534122265131498,
      "loss": 4.8924,
      "mean_token_accuracy": 0.20403113067150117,
      "num_tokens": 47878395.0,
      "step": 27595
    },
    {
      "entropy": 5.635064888000488,
      "epoch": 2.1473643260066133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045339533817613807,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21321532279253005,
      "num_tokens": 47887314.0,
      "step": 27600
    },
    {
      "entropy": 5.670640325546264,
      "epoch": 2.147753355378331,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045337844713226106,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1950090542435646,
      "num_tokens": 47896872.0,
      "step": 27605
    },
    {
      "entropy": 5.637800312042236,
      "epoch": 2.1481423847500487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045336155338177497,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21361474096775054,
      "num_tokens": 47905291.0,
      "step": 27610
    },
    {
      "entropy": 5.678752422332764,
      "epoch": 2.1485314141217664,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004533446569249362,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19896254390478135,
      "num_tokens": 47913872.0,
      "step": 27615
    },
    {
      "entropy": 5.617026662826538,
      "epoch": 2.1489204434934837,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045332775776200116,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21164336204528808,
      "num_tokens": 47921066.0,
      "step": 27620
    },
    {
      "entropy": 5.648990964889526,
      "epoch": 2.1493094728652014,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045331085589322626,
      "loss": 5.0484,
      "mean_token_accuracy": 0.2014581799507141,
      "num_tokens": 47930060.0,
      "step": 27625
    },
    {
      "entropy": 5.705538272857666,
      "epoch": 2.149698502236919,
      "grad_norm": 1.25,
      "learning_rate": 0.0004532939513188679,
      "loss": 5.0475,
      "mean_token_accuracy": 0.19529576152563094,
      "num_tokens": 47938299.0,
      "step": 27630
    },
    {
      "entropy": 5.677837085723877,
      "epoch": 2.1500875316086363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004532770440391826,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2057972565293312,
      "num_tokens": 47947552.0,
      "step": 27635
    },
    {
      "entropy": 5.647033452987671,
      "epoch": 2.150476560980354,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004532601340544269,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20253120511770248,
      "num_tokens": 47957064.0,
      "step": 27640
    },
    {
      "entropy": 5.733011865615845,
      "epoch": 2.1508655903520717,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004532432213648574,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19407415390014648,
      "num_tokens": 47966310.0,
      "step": 27645
    },
    {
      "entropy": 5.66233606338501,
      "epoch": 2.151254619723789,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045322630597073067,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2051491215825081,
      "num_tokens": 47975136.0,
      "step": 27650
    },
    {
      "entropy": 5.72068772315979,
      "epoch": 2.1516436490955066,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045320938787230355,
      "loss": 5.062,
      "mean_token_accuracy": 0.19990981072187425,
      "num_tokens": 47984386.0,
      "step": 27655
    },
    {
      "entropy": 5.582844495773315,
      "epoch": 2.1520326784672243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045319246706983257,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20640043914318085,
      "num_tokens": 47993297.0,
      "step": 27660
    },
    {
      "entropy": 5.683313035964966,
      "epoch": 2.152421707838942,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045317554356357446,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21300583928823472,
      "num_tokens": 48001489.0,
      "step": 27665
    },
    {
      "entropy": 5.68531289100647,
      "epoch": 2.1528107372106593,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045315861735378615,
      "loss": 5.0068,
      "mean_token_accuracy": 0.20370992869138718,
      "num_tokens": 48010041.0,
      "step": 27670
    },
    {
      "entropy": 5.686743545532226,
      "epoch": 2.153199766582377,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045314168844072435,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20503681749105454,
      "num_tokens": 48018551.0,
      "step": 27675
    },
    {
      "entropy": 5.693939065933227,
      "epoch": 2.1535887959540947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045312475682464604,
      "loss": 5.0623,
      "mean_token_accuracy": 0.19794443100690842,
      "num_tokens": 48026471.0,
      "step": 27680
    },
    {
      "entropy": 5.628991746902466,
      "epoch": 2.153977825325812,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045310782250580794,
      "loss": 5.0334,
      "mean_token_accuracy": 0.2054133117198944,
      "num_tokens": 48036241.0,
      "step": 27685
    },
    {
      "entropy": 5.57493486404419,
      "epoch": 2.1543668546975296,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004530908854844672,
      "loss": 4.8772,
      "mean_token_accuracy": 0.20684104561805725,
      "num_tokens": 48044730.0,
      "step": 27690
    },
    {
      "entropy": 5.68789005279541,
      "epoch": 2.1547558840692473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045307394576088076,
      "loss": 5.0031,
      "mean_token_accuracy": 0.19947132617235183,
      "num_tokens": 48052784.0,
      "step": 27695
    },
    {
      "entropy": 5.719441509246826,
      "epoch": 2.1551449134409646,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004530570033353056,
      "loss": 5.028,
      "mean_token_accuracy": 0.20202393531799318,
      "num_tokens": 48061741.0,
      "step": 27700
    },
    {
      "entropy": 5.687761831283569,
      "epoch": 2.1555339428126823,
      "grad_norm": 1.25,
      "learning_rate": 0.00045304005820799874,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2039281979203224,
      "num_tokens": 48070704.0,
      "step": 27705
    },
    {
      "entropy": 5.571134233474732,
      "epoch": 2.1559229721844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004530231103792175,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2129618600010872,
      "num_tokens": 48079475.0,
      "step": 27710
    },
    {
      "entropy": 5.600165891647339,
      "epoch": 2.1563120015561177,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004530061598492188,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2139853447675705,
      "num_tokens": 48088574.0,
      "step": 27715
    },
    {
      "entropy": 5.670781517028809,
      "epoch": 2.156701030927835,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045298920661826004,
      "loss": 4.952,
      "mean_token_accuracy": 0.20930106341838836,
      "num_tokens": 48097203.0,
      "step": 27720
    },
    {
      "entropy": 5.6631760597229,
      "epoch": 2.1570900602995526,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004529722506865984,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20853367447853088,
      "num_tokens": 48105950.0,
      "step": 27725
    },
    {
      "entropy": 5.613159990310669,
      "epoch": 2.1574790896712703,
      "grad_norm": 1.140625,
      "learning_rate": 0.000452955292054491,
      "loss": 4.9915,
      "mean_token_accuracy": 0.1985842227935791,
      "num_tokens": 48115034.0,
      "step": 27730
    },
    {
      "entropy": 5.6286862850189205,
      "epoch": 2.1578681190429876,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045293833072219535,
      "loss": 4.989,
      "mean_token_accuracy": 0.20423654168844224,
      "num_tokens": 48124317.0,
      "step": 27735
    },
    {
      "entropy": 5.70945897102356,
      "epoch": 2.1582571484147053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045292136668996876,
      "loss": 4.9944,
      "mean_token_accuracy": 0.202090260386467,
      "num_tokens": 48133075.0,
      "step": 27740
    },
    {
      "entropy": 5.650589370727539,
      "epoch": 2.158646177786423,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004529043999580686,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2035435453057289,
      "num_tokens": 48142565.0,
      "step": 27745
    },
    {
      "entropy": 5.662950706481934,
      "epoch": 2.1590352071581402,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045288743052675234,
      "loss": 4.9443,
      "mean_token_accuracy": 0.20915111899375916,
      "num_tokens": 48150560.0,
      "step": 27750
    },
    {
      "entropy": 5.604933214187622,
      "epoch": 2.159424236529858,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045287045839627744,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2056383028626442,
      "num_tokens": 48159512.0,
      "step": 27755
    },
    {
      "entropy": 5.677662992477417,
      "epoch": 2.1598132659015756,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045285348356690155,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2051140174269676,
      "num_tokens": 48167774.0,
      "step": 27760
    },
    {
      "entropy": 5.641291570663452,
      "epoch": 2.1602022952732933,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000452836506038882,
      "loss": 5.024,
      "mean_token_accuracy": 0.2028786465525627,
      "num_tokens": 48176441.0,
      "step": 27765
    },
    {
      "entropy": 5.706268978118897,
      "epoch": 2.1605913246450106,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045281952581247654,
      "loss": 5.008,
      "mean_token_accuracy": 0.21060073673725127,
      "num_tokens": 48184515.0,
      "step": 27770
    },
    {
      "entropy": 5.658855247497558,
      "epoch": 2.1609803540167283,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045280254288794286,
      "loss": 4.9779,
      "mean_token_accuracy": 0.19504284709692002,
      "num_tokens": 48192825.0,
      "step": 27775
    },
    {
      "entropy": 5.5917449474334715,
      "epoch": 2.161369383388446,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045278555726553855,
      "loss": 4.9265,
      "mean_token_accuracy": 0.20929719060659407,
      "num_tokens": 48200665.0,
      "step": 27780
    },
    {
      "entropy": 5.64736852645874,
      "epoch": 2.1617584127601632,
      "grad_norm": 1.25,
      "learning_rate": 0.00045276856894552143,
      "loss": 5.011,
      "mean_token_accuracy": 0.21266822069883345,
      "num_tokens": 48209097.0,
      "step": 27785
    },
    {
      "entropy": 5.58549542427063,
      "epoch": 2.162147442131881,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004527515779281492,
      "loss": 4.8222,
      "mean_token_accuracy": 0.21826852411031722,
      "num_tokens": 48217699.0,
      "step": 27790
    },
    {
      "entropy": 5.604995489120483,
      "epoch": 2.1625364715035986,
      "grad_norm": 1.25,
      "learning_rate": 0.00045273458421367976,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2111889749765396,
      "num_tokens": 48225461.0,
      "step": 27795
    },
    {
      "entropy": 5.638203573226929,
      "epoch": 2.162925500875316,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004527175878023708,
      "loss": 5.0479,
      "mean_token_accuracy": 0.1999452829360962,
      "num_tokens": 48234371.0,
      "step": 27800
    },
    {
      "entropy": 5.65357723236084,
      "epoch": 2.1633145302470336,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004527005886944803,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2018502250313759,
      "num_tokens": 48243095.0,
      "step": 27805
    },
    {
      "entropy": 5.648229312896729,
      "epoch": 2.1637035596187513,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045268358689026626,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20550282001495362,
      "num_tokens": 48251293.0,
      "step": 27810
    },
    {
      "entropy": 5.516548347473145,
      "epoch": 2.164092588990469,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004526665823899866,
      "loss": 4.8364,
      "mean_token_accuracy": 0.2094391867518425,
      "num_tokens": 48260021.0,
      "step": 27815
    },
    {
      "entropy": 5.563632202148438,
      "epoch": 2.164481618362186,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045264957519389936,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21243079453706742,
      "num_tokens": 48268370.0,
      "step": 27820
    },
    {
      "entropy": 5.70854868888855,
      "epoch": 2.164870647733904,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045263256530226253,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20449393689632417,
      "num_tokens": 48277623.0,
      "step": 27825
    },
    {
      "entropy": 5.73487434387207,
      "epoch": 2.1652596771056216,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004526155527153343,
      "loss": 5.05,
      "mean_token_accuracy": 0.1914417713880539,
      "num_tokens": 48286555.0,
      "step": 27830
    },
    {
      "entropy": 5.641337156295776,
      "epoch": 2.165648706477339,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004525985374333727,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2129783645272255,
      "num_tokens": 48295312.0,
      "step": 27835
    },
    {
      "entropy": 5.681504487991333,
      "epoch": 2.1660377358490566,
      "grad_norm": 1.296875,
      "learning_rate": 0.000452581519456636,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2011671707034111,
      "num_tokens": 48304061.0,
      "step": 27840
    },
    {
      "entropy": 5.6578826904296875,
      "epoch": 2.1664267652207743,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045256449878538243,
      "loss": 4.956,
      "mean_token_accuracy": 0.20110211074352263,
      "num_tokens": 48311935.0,
      "step": 27845
    },
    {
      "entropy": 5.636744213104248,
      "epoch": 2.166815794592492,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045254747541987017,
      "loss": 5.0488,
      "mean_token_accuracy": 0.20137927830219268,
      "num_tokens": 48320298.0,
      "step": 27850
    },
    {
      "entropy": 5.590518283843994,
      "epoch": 2.167204823964209,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004525304493603576,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21312302649021148,
      "num_tokens": 48328729.0,
      "step": 27855
    },
    {
      "entropy": 5.659118127822876,
      "epoch": 2.167593853335927,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045251342060710295,
      "loss": 5.0389,
      "mean_token_accuracy": 0.2011456772685051,
      "num_tokens": 48336935.0,
      "step": 27860
    },
    {
      "entropy": 5.624445629119873,
      "epoch": 2.1679828827076446,
      "grad_norm": 1.25,
      "learning_rate": 0.0004524963891603647,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21389816999435424,
      "num_tokens": 48345553.0,
      "step": 27865
    },
    {
      "entropy": 5.568893194198608,
      "epoch": 2.168371912079362,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045247935502040136,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2226459190249443,
      "num_tokens": 48354209.0,
      "step": 27870
    },
    {
      "entropy": 5.60405912399292,
      "epoch": 2.1687609414510796,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004524623181874712,
      "loss": 4.99,
      "mean_token_accuracy": 0.20411522090435028,
      "num_tokens": 48362612.0,
      "step": 27875
    },
    {
      "entropy": 5.66093339920044,
      "epoch": 2.1691499708227973,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004524452786618329,
      "loss": 4.9528,
      "mean_token_accuracy": 0.20321442186832428,
      "num_tokens": 48371089.0,
      "step": 27880
    },
    {
      "entropy": 5.5390753746032715,
      "epoch": 2.1695390001945145,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045242823644374484,
      "loss": 4.8443,
      "mean_token_accuracy": 0.21588192582130433,
      "num_tokens": 48379883.0,
      "step": 27885
    },
    {
      "entropy": 5.585571765899658,
      "epoch": 2.169928029566232,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004524111915334658,
      "loss": 4.9858,
      "mean_token_accuracy": 0.201564721763134,
      "num_tokens": 48388742.0,
      "step": 27890
    },
    {
      "entropy": 5.6676877498626705,
      "epoch": 2.17031705893795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045239414393125424,
      "loss": 5.0127,
      "mean_token_accuracy": 0.19953570067882537,
      "num_tokens": 48398563.0,
      "step": 27895
    },
    {
      "entropy": 5.671126985549927,
      "epoch": 2.170706088309667,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004523770936373689,
      "loss": 5.0498,
      "mean_token_accuracy": 0.19919685870409012,
      "num_tokens": 48407558.0,
      "step": 27900
    },
    {
      "entropy": 5.745979928970337,
      "epoch": 2.171095117681385,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004523600406520685,
      "loss": 5.033,
      "mean_token_accuracy": 0.20019682198762895,
      "num_tokens": 48416139.0,
      "step": 27905
    },
    {
      "entropy": 5.660734415054321,
      "epoch": 2.1714841470531026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004523429849756118,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20504415333271026,
      "num_tokens": 48425096.0,
      "step": 27910
    },
    {
      "entropy": 5.5979372501373295,
      "epoch": 2.1718731764248203,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004523259266082576,
      "loss": 4.919,
      "mean_token_accuracy": 0.19872606098651885,
      "num_tokens": 48433890.0,
      "step": 27915
    },
    {
      "entropy": 5.577544355392456,
      "epoch": 2.1722622057965375,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045230886555026455,
      "loss": 4.9358,
      "mean_token_accuracy": 0.20766192078590393,
      "num_tokens": 48442456.0,
      "step": 27920
    },
    {
      "entropy": 5.665525913238525,
      "epoch": 2.172651235168255,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045229180180189176,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2067467004060745,
      "num_tokens": 48450374.0,
      "step": 27925
    },
    {
      "entropy": 5.733412551879883,
      "epoch": 2.173040264539973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045227473536339817,
      "loss": 5.0713,
      "mean_token_accuracy": 0.19755268841981888,
      "num_tokens": 48459026.0,
      "step": 27930
    },
    {
      "entropy": 5.645839405059815,
      "epoch": 2.17342929391169,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004522576662350425,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20401676297187804,
      "num_tokens": 48467953.0,
      "step": 27935
    },
    {
      "entropy": 5.724455976486206,
      "epoch": 2.173818323283408,
      "grad_norm": 1.234375,
      "learning_rate": 0.000452240594417084,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20372615605592728,
      "num_tokens": 48476460.0,
      "step": 27940
    },
    {
      "entropy": 5.694390916824341,
      "epoch": 2.1742073526551255,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004522235199097815,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2027232229709625,
      "num_tokens": 48484945.0,
      "step": 27945
    },
    {
      "entropy": 5.652435398101806,
      "epoch": 2.1745963820268432,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004522064427133942,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2093926340341568,
      "num_tokens": 48492891.0,
      "step": 27950
    },
    {
      "entropy": 5.602068710327148,
      "epoch": 2.1749854113985605,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045218936282818116,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2158997103571892,
      "num_tokens": 48501370.0,
      "step": 27955
    },
    {
      "entropy": 5.569188213348388,
      "epoch": 2.175374440770278,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004521722802544016,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20702164024114608,
      "num_tokens": 48510518.0,
      "step": 27960
    },
    {
      "entropy": 5.760787773132324,
      "epoch": 2.175763470141996,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045215519499231465,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19750125110149383,
      "num_tokens": 48520173.0,
      "step": 27965
    },
    {
      "entropy": 5.744514274597168,
      "epoch": 2.176152499513713,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045213810704217963,
      "loss": 5.056,
      "mean_token_accuracy": 0.19875043034553527,
      "num_tokens": 48529486.0,
      "step": 27970
    },
    {
      "entropy": 5.649371767044068,
      "epoch": 2.176541528885431,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004521210164042557,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20206229388713837,
      "num_tokens": 48538725.0,
      "step": 27975
    },
    {
      "entropy": 5.618534326553345,
      "epoch": 2.1769305582571485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004521039230788024,
      "loss": 5.0204,
      "mean_token_accuracy": 0.19740262925624846,
      "num_tokens": 48547578.0,
      "step": 27980
    },
    {
      "entropy": 5.572231674194336,
      "epoch": 2.177319587628866,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004520868270660789,
      "loss": 4.8956,
      "mean_token_accuracy": 0.209757524728775,
      "num_tokens": 48556011.0,
      "step": 27985
    },
    {
      "entropy": 5.622557783126831,
      "epoch": 2.1777086170005835,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004520697283663446,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20390485525131224,
      "num_tokens": 48564844.0,
      "step": 27990
    },
    {
      "entropy": 5.555831050872802,
      "epoch": 2.178097646372301,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045205262697985897,
      "loss": 4.9268,
      "mean_token_accuracy": 0.20634517520666124,
      "num_tokens": 48574079.0,
      "step": 27995
    },
    {
      "entropy": 5.652505159378052,
      "epoch": 2.1784866757440184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004520355229068816,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20242383182048798,
      "num_tokens": 48584357.0,
      "step": 28000
    },
    {
      "entropy": 5.568860960006714,
      "epoch": 2.178875705115736,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045201841614767203,
      "loss": 4.9607,
      "mean_token_accuracy": 0.19928446263074875,
      "num_tokens": 48593001.0,
      "step": 28005
    },
    {
      "entropy": 5.630989360809326,
      "epoch": 2.179264734487454,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045200130670248974,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2104955717921257,
      "num_tokens": 48601883.0,
      "step": 28010
    },
    {
      "entropy": 5.573348617553711,
      "epoch": 2.1796537638591715,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045198419457159424,
      "loss": 4.9358,
      "mean_token_accuracy": 0.211085906624794,
      "num_tokens": 48610765.0,
      "step": 28015
    },
    {
      "entropy": 5.627802467346191,
      "epoch": 2.180042793230889,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004519670797552453,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2113234966993332,
      "num_tokens": 48618950.0,
      "step": 28020
    },
    {
      "entropy": 5.713563108444214,
      "epoch": 2.1804318226026065,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045194996225370266,
      "loss": 5.0978,
      "mean_token_accuracy": 0.19300964176654817,
      "num_tokens": 48627654.0,
      "step": 28025
    },
    {
      "entropy": 5.695786428451538,
      "epoch": 2.180820851974324,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045193284206722595,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20657435953617095,
      "num_tokens": 48636295.0,
      "step": 28030
    },
    {
      "entropy": 5.673643779754639,
      "epoch": 2.1812098813460414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004519157191960749,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21093812137842177,
      "num_tokens": 48644543.0,
      "step": 28035
    },
    {
      "entropy": 5.652819108963013,
      "epoch": 2.181598910717759,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004518985936405096,
      "loss": 4.9467,
      "mean_token_accuracy": 0.20411646366119385,
      "num_tokens": 48653873.0,
      "step": 28040
    },
    {
      "entropy": 5.581996870040894,
      "epoch": 2.181987940089477,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004518814654007896,
      "loss": 4.867,
      "mean_token_accuracy": 0.21281438171863556,
      "num_tokens": 48663083.0,
      "step": 28045
    },
    {
      "entropy": 5.672663784027099,
      "epoch": 2.1823769694611945,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045186433447717486,
      "loss": 5.0178,
      "mean_token_accuracy": 0.19615076333284379,
      "num_tokens": 48671093.0,
      "step": 28050
    },
    {
      "entropy": 5.551146984100342,
      "epoch": 2.182765998832912,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045184720086992536,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2087004467844963,
      "num_tokens": 48680410.0,
      "step": 28055
    },
    {
      "entropy": 5.637509107589722,
      "epoch": 2.1831550282046295,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045183006457930107,
      "loss": 4.9366,
      "mean_token_accuracy": 0.19815604984760285,
      "num_tokens": 48688928.0,
      "step": 28060
    },
    {
      "entropy": 5.6413867473602295,
      "epoch": 2.183544057576347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004518129256055621,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20229238867759705,
      "num_tokens": 48697462.0,
      "step": 28065
    },
    {
      "entropy": 5.636691904067993,
      "epoch": 2.1839330869480644,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004517957839489683,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21011649072170258,
      "num_tokens": 48706234.0,
      "step": 28070
    },
    {
      "entropy": 5.6010349750518795,
      "epoch": 2.184322116319782,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045177863960977986,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2107345014810562,
      "num_tokens": 48714694.0,
      "step": 28075
    },
    {
      "entropy": 5.688062858581543,
      "epoch": 2.1847111456915,
      "grad_norm": 1.1875,
      "learning_rate": 0.000451761492588257,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20155843943357468,
      "num_tokens": 48722762.0,
      "step": 28080
    },
    {
      "entropy": 5.690778017044067,
      "epoch": 2.185100175063217,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045174434288465986,
      "loss": 5.0323,
      "mean_token_accuracy": 0.19827434569597244,
      "num_tokens": 48731986.0,
      "step": 28085
    },
    {
      "entropy": 5.666117525100708,
      "epoch": 2.1854892044349348,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004517271904992486,
      "loss": 5.0166,
      "mean_token_accuracy": 0.204315946996212,
      "num_tokens": 48741740.0,
      "step": 28090
    },
    {
      "entropy": 5.6096426486969,
      "epoch": 2.1858782338066525,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004517100354322835,
      "loss": 4.939,
      "mean_token_accuracy": 0.20826371759176254,
      "num_tokens": 48751043.0,
      "step": 28095
    },
    {
      "entropy": 5.6830010414123535,
      "epoch": 2.18626726317837,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045169287768402497,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20770698934793472,
      "num_tokens": 48760004.0,
      "step": 28100
    },
    {
      "entropy": 5.671423816680909,
      "epoch": 2.1866562925500874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004516757172547331,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20075724720954896,
      "num_tokens": 48768829.0,
      "step": 28105
    },
    {
      "entropy": 5.726738739013672,
      "epoch": 2.187045321921805,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004516585541446685,
      "loss": 5.0822,
      "mean_token_accuracy": 0.19186793863773347,
      "num_tokens": 48777508.0,
      "step": 28110
    },
    {
      "entropy": 5.729541301727295,
      "epoch": 2.187434351293523,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045164138835409153,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2060468852519989,
      "num_tokens": 48785809.0,
      "step": 28115
    },
    {
      "entropy": 5.685092926025391,
      "epoch": 2.18782338066524,
      "grad_norm": 1.25,
      "learning_rate": 0.00045162421988326266,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20507350713014602,
      "num_tokens": 48795078.0,
      "step": 28120
    },
    {
      "entropy": 5.60854206085205,
      "epoch": 2.1882124100369578,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004516070487324424,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2087715446949005,
      "num_tokens": 48803699.0,
      "step": 28125
    },
    {
      "entropy": 5.646251440048218,
      "epoch": 2.1886014394086755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045158987490189124,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20338423550128937,
      "num_tokens": 48811738.0,
      "step": 28130
    },
    {
      "entropy": 5.644844913482666,
      "epoch": 2.1889904687803927,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004515726983918697,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2185904413461685,
      "num_tokens": 48819386.0,
      "step": 28135
    },
    {
      "entropy": 5.675701761245728,
      "epoch": 2.1893794981521104,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045155551920263864,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20035598874092103,
      "num_tokens": 48827862.0,
      "step": 28140
    },
    {
      "entropy": 5.597946262359619,
      "epoch": 2.189768527523828,
      "grad_norm": 1.125,
      "learning_rate": 0.00045153833733445857,
      "loss": 4.9175,
      "mean_token_accuracy": 0.20790434926748275,
      "num_tokens": 48836423.0,
      "step": 28145
    },
    {
      "entropy": 5.646891975402832,
      "epoch": 2.190157556895546,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004515211527875903,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20711836367845535,
      "num_tokens": 48844572.0,
      "step": 28150
    },
    {
      "entropy": 5.658052921295166,
      "epoch": 2.190546586267263,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004515039655622943,
      "loss": 5.0197,
      "mean_token_accuracy": 0.1975495085120201,
      "num_tokens": 48853632.0,
      "step": 28155
    },
    {
      "entropy": 5.574565029144287,
      "epoch": 2.1909356156389808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045148677565883167,
      "loss": 4.8674,
      "mean_token_accuracy": 0.21655509024858474,
      "num_tokens": 48862306.0,
      "step": 28160
    },
    {
      "entropy": 5.64834508895874,
      "epoch": 2.1913246450106985,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004514695830774631,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20402767062187194,
      "num_tokens": 48870749.0,
      "step": 28165
    },
    {
      "entropy": 5.657930946350097,
      "epoch": 2.1917136743824157,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045145238781844953,
      "loss": 4.968,
      "mean_token_accuracy": 0.20569077581167222,
      "num_tokens": 48879622.0,
      "step": 28170
    },
    {
      "entropy": 5.56830062866211,
      "epoch": 2.1921027037541334,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045143518988205184,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20317641645669937,
      "num_tokens": 48889766.0,
      "step": 28175
    },
    {
      "entropy": 5.662889671325684,
      "epoch": 2.192491733125851,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045141798926853094,
      "loss": 5.0422,
      "mean_token_accuracy": 0.19983337670564652,
      "num_tokens": 48898791.0,
      "step": 28180
    },
    {
      "entropy": 5.652655935287475,
      "epoch": 2.1928807624975684,
      "grad_norm": 1.125,
      "learning_rate": 0.0004514007859781478,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20582958906888962,
      "num_tokens": 48907418.0,
      "step": 28185
    },
    {
      "entropy": 5.595499563217163,
      "epoch": 2.193269791869286,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004513835800111635,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2061915174126625,
      "num_tokens": 48915510.0,
      "step": 28190
    },
    {
      "entropy": 5.5971923828125,
      "epoch": 2.1936588212410038,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004513663713678392,
      "loss": 4.9811,
      "mean_token_accuracy": 0.19574748128652572,
      "num_tokens": 48924067.0,
      "step": 28195
    },
    {
      "entropy": 5.6234969139099125,
      "epoch": 2.1940478506127215,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045134916004843596,
      "loss": 4.9186,
      "mean_token_accuracy": 0.20731496661901475,
      "num_tokens": 48932587.0,
      "step": 28200
    },
    {
      "entropy": 5.6360002040863035,
      "epoch": 2.1944368799844387,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004513319460532148,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21174990385770798,
      "num_tokens": 48941237.0,
      "step": 28205
    },
    {
      "entropy": 5.656458759307862,
      "epoch": 2.1948259093561564,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045131472938243707,
      "loss": 5.0313,
      "mean_token_accuracy": 0.2062858909368515,
      "num_tokens": 48949982.0,
      "step": 28210
    },
    {
      "entropy": 5.613702964782715,
      "epoch": 2.195214938727874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004512975100363639,
      "loss": 5.0148,
      "mean_token_accuracy": 0.19961462020874024,
      "num_tokens": 48959039.0,
      "step": 28215
    },
    {
      "entropy": 5.686495351791382,
      "epoch": 2.1956039680995914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045128028801525675,
      "loss": 4.9109,
      "mean_token_accuracy": 0.21068979650735856,
      "num_tokens": 48968743.0,
      "step": 28220
    },
    {
      "entropy": 5.741949653625488,
      "epoch": 2.195992997471309,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045126306331937674,
      "loss": 5.0264,
      "mean_token_accuracy": 0.19319188445806504,
      "num_tokens": 48976723.0,
      "step": 28225
    },
    {
      "entropy": 5.690803527832031,
      "epoch": 2.1963820268430267,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004512458359489853,
      "loss": 5.039,
      "mean_token_accuracy": 0.20466383993625642,
      "num_tokens": 48985362.0,
      "step": 28230
    },
    {
      "entropy": 5.587969875335693,
      "epoch": 2.196771056214744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004512286059043437,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21002357602119445,
      "num_tokens": 48993838.0,
      "step": 28235
    },
    {
      "entropy": 5.658978128433228,
      "epoch": 2.1971600855864617,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045121137318571364,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20257118046283723,
      "num_tokens": 49002054.0,
      "step": 28240
    },
    {
      "entropy": 5.717769050598145,
      "epoch": 2.1975491149581794,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045119413779335645,
      "loss": 4.9742,
      "mean_token_accuracy": 0.19996379762887956,
      "num_tokens": 49010345.0,
      "step": 28245
    },
    {
      "entropy": 5.537086296081543,
      "epoch": 2.197938144329897,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004511768997275336,
      "loss": 4.8652,
      "mean_token_accuracy": 0.21713197976350784,
      "num_tokens": 49018152.0,
      "step": 28250
    },
    {
      "entropy": 5.584592485427857,
      "epoch": 2.1983271737016143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004511596589885068,
      "loss": 4.932,
      "mean_token_accuracy": 0.20973170101642608,
      "num_tokens": 49027876.0,
      "step": 28255
    },
    {
      "entropy": 5.681524658203125,
      "epoch": 2.198716203073332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004511424155765375,
      "loss": 4.9452,
      "mean_token_accuracy": 0.20627468675374985,
      "num_tokens": 49036048.0,
      "step": 28260
    },
    {
      "entropy": 5.704376173019409,
      "epoch": 2.1991052324450497,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045112516949188735,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21067523807287217,
      "num_tokens": 49044705.0,
      "step": 28265
    },
    {
      "entropy": 5.588010978698731,
      "epoch": 2.199494261816767,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004511079207348182,
      "loss": 4.898,
      "mean_token_accuracy": 0.2084199532866478,
      "num_tokens": 49053281.0,
      "step": 28270
    },
    {
      "entropy": 5.582923078536988,
      "epoch": 2.1998832911884847,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004510906693055916,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21140260249376297,
      "num_tokens": 49061681.0,
      "step": 28275
    },
    {
      "entropy": 5.572782945632935,
      "epoch": 2.2002723205602024,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004510734152044694,
      "loss": 4.8133,
      "mean_token_accuracy": 0.21331465393304824,
      "num_tokens": 49069671.0,
      "step": 28280
    },
    {
      "entropy": 5.664387369155884,
      "epoch": 2.20066134993192,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045105615843171336,
      "loss": 5.0752,
      "mean_token_accuracy": 0.19430896788835525,
      "num_tokens": 49078190.0,
      "step": 28285
    },
    {
      "entropy": 5.67475848197937,
      "epoch": 2.2010503793036373,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045103889898758534,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2026989832520485,
      "num_tokens": 49087525.0,
      "step": 28290
    },
    {
      "entropy": 5.679633522033692,
      "epoch": 2.201439408675355,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004510216368723472,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2060781016945839,
      "num_tokens": 49095992.0,
      "step": 28295
    },
    {
      "entropy": 5.706890678405761,
      "epoch": 2.2018284380470727,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045100437208626086,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1951887995004654,
      "num_tokens": 49104195.0,
      "step": 28300
    },
    {
      "entropy": 5.624286413192749,
      "epoch": 2.20221746741879,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045098710462958844,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20303198099136352,
      "num_tokens": 49113070.0,
      "step": 28305
    },
    {
      "entropy": 5.571534967422485,
      "epoch": 2.2026064967905077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045096983450259166,
      "loss": 4.941,
      "mean_token_accuracy": 0.2074427217245102,
      "num_tokens": 49121686.0,
      "step": 28310
    },
    {
      "entropy": 5.6869526386260985,
      "epoch": 2.2029955261622254,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045095256170553285,
      "loss": 5.0524,
      "mean_token_accuracy": 0.1975306436419487,
      "num_tokens": 49129920.0,
      "step": 28315
    },
    {
      "entropy": 5.696602821350098,
      "epoch": 2.2033845555339426,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045093528623867384,
      "loss": 5.028,
      "mean_token_accuracy": 0.2024422511458397,
      "num_tokens": 49138314.0,
      "step": 28320
    },
    {
      "entropy": 5.669262170791626,
      "epoch": 2.2037735849056603,
      "grad_norm": 1.25,
      "learning_rate": 0.00045091800810227706,
      "loss": 4.9245,
      "mean_token_accuracy": 0.20495891571044922,
      "num_tokens": 49146689.0,
      "step": 28325
    },
    {
      "entropy": 5.628743314743042,
      "epoch": 2.204162614277378,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004509007272966044,
      "loss": 4.9409,
      "mean_token_accuracy": 0.20222798138856887,
      "num_tokens": 49155374.0,
      "step": 28330
    },
    {
      "entropy": 5.640137338638306,
      "epoch": 2.2045516436490953,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004508834438219182,
      "loss": 4.9793,
      "mean_token_accuracy": 0.20476195067167283,
      "num_tokens": 49163535.0,
      "step": 28335
    },
    {
      "entropy": 5.631760120391846,
      "epoch": 2.204940673020813,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004508661576784807,
      "loss": 4.929,
      "mean_token_accuracy": 0.20172805190086365,
      "num_tokens": 49171970.0,
      "step": 28340
    },
    {
      "entropy": 5.62496657371521,
      "epoch": 2.2053297023925307,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004508488688665541,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1935613453388214,
      "num_tokens": 49180232.0,
      "step": 28345
    },
    {
      "entropy": 5.614452266693116,
      "epoch": 2.2057187317642484,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004508315773864009,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21733712553977966,
      "num_tokens": 49188611.0,
      "step": 28350
    },
    {
      "entropy": 5.632849931716919,
      "epoch": 2.2061077611359656,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045081428323828337,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20272507518529892,
      "num_tokens": 49197143.0,
      "step": 28355
    },
    {
      "entropy": 5.600899505615234,
      "epoch": 2.2064967905076833,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004507969864224639,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21393770277500151,
      "num_tokens": 49205672.0,
      "step": 28360
    },
    {
      "entropy": 5.596367454528808,
      "epoch": 2.206885819879401,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000450779686939205,
      "loss": 4.9109,
      "mean_token_accuracy": 0.21028747856616975,
      "num_tokens": 49214917.0,
      "step": 28365
    },
    {
      "entropy": 5.651163816452026,
      "epoch": 2.2072748492511183,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045076238478876916,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2006843537092209,
      "num_tokens": 49223768.0,
      "step": 28370
    },
    {
      "entropy": 5.663599252700806,
      "epoch": 2.207663878622836,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004507450799714188,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20090432465076447,
      "num_tokens": 49232075.0,
      "step": 28375
    },
    {
      "entropy": 5.613945341110229,
      "epoch": 2.2080529079945537,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045072777248741664,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2083156570792198,
      "num_tokens": 49240266.0,
      "step": 28380
    },
    {
      "entropy": 5.67335844039917,
      "epoch": 2.2084419373662714,
      "grad_norm": 1.25,
      "learning_rate": 0.0004507104623370252,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20498573184013366,
      "num_tokens": 49248308.0,
      "step": 28385
    },
    {
      "entropy": 5.683750581741333,
      "epoch": 2.2088309667379886,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004506931495205072,
      "loss": 5.0354,
      "mean_token_accuracy": 0.19959783554077148,
      "num_tokens": 49257353.0,
      "step": 28390
    },
    {
      "entropy": 5.715412044525147,
      "epoch": 2.2092199961097063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004506758340381253,
      "loss": 5.0086,
      "mean_token_accuracy": 0.19767916798591614,
      "num_tokens": 49266652.0,
      "step": 28395
    },
    {
      "entropy": 5.578192234039307,
      "epoch": 2.209609025481424,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004506585158901422,
      "loss": 4.8424,
      "mean_token_accuracy": 0.2168688878417015,
      "num_tokens": 49274512.0,
      "step": 28400
    },
    {
      "entropy": 5.5548155307769775,
      "epoch": 2.2099980548531413,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004506411950768207,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2097577691078186,
      "num_tokens": 49282734.0,
      "step": 28405
    },
    {
      "entropy": 5.574290227890015,
      "epoch": 2.210387084224859,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045062387159842363,
      "loss": 4.8699,
      "mean_token_accuracy": 0.21610298454761506,
      "num_tokens": 49291615.0,
      "step": 28410
    },
    {
      "entropy": 5.743170738220215,
      "epoch": 2.2107761135965767,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045060654545521386,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20171353071928025,
      "num_tokens": 49300865.0,
      "step": 28415
    },
    {
      "entropy": 5.651709794998169,
      "epoch": 2.211165142968294,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045058921664745436,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21062974333763124,
      "num_tokens": 49309118.0,
      "step": 28420
    },
    {
      "entropy": 5.6253408908844,
      "epoch": 2.2115541723400116,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045057188517540783,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2016489878296852,
      "num_tokens": 49317954.0,
      "step": 28425
    },
    {
      "entropy": 5.692108821868897,
      "epoch": 2.2119432017117293,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004505545510393374,
      "loss": 5.0321,
      "mean_token_accuracy": 0.19788238853216172,
      "num_tokens": 49327629.0,
      "step": 28430
    },
    {
      "entropy": 5.719353246688843,
      "epoch": 2.2123322310834466,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045053721423950617,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20825668424367905,
      "num_tokens": 49336149.0,
      "step": 28435
    },
    {
      "entropy": 5.623496770858765,
      "epoch": 2.2127212604551643,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004505198747761771,
      "loss": 4.9163,
      "mean_token_accuracy": 0.206802037358284,
      "num_tokens": 49343801.0,
      "step": 28440
    },
    {
      "entropy": 5.57391209602356,
      "epoch": 2.213110289826882,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004505025326496132,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21114057451486587,
      "num_tokens": 49352095.0,
      "step": 28445
    },
    {
      "entropy": 5.665153932571411,
      "epoch": 2.2134993191985997,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045048518786007784,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20099566727876664,
      "num_tokens": 49361313.0,
      "step": 28450
    },
    {
      "entropy": 5.692573308944702,
      "epoch": 2.213888348570317,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045046784040783395,
      "loss": 4.982,
      "mean_token_accuracy": 0.2068189024925232,
      "num_tokens": 49369143.0,
      "step": 28455
    },
    {
      "entropy": 5.639870023727417,
      "epoch": 2.2142773779420346,
      "grad_norm": 1.09375,
      "learning_rate": 0.000450450490293145,
      "loss": 5.0071,
      "mean_token_accuracy": 0.2033908486366272,
      "num_tokens": 49378070.0,
      "step": 28460
    },
    {
      "entropy": 5.700644636154175,
      "epoch": 2.2146664073137523,
      "grad_norm": 1.15625,
      "learning_rate": 0.000450433137516274,
      "loss": 5.0696,
      "mean_token_accuracy": 0.19819185584783555,
      "num_tokens": 49387384.0,
      "step": 28465
    },
    {
      "entropy": 5.530457878112793,
      "epoch": 2.2150554366854696,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004504157820774844,
      "loss": 4.8428,
      "mean_token_accuracy": 0.21621347665786744,
      "num_tokens": 49395963.0,
      "step": 28470
    },
    {
      "entropy": 5.590002822875976,
      "epoch": 2.2154444660571873,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004503984239770396,
      "loss": 4.9684,
      "mean_token_accuracy": 0.205004546046257,
      "num_tokens": 49404319.0,
      "step": 28475
    },
    {
      "entropy": 5.607525634765625,
      "epoch": 2.215833495428905,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004503810632152028,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20632750540971756,
      "num_tokens": 49413492.0,
      "step": 28480
    },
    {
      "entropy": 5.603136157989502,
      "epoch": 2.2162225248006227,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004503636997922375,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20214542150497436,
      "num_tokens": 49422272.0,
      "step": 28485
    },
    {
      "entropy": 5.6705221176147464,
      "epoch": 2.21661155417234,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004503463337084072,
      "loss": 4.9887,
      "mean_token_accuracy": 0.2098117232322693,
      "num_tokens": 49430987.0,
      "step": 28490
    },
    {
      "entropy": 5.5846397399902346,
      "epoch": 2.2170005835440576,
      "grad_norm": 1.25,
      "learning_rate": 0.0004503289649639754,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2093619167804718,
      "num_tokens": 49439951.0,
      "step": 28495
    },
    {
      "entropy": 5.635721492767334,
      "epoch": 2.2173896129157753,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004503115935592056,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20393471270799637,
      "num_tokens": 49448619.0,
      "step": 28500
    },
    {
      "entropy": 5.645308876037598,
      "epoch": 2.2177786422874926,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004502942194943614,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20709340125322342,
      "num_tokens": 49458492.0,
      "step": 28505
    },
    {
      "entropy": 5.675280618667602,
      "epoch": 2.2181676716592102,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045027684276970645,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2071888968348503,
      "num_tokens": 49467022.0,
      "step": 28510
    },
    {
      "entropy": 5.642430019378662,
      "epoch": 2.218556701030928,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045025946338550434,
      "loss": 4.9832,
      "mean_token_accuracy": 0.20372161120176316,
      "num_tokens": 49475494.0,
      "step": 28515
    },
    {
      "entropy": 5.6486741065979,
      "epoch": 2.218945730402645,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004502420813420189,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21352628022432327,
      "num_tokens": 49484124.0,
      "step": 28520
    },
    {
      "entropy": 5.640975856781006,
      "epoch": 2.219334759774363,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004502246966395137,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20186113715171813,
      "num_tokens": 49493584.0,
      "step": 28525
    },
    {
      "entropy": 5.64012131690979,
      "epoch": 2.2197237891460806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045020730927825274,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2090249627828598,
      "num_tokens": 49502635.0,
      "step": 28530
    },
    {
      "entropy": 5.743996047973633,
      "epoch": 2.2201128185177983,
      "grad_norm": 1.25,
      "learning_rate": 0.0004501899192584997,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2103131338953972,
      "num_tokens": 49512166.0,
      "step": 28535
    },
    {
      "entropy": 5.652443504333496,
      "epoch": 2.2205018478895155,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004501725265805185,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20215335190296174,
      "num_tokens": 49521141.0,
      "step": 28540
    },
    {
      "entropy": 5.664021062850952,
      "epoch": 2.2208908772612332,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045015513124457303,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20159000605344773,
      "num_tokens": 49530516.0,
      "step": 28545
    },
    {
      "entropy": 5.620539855957031,
      "epoch": 2.221279906632951,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004501377332509272,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21058933138847352,
      "num_tokens": 49538235.0,
      "step": 28550
    },
    {
      "entropy": 5.618322372436523,
      "epoch": 2.221668936004668,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045012033259984494,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20981960743665695,
      "num_tokens": 49547303.0,
      "step": 28555
    },
    {
      "entropy": 5.655894804000854,
      "epoch": 2.222057965376386,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004501029292915905,
      "loss": 4.9358,
      "mean_token_accuracy": 0.20866496711969376,
      "num_tokens": 49555280.0,
      "step": 28560
    },
    {
      "entropy": 5.733003616333008,
      "epoch": 2.2224469947481036,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045008552332642774,
      "loss": 5.0638,
      "mean_token_accuracy": 0.19714297652244567,
      "num_tokens": 49563673.0,
      "step": 28565
    },
    {
      "entropy": 5.650975513458252,
      "epoch": 2.222836024119821,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045006811470462083,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2033947378396988,
      "num_tokens": 49572253.0,
      "step": 28570
    },
    {
      "entropy": 5.613274669647216,
      "epoch": 2.2232250534915385,
      "grad_norm": 1.234375,
      "learning_rate": 0.000450050703426434,
      "loss": 4.9079,
      "mean_token_accuracy": 0.20420370101928711,
      "num_tokens": 49580780.0,
      "step": 28575
    },
    {
      "entropy": 5.698044395446777,
      "epoch": 2.2236140828632562,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004500332894921313,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21119306832551957,
      "num_tokens": 49589516.0,
      "step": 28580
    },
    {
      "entropy": 5.705364418029785,
      "epoch": 2.224003112234974,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004500158729019771,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20327957570552826,
      "num_tokens": 49597893.0,
      "step": 28585
    },
    {
      "entropy": 5.687286424636841,
      "epoch": 2.224392141606691,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004499984536562355,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2004205361008644,
      "num_tokens": 49606367.0,
      "step": 28590
    },
    {
      "entropy": 5.693738412857056,
      "epoch": 2.224781170978409,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000449981031755171,
      "loss": 4.964,
      "mean_token_accuracy": 0.20433780401945115,
      "num_tokens": 49615121.0,
      "step": 28595
    },
    {
      "entropy": 5.671412086486816,
      "epoch": 2.2251702003501266,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044996360719904777,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20793501734733583,
      "num_tokens": 49623977.0,
      "step": 28600
    },
    {
      "entropy": 5.637563228607178,
      "epoch": 2.225559229721844,
      "grad_norm": 1.125,
      "learning_rate": 0.0004499461799881303,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19995869547128678,
      "num_tokens": 49633398.0,
      "step": 28605
    },
    {
      "entropy": 5.705842447280884,
      "epoch": 2.2259482590935615,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004499287501226831,
      "loss": 5.0484,
      "mean_token_accuracy": 0.2065889835357666,
      "num_tokens": 49641930.0,
      "step": 28610
    },
    {
      "entropy": 5.656173992156982,
      "epoch": 2.2263372884652792,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044991131760297045,
      "loss": 4.961,
      "mean_token_accuracy": 0.20247951447963713,
      "num_tokens": 49650769.0,
      "step": 28615
    },
    {
      "entropy": 5.579816913604736,
      "epoch": 2.226726317836997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044989388242925693,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2084539234638214,
      "num_tokens": 49659057.0,
      "step": 28620
    },
    {
      "entropy": 5.602654314041137,
      "epoch": 2.227115347208714,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004498764446018073,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20144496262073516,
      "num_tokens": 49668016.0,
      "step": 28625
    },
    {
      "entropy": 5.745183086395263,
      "epoch": 2.227504376580432,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498590041208857,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20160751193761825,
      "num_tokens": 49676822.0,
      "step": 28630
    },
    {
      "entropy": 5.671241044998169,
      "epoch": 2.2278934059521496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004498415609867572,
      "loss": 4.9246,
      "mean_token_accuracy": 0.20989976078271866,
      "num_tokens": 49685654.0,
      "step": 28635
    },
    {
      "entropy": 5.718523693084717,
      "epoch": 2.228282435323867,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044982411519968625,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20201894491910935,
      "num_tokens": 49694216.0,
      "step": 28640
    },
    {
      "entropy": 5.597354698181152,
      "epoch": 2.2286714646955845,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004498066667599377,
      "loss": 4.974,
      "mean_token_accuracy": 0.2120068445801735,
      "num_tokens": 49702149.0,
      "step": 28645
    },
    {
      "entropy": 5.630117893218994,
      "epoch": 2.2290604940673022,
      "grad_norm": 1.25,
      "learning_rate": 0.00044978921566777604,
      "loss": 4.9161,
      "mean_token_accuracy": 0.20473225861787797,
      "num_tokens": 49710137.0,
      "step": 28650
    },
    {
      "entropy": 5.587685823440552,
      "epoch": 2.2294495234390195,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004497717619234664,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2163330614566803,
      "num_tokens": 49719577.0,
      "step": 28655
    },
    {
      "entropy": 5.569232845306397,
      "epoch": 2.229838552810737,
      "grad_norm": 1.25,
      "learning_rate": 0.00044975430552727337,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2144857317209244,
      "num_tokens": 49728126.0,
      "step": 28660
    },
    {
      "entropy": 5.643462800979615,
      "epoch": 2.230227582182455,
      "grad_norm": 1.25,
      "learning_rate": 0.0004497368464794619,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20651634484529496,
      "num_tokens": 49736641.0,
      "step": 28665
    },
    {
      "entropy": 5.689315986633301,
      "epoch": 2.230616611554172,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044971938478029693,
      "loss": 4.984,
      "mean_token_accuracy": 0.2026783674955368,
      "num_tokens": 49745090.0,
      "step": 28670
    },
    {
      "entropy": 5.72116436958313,
      "epoch": 2.23100564092589,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044970192043004343,
      "loss": 5.024,
      "mean_token_accuracy": 0.1980004832148552,
      "num_tokens": 49753014.0,
      "step": 28675
    },
    {
      "entropy": 5.594558191299439,
      "epoch": 2.2313946702976075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004496844534289663,
      "loss": 4.9308,
      "mean_token_accuracy": 0.203846837580204,
      "num_tokens": 49761301.0,
      "step": 28680
    },
    {
      "entropy": 5.650203609466553,
      "epoch": 2.231783699669325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044966698377733064,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2100025087594986,
      "num_tokens": 49770004.0,
      "step": 28685
    },
    {
      "entropy": 5.638822555541992,
      "epoch": 2.2321727290410425,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004496495114754015,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2113855615258217,
      "num_tokens": 49778229.0,
      "step": 28690
    },
    {
      "entropy": 5.644185256958008,
      "epoch": 2.23256175841276,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044963203652344404,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20220270454883577,
      "num_tokens": 49786624.0,
      "step": 28695
    },
    {
      "entropy": 5.610005474090576,
      "epoch": 2.232950787784478,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004496145589217234,
      "loss": 4.861,
      "mean_token_accuracy": 0.2151682272553444,
      "num_tokens": 49794487.0,
      "step": 28700
    },
    {
      "entropy": 5.6170085906982425,
      "epoch": 2.233339817156195,
      "grad_norm": 1.25,
      "learning_rate": 0.00044959707867050467,
      "loss": 4.9069,
      "mean_token_accuracy": 0.20929718017578125,
      "num_tokens": 49803141.0,
      "step": 28705
    },
    {
      "entropy": 5.606162405014038,
      "epoch": 2.233728846527913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004495795957700532,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20854750871658326,
      "num_tokens": 49811539.0,
      "step": 28710
    },
    {
      "entropy": 5.554507923126221,
      "epoch": 2.2341178758996305,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044956211022063424,
      "loss": 4.868,
      "mean_token_accuracy": 0.20275925546884538,
      "num_tokens": 49820539.0,
      "step": 28715
    },
    {
      "entropy": 5.623787307739258,
      "epoch": 2.234506905271348,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044954462202251316,
      "loss": 4.9118,
      "mean_token_accuracy": 0.21089418828487397,
      "num_tokens": 49829256.0,
      "step": 28720
    },
    {
      "entropy": 5.67448844909668,
      "epoch": 2.2348959346430655,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044952713117595516,
      "loss": 5.0286,
      "mean_token_accuracy": 0.19897452741861343,
      "num_tokens": 49837631.0,
      "step": 28725
    },
    {
      "entropy": 5.6379838466644285,
      "epoch": 2.235284964014783,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044950963768122576,
      "loss": 5.0177,
      "mean_token_accuracy": 0.19833195954561234,
      "num_tokens": 49846431.0,
      "step": 28730
    },
    {
      "entropy": 5.675135946273803,
      "epoch": 2.235673993386501,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004494921415385905,
      "loss": 5.005,
      "mean_token_accuracy": 0.20128373205661773,
      "num_tokens": 49855346.0,
      "step": 28735
    },
    {
      "entropy": 5.706869983673096,
      "epoch": 2.236063022758218,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044947464274831464,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20038591772317887,
      "num_tokens": 49863993.0,
      "step": 28740
    },
    {
      "entropy": 5.6477807521820065,
      "epoch": 2.236452052129936,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004494571413106637,
      "loss": 4.9011,
      "mean_token_accuracy": 0.20405910462141036,
      "num_tokens": 49872028.0,
      "step": 28745
    },
    {
      "entropy": 5.5834815979003904,
      "epoch": 2.2368410815016535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044943963722590344,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2086910605430603,
      "num_tokens": 49880684.0,
      "step": 28750
    },
    {
      "entropy": 5.640209341049195,
      "epoch": 2.2372301108733708,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044942213049429917,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2111585557460785,
      "num_tokens": 49888818.0,
      "step": 28755
    },
    {
      "entropy": 5.676238918304444,
      "epoch": 2.2376191402450885,
      "grad_norm": 1.125,
      "learning_rate": 0.00044940462111611685,
      "loss": 4.9998,
      "mean_token_accuracy": 0.20719074308872223,
      "num_tokens": 49897558.0,
      "step": 28760
    },
    {
      "entropy": 5.656907081604004,
      "epoch": 2.238008169616806,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044938710909162195,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20505982488393784,
      "num_tokens": 49905652.0,
      "step": 28765
    },
    {
      "entropy": 5.579398202896118,
      "epoch": 2.2383971989885234,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004493695944210802,
      "loss": 4.964,
      "mean_token_accuracy": 0.2014552906155586,
      "num_tokens": 49913911.0,
      "step": 28770
    },
    {
      "entropy": 5.615656137466431,
      "epoch": 2.238786228360241,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004493520771047574,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2068767264485359,
      "num_tokens": 49921794.0,
      "step": 28775
    },
    {
      "entropy": 5.6582183837890625,
      "epoch": 2.239175257731959,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044933455714291944,
      "loss": 4.9943,
      "mean_token_accuracy": 0.19129490554332734,
      "num_tokens": 49930230.0,
      "step": 28780
    },
    {
      "entropy": 5.682102489471435,
      "epoch": 2.2395642871036765,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044931703453583195,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21289683729410172,
      "num_tokens": 49937995.0,
      "step": 28785
    },
    {
      "entropy": 5.688552188873291,
      "epoch": 2.2399533164753938,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044929950928376095,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20415483862161637,
      "num_tokens": 49945456.0,
      "step": 28790
    },
    {
      "entropy": 5.594967460632324,
      "epoch": 2.2403423458471114,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004492819813869723,
      "loss": 4.964,
      "mean_token_accuracy": 0.20929117500782013,
      "num_tokens": 49953679.0,
      "step": 28795
    },
    {
      "entropy": 5.625983047485351,
      "epoch": 2.240731375218829,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044926445084573206,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21350717097520827,
      "num_tokens": 49961771.0,
      "step": 28800
    },
    {
      "entropy": 5.655473279953003,
      "epoch": 2.2411204045905464,
      "grad_norm": 1.21875,
      "learning_rate": 0.000449246917660306,
      "loss": 4.91,
      "mean_token_accuracy": 0.2106430396437645,
      "num_tokens": 49970157.0,
      "step": 28805
    },
    {
      "entropy": 5.706585121154785,
      "epoch": 2.241509433962264,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004492293818309604,
      "loss": 5.0368,
      "mean_token_accuracy": 0.19362106621265412,
      "num_tokens": 49978586.0,
      "step": 28810
    },
    {
      "entropy": 5.573245573043823,
      "epoch": 2.241898463333982,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004492118433579613,
      "loss": 4.9438,
      "mean_token_accuracy": 0.20379832088947297,
      "num_tokens": 49987686.0,
      "step": 28815
    },
    {
      "entropy": 5.663925743103027,
      "epoch": 2.2422874927056995,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044919430224157463,
      "loss": 4.9507,
      "mean_token_accuracy": 0.20690799206495286,
      "num_tokens": 49996529.0,
      "step": 28820
    },
    {
      "entropy": 5.709371519088745,
      "epoch": 2.2426765220774167,
      "grad_norm": 1.125,
      "learning_rate": 0.00044917675848206684,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20729530453681946,
      "num_tokens": 50005506.0,
      "step": 28825
    },
    {
      "entropy": 5.634599018096924,
      "epoch": 2.2430655514491344,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044915921207970387,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21217499375343324,
      "num_tokens": 50013424.0,
      "step": 28830
    },
    {
      "entropy": 5.633788967132569,
      "epoch": 2.243454580820852,
      "grad_norm": 1.25,
      "learning_rate": 0.00044914166303475206,
      "loss": 4.9338,
      "mean_token_accuracy": 0.20475925505161285,
      "num_tokens": 50021702.0,
      "step": 28835
    },
    {
      "entropy": 5.630589437484741,
      "epoch": 2.2438436101925694,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044912411134747764,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20542633384466172,
      "num_tokens": 50030722.0,
      "step": 28840
    },
    {
      "entropy": 5.695249700546265,
      "epoch": 2.244232639564287,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044910655701814705,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2023451194167137,
      "num_tokens": 50038829.0,
      "step": 28845
    },
    {
      "entropy": 5.706965208053589,
      "epoch": 2.244621668936005,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004490890000470266,
      "loss": 4.9453,
      "mean_token_accuracy": 0.20820437520742416,
      "num_tokens": 50046461.0,
      "step": 28850
    },
    {
      "entropy": 5.59414849281311,
      "epoch": 2.245010698307722,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044907144043438265,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20457579791545868,
      "num_tokens": 50055087.0,
      "step": 28855
    },
    {
      "entropy": 5.6637248516082765,
      "epoch": 2.2453997276794397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004490538781804816,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20287902355194093,
      "num_tokens": 50064802.0,
      "step": 28860
    },
    {
      "entropy": 5.721163654327393,
      "epoch": 2.2457887570511574,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044903631328559,
      "loss": 5.0728,
      "mean_token_accuracy": 0.19541071802377702,
      "num_tokens": 50073617.0,
      "step": 28865
    },
    {
      "entropy": 5.642386865615845,
      "epoch": 2.246177786422875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044901874574997444,
      "loss": 4.9164,
      "mean_token_accuracy": 0.21159314215183259,
      "num_tokens": 50081705.0,
      "step": 28870
    },
    {
      "entropy": 5.57199912071228,
      "epoch": 2.2465668157945924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044900117557390137,
      "loss": 4.9109,
      "mean_token_accuracy": 0.21109004318714142,
      "num_tokens": 50089982.0,
      "step": 28875
    },
    {
      "entropy": 5.612879991531372,
      "epoch": 2.24695584516631,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044898360275763733,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21332501769065856,
      "num_tokens": 50098837.0,
      "step": 28880
    },
    {
      "entropy": 5.695000505447387,
      "epoch": 2.247344874538028,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044896602730144905,
      "loss": 5.022,
      "mean_token_accuracy": 0.1956079587340355,
      "num_tokens": 50108109.0,
      "step": 28885
    },
    {
      "entropy": 5.612943506240844,
      "epoch": 2.247733903909745,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044894844920560323,
      "loss": 4.9309,
      "mean_token_accuracy": 0.20680210292339324,
      "num_tokens": 50116758.0,
      "step": 28890
    },
    {
      "entropy": 5.6531096458435055,
      "epoch": 2.2481229332814627,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004489308684703666,
      "loss": 4.948,
      "mean_token_accuracy": 0.20964026898145677,
      "num_tokens": 50125014.0,
      "step": 28895
    },
    {
      "entropy": 5.715013647079468,
      "epoch": 2.2485119626531804,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004489132850960059,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1944926053285599,
      "num_tokens": 50132829.0,
      "step": 28900
    },
    {
      "entropy": 5.635134077072143,
      "epoch": 2.2489009920248977,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004488956990827878,
      "loss": 4.9421,
      "mean_token_accuracy": 0.20699441879987718,
      "num_tokens": 50141161.0,
      "step": 28905
    },
    {
      "entropy": 5.635026216506958,
      "epoch": 2.2492900213966154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004488781104309793,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20617321878671646,
      "num_tokens": 50150356.0,
      "step": 28910
    },
    {
      "entropy": 5.608652782440186,
      "epoch": 2.249679050768333,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044886051914084726,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21375665813684464,
      "num_tokens": 50159235.0,
      "step": 28915
    },
    {
      "entropy": 5.655436229705811,
      "epoch": 2.2500680801400508,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044884292521265846,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2000218391418457,
      "num_tokens": 50168420.0,
      "step": 28920
    },
    {
      "entropy": 5.648804759979248,
      "epoch": 2.250457109511768,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004488253286466801,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21292107850313186,
      "num_tokens": 50177432.0,
      "step": 28925
    },
    {
      "entropy": 5.763665866851807,
      "epoch": 2.2508461388834857,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044880772944317905,
      "loss": 5.042,
      "mean_token_accuracy": 0.20566439330577851,
      "num_tokens": 50185610.0,
      "step": 28930
    },
    {
      "entropy": 5.688184309005737,
      "epoch": 2.2512351682552034,
      "grad_norm": 1.25,
      "learning_rate": 0.00044879012760242224,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2054459571838379,
      "num_tokens": 50194129.0,
      "step": 28935
    },
    {
      "entropy": 5.6408792495727536,
      "epoch": 2.2516241976269207,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044877252312467694,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20717571973800658,
      "num_tokens": 50202732.0,
      "step": 28940
    },
    {
      "entropy": 5.6717921733856205,
      "epoch": 2.2520132269986384,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004487549160102101,
      "loss": 5.0398,
      "mean_token_accuracy": 0.1972149655222893,
      "num_tokens": 50211060.0,
      "step": 28945
    },
    {
      "entropy": 5.733799839019776,
      "epoch": 2.252402256370356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044873730625928914,
      "loss": 5.097,
      "mean_token_accuracy": 0.20007816702127457,
      "num_tokens": 50220664.0,
      "step": 28950
    },
    {
      "entropy": 5.737061309814453,
      "epoch": 2.2527912857420738,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044871969387218094,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2076158806681633,
      "num_tokens": 50229425.0,
      "step": 28955
    },
    {
      "entropy": 5.674380588531494,
      "epoch": 2.253180315113791,
      "grad_norm": 1.234375,
      "learning_rate": 0.000448702078849153,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1962573930621147,
      "num_tokens": 50237992.0,
      "step": 28960
    },
    {
      "entropy": 5.5988188743591305,
      "epoch": 2.2535693444855087,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044868446119047234,
      "loss": 4.9725,
      "mean_token_accuracy": 0.20356984585523605,
      "num_tokens": 50246480.0,
      "step": 28965
    },
    {
      "entropy": 5.6960889339447025,
      "epoch": 2.253958373857226,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004486668408964065,
      "loss": 4.9935,
      "mean_token_accuracy": 0.19714525938034058,
      "num_tokens": 50256250.0,
      "step": 28970
    },
    {
      "entropy": 5.63188591003418,
      "epoch": 2.2543474032289437,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004486492179672228,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2085650682449341,
      "num_tokens": 50265126.0,
      "step": 28975
    },
    {
      "entropy": 5.625449132919312,
      "epoch": 2.2547364326006614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044863159240318863,
      "loss": 4.931,
      "mean_token_accuracy": 0.20666096955537797,
      "num_tokens": 50273636.0,
      "step": 28980
    },
    {
      "entropy": 5.72828860282898,
      "epoch": 2.255125461972379,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004486139642045714,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1994878277182579,
      "num_tokens": 50281971.0,
      "step": 28985
    },
    {
      "entropy": 5.707199335098267,
      "epoch": 2.2555144913440963,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004485963333716385,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2053772211074829,
      "num_tokens": 50291108.0,
      "step": 28990
    },
    {
      "entropy": 5.700412845611572,
      "epoch": 2.255903520715814,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004485786999046576,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2057879537343979,
      "num_tokens": 50300153.0,
      "step": 28995
    },
    {
      "entropy": 5.652979230880737,
      "epoch": 2.2562925500875317,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044856106380389624,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20744085013866426,
      "num_tokens": 50308887.0,
      "step": 29000
    },
    {
      "entropy": 5.597199583053589,
      "epoch": 2.256681579459249,
      "grad_norm": 1.203125,
      "learning_rate": 0.000448543425069622,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21122995316982268,
      "num_tokens": 50316774.0,
      "step": 29005
    },
    {
      "entropy": 5.669611978530884,
      "epoch": 2.2570706088309667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004485257837021025,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20373129546642305,
      "num_tokens": 50325943.0,
      "step": 29010
    },
    {
      "entropy": 5.706617116928101,
      "epoch": 2.2574596382026844,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044850813970160534,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20829411447048188,
      "num_tokens": 50335043.0,
      "step": 29015
    },
    {
      "entropy": 5.726704168319702,
      "epoch": 2.257848667574402,
      "grad_norm": 1.25,
      "learning_rate": 0.0004484904930683984,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2086644947528839,
      "num_tokens": 50343108.0,
      "step": 29020
    },
    {
      "entropy": 5.619496583938599,
      "epoch": 2.2582376969461193,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004484728438027494,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20237225592136382,
      "num_tokens": 50352177.0,
      "step": 29025
    },
    {
      "entropy": 5.659099388122558,
      "epoch": 2.258626726317837,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004484551919049261,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21082679629325868,
      "num_tokens": 50360986.0,
      "step": 29030
    },
    {
      "entropy": 5.731006336212158,
      "epoch": 2.2590157556895547,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044843753737519625,
      "loss": 5.0435,
      "mean_token_accuracy": 0.194256229698658,
      "num_tokens": 50369871.0,
      "step": 29035
    },
    {
      "entropy": 5.6692193984985355,
      "epoch": 2.259404785061272,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004484198802138279,
      "loss": 4.9722,
      "mean_token_accuracy": 0.20525528341531754,
      "num_tokens": 50378482.0,
      "step": 29040
    },
    {
      "entropy": 5.658724546432495,
      "epoch": 2.2597938144329897,
      "grad_norm": 1.25,
      "learning_rate": 0.0004484022204210889,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21005659848451613,
      "num_tokens": 50387343.0,
      "step": 29045
    },
    {
      "entropy": 5.583187532424927,
      "epoch": 2.2601828438047074,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044838455799724717,
      "loss": 4.8885,
      "mean_token_accuracy": 0.20845258235931396,
      "num_tokens": 50395704.0,
      "step": 29050
    },
    {
      "entropy": 5.701115083694458,
      "epoch": 2.260571873176425,
      "grad_norm": 1.25,
      "learning_rate": 0.00044836689294257075,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2048500195145607,
      "num_tokens": 50404719.0,
      "step": 29055
    },
    {
      "entropy": 5.6695143699646,
      "epoch": 2.2609609025481423,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004483492252573276,
      "loss": 4.9334,
      "mean_token_accuracy": 0.20412212312221528,
      "num_tokens": 50413167.0,
      "step": 29060
    },
    {
      "entropy": 5.718425273895264,
      "epoch": 2.26134993191986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000448331554941786,
      "loss": 5.1356,
      "mean_token_accuracy": 0.19580833613872528,
      "num_tokens": 50422609.0,
      "step": 29065
    },
    {
      "entropy": 5.665090751647949,
      "epoch": 2.2617389612915777,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044831388199621385,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20689670294523238,
      "num_tokens": 50431841.0,
      "step": 29070
    },
    {
      "entropy": 5.632295608520508,
      "epoch": 2.262127990663295,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044829620642087946,
      "loss": 4.966,
      "mean_token_accuracy": 0.20480636209249498,
      "num_tokens": 50440625.0,
      "step": 29075
    },
    {
      "entropy": 5.6660562515258786,
      "epoch": 2.2625170200350126,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004482785282160509,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19362804740667344,
      "num_tokens": 50449577.0,
      "step": 29080
    },
    {
      "entropy": 5.714691209793091,
      "epoch": 2.2629060494067303,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044826084738199657,
      "loss": 5.022,
      "mean_token_accuracy": 0.20577373206615449,
      "num_tokens": 50457817.0,
      "step": 29085
    },
    {
      "entropy": 5.6106161117553714,
      "epoch": 2.2632950787784476,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044824316391898464,
      "loss": 4.9614,
      "mean_token_accuracy": 0.20961310863494872,
      "num_tokens": 50466215.0,
      "step": 29090
    },
    {
      "entropy": 5.609082794189453,
      "epoch": 2.2636841081501653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044822547782728334,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2073902204632759,
      "num_tokens": 50475415.0,
      "step": 29095
    },
    {
      "entropy": 5.6919639110565186,
      "epoch": 2.264073137521883,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004482077891071612,
      "loss": 4.992,
      "mean_token_accuracy": 0.20399350970983504,
      "num_tokens": 50483883.0,
      "step": 29100
    },
    {
      "entropy": 5.669452810287476,
      "epoch": 2.2644621668936002,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044819009775888655,
      "loss": 4.9117,
      "mean_token_accuracy": 0.21183478087186813,
      "num_tokens": 50493541.0,
      "step": 29105
    },
    {
      "entropy": 5.636380529403686,
      "epoch": 2.264851196265318,
      "grad_norm": 1.25,
      "learning_rate": 0.00044817240378272784,
      "loss": 5.0321,
      "mean_token_accuracy": 0.20962298810482025,
      "num_tokens": 50502063.0,
      "step": 29110
    },
    {
      "entropy": 5.623637008666992,
      "epoch": 2.2652402256370356,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044815470717895345,
      "loss": 4.8656,
      "mean_token_accuracy": 0.21597001999616622,
      "num_tokens": 50510956.0,
      "step": 29115
    },
    {
      "entropy": 5.716809415817261,
      "epoch": 2.2656292550087533,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004481370079478321,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20452577769756317,
      "num_tokens": 50519383.0,
      "step": 29120
    },
    {
      "entropy": 5.662494039535522,
      "epoch": 2.2660182843804706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004481193060896322,
      "loss": 4.977,
      "mean_token_accuracy": 0.20429450422525405,
      "num_tokens": 50528172.0,
      "step": 29125
    },
    {
      "entropy": 5.639087343215943,
      "epoch": 2.2664073137521883,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004481016016046223,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20446632504463197,
      "num_tokens": 50536720.0,
      "step": 29130
    },
    {
      "entropy": 5.616045141220093,
      "epoch": 2.266796343123906,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004480838944930712,
      "loss": 4.9111,
      "mean_token_accuracy": 0.20678861141204835,
      "num_tokens": 50545502.0,
      "step": 29135
    },
    {
      "entropy": 5.62020354270935,
      "epoch": 2.2671853724956232,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004480661847552474,
      "loss": 4.9237,
      "mean_token_accuracy": 0.20732140988111497,
      "num_tokens": 50553262.0,
      "step": 29140
    },
    {
      "entropy": 5.594115591049194,
      "epoch": 2.267574401867341,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004480484723914198,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20471178293228148,
      "num_tokens": 50562153.0,
      "step": 29145
    },
    {
      "entropy": 5.653083038330078,
      "epoch": 2.2679634312390586,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004480307574018571,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20297963470220565,
      "num_tokens": 50571476.0,
      "step": 29150
    },
    {
      "entropy": 5.69712381362915,
      "epoch": 2.2683524606107763,
      "grad_norm": 1.171875,
      "learning_rate": 0.000448013039786828,
      "loss": 5.0734,
      "mean_token_accuracy": 0.19496839493513107,
      "num_tokens": 50579371.0,
      "step": 29155
    },
    {
      "entropy": 5.624006080627441,
      "epoch": 2.2687414899824936,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044799531954660133,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20544691979885102,
      "num_tokens": 50588214.0,
      "step": 29160
    },
    {
      "entropy": 5.592681980133056,
      "epoch": 2.2691305193542113,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004479775966814461,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20001684725284577,
      "num_tokens": 50596686.0,
      "step": 29165
    },
    {
      "entropy": 5.658980464935302,
      "epoch": 2.269519548725929,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044795987119163115,
      "loss": 4.9093,
      "mean_token_accuracy": 0.20652946084737778,
      "num_tokens": 50604843.0,
      "step": 29170
    },
    {
      "entropy": 5.701581954956055,
      "epoch": 2.2699085780976462,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004479421430774255,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20385261625051498,
      "num_tokens": 50613722.0,
      "step": 29175
    },
    {
      "entropy": 5.6947612285614015,
      "epoch": 2.270297607469364,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004479244123390981,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2040952757000923,
      "num_tokens": 50622708.0,
      "step": 29180
    },
    {
      "entropy": 5.654938077926635,
      "epoch": 2.2706866368410816,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044790667897691796,
      "loss": 4.9177,
      "mean_token_accuracy": 0.20391227900981904,
      "num_tokens": 50630516.0,
      "step": 29185
    },
    {
      "entropy": 5.6324304103851315,
      "epoch": 2.271075666212799,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044788894299115415,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2124847486615181,
      "num_tokens": 50639695.0,
      "step": 29190
    },
    {
      "entropy": 5.646047687530517,
      "epoch": 2.2714646955845166,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004478712043820758,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2068116158246994,
      "num_tokens": 50648600.0,
      "step": 29195
    },
    {
      "entropy": 5.626090240478516,
      "epoch": 2.2718537249562343,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004478534631499521,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21525303274393082,
      "num_tokens": 50657020.0,
      "step": 29200
    },
    {
      "entropy": 5.687259197235107,
      "epoch": 2.2722427543279515,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044783571929505223,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2123490482568741,
      "num_tokens": 50665909.0,
      "step": 29205
    },
    {
      "entropy": 5.61186146736145,
      "epoch": 2.2726317836996692,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004478179728176454,
      "loss": 4.8718,
      "mean_token_accuracy": 0.21285301595926284,
      "num_tokens": 50674152.0,
      "step": 29210
    },
    {
      "entropy": 5.576617860794068,
      "epoch": 2.273020813071387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004478002237180009,
      "loss": 4.8613,
      "mean_token_accuracy": 0.21440641582012177,
      "num_tokens": 50682447.0,
      "step": 29215
    },
    {
      "entropy": 5.603819513320923,
      "epoch": 2.2734098424431046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000447782471996388,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2060973435640335,
      "num_tokens": 50691253.0,
      "step": 29220
    },
    {
      "entropy": 5.606770277023315,
      "epoch": 2.273798871814822,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004477647176530762,
      "loss": 4.999,
      "mean_token_accuracy": 0.20557631999254228,
      "num_tokens": 50700336.0,
      "step": 29225
    },
    {
      "entropy": 5.6338193893432615,
      "epoch": 2.2741879011865396,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004477469606883347,
      "loss": 4.906,
      "mean_token_accuracy": 0.21142034381628036,
      "num_tokens": 50709231.0,
      "step": 29230
    },
    {
      "entropy": 5.604260683059692,
      "epoch": 2.2745769305582573,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004477292011024331,
      "loss": 4.8738,
      "mean_token_accuracy": 0.21137246787548064,
      "num_tokens": 50718604.0,
      "step": 29235
    },
    {
      "entropy": 5.643488550186158,
      "epoch": 2.2749659599299745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044771143889564075,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20687718987464904,
      "num_tokens": 50727146.0,
      "step": 29240
    },
    {
      "entropy": 5.58636245727539,
      "epoch": 2.275354989301692,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044769367406822727,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2096048563718796,
      "num_tokens": 50736672.0,
      "step": 29245
    },
    {
      "entropy": 5.627687692642212,
      "epoch": 2.27574401867341,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004476759066204621,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20254482328891754,
      "num_tokens": 50744609.0,
      "step": 29250
    },
    {
      "entropy": 5.6986829280853275,
      "epoch": 2.2761330480451276,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004476581365526149,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21126613169908523,
      "num_tokens": 50753286.0,
      "step": 29255
    },
    {
      "entropy": 5.65384612083435,
      "epoch": 2.276522077416845,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004476403638649553,
      "loss": 5.0104,
      "mean_token_accuracy": 0.19101709872484207,
      "num_tokens": 50761554.0,
      "step": 29260
    },
    {
      "entropy": 5.584729909896851,
      "epoch": 2.2769111067885626,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044762258855775294,
      "loss": 4.9505,
      "mean_token_accuracy": 0.20428793728351594,
      "num_tokens": 50770628.0,
      "step": 29265
    },
    {
      "entropy": 5.658384323120117,
      "epoch": 2.2773001361602803,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044760481063127754,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2094045490026474,
      "num_tokens": 50779159.0,
      "step": 29270
    },
    {
      "entropy": 5.748681259155274,
      "epoch": 2.2776891655319975,
      "grad_norm": 1.328125,
      "learning_rate": 0.000447587030085799,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20831007957458497,
      "num_tokens": 50787192.0,
      "step": 29275
    },
    {
      "entropy": 5.594430685043335,
      "epoch": 2.278078194903715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044756924692158684,
      "loss": 4.8096,
      "mean_token_accuracy": 0.21871258318424225,
      "num_tokens": 50795441.0,
      "step": 29280
    },
    {
      "entropy": 5.689841413497925,
      "epoch": 2.278467224275433,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004475514611389111,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2078137919306755,
      "num_tokens": 50804178.0,
      "step": 29285
    },
    {
      "entropy": 5.732893085479736,
      "epoch": 2.2788562536471506,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004475336727380415,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20614046454429627,
      "num_tokens": 50813155.0,
      "step": 29290
    },
    {
      "entropy": 5.688362884521484,
      "epoch": 2.279245283018868,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004475158817192481,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21584997624158858,
      "num_tokens": 50822131.0,
      "step": 29295
    },
    {
      "entropy": 5.6883501529693605,
      "epoch": 2.2796343123905856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044749808808280077,
      "loss": 5.0131,
      "mean_token_accuracy": 0.2065530776977539,
      "num_tokens": 50830821.0,
      "step": 29300
    },
    {
      "entropy": 5.641874074935913,
      "epoch": 2.280023341762303,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044748029182896956,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20067632496356963,
      "num_tokens": 50839928.0,
      "step": 29305
    },
    {
      "entropy": 5.659412860870361,
      "epoch": 2.2804123711340205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004474624929580243,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21475536227226258,
      "num_tokens": 50848674.0,
      "step": 29310
    },
    {
      "entropy": 5.6953552722930905,
      "epoch": 2.280801400505738,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004474446914702354,
      "loss": 5.018,
      "mean_token_accuracy": 0.20654019713401794,
      "num_tokens": 50857376.0,
      "step": 29315
    },
    {
      "entropy": 5.715448474884033,
      "epoch": 2.281190429877456,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004474268873658727,
      "loss": 5.0755,
      "mean_token_accuracy": 0.1939623937010765,
      "num_tokens": 50866214.0,
      "step": 29320
    },
    {
      "entropy": 5.663600587844849,
      "epoch": 2.281579459249173,
      "grad_norm": 1.25,
      "learning_rate": 0.0004474090806452064,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21003350764513015,
      "num_tokens": 50875151.0,
      "step": 29325
    },
    {
      "entropy": 5.6557701587677,
      "epoch": 2.281968488620891,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044739127130850675,
      "loss": 4.9208,
      "mean_token_accuracy": 0.20873579978942872,
      "num_tokens": 50882600.0,
      "step": 29330
    },
    {
      "entropy": 5.628037929534912,
      "epoch": 2.2823575179926086,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044737345935604397,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20092481672763823,
      "num_tokens": 50891348.0,
      "step": 29335
    },
    {
      "entropy": 5.602987289428711,
      "epoch": 2.282746547364326,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004473556447880883,
      "loss": 4.9183,
      "mean_token_accuracy": 0.20732947140932084,
      "num_tokens": 50900452.0,
      "step": 29340
    },
    {
      "entropy": 5.644867181777954,
      "epoch": 2.2831355767360435,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044733782760490997,
      "loss": 4.99,
      "mean_token_accuracy": 0.2038433775305748,
      "num_tokens": 50909599.0,
      "step": 29345
    },
    {
      "entropy": 5.652254295349121,
      "epoch": 2.283524606107761,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004473200078067795,
      "loss": 4.9377,
      "mean_token_accuracy": 0.20353127717971803,
      "num_tokens": 50918864.0,
      "step": 29350
    },
    {
      "entropy": 5.649256896972656,
      "epoch": 2.283913635479479,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044730218539396717,
      "loss": 5.0462,
      "mean_token_accuracy": 0.20378716737031938,
      "num_tokens": 50927851.0,
      "step": 29355
    },
    {
      "entropy": 5.664357137680054,
      "epoch": 2.284302664851196,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004472843603667434,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21372474133968353,
      "num_tokens": 50936974.0,
      "step": 29360
    },
    {
      "entropy": 5.757056379318238,
      "epoch": 2.284691694222914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044726653272537866,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2000075623393059,
      "num_tokens": 50945482.0,
      "step": 29365
    },
    {
      "entropy": 5.760507822036743,
      "epoch": 2.2850807235946315,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004472487024701435,
      "loss": 5.0858,
      "mean_token_accuracy": 0.19825949221849443,
      "num_tokens": 50954961.0,
      "step": 29370
    },
    {
      "entropy": 5.688434171676636,
      "epoch": 2.285469752966349,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004472308696013085,
      "loss": 4.9357,
      "mean_token_accuracy": 0.20900008380413054,
      "num_tokens": 50963666.0,
      "step": 29375
    },
    {
      "entropy": 5.65955286026001,
      "epoch": 2.2858587823380665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004472130341191441,
      "loss": 4.9269,
      "mean_token_accuracy": 0.20942501574754716,
      "num_tokens": 50972761.0,
      "step": 29380
    },
    {
      "entropy": 5.676043176651001,
      "epoch": 2.286247811709784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004471951960239211,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21027719229459763,
      "num_tokens": 50981755.0,
      "step": 29385
    },
    {
      "entropy": 5.696746730804444,
      "epoch": 2.286636841081502,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044717735531591,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20573717653751372,
      "num_tokens": 50990381.0,
      "step": 29390
    },
    {
      "entropy": 5.638506841659546,
      "epoch": 2.287025870453219,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044715951199538156,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20613723546266555,
      "num_tokens": 50999076.0,
      "step": 29395
    },
    {
      "entropy": 5.62277045249939,
      "epoch": 2.287414899824937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044714166606260657,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21194605976343156,
      "num_tokens": 51007626.0,
      "step": 29400
    },
    {
      "entropy": 5.630076932907104,
      "epoch": 2.2878039291966545,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004471238175178559,
      "loss": 4.9331,
      "mean_token_accuracy": 0.20930130779743195,
      "num_tokens": 51016120.0,
      "step": 29405
    },
    {
      "entropy": 5.646851444244385,
      "epoch": 2.288192958568372,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004471059663614002,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20247174054384232,
      "num_tokens": 51024558.0,
      "step": 29410
    },
    {
      "entropy": 5.611168098449707,
      "epoch": 2.2885819879400895,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004470881125935103,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21367865204811096,
      "num_tokens": 51032828.0,
      "step": 29415
    },
    {
      "entropy": 5.690545320510864,
      "epoch": 2.288971017311807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044707025621445735,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2052918031811714,
      "num_tokens": 51040966.0,
      "step": 29420
    },
    {
      "entropy": 5.6816990852355955,
      "epoch": 2.2893600466835244,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044705239722451194,
      "loss": 5.033,
      "mean_token_accuracy": 0.20240819454193115,
      "num_tokens": 51049611.0,
      "step": 29425
    },
    {
      "entropy": 5.6392981052398685,
      "epoch": 2.289749076055242,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044703453562394544,
      "loss": 4.929,
      "mean_token_accuracy": 0.20599979013204575,
      "num_tokens": 51058446.0,
      "step": 29430
    },
    {
      "entropy": 5.776081323623657,
      "epoch": 2.29013810542696,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004470166714130286,
      "loss": 5.0831,
      "mean_token_accuracy": 0.19447289854288102,
      "num_tokens": 51067526.0,
      "step": 29435
    },
    {
      "entropy": 5.692907476425171,
      "epoch": 2.290527134798677,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044699880459203264,
      "loss": 5.0141,
      "mean_token_accuracy": 0.2093735709786415,
      "num_tokens": 51076154.0,
      "step": 29440
    },
    {
      "entropy": 5.692427301406861,
      "epoch": 2.290916164170395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044698093516122844,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20358216911554336,
      "num_tokens": 51084845.0,
      "step": 29445
    },
    {
      "entropy": 5.655909681320191,
      "epoch": 2.2913051935421125,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004469630631208874,
      "loss": 4.9634,
      "mean_token_accuracy": 0.20862861573696137,
      "num_tokens": 51093436.0,
      "step": 29450
    },
    {
      "entropy": 5.664717054367065,
      "epoch": 2.29169422291383,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004469451884712805,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20271194726228714,
      "num_tokens": 51102369.0,
      "step": 29455
    },
    {
      "entropy": 5.658638286590576,
      "epoch": 2.2920832522855474,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004469273112126791,
      "loss": 4.98,
      "mean_token_accuracy": 0.20381619781255722,
      "num_tokens": 51111335.0,
      "step": 29460
    },
    {
      "entropy": 5.6795759201049805,
      "epoch": 2.292472281657265,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044690943134535445,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1995371997356415,
      "num_tokens": 51120161.0,
      "step": 29465
    },
    {
      "entropy": 5.633863067626953,
      "epoch": 2.292861311028983,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004468915488695777,
      "loss": 4.962,
      "mean_token_accuracy": 0.2028233990073204,
      "num_tokens": 51128840.0,
      "step": 29470
    },
    {
      "entropy": 5.674533987045288,
      "epoch": 2.2932503404007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044687366378562036,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21068734675645828,
      "num_tokens": 51137321.0,
      "step": 29475
    },
    {
      "entropy": 5.722371530532837,
      "epoch": 2.293639369772418,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044685577609375373,
      "loss": 5.063,
      "mean_token_accuracy": 0.19566056728363038,
      "num_tokens": 51145564.0,
      "step": 29480
    },
    {
      "entropy": 5.658788681030273,
      "epoch": 2.2940283991441355,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004468378857942492,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2070489838719368,
      "num_tokens": 51153983.0,
      "step": 29485
    },
    {
      "entropy": 5.598476409912109,
      "epoch": 2.294417428515853,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044681999288737826,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20144178420305253,
      "num_tokens": 51162547.0,
      "step": 29490
    },
    {
      "entropy": 5.642254495620728,
      "epoch": 2.2948064578875704,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044680209737341244,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20773879289627076,
      "num_tokens": 51171268.0,
      "step": 29495
    },
    {
      "entropy": 5.664510345458984,
      "epoch": 2.295195487259288,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004467841992526233,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2022454023361206,
      "num_tokens": 51179807.0,
      "step": 29500
    },
    {
      "entropy": 5.63982195854187,
      "epoch": 2.295584516631006,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044676629852528226,
      "loss": 4.9841,
      "mean_token_accuracy": 0.20074319690465928,
      "num_tokens": 51187670.0,
      "step": 29505
    },
    {
      "entropy": 5.728522253036499,
      "epoch": 2.295973546002723,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044674839519166104,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1998781830072403,
      "num_tokens": 51196512.0,
      "step": 29510
    },
    {
      "entropy": 5.652633762359619,
      "epoch": 2.2963625753744408,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044673048925203136,
      "loss": 4.8892,
      "mean_token_accuracy": 0.20725188553333282,
      "num_tokens": 51204985.0,
      "step": 29515
    },
    {
      "entropy": 5.6189874649047855,
      "epoch": 2.2967516047461585,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044671258070666476,
      "loss": 4.977,
      "mean_token_accuracy": 0.20666809678077697,
      "num_tokens": 51214472.0,
      "step": 29520
    },
    {
      "entropy": 5.729306030273437,
      "epoch": 2.2971406341178757,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004466946695558331,
      "loss": 5.1343,
      "mean_token_accuracy": 0.19044889360666276,
      "num_tokens": 51223836.0,
      "step": 29525
    },
    {
      "entropy": 5.632682037353516,
      "epoch": 2.2975296634895934,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044667675579980817,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20354359745979309,
      "num_tokens": 51232060.0,
      "step": 29530
    },
    {
      "entropy": 5.621022653579712,
      "epoch": 2.297918692861311,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004466588394388616,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2084936812520027,
      "num_tokens": 51240273.0,
      "step": 29535
    },
    {
      "entropy": 5.655084705352783,
      "epoch": 2.2983077222330284,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004466409204732654,
      "loss": 4.8421,
      "mean_token_accuracy": 0.21260540932416916,
      "num_tokens": 51248892.0,
      "step": 29540
    },
    {
      "entropy": 5.671745109558105,
      "epoch": 2.298696751604746,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044662299890329144,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2111931025981903,
      "num_tokens": 51258028.0,
      "step": 29545
    },
    {
      "entropy": 5.627292537689209,
      "epoch": 2.2990857809764638,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004466050747292116,
      "loss": 4.9282,
      "mean_token_accuracy": 0.20538226068019866,
      "num_tokens": 51266624.0,
      "step": 29550
    },
    {
      "entropy": 5.6389298915863035,
      "epoch": 2.2994748103481815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000446587147951298,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19966224431991578,
      "num_tokens": 51275094.0,
      "step": 29555
    },
    {
      "entropy": 5.625348806381226,
      "epoch": 2.2998638397198987,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004465692185698224,
      "loss": 5.0159,
      "mean_token_accuracy": 0.20186548084020614,
      "num_tokens": 51283675.0,
      "step": 29560
    },
    {
      "entropy": 5.69339075088501,
      "epoch": 2.3002528690916164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044655128658505717,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20369889587163925,
      "num_tokens": 51292208.0,
      "step": 29565
    },
    {
      "entropy": 5.613300418853759,
      "epoch": 2.300641898463334,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004465333519972741,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2135475441813469,
      "num_tokens": 51301055.0,
      "step": 29570
    },
    {
      "entropy": 5.693405961990356,
      "epoch": 2.3010309278350514,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004465154148067454,
      "loss": 5.0851,
      "mean_token_accuracy": 0.19638503938913346,
      "num_tokens": 51310207.0,
      "step": 29575
    },
    {
      "entropy": 5.601260423660278,
      "epoch": 2.301419957206769,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044649747501374336,
      "loss": 4.8714,
      "mean_token_accuracy": 0.20818833708763124,
      "num_tokens": 51319284.0,
      "step": 29580
    },
    {
      "entropy": 5.764244794845581,
      "epoch": 2.3018089865784868,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004464795326185401,
      "loss": 5.0513,
      "mean_token_accuracy": 0.19652780890464783,
      "num_tokens": 51328217.0,
      "step": 29585
    },
    {
      "entropy": 5.7103053569793705,
      "epoch": 2.3021980159502045,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004464615876214078,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2044257253408432,
      "num_tokens": 51336517.0,
      "step": 29590
    },
    {
      "entropy": 5.627273750305176,
      "epoch": 2.3025870453219217,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004464436400226188,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20412445664405823,
      "num_tokens": 51344776.0,
      "step": 29595
    },
    {
      "entropy": 5.557971334457397,
      "epoch": 2.3029760746936394,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004464256898224455,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21170520931482315,
      "num_tokens": 51353315.0,
      "step": 29600
    },
    {
      "entropy": 5.643322706222534,
      "epoch": 2.303365104065357,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004464077370211602,
      "loss": 4.9977,
      "mean_token_accuracy": 0.20326978862285613,
      "num_tokens": 51362204.0,
      "step": 29605
    },
    {
      "entropy": 5.572416639328003,
      "epoch": 2.3037541334370744,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044638978161903533,
      "loss": 4.825,
      "mean_token_accuracy": 0.20783982127904893,
      "num_tokens": 51370869.0,
      "step": 29610
    },
    {
      "entropy": 5.634156799316406,
      "epoch": 2.304143162808792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004463718236163433,
      "loss": 4.9527,
      "mean_token_accuracy": 0.20529302060604096,
      "num_tokens": 51379548.0,
      "step": 29615
    },
    {
      "entropy": 5.677594804763794,
      "epoch": 2.3045321921805098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044635386301335666,
      "loss": 5.0307,
      "mean_token_accuracy": 0.19620249122381211,
      "num_tokens": 51388095.0,
      "step": 29620
    },
    {
      "entropy": 5.676456451416016,
      "epoch": 2.304921221552227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004463358998103478,
      "loss": 4.9876,
      "mean_token_accuracy": 0.19870616495609283,
      "num_tokens": 51397195.0,
      "step": 29625
    },
    {
      "entropy": 5.715763235092163,
      "epoch": 2.3053102509239447,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004463179340075894,
      "loss": 5.0451,
      "mean_token_accuracy": 0.2000088080763817,
      "num_tokens": 51406012.0,
      "step": 29630
    },
    {
      "entropy": 5.664521884918213,
      "epoch": 2.3056992802956624,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004462999656053541,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22030634433031082,
      "num_tokens": 51414059.0,
      "step": 29635
    },
    {
      "entropy": 5.625604963302612,
      "epoch": 2.3060883096673797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044628199460391445,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20533144026994704,
      "num_tokens": 51422444.0,
      "step": 29640
    },
    {
      "entropy": 5.589230680465699,
      "epoch": 2.3064773390390974,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044626402100354307,
      "loss": 4.926,
      "mean_token_accuracy": 0.19999572932720183,
      "num_tokens": 51431858.0,
      "step": 29645
    },
    {
      "entropy": 5.615210771560669,
      "epoch": 2.306866368410815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004462460448045129,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21358368545770645,
      "num_tokens": 51440295.0,
      "step": 29650
    },
    {
      "entropy": 5.67876935005188,
      "epoch": 2.3072553977825327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044622806600709645,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20233010798692702,
      "num_tokens": 51448949.0,
      "step": 29655
    },
    {
      "entropy": 5.6314489364624025,
      "epoch": 2.30764442715425,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004462100846115667,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20391274988651276,
      "num_tokens": 51458228.0,
      "step": 29660
    },
    {
      "entropy": 5.618000316619873,
      "epoch": 2.3080334565259677,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004461921006181964,
      "loss": 4.9385,
      "mean_token_accuracy": 0.20904507786035537,
      "num_tokens": 51466348.0,
      "step": 29665
    },
    {
      "entropy": 5.608206748962402,
      "epoch": 2.3084224858976854,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004461741140272584,
      "loss": 4.891,
      "mean_token_accuracy": 0.21404290050268174,
      "num_tokens": 51474854.0,
      "step": 29670
    },
    {
      "entropy": 5.726217031478882,
      "epoch": 2.3088115152694026,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004461561248390257,
      "loss": 5.1439,
      "mean_token_accuracy": 0.19199473410844803,
      "num_tokens": 51483677.0,
      "step": 29675
    },
    {
      "entropy": 5.671559381484985,
      "epoch": 2.3092005446411203,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004461381330537713,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20263229608535765,
      "num_tokens": 51493068.0,
      "step": 29680
    },
    {
      "entropy": 5.740030002593994,
      "epoch": 2.309589574012838,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000446120138671768,
      "loss": 4.9693,
      "mean_token_accuracy": 0.20346267521381378,
      "num_tokens": 51501299.0,
      "step": 29685
    },
    {
      "entropy": 5.760293626785279,
      "epoch": 2.3099786033845557,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044610214169328913,
      "loss": 5.0282,
      "mean_token_accuracy": 0.20377151519060135,
      "num_tokens": 51509806.0,
      "step": 29690
    },
    {
      "entropy": 5.629344701766968,
      "epoch": 2.310367632756273,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004460841421186075,
      "loss": 4.9511,
      "mean_token_accuracy": 0.20478152334690095,
      "num_tokens": 51518573.0,
      "step": 29695
    },
    {
      "entropy": 5.69682502746582,
      "epoch": 2.3107566621279907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004460661399479963,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2121780261397362,
      "num_tokens": 51527370.0,
      "step": 29700
    },
    {
      "entropy": 5.661144971847534,
      "epoch": 2.3111456914997084,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044604813518172876,
      "loss": 4.97,
      "mean_token_accuracy": 0.20920703709125518,
      "num_tokens": 51536325.0,
      "step": 29705
    },
    {
      "entropy": 5.737629318237305,
      "epoch": 2.3115347208714256,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044603012782007796,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20193764120340346,
      "num_tokens": 51544447.0,
      "step": 29710
    },
    {
      "entropy": 5.644625568389893,
      "epoch": 2.3119237502431433,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004460121178633172,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20472848862409593,
      "num_tokens": 51552345.0,
      "step": 29715
    },
    {
      "entropy": 5.661270332336426,
      "epoch": 2.312312779614861,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044599410531171986,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2083461180329323,
      "num_tokens": 51560905.0,
      "step": 29720
    },
    {
      "entropy": 5.701903486251831,
      "epoch": 2.3127018089865787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000445976090165559,
      "loss": 4.9822,
      "mean_token_accuracy": 0.20884294509887696,
      "num_tokens": 51569857.0,
      "step": 29725
    },
    {
      "entropy": 5.6133476257324215,
      "epoch": 2.313090838358296,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004459580724251082,
      "loss": 4.8991,
      "mean_token_accuracy": 0.20771622955799102,
      "num_tokens": 51578921.0,
      "step": 29730
    },
    {
      "entropy": 5.629269027709961,
      "epoch": 2.3134798677300137,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044594005209064064,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21207161843776703,
      "num_tokens": 51587607.0,
      "step": 29735
    },
    {
      "entropy": 5.707592391967774,
      "epoch": 2.313868897101731,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044592202916243003,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20030743777751922,
      "num_tokens": 51597064.0,
      "step": 29740
    },
    {
      "entropy": 5.688130807876587,
      "epoch": 2.3142579264734486,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004459040036407495,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20587556362152098,
      "num_tokens": 51606063.0,
      "step": 29745
    },
    {
      "entropy": 5.758700799942017,
      "epoch": 2.3146469558451663,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004458859755258729,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1978215217590332,
      "num_tokens": 51615555.0,
      "step": 29750
    },
    {
      "entropy": 5.715987873077393,
      "epoch": 2.315035985216884,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044586794481807357,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2007552281022072,
      "num_tokens": 51624223.0,
      "step": 29755
    },
    {
      "entropy": 5.734735298156738,
      "epoch": 2.3154250145886013,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044584991151762506,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21247360706329346,
      "num_tokens": 51633216.0,
      "step": 29760
    },
    {
      "entropy": 5.647370100021362,
      "epoch": 2.315814043960319,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004458318756248011,
      "loss": 4.9113,
      "mean_token_accuracy": 0.20937916338443757,
      "num_tokens": 51641105.0,
      "step": 29765
    },
    {
      "entropy": 5.656595849990845,
      "epoch": 2.3162030733320367,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044581383713987547,
      "loss": 4.9626,
      "mean_token_accuracy": 0.20464709401130676,
      "num_tokens": 51649452.0,
      "step": 29770
    },
    {
      "entropy": 5.687713479995727,
      "epoch": 2.316592102703754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004457957960631216,
      "loss": 5.0227,
      "mean_token_accuracy": 0.19888788163661958,
      "num_tokens": 51658268.0,
      "step": 29775
    },
    {
      "entropy": 5.600648593902588,
      "epoch": 2.3169811320754716,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004457777523948134,
      "loss": 4.801,
      "mean_token_accuracy": 0.21936177015304564,
      "num_tokens": 51666465.0,
      "step": 29780
    },
    {
      "entropy": 5.59742431640625,
      "epoch": 2.3173701614471893,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004457597061352247,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2092460572719574,
      "num_tokens": 51674912.0,
      "step": 29785
    },
    {
      "entropy": 5.610591554641724,
      "epoch": 2.317759190818907,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004457416572846291,
      "loss": 4.9497,
      "mean_token_accuracy": 0.19611132740974427,
      "num_tokens": 51683496.0,
      "step": 29790
    },
    {
      "entropy": 5.714025926589966,
      "epoch": 2.3181482201906243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004457236058433008,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20570492893457412,
      "num_tokens": 51692414.0,
      "step": 29795
    },
    {
      "entropy": 5.628353071212769,
      "epoch": 2.318537249562342,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044570555181151333,
      "loss": 4.9518,
      "mean_token_accuracy": 0.20527004152536393,
      "num_tokens": 51700643.0,
      "step": 29800
    },
    {
      "entropy": 5.637428283691406,
      "epoch": 2.3189262789340597,
      "grad_norm": 1.15625,
      "learning_rate": 0.000445687495189541,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20438080728054048,
      "num_tokens": 51708882.0,
      "step": 29805
    },
    {
      "entropy": 5.7315562725067135,
      "epoch": 2.319315308305777,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044566943597765745,
      "loss": 5.0478,
      "mean_token_accuracy": 0.2012246295809746,
      "num_tokens": 51718191.0,
      "step": 29810
    },
    {
      "entropy": 5.748530101776123,
      "epoch": 2.3197043376774946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044565137417613694,
      "loss": 5.006,
      "mean_token_accuracy": 0.21281303018331527,
      "num_tokens": 51726622.0,
      "step": 29815
    },
    {
      "entropy": 5.605900621414184,
      "epoch": 2.3200933670492123,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004456333097852534,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21654877215623855,
      "num_tokens": 51734544.0,
      "step": 29820
    },
    {
      "entropy": 5.63562707901001,
      "epoch": 2.32048239642093,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044561524280528104,
      "loss": 4.9812,
      "mean_token_accuracy": 0.19827888906002045,
      "num_tokens": 51743481.0,
      "step": 29825
    },
    {
      "entropy": 5.668291759490967,
      "epoch": 2.3208714257926473,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044559717323649383,
      "loss": 4.9772,
      "mean_token_accuracy": 0.20335174202919007,
      "num_tokens": 51752365.0,
      "step": 29830
    },
    {
      "entropy": 5.809470129013062,
      "epoch": 2.321260455164365,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004455791010791662,
      "loss": 5.1298,
      "mean_token_accuracy": 0.1978054478764534,
      "num_tokens": 51760729.0,
      "step": 29835
    },
    {
      "entropy": 5.648795032501221,
      "epoch": 2.3216494845360827,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004455610263335721,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2135866940021515,
      "num_tokens": 51769250.0,
      "step": 29840
    },
    {
      "entropy": 5.667315816879272,
      "epoch": 2.3220385139078,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004455429489999859,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20472099930047988,
      "num_tokens": 51778206.0,
      "step": 29845
    },
    {
      "entropy": 5.63917727470398,
      "epoch": 2.3224275432795176,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044552486907868194,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2087121069431305,
      "num_tokens": 51786364.0,
      "step": 29850
    },
    {
      "entropy": 5.615413999557495,
      "epoch": 2.3228165726512353,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044550678656993454,
      "loss": 4.9522,
      "mean_token_accuracy": 0.20492195188999177,
      "num_tokens": 51794907.0,
      "step": 29855
    },
    {
      "entropy": 5.642756843566895,
      "epoch": 2.3232056020229526,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000445488701474018,
      "loss": 4.9016,
      "mean_token_accuracy": 0.20731325149536134,
      "num_tokens": 51803290.0,
      "step": 29860
    },
    {
      "entropy": 5.679126167297364,
      "epoch": 2.3235946313946703,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004454706137912067,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20499469041824342,
      "num_tokens": 51812065.0,
      "step": 29865
    },
    {
      "entropy": 5.641716766357422,
      "epoch": 2.323983660766388,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004454525235217753,
      "loss": 4.8641,
      "mean_token_accuracy": 0.208302541077137,
      "num_tokens": 51819811.0,
      "step": 29870
    },
    {
      "entropy": 5.638849258422852,
      "epoch": 2.324372690138105,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044543443066599807,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21464349925518036,
      "num_tokens": 51828373.0,
      "step": 29875
    },
    {
      "entropy": 5.645402526855468,
      "epoch": 2.324761719509823,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004454163352241498,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20826891660690308,
      "num_tokens": 51836922.0,
      "step": 29880
    },
    {
      "entropy": 5.726996326446534,
      "epoch": 2.3251507488815406,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044539823719650463,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2015913337469101,
      "num_tokens": 51846550.0,
      "step": 29885
    },
    {
      "entropy": 5.640140867233276,
      "epoch": 2.3255397782532583,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004453801365833376,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2092990681529045,
      "num_tokens": 51854807.0,
      "step": 29890
    },
    {
      "entropy": 5.639181280136109,
      "epoch": 2.3259288076249756,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044536203338492317,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21344771683216096,
      "num_tokens": 51863723.0,
      "step": 29895
    },
    {
      "entropy": 5.606798124313355,
      "epoch": 2.3263178369966933,
      "grad_norm": 1.125,
      "learning_rate": 0.00044534392760153596,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20845917165279387,
      "num_tokens": 51872545.0,
      "step": 29900
    },
    {
      "entropy": 5.640400981903076,
      "epoch": 2.326706866368411,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004453258192334508,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2058084636926651,
      "num_tokens": 51880669.0,
      "step": 29905
    },
    {
      "entropy": 5.695552778244019,
      "epoch": 2.327095895740128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004453077082809425,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20020557790994645,
      "num_tokens": 51888851.0,
      "step": 29910
    },
    {
      "entropy": 5.586156034469605,
      "epoch": 2.327484925111846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044528959474428575,
      "loss": 4.913,
      "mean_token_accuracy": 0.2036057949066162,
      "num_tokens": 51897502.0,
      "step": 29915
    },
    {
      "entropy": 5.63563175201416,
      "epoch": 2.3278739544835636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004452714786237555,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20563496947288512,
      "num_tokens": 51906087.0,
      "step": 29920
    },
    {
      "entropy": 5.635730838775634,
      "epoch": 2.3282629838552813,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004452533599196265,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2064405471086502,
      "num_tokens": 51914610.0,
      "step": 29925
    },
    {
      "entropy": 5.645227384567261,
      "epoch": 2.3286520132269986,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044523523863217374,
      "loss": 4.9056,
      "mean_token_accuracy": 0.20647140890359877,
      "num_tokens": 51922914.0,
      "step": 29930
    },
    {
      "entropy": 5.5913698196411135,
      "epoch": 2.3290410425987162,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004452171147616723,
      "loss": 4.8632,
      "mean_token_accuracy": 0.21464110910892487,
      "num_tokens": 51931589.0,
      "step": 29935
    },
    {
      "entropy": 5.599503660202027,
      "epoch": 2.329430071970434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000445198988308397,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21458784192800523,
      "num_tokens": 51939110.0,
      "step": 29940
    },
    {
      "entropy": 5.658889722824097,
      "epoch": 2.329819101342151,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044518085927262293,
      "loss": 4.9363,
      "mean_token_accuracy": 0.20441262871026994,
      "num_tokens": 51946850.0,
      "step": 29945
    },
    {
      "entropy": 5.687842178344726,
      "epoch": 2.330208130713869,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004451627276546252,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20744600743055344,
      "num_tokens": 51955458.0,
      "step": 29950
    },
    {
      "entropy": 5.627331018447876,
      "epoch": 2.3305971600855866,
      "grad_norm": 1.265625,
      "learning_rate": 0.000445144593454679,
      "loss": 4.9402,
      "mean_token_accuracy": 0.20958840250968933,
      "num_tokens": 51964462.0,
      "step": 29955
    },
    {
      "entropy": 5.671924495697022,
      "epoch": 2.330986189457304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004451264566730593,
      "loss": 4.9797,
      "mean_token_accuracy": 0.19699678868055343,
      "num_tokens": 51973309.0,
      "step": 29960
    },
    {
      "entropy": 5.733008098602295,
      "epoch": 2.3313752188290215,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044510831731004146,
      "loss": 5.0992,
      "mean_token_accuracy": 0.19489036947488786,
      "num_tokens": 51980938.0,
      "step": 29965
    },
    {
      "entropy": 5.630222606658935,
      "epoch": 2.3317642482007392,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004450901753659007,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21468451321125032,
      "num_tokens": 51989104.0,
      "step": 29970
    },
    {
      "entropy": 5.554630374908447,
      "epoch": 2.3321532775724565,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044507203084091215,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21130727380514144,
      "num_tokens": 51997968.0,
      "step": 29975
    },
    {
      "entropy": 5.790522289276123,
      "epoch": 2.332542306944174,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004450538837353513,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20107235312461852,
      "num_tokens": 52005972.0,
      "step": 29980
    },
    {
      "entropy": 5.747951793670654,
      "epoch": 2.332931336315892,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044503573404949343,
      "loss": 5.113,
      "mean_token_accuracy": 0.19359252750873565,
      "num_tokens": 52015178.0,
      "step": 29985
    },
    {
      "entropy": 5.669077062606812,
      "epoch": 2.3333203656876096,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004450175817836139,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20522669851779937,
      "num_tokens": 52023186.0,
      "step": 29990
    },
    {
      "entropy": 5.726346206665039,
      "epoch": 2.333709395059327,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044499942693798823,
      "loss": 5.0683,
      "mean_token_accuracy": 0.19782310128211975,
      "num_tokens": 52031432.0,
      "step": 29995
    },
    {
      "entropy": 5.627503681182861,
      "epoch": 2.3340984244310445,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004449812695128918,
      "loss": 4.8796,
      "mean_token_accuracy": 0.20966476202011108,
      "num_tokens": 52039499.0,
      "step": 30000
    },
    {
      "epoch": 2.3340984244310445,
      "eval_entropy": 5.444319304740693,
      "eval_loss": 5.0652756690979,
      "eval_mean_token_accuracy": 0.2104744389065191,
      "eval_num_tokens": 52039499.0,
      "eval_runtime": 21.7118,
      "eval_samples_per_second": 1914.075,
      "eval_steps_per_second": 239.271,
      "step": 30000
    },
    {
      "entropy": 5.6786071300506595,
      "epoch": 2.3344874538027622,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044496310950860015,
      "loss": 4.8857,
      "mean_token_accuracy": 0.217248597741127,
      "num_tokens": 52047532.0,
      "step": 30005
    },
    {
      "entropy": 5.665977573394775,
      "epoch": 2.3348764831744795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044494494692538886,
      "loss": 5.0979,
      "mean_token_accuracy": 0.19900950640439988,
      "num_tokens": 52056504.0,
      "step": 30010
    },
    {
      "entropy": 5.624105548858642,
      "epoch": 2.335265512546197,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044492678176353347,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21607339531183242,
      "num_tokens": 52065048.0,
      "step": 30015
    },
    {
      "entropy": 5.688593816757202,
      "epoch": 2.335654541917915,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044490861402330967,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20257134288549422,
      "num_tokens": 52073794.0,
      "step": 30020
    },
    {
      "entropy": 5.671488475799561,
      "epoch": 2.3360435712896326,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004448904437049931,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1986183390021324,
      "num_tokens": 52083193.0,
      "step": 30025
    },
    {
      "entropy": 5.62253885269165,
      "epoch": 2.33643260066135,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004448722708088594,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21247195601463317,
      "num_tokens": 52091297.0,
      "step": 30030
    },
    {
      "entropy": 5.5956658840179445,
      "epoch": 2.3368216300330675,
      "grad_norm": 1.125,
      "learning_rate": 0.0004448540953351844,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21328657120466232,
      "num_tokens": 52099846.0,
      "step": 30035
    },
    {
      "entropy": 5.668247175216675,
      "epoch": 2.3372106594047852,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004448359172842439,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2038309469819069,
      "num_tokens": 52108291.0,
      "step": 30040
    },
    {
      "entropy": 5.639101457595825,
      "epoch": 2.3375996887765025,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044481773665631355,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2141914486885071,
      "num_tokens": 52116419.0,
      "step": 30045
    },
    {
      "entropy": 5.705658102035523,
      "epoch": 2.33798871814822,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004447995534516695,
      "loss": 4.9247,
      "mean_token_accuracy": 0.20504148155450821,
      "num_tokens": 52124820.0,
      "step": 30050
    },
    {
      "entropy": 5.709504318237305,
      "epoch": 2.338377747519938,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044478136767058733,
      "loss": 5.0463,
      "mean_token_accuracy": 0.2011807456612587,
      "num_tokens": 52134092.0,
      "step": 30055
    },
    {
      "entropy": 5.589872646331787,
      "epoch": 2.338766776891655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004447631793133432,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21291635930538177,
      "num_tokens": 52143013.0,
      "step": 30060
    },
    {
      "entropy": 5.609470176696777,
      "epoch": 2.339155806263373,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004447449883802131,
      "loss": 4.9037,
      "mean_token_accuracy": 0.20822709202766418,
      "num_tokens": 52152115.0,
      "step": 30065
    },
    {
      "entropy": 5.7298084735870365,
      "epoch": 2.3395448356350905,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044472679487147295,
      "loss": 5.0653,
      "mean_token_accuracy": 0.19188129305839538,
      "num_tokens": 52161066.0,
      "step": 30070
    },
    {
      "entropy": 5.7496367454528805,
      "epoch": 2.339933865006808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044470859878739877,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1954355791211128,
      "num_tokens": 52169904.0,
      "step": 30075
    },
    {
      "entropy": 5.6612138748168945,
      "epoch": 2.3403228943785255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044469040012826676,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20932124704122543,
      "num_tokens": 52178179.0,
      "step": 30080
    },
    {
      "entropy": 5.660474348068237,
      "epoch": 2.340711923750243,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044467219889435304,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20955818444490432,
      "num_tokens": 52187025.0,
      "step": 30085
    },
    {
      "entropy": 5.668624544143677,
      "epoch": 2.341100953121961,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004446539950859337,
      "loss": 5.0482,
      "mean_token_accuracy": 0.19847813844680787,
      "num_tokens": 52196019.0,
      "step": 30090
    },
    {
      "entropy": 5.700398015975952,
      "epoch": 2.341489982493678,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044463578870328506,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20149759352207183,
      "num_tokens": 52204403.0,
      "step": 30095
    },
    {
      "entropy": 5.706373071670532,
      "epoch": 2.341879011865396,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004446175797466834,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20962212979793549,
      "num_tokens": 52213425.0,
      "step": 30100
    },
    {
      "entropy": 5.620306015014648,
      "epoch": 2.3422680412371135,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044459936821640485,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21124411821365358,
      "num_tokens": 52222454.0,
      "step": 30105
    },
    {
      "entropy": 5.607381200790405,
      "epoch": 2.3426570706088308,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004445811541127258,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20756374299526215,
      "num_tokens": 52230994.0,
      "step": 30110
    },
    {
      "entropy": 5.647941398620605,
      "epoch": 2.3430460999805485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044456293743592274,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2015425145626068,
      "num_tokens": 52240532.0,
      "step": 30115
    },
    {
      "entropy": 5.706516933441162,
      "epoch": 2.343435129352266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044454471818627196,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20278163254261017,
      "num_tokens": 52249461.0,
      "step": 30120
    },
    {
      "entropy": 5.643615436553955,
      "epoch": 2.343824158723984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044452649636404995,
      "loss": 4.909,
      "mean_token_accuracy": 0.21659154891967775,
      "num_tokens": 52258982.0,
      "step": 30125
    },
    {
      "entropy": 5.643262481689453,
      "epoch": 2.344213188095701,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004445082719695332,
      "loss": 4.97,
      "mean_token_accuracy": 0.2056376412510872,
      "num_tokens": 52267856.0,
      "step": 30130
    },
    {
      "entropy": 5.588414287567138,
      "epoch": 2.344602217467419,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044449004500299815,
      "loss": 4.8659,
      "mean_token_accuracy": 0.21830131709575654,
      "num_tokens": 52275936.0,
      "step": 30135
    },
    {
      "entropy": 5.652890920639038,
      "epoch": 2.3449912468391365,
      "grad_norm": 1.25,
      "learning_rate": 0.00044447181546472146,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20313683152198792,
      "num_tokens": 52284323.0,
      "step": 30140
    },
    {
      "entropy": 5.669830656051635,
      "epoch": 2.3453802762108538,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044445358335497975,
      "loss": 4.958,
      "mean_token_accuracy": 0.20942258834838867,
      "num_tokens": 52293155.0,
      "step": 30145
    },
    {
      "entropy": 5.6569417953491214,
      "epoch": 2.3457693055825715,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004444353486740496,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20260220021009445,
      "num_tokens": 52302613.0,
      "step": 30150
    },
    {
      "entropy": 5.629034996032715,
      "epoch": 2.346158334954289,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004444171114222078,
      "loss": 4.892,
      "mean_token_accuracy": 0.2092030167579651,
      "num_tokens": 52311418.0,
      "step": 30155
    },
    {
      "entropy": 5.650941228866577,
      "epoch": 2.346547364326007,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044439887159973096,
      "loss": 4.9699,
      "mean_token_accuracy": 0.2035482719540596,
      "num_tokens": 52319688.0,
      "step": 30160
    },
    {
      "entropy": 5.607905149459839,
      "epoch": 2.346936393697724,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004443806292068958,
      "loss": 4.9062,
      "mean_token_accuracy": 0.20824539214372634,
      "num_tokens": 52328373.0,
      "step": 30165
    },
    {
      "entropy": 5.655202198028564,
      "epoch": 2.347325423069442,
      "grad_norm": 1.25,
      "learning_rate": 0.0004443623842439792,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2034092739224434,
      "num_tokens": 52336822.0,
      "step": 30170
    },
    {
      "entropy": 5.717740154266357,
      "epoch": 2.347714452441159,
      "grad_norm": 1.15625,
      "learning_rate": 0.000444344136711258,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20717599987983704,
      "num_tokens": 52345420.0,
      "step": 30175
    },
    {
      "entropy": 5.69046573638916,
      "epoch": 2.3481034818128768,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044432588660900905,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20402906686067582,
      "num_tokens": 52353959.0,
      "step": 30180
    },
    {
      "entropy": 5.588909435272217,
      "epoch": 2.3484925111845945,
      "grad_norm": 1.25,
      "learning_rate": 0.0004443076339375093,
      "loss": 4.9374,
      "mean_token_accuracy": 0.20628469437360764,
      "num_tokens": 52362904.0,
      "step": 30185
    },
    {
      "entropy": 5.642540836334229,
      "epoch": 2.348881540556312,
      "grad_norm": 1.125,
      "learning_rate": 0.00044428937869703575,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20060611367225648,
      "num_tokens": 52371587.0,
      "step": 30190
    },
    {
      "entropy": 5.624602794647217,
      "epoch": 2.3492705699280294,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044427112088786525,
      "loss": 4.8125,
      "mean_token_accuracy": 0.21473609358072282,
      "num_tokens": 52380603.0,
      "step": 30195
    },
    {
      "entropy": 5.6211175441741945,
      "epoch": 2.349659599299747,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004442528605102749,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20671709775924682,
      "num_tokens": 52389477.0,
      "step": 30200
    },
    {
      "entropy": 5.630314207077026,
      "epoch": 2.350048628671465,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004442345975645418,
      "loss": 4.9163,
      "mean_token_accuracy": 0.20832363963127137,
      "num_tokens": 52398039.0,
      "step": 30205
    },
    {
      "entropy": 5.656469392776489,
      "epoch": 2.350437658043182,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004442163320509431,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21114147007465361,
      "num_tokens": 52406358.0,
      "step": 30210
    },
    {
      "entropy": 5.609855699539184,
      "epoch": 2.3508266874148998,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441980639697558,
      "loss": 4.9707,
      "mean_token_accuracy": 0.20978487581014632,
      "num_tokens": 52414873.0,
      "step": 30215
    },
    {
      "entropy": 5.685354566574096,
      "epoch": 2.3512157167866174,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004441797933212573,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20147286355495453,
      "num_tokens": 52423848.0,
      "step": 30220
    },
    {
      "entropy": 5.658399772644043,
      "epoch": 2.351604746158335,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004441615201057247,
      "loss": 5.0003,
      "mean_token_accuracy": 0.20537327229976654,
      "num_tokens": 52432153.0,
      "step": 30225
    },
    {
      "entropy": 5.670348072052002,
      "epoch": 2.3519937755300524,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441432443234352,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20375650823116304,
      "num_tokens": 52440810.0,
      "step": 30230
    },
    {
      "entropy": 5.643746376037598,
      "epoch": 2.35238280490177,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044412496597466626,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21075344532728196,
      "num_tokens": 52449544.0,
      "step": 30235
    },
    {
      "entropy": 5.674884033203125,
      "epoch": 2.352771834273488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004441066850596951,
      "loss": 4.972,
      "mean_token_accuracy": 0.2112704262137413,
      "num_tokens": 52457903.0,
      "step": 30240
    },
    {
      "entropy": 5.6133363246917725,
      "epoch": 2.353160863645205,
      "grad_norm": 1.25,
      "learning_rate": 0.0004440884015787992,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21285447478294373,
      "num_tokens": 52466336.0,
      "step": 30245
    },
    {
      "entropy": 5.722786998748779,
      "epoch": 2.3535498930169227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004440701155322558,
      "loss": 5.0656,
      "mean_token_accuracy": 0.19681797325611114,
      "num_tokens": 52475535.0,
      "step": 30250
    },
    {
      "entropy": 5.696746444702148,
      "epoch": 2.3539389223886404,
      "grad_norm": 1.125,
      "learning_rate": 0.0004440518269203427,
      "loss": 4.9329,
      "mean_token_accuracy": 0.20951234847307204,
      "num_tokens": 52483929.0,
      "step": 30255
    },
    {
      "entropy": 5.660293912887573,
      "epoch": 2.354327951760358,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044403353574333715,
      "loss": 4.9199,
      "mean_token_accuracy": 0.20590932965278624,
      "num_tokens": 52493559.0,
      "step": 30260
    },
    {
      "entropy": 5.593469333648682,
      "epoch": 2.3547169811320754,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044401524200151667,
      "loss": 4.902,
      "mean_token_accuracy": 0.2048324391245842,
      "num_tokens": 52502256.0,
      "step": 30265
    },
    {
      "entropy": 5.735697507858276,
      "epoch": 2.355106010503793,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044399694569515895,
      "loss": 5.0752,
      "mean_token_accuracy": 0.2027621239423752,
      "num_tokens": 52510572.0,
      "step": 30270
    },
    {
      "entropy": 5.714899110794067,
      "epoch": 2.355495039875511,
      "grad_norm": 1.125,
      "learning_rate": 0.00044397864682454156,
      "loss": 4.9578,
      "mean_token_accuracy": 0.20717235654592514,
      "num_tokens": 52518908.0,
      "step": 30275
    },
    {
      "entropy": 5.716857385635376,
      "epoch": 2.355884069247228,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004439603453899421,
      "loss": 5.1162,
      "mean_token_accuracy": 0.195631705224514,
      "num_tokens": 52527471.0,
      "step": 30280
    },
    {
      "entropy": 5.622205352783203,
      "epoch": 2.3562730986189457,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004439420413916384,
      "loss": 4.939,
      "mean_token_accuracy": 0.2113445907831192,
      "num_tokens": 52536873.0,
      "step": 30285
    },
    {
      "entropy": 5.759760141372681,
      "epoch": 2.3566621279906634,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044392373482990817,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20190776139497757,
      "num_tokens": 52545322.0,
      "step": 30290
    },
    {
      "entropy": 5.685457944869995,
      "epoch": 2.3570511573623807,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004439054257050291,
      "loss": 4.9794,
      "mean_token_accuracy": 0.2067481592297554,
      "num_tokens": 52553870.0,
      "step": 30295
    },
    {
      "entropy": 5.611184120178223,
      "epoch": 2.3574401867340984,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004438871140172789,
      "loss": 5.0069,
      "mean_token_accuracy": 0.20768812596797942,
      "num_tokens": 52563362.0,
      "step": 30300
    },
    {
      "entropy": 5.629876804351807,
      "epoch": 2.357829216105816,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004438687997669357,
      "loss": 4.9618,
      "mean_token_accuracy": 0.20594909340143203,
      "num_tokens": 52571456.0,
      "step": 30305
    },
    {
      "entropy": 5.602367687225342,
      "epoch": 2.3582182454775333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044385048295427724,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2070243775844574,
      "num_tokens": 52580555.0,
      "step": 30310
    },
    {
      "entropy": 5.733325433731079,
      "epoch": 2.358607274849251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004438321635795814,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19958677589893342,
      "num_tokens": 52588641.0,
      "step": 30315
    },
    {
      "entropy": 5.837975072860718,
      "epoch": 2.3589963042209687,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004438138416431262,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19127443581819534,
      "num_tokens": 52597501.0,
      "step": 30320
    },
    {
      "entropy": 5.743187475204468,
      "epoch": 2.3593853335926864,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044379551714518966,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21107970625162126,
      "num_tokens": 52605639.0,
      "step": 30325
    },
    {
      "entropy": 5.664077949523926,
      "epoch": 2.3597743629644037,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044377719008604984,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2055906116962433,
      "num_tokens": 52614120.0,
      "step": 30330
    },
    {
      "entropy": 5.66447958946228,
      "epoch": 2.3601633923361214,
      "grad_norm": 1.25,
      "learning_rate": 0.0004437588604659848,
      "loss": 4.9955,
      "mean_token_accuracy": 0.210799939930439,
      "num_tokens": 52622782.0,
      "step": 30335
    },
    {
      "entropy": 5.7443195343017575,
      "epoch": 2.360552421707839,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004437405282852726,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20217103362083436,
      "num_tokens": 52631563.0,
      "step": 30340
    },
    {
      "entropy": 5.670605230331421,
      "epoch": 2.3609414510795563,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044372219354419154,
      "loss": 4.9028,
      "mean_token_accuracy": 0.21183837950229645,
      "num_tokens": 52639776.0,
      "step": 30345
    },
    {
      "entropy": 5.654925060272217,
      "epoch": 2.361330480451274,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004437038562430197,
      "loss": 4.9332,
      "mean_token_accuracy": 0.20609361678361893,
      "num_tokens": 52648361.0,
      "step": 30350
    },
    {
      "entropy": 5.602179718017578,
      "epoch": 2.3617195098229917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044368551638203536,
      "loss": 4.9111,
      "mean_token_accuracy": 0.20823199152946473,
      "num_tokens": 52656609.0,
      "step": 30355
    },
    {
      "entropy": 5.692745113372803,
      "epoch": 2.3621085391947094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004436671739615168,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2051967278122902,
      "num_tokens": 52665294.0,
      "step": 30360
    },
    {
      "entropy": 5.699210119247437,
      "epoch": 2.3624975685664267,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044364882898174235,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21391389071941375,
      "num_tokens": 52673647.0,
      "step": 30365
    },
    {
      "entropy": 5.644039678573608,
      "epoch": 2.3628865979381444,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004436304814429903,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20067928582429886,
      "num_tokens": 52683010.0,
      "step": 30370
    },
    {
      "entropy": 5.654058361053467,
      "epoch": 2.363275627309862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004436121313455391,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20631351619958876,
      "num_tokens": 52692045.0,
      "step": 30375
    },
    {
      "entropy": 5.66507511138916,
      "epoch": 2.3636646566815793,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004435937786896673,
      "loss": 4.8217,
      "mean_token_accuracy": 0.21593714207410813,
      "num_tokens": 52700246.0,
      "step": 30380
    },
    {
      "entropy": 5.689355087280274,
      "epoch": 2.364053686053297,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044357542347565323,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20187070369720458,
      "num_tokens": 52708748.0,
      "step": 30385
    },
    {
      "entropy": 5.592563676834106,
      "epoch": 2.3644427154250147,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004435570657037753,
      "loss": 4.9158,
      "mean_token_accuracy": 0.20747235268354416,
      "num_tokens": 52717586.0,
      "step": 30390
    },
    {
      "entropy": 5.614567470550537,
      "epoch": 2.364831744796732,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004435387053743123,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2057468995451927,
      "num_tokens": 52726301.0,
      "step": 30395
    },
    {
      "entropy": 5.669002962112427,
      "epoch": 2.3652207741684497,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044352034248754265,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2103371039032936,
      "num_tokens": 52734082.0,
      "step": 30400
    },
    {
      "entropy": 5.7148209571838375,
      "epoch": 2.3656098035401674,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000443501977043745,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2054022178053856,
      "num_tokens": 52742379.0,
      "step": 30405
    },
    {
      "entropy": 5.64781551361084,
      "epoch": 2.3659988329118846,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004434836090431981,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21018406003713608,
      "num_tokens": 52751339.0,
      "step": 30410
    },
    {
      "entropy": 5.602663230895996,
      "epoch": 2.3663878622836023,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004434652384861806,
      "loss": 4.9012,
      "mean_token_accuracy": 0.20308338850736618,
      "num_tokens": 52760665.0,
      "step": 30415
    },
    {
      "entropy": 5.6194983959198,
      "epoch": 2.36677689165532,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004434468653729712,
      "loss": 4.9166,
      "mean_token_accuracy": 0.21178395599126815,
      "num_tokens": 52769650.0,
      "step": 30420
    },
    {
      "entropy": 5.728986597061157,
      "epoch": 2.3671659210270377,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044342848970384876,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2053503543138504,
      "num_tokens": 52779258.0,
      "step": 30425
    },
    {
      "entropy": 5.743943595886231,
      "epoch": 2.367554950398755,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044341011147909206,
      "loss": 4.9885,
      "mean_token_accuracy": 0.205703067779541,
      "num_tokens": 52788067.0,
      "step": 30430
    },
    {
      "entropy": 5.650100517272949,
      "epoch": 2.3679439797704727,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044339173069897996,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20623218417167663,
      "num_tokens": 52797135.0,
      "step": 30435
    },
    {
      "entropy": 5.615898323059082,
      "epoch": 2.3683330091421904,
      "grad_norm": 1.125,
      "learning_rate": 0.00044337334736379143,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2009732633829117,
      "num_tokens": 52805854.0,
      "step": 30440
    },
    {
      "entropy": 5.723448181152344,
      "epoch": 2.3687220385139076,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004433549614738053,
      "loss": 4.9934,
      "mean_token_accuracy": 0.19996853172779083,
      "num_tokens": 52814711.0,
      "step": 30445
    },
    {
      "entropy": 5.717904090881348,
      "epoch": 2.3691110678856253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044333657302930056,
      "loss": 4.9195,
      "mean_token_accuracy": 0.20715565830469132,
      "num_tokens": 52824399.0,
      "step": 30450
    },
    {
      "entropy": 5.621094799041748,
      "epoch": 2.369500097257343,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004433181820305564,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21408006995916368,
      "num_tokens": 52833089.0,
      "step": 30455
    },
    {
      "entropy": 5.652538013458252,
      "epoch": 2.3698891266290607,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044329978847785156,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20665470957756044,
      "num_tokens": 52841546.0,
      "step": 30460
    },
    {
      "entropy": 5.671103525161743,
      "epoch": 2.370278156000778,
      "grad_norm": 1.125,
      "learning_rate": 0.0004432813923714654,
      "loss": 4.9819,
      "mean_token_accuracy": 0.19868847578763962,
      "num_tokens": 52851011.0,
      "step": 30465
    },
    {
      "entropy": 5.673946046829224,
      "epoch": 2.3706671853724957,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044326299371167695,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20656031668186187,
      "num_tokens": 52859276.0,
      "step": 30470
    },
    {
      "entropy": 5.764700937271118,
      "epoch": 2.3710562147442134,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044324459249876536,
      "loss": 5.1552,
      "mean_token_accuracy": 0.19544880092144012,
      "num_tokens": 52867759.0,
      "step": 30475
    },
    {
      "entropy": 5.6534895420074465,
      "epoch": 2.3714452441159306,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004432261887330099,
      "loss": 4.9321,
      "mean_token_accuracy": 0.20658342391252518,
      "num_tokens": 52876275.0,
      "step": 30480
    },
    {
      "entropy": 5.5898336410522464,
      "epoch": 2.3718342734876483,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004432077824146898,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2036968305706978,
      "num_tokens": 52884768.0,
      "step": 30485
    },
    {
      "entropy": 5.735194396972656,
      "epoch": 2.372223302859366,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004431893735440843,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20098817497491836,
      "num_tokens": 52893040.0,
      "step": 30490
    },
    {
      "entropy": 5.70246434211731,
      "epoch": 2.3726123322310833,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004431709621214727,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2072102516889572,
      "num_tokens": 52901411.0,
      "step": 30495
    },
    {
      "entropy": 5.579111766815186,
      "epoch": 2.373001361602801,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044315254814713455,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21395539194345475,
      "num_tokens": 52909751.0,
      "step": 30500
    },
    {
      "entropy": 5.600125932693482,
      "epoch": 2.3733903909745186,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044313413162134894,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2086743801832199,
      "num_tokens": 52919022.0,
      "step": 30505
    },
    {
      "entropy": 5.706230688095093,
      "epoch": 2.373779420346236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004431157125443958,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2137885183095932,
      "num_tokens": 52927927.0,
      "step": 30510
    },
    {
      "entropy": 5.723325490951538,
      "epoch": 2.3741684497179536,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000443097290916554,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2002118319272995,
      "num_tokens": 52936170.0,
      "step": 30515
    },
    {
      "entropy": 5.623728513717651,
      "epoch": 2.3745574790896713,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004430788667381035,
      "loss": 4.9738,
      "mean_token_accuracy": 0.20719371289014815,
      "num_tokens": 52946007.0,
      "step": 30520
    },
    {
      "entropy": 5.652911758422851,
      "epoch": 2.374946508461389,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044306044000932374,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2021512970328331,
      "num_tokens": 52955286.0,
      "step": 30525
    },
    {
      "entropy": 5.6979272842407225,
      "epoch": 2.3753355378331062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430420107304943,
      "loss": 4.9519,
      "mean_token_accuracy": 0.20396287143230438,
      "num_tokens": 52963608.0,
      "step": 30530
    },
    {
      "entropy": 5.663862562179565,
      "epoch": 2.375724567204824,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044302357890189475,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20785980373620988,
      "num_tokens": 52973068.0,
      "step": 30535
    },
    {
      "entropy": 5.660850906372071,
      "epoch": 2.3761135965765416,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004430051445238049,
      "loss": 4.9782,
      "mean_token_accuracy": 0.20919410586357118,
      "num_tokens": 52981571.0,
      "step": 30540
    },
    {
      "entropy": 5.66081657409668,
      "epoch": 2.376502625948259,
      "grad_norm": 1.25,
      "learning_rate": 0.00044298670759650426,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20958238542079927,
      "num_tokens": 52990298.0,
      "step": 30545
    },
    {
      "entropy": 5.6958517074584964,
      "epoch": 2.3768916553199766,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004429682681202728,
      "loss": 5.012,
      "mean_token_accuracy": 0.20475508272647858,
      "num_tokens": 52999270.0,
      "step": 30550
    },
    {
      "entropy": 5.665611314773559,
      "epoch": 2.3772806846916943,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044294982609539027,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20332012325525284,
      "num_tokens": 53008278.0,
      "step": 30555
    },
    {
      "entropy": 5.707945156097412,
      "epoch": 2.377669714063412,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004429313815221363,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20458045601844788,
      "num_tokens": 53016992.0,
      "step": 30560
    },
    {
      "entropy": 5.712107086181641,
      "epoch": 2.3780587434351292,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044291293440079107,
      "loss": 4.9737,
      "mean_token_accuracy": 0.20201164931058885,
      "num_tokens": 53025470.0,
      "step": 30565
    },
    {
      "entropy": 5.652855348587036,
      "epoch": 2.378447772806847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004428944847316342,
      "loss": 4.9933,
      "mean_token_accuracy": 0.2060358628630638,
      "num_tokens": 53034588.0,
      "step": 30570
    },
    {
      "entropy": 5.648515415191651,
      "epoch": 2.3788368021785646,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004428760325149458,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21969860345125197,
      "num_tokens": 53043312.0,
      "step": 30575
    },
    {
      "entropy": 5.692320394515991,
      "epoch": 2.379225831550282,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044285757775100584,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2106770619750023,
      "num_tokens": 53051242.0,
      "step": 30580
    },
    {
      "entropy": 5.6462421894073485,
      "epoch": 2.3796148609219996,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044283912044009436,
      "loss": 4.9624,
      "mean_token_accuracy": 0.20744537562131882,
      "num_tokens": 53060595.0,
      "step": 30585
    },
    {
      "entropy": 5.678337907791137,
      "epoch": 2.3800038902937173,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044282066058249123,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2067235067486763,
      "num_tokens": 53068920.0,
      "step": 30590
    },
    {
      "entropy": 5.641405010223389,
      "epoch": 2.380392919665435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004428021981784768,
      "loss": 4.981,
      "mean_token_accuracy": 0.2094552829861641,
      "num_tokens": 53077161.0,
      "step": 30595
    },
    {
      "entropy": 5.6245434284210205,
      "epoch": 2.3807819490371522,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044278373322833106,
      "loss": 5.0001,
      "mean_token_accuracy": 0.20450412333011628,
      "num_tokens": 53085959.0,
      "step": 30600
    },
    {
      "entropy": 5.69985146522522,
      "epoch": 2.38117097840887,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044276526573233416,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2057655557990074,
      "num_tokens": 53094117.0,
      "step": 30605
    },
    {
      "entropy": 5.679349088668824,
      "epoch": 2.381560007780587,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004427467956907664,
      "loss": 5.1018,
      "mean_token_accuracy": 0.1885227456688881,
      "num_tokens": 53103122.0,
      "step": 30610
    },
    {
      "entropy": 5.649148321151733,
      "epoch": 2.381949037152305,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044272832310390814,
      "loss": 4.9567,
      "mean_token_accuracy": 0.20275744050741196,
      "num_tokens": 53111780.0,
      "step": 30615
    },
    {
      "entropy": 5.741524076461792,
      "epoch": 2.3823380665240226,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004427098479720394,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2011920064687729,
      "num_tokens": 53120146.0,
      "step": 30620
    },
    {
      "entropy": 5.685014152526856,
      "epoch": 2.3827270958957403,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044269137029544073,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2088218078017235,
      "num_tokens": 53129431.0,
      "step": 30625
    },
    {
      "entropy": 5.6351847648620605,
      "epoch": 2.3831161252674575,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004426728900743924,
      "loss": 4.9518,
      "mean_token_accuracy": 0.20892640352249145,
      "num_tokens": 53138324.0,
      "step": 30630
    },
    {
      "entropy": 5.647952604293823,
      "epoch": 2.3835051546391752,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044265440730917484,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2047274276614189,
      "num_tokens": 53147249.0,
      "step": 30635
    },
    {
      "entropy": 5.704998445510864,
      "epoch": 2.383894184010893,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044263592200006845,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19914232343435287,
      "num_tokens": 53156126.0,
      "step": 30640
    },
    {
      "entropy": 5.61530876159668,
      "epoch": 2.38428321338261,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044261743414735383,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21479441076517106,
      "num_tokens": 53164959.0,
      "step": 30645
    },
    {
      "entropy": 5.7180884838104244,
      "epoch": 2.384672242754328,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004425989437513114,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20787761062383653,
      "num_tokens": 53173885.0,
      "step": 30650
    },
    {
      "entropy": 5.667928981781006,
      "epoch": 2.3850612721260456,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004425804508122218,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20738785415887834,
      "num_tokens": 53182618.0,
      "step": 30655
    },
    {
      "entropy": 5.608526420593262,
      "epoch": 2.3854503014977633,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044256195533036566,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20353079885244368,
      "num_tokens": 53191130.0,
      "step": 30660
    },
    {
      "entropy": 5.673959255218506,
      "epoch": 2.3858393308694805,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044254345730602345,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2100020706653595,
      "num_tokens": 53199480.0,
      "step": 30665
    },
    {
      "entropy": 5.674824523925781,
      "epoch": 2.386228360241198,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004425249567394759,
      "loss": 5.0603,
      "mean_token_accuracy": 0.20057784467935563,
      "num_tokens": 53207912.0,
      "step": 30670
    },
    {
      "entropy": 5.738715600967407,
      "epoch": 2.386617389612916,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044250645363100386,
      "loss": 5.0751,
      "mean_token_accuracy": 0.19629358649253845,
      "num_tokens": 53216905.0,
      "step": 30675
    },
    {
      "entropy": 5.71388840675354,
      "epoch": 2.387006418984633,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044248794798088805,
      "loss": 4.9973,
      "mean_token_accuracy": 0.20913136154413223,
      "num_tokens": 53225657.0,
      "step": 30680
    },
    {
      "entropy": 5.692236280441284,
      "epoch": 2.387395448356351,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004424694397894091,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20300091654062272,
      "num_tokens": 53234637.0,
      "step": 30685
    },
    {
      "entropy": 5.619132614135742,
      "epoch": 2.3877844777280686,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044245092905684794,
      "loss": 5.0257,
      "mean_token_accuracy": 0.19566274136304856,
      "num_tokens": 53243047.0,
      "step": 30690
    },
    {
      "entropy": 5.693881702423096,
      "epoch": 2.3881735070997863,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044243241578348553,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20847547203302383,
      "num_tokens": 53251955.0,
      "step": 30695
    },
    {
      "entropy": 5.753620052337647,
      "epoch": 2.3885625364715035,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004424138999696027,
      "loss": 5.0678,
      "mean_token_accuracy": 0.19815507382154465,
      "num_tokens": 53260745.0,
      "step": 30700
    },
    {
      "entropy": 5.734772109985352,
      "epoch": 2.388951565843221,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004423953816154804,
      "loss": 5.0909,
      "mean_token_accuracy": 0.1996892660856247,
      "num_tokens": 53270324.0,
      "step": 30705
    },
    {
      "entropy": 5.623935508728027,
      "epoch": 2.389340595214939,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044237686072139967,
      "loss": 4.9398,
      "mean_token_accuracy": 0.20358007699251174,
      "num_tokens": 53279785.0,
      "step": 30710
    },
    {
      "entropy": 5.685362148284912,
      "epoch": 2.389729624586656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004423583372876414,
      "loss": 5.0399,
      "mean_token_accuracy": 0.1991894066333771,
      "num_tokens": 53289685.0,
      "step": 30715
    },
    {
      "entropy": 5.679453325271607,
      "epoch": 2.390118653958374,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044233981131448677,
      "loss": 4.9806,
      "mean_token_accuracy": 0.19991449415683746,
      "num_tokens": 53298165.0,
      "step": 30720
    },
    {
      "entropy": 5.706236934661865,
      "epoch": 2.3905076833300916,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044232128280221683,
      "loss": 5.0747,
      "mean_token_accuracy": 0.195381698012352,
      "num_tokens": 53307009.0,
      "step": 30725
    },
    {
      "entropy": 5.640836954116821,
      "epoch": 2.390896712701809,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004423027517511128,
      "loss": 4.9771,
      "mean_token_accuracy": 0.19873403161764144,
      "num_tokens": 53315686.0,
      "step": 30730
    },
    {
      "entropy": 5.721861267089844,
      "epoch": 2.3912857420735265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044228421816145573,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20062563419342042,
      "num_tokens": 53325261.0,
      "step": 30735
    },
    {
      "entropy": 5.68454942703247,
      "epoch": 2.391674771445244,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044226568203352694,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21012009382247926,
      "num_tokens": 53333492.0,
      "step": 30740
    },
    {
      "entropy": 5.6638720512390135,
      "epoch": 2.3920638008169615,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044224714336760766,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20526371151208878,
      "num_tokens": 53342205.0,
      "step": 30745
    },
    {
      "entropy": 5.703261137008667,
      "epoch": 2.392452830188679,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004422286021639792,
      "loss": 5.093,
      "mean_token_accuracy": 0.19288440197706222,
      "num_tokens": 53350564.0,
      "step": 30750
    },
    {
      "entropy": 5.615387582778931,
      "epoch": 2.392841859560397,
      "grad_norm": 1.125,
      "learning_rate": 0.0004422100584229228,
      "loss": 4.8819,
      "mean_token_accuracy": 0.20881712883710862,
      "num_tokens": 53359330.0,
      "step": 30755
    },
    {
      "entropy": 5.601603555679321,
      "epoch": 2.3932308889321146,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044219151214472,
      "loss": 4.96,
      "mean_token_accuracy": 0.20350924432277678,
      "num_tokens": 53367841.0,
      "step": 30760
    },
    {
      "entropy": 5.657927083969116,
      "epoch": 2.393619918303832,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004421729633296521,
      "loss": 5.0639,
      "mean_token_accuracy": 0.19894129484891893,
      "num_tokens": 53376108.0,
      "step": 30765
    },
    {
      "entropy": 5.691621971130371,
      "epoch": 2.3940089476755495,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044215441197800054,
      "loss": 4.9263,
      "mean_token_accuracy": 0.20938316583633423,
      "num_tokens": 53384637.0,
      "step": 30770
    },
    {
      "entropy": 5.641753721237182,
      "epoch": 2.394397977047267,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044213585809004685,
      "loss": 4.8731,
      "mean_token_accuracy": 0.21489685773849487,
      "num_tokens": 53393874.0,
      "step": 30775
    },
    {
      "entropy": 5.713113737106323,
      "epoch": 2.3947870064189845,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004421173016660725,
      "loss": 5.0892,
      "mean_token_accuracy": 0.2009956121444702,
      "num_tokens": 53402171.0,
      "step": 30780
    },
    {
      "entropy": 5.68362979888916,
      "epoch": 2.395176035790702,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004420987427063592,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20414253771305085,
      "num_tokens": 53410889.0,
      "step": 30785
    },
    {
      "entropy": 5.728204393386841,
      "epoch": 2.39556506516242,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044208018121118833,
      "loss": 4.9993,
      "mean_token_accuracy": 0.20620881170034408,
      "num_tokens": 53418943.0,
      "step": 30790
    },
    {
      "entropy": 5.64262056350708,
      "epoch": 2.3959540945341375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044206161718084164,
      "loss": 4.889,
      "mean_token_accuracy": 0.2118679851293564,
      "num_tokens": 53427997.0,
      "step": 30795
    },
    {
      "entropy": 5.603597784042359,
      "epoch": 2.396343123905855,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004420430506156009,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2154457911849022,
      "num_tokens": 53436403.0,
      "step": 30800
    },
    {
      "entropy": 5.71500415802002,
      "epoch": 2.3967321532775725,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044202448151574764,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20594595819711686,
      "num_tokens": 53444690.0,
      "step": 30805
    },
    {
      "entropy": 5.71856803894043,
      "epoch": 2.39712118264929,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004420059098815638,
      "loss": 4.9805,
      "mean_token_accuracy": 0.20345345735549927,
      "num_tokens": 53453807.0,
      "step": 30810
    },
    {
      "entropy": 5.676197338104248,
      "epoch": 2.3975102120210074,
      "grad_norm": 1.125,
      "learning_rate": 0.0004419873357133311,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20212040096521378,
      "num_tokens": 53462702.0,
      "step": 30815
    },
    {
      "entropy": 5.689618158340454,
      "epoch": 2.397899241392725,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004419687590113313,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1965423196554184,
      "num_tokens": 53471851.0,
      "step": 30820
    },
    {
      "entropy": 5.669615268707275,
      "epoch": 2.398288270764443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044195017977584637,
      "loss": 4.9909,
      "mean_token_accuracy": 0.205000664293766,
      "num_tokens": 53480844.0,
      "step": 30825
    },
    {
      "entropy": 5.660860157012939,
      "epoch": 2.39867730013616,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044193159800715823,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20751722007989884,
      "num_tokens": 53489600.0,
      "step": 30830
    },
    {
      "entropy": 5.655732583999634,
      "epoch": 2.399066329507878,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044191301370554874,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21011210530996322,
      "num_tokens": 53498008.0,
      "step": 30835
    },
    {
      "entropy": 5.661489582061767,
      "epoch": 2.3994553588795955,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044189442687129994,
      "loss": 4.9387,
      "mean_token_accuracy": 0.20694252997636794,
      "num_tokens": 53506769.0,
      "step": 30840
    },
    {
      "entropy": 5.721233749389649,
      "epoch": 2.3998443882513127,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004418758375046939,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21174127459526063,
      "num_tokens": 53515311.0,
      "step": 30845
    },
    {
      "entropy": 5.618696975708008,
      "epoch": 2.4002334176230304,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044185724560601267,
      "loss": 4.88,
      "mean_token_accuracy": 0.21530278772115707,
      "num_tokens": 53523644.0,
      "step": 30850
    },
    {
      "entropy": 5.644580745697022,
      "epoch": 2.400622446994748,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004418386511755382,
      "loss": 4.9594,
      "mean_token_accuracy": 0.20411509573459624,
      "num_tokens": 53532171.0,
      "step": 30855
    },
    {
      "entropy": 5.686641788482666,
      "epoch": 2.401011476366466,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004418200542135528,
      "loss": 4.965,
      "mean_token_accuracy": 0.19877372682094574,
      "num_tokens": 53540389.0,
      "step": 30860
    },
    {
      "entropy": 5.701204442977906,
      "epoch": 2.401400505738183,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004418014547203386,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21128827333450317,
      "num_tokens": 53548473.0,
      "step": 30865
    },
    {
      "entropy": 5.664282274246216,
      "epoch": 2.401789535109901,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004417828526961778,
      "loss": 5.0085,
      "mean_token_accuracy": 0.19700499773025512,
      "num_tokens": 53557149.0,
      "step": 30870
    },
    {
      "entropy": 5.658229684829712,
      "epoch": 2.4021785644816185,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044176424814135266,
      "loss": 4.9265,
      "mean_token_accuracy": 0.20826603174209596,
      "num_tokens": 53565385.0,
      "step": 30875
    },
    {
      "entropy": 5.576272439956665,
      "epoch": 2.4025675938533357,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004417456410561455,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21350942850112914,
      "num_tokens": 53573458.0,
      "step": 30880
    },
    {
      "entropy": 5.628418493270874,
      "epoch": 2.4029566232250534,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004417270314408387,
      "loss": 5.084,
      "mean_token_accuracy": 0.19646272659301758,
      "num_tokens": 53582465.0,
      "step": 30885
    },
    {
      "entropy": 5.6883800506591795,
      "epoch": 2.403345652596771,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044170841929571454,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20823460519313813,
      "num_tokens": 53591028.0,
      "step": 30890
    },
    {
      "entropy": 5.762244367599488,
      "epoch": 2.403734681968489,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044168980462105543,
      "loss": 5.094,
      "mean_token_accuracy": 0.1965422660112381,
      "num_tokens": 53600401.0,
      "step": 30895
    },
    {
      "entropy": 5.762526988983154,
      "epoch": 2.404123711340206,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004416711874171439,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19669243693351746,
      "num_tokens": 53608605.0,
      "step": 30900
    },
    {
      "entropy": 5.666051626205444,
      "epoch": 2.404512740711924,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004416525676842624,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2089785650372505,
      "num_tokens": 53617499.0,
      "step": 30905
    },
    {
      "entropy": 5.680056190490722,
      "epoch": 2.4049017700836415,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004416339454226933,
      "loss": 4.9372,
      "mean_token_accuracy": 0.208861742913723,
      "num_tokens": 53625545.0,
      "step": 30910
    },
    {
      "entropy": 5.699866247177124,
      "epoch": 2.4052907994553587,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004416153206327194,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20193272233009338,
      "num_tokens": 53634515.0,
      "step": 30915
    },
    {
      "entropy": 5.646988010406494,
      "epoch": 2.4056798288270764,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044159669331462326,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21122800409793854,
      "num_tokens": 53642999.0,
      "step": 30920
    },
    {
      "entropy": 5.6742918491363525,
      "epoch": 2.406068858198794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044157806346868737,
      "loss": 4.91,
      "mean_token_accuracy": 0.21074799150228501,
      "num_tokens": 53651921.0,
      "step": 30925
    },
    {
      "entropy": 5.756227922439575,
      "epoch": 2.406457887570512,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044155943109519454,
      "loss": 5.0018,
      "mean_token_accuracy": 0.20670194774866105,
      "num_tokens": 53660411.0,
      "step": 30930
    },
    {
      "entropy": 5.657260513305664,
      "epoch": 2.406846916942229,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004415407961944274,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20527012795209884,
      "num_tokens": 53668240.0,
      "step": 30935
    },
    {
      "entropy": 5.616481161117553,
      "epoch": 2.4072359463139468,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044152215876666883,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20727506875991822,
      "num_tokens": 53677193.0,
      "step": 30940
    },
    {
      "entropy": 5.697905206680298,
      "epoch": 2.407624975685664,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004415035188122016,
      "loss": 5.0187,
      "mean_token_accuracy": 0.20443903505802155,
      "num_tokens": 53686747.0,
      "step": 30945
    },
    {
      "entropy": 5.751395225524902,
      "epoch": 2.4080140050573817,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044148487633130837,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2018015757203102,
      "num_tokens": 53695566.0,
      "step": 30950
    },
    {
      "entropy": 5.605574083328247,
      "epoch": 2.4084030344290994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044146623132427213,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21648793816566467,
      "num_tokens": 53704659.0,
      "step": 30955
    },
    {
      "entropy": 5.633669376373291,
      "epoch": 2.408792063800817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044144758379137584,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21182724535465242,
      "num_tokens": 53713908.0,
      "step": 30960
    },
    {
      "entropy": 5.641108703613281,
      "epoch": 2.4091810931725344,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044142893373290236,
      "loss": 4.944,
      "mean_token_accuracy": 0.20871031433343887,
      "num_tokens": 53722038.0,
      "step": 30965
    },
    {
      "entropy": 5.653949308395386,
      "epoch": 2.409570122544252,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004414102811491348,
      "loss": 5.0895,
      "mean_token_accuracy": 0.19648343175649643,
      "num_tokens": 53730658.0,
      "step": 30970
    },
    {
      "entropy": 5.665880298614502,
      "epoch": 2.4099591519159698,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044139162604035597,
      "loss": 4.9035,
      "mean_token_accuracy": 0.20740678906440735,
      "num_tokens": 53738999.0,
      "step": 30975
    },
    {
      "entropy": 5.701525115966797,
      "epoch": 2.410348181287687,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044137296840684917,
      "loss": 4.9916,
      "mean_token_accuracy": 0.20206726491451263,
      "num_tokens": 53747448.0,
      "step": 30980
    },
    {
      "entropy": 5.75875039100647,
      "epoch": 2.4107372106594047,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004413543082488973,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20116353183984756,
      "num_tokens": 53756022.0,
      "step": 30985
    },
    {
      "entropy": 5.683370637893677,
      "epoch": 2.4111262400311224,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004413356455667836,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2081265226006508,
      "num_tokens": 53764105.0,
      "step": 30990
    },
    {
      "entropy": 5.573893976211548,
      "epoch": 2.41151526940284,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004413169803607913,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21975770592689514,
      "num_tokens": 53772692.0,
      "step": 30995
    },
    {
      "entropy": 5.699776601791382,
      "epoch": 2.4119042987745574,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044129831263120344,
      "loss": 5.0234,
      "mean_token_accuracy": 0.20483534038066864,
      "num_tokens": 53781432.0,
      "step": 31000
    },
    {
      "entropy": 5.696863889694214,
      "epoch": 2.412293328146275,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004412796423783034,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20291151702404023,
      "num_tokens": 53789685.0,
      "step": 31005
    },
    {
      "entropy": 5.668190383911133,
      "epoch": 2.4126823575179928,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004412609696023745,
      "loss": 4.9953,
      "mean_token_accuracy": 0.2071694925427437,
      "num_tokens": 53798383.0,
      "step": 31010
    },
    {
      "entropy": 5.71931562423706,
      "epoch": 2.41307138688971,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004412422943037,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21233827620744705,
      "num_tokens": 53805841.0,
      "step": 31015
    },
    {
      "entropy": 5.684670782089233,
      "epoch": 2.4134604162614277,
      "grad_norm": 1.125,
      "learning_rate": 0.00044122361648256327,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21262180656194687,
      "num_tokens": 53814949.0,
      "step": 31020
    },
    {
      "entropy": 5.645623826980591,
      "epoch": 2.4138494456331454,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044120493613924766,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21578270941972733,
      "num_tokens": 53823421.0,
      "step": 31025
    },
    {
      "entropy": 5.663108015060425,
      "epoch": 2.414238475004863,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044118625327403677,
      "loss": 4.957,
      "mean_token_accuracy": 0.21846263706684113,
      "num_tokens": 53831678.0,
      "step": 31030
    },
    {
      "entropy": 5.693938207626343,
      "epoch": 2.4146275043765804,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000441167567887214,
      "loss": 5.0654,
      "mean_token_accuracy": 0.195979742705822,
      "num_tokens": 53840696.0,
      "step": 31035
    },
    {
      "entropy": 5.7445894241333,
      "epoch": 2.415016533748298,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004411488799790629,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20039933919906616,
      "num_tokens": 53849054.0,
      "step": 31040
    },
    {
      "entropy": 5.685613679885864,
      "epoch": 2.4154055631200158,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044113018954986693,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2040001779794693,
      "num_tokens": 53857754.0,
      "step": 31045
    },
    {
      "entropy": 5.690042161941529,
      "epoch": 2.415794592491733,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004411114965999098,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2036678746342659,
      "num_tokens": 53866522.0,
      "step": 31050
    },
    {
      "entropy": 5.615806722640992,
      "epoch": 2.4161836218634507,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000441092801129475,
      "loss": 4.8934,
      "mean_token_accuracy": 0.20919670164585114,
      "num_tokens": 53874654.0,
      "step": 31055
    },
    {
      "entropy": 5.648542213439941,
      "epoch": 2.4165726512351684,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044107410313884643,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2074056699872017,
      "num_tokens": 53882918.0,
      "step": 31060
    },
    {
      "entropy": 5.648423290252685,
      "epoch": 2.4169616806068857,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044105540262830763,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2030644878745079,
      "num_tokens": 53891334.0,
      "step": 31065
    },
    {
      "entropy": 5.63470778465271,
      "epoch": 2.4173507099786034,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004410366995981424,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21252783089876176,
      "num_tokens": 53900047.0,
      "step": 31070
    },
    {
      "entropy": 5.61437726020813,
      "epoch": 2.417739739350321,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044101799404863457,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20459308922290803,
      "num_tokens": 53908764.0,
      "step": 31075
    },
    {
      "entropy": 5.7168478012084964,
      "epoch": 2.4181287687220383,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004409992859800679,
      "loss": 5.0152,
      "mean_token_accuracy": 0.20090483874082565,
      "num_tokens": 53916634.0,
      "step": 31080
    },
    {
      "entropy": 5.7492564678192135,
      "epoch": 2.418517798093756,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004409805753927263,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1995559811592102,
      "num_tokens": 53925458.0,
      "step": 31085
    },
    {
      "entropy": 5.687049198150635,
      "epoch": 2.4189068274654737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004409618622868936,
      "loss": 4.9928,
      "mean_token_accuracy": 0.20529905408620835,
      "num_tokens": 53934385.0,
      "step": 31090
    },
    {
      "entropy": 5.652777099609375,
      "epoch": 2.4192958568371914,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044094314666285385,
      "loss": 4.952,
      "mean_token_accuracy": 0.2059738501906395,
      "num_tokens": 53943674.0,
      "step": 31095
    },
    {
      "entropy": 5.681382846832276,
      "epoch": 2.4196848862089086,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044092442852089095,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2024369552731514,
      "num_tokens": 53952455.0,
      "step": 31100
    },
    {
      "entropy": 5.641343498229981,
      "epoch": 2.4200739155806263,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004409057078612889,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2159601241350174,
      "num_tokens": 53961344.0,
      "step": 31105
    },
    {
      "entropy": 5.680008125305176,
      "epoch": 2.420462944952344,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044088698468433194,
      "loss": 4.9495,
      "mean_token_accuracy": 0.2131201907992363,
      "num_tokens": 53969277.0,
      "step": 31110
    },
    {
      "entropy": 5.635416030883789,
      "epoch": 2.4208519743240613,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044086825899030383,
      "loss": 4.946,
      "mean_token_accuracy": 0.20070736855268478,
      "num_tokens": 53977830.0,
      "step": 31115
    },
    {
      "entropy": 5.648839902877808,
      "epoch": 2.421241003695779,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000440849530779489,
      "loss": 4.9516,
      "mean_token_accuracy": 0.20884760320186616,
      "num_tokens": 53986273.0,
      "step": 31120
    },
    {
      "entropy": 5.675412130355835,
      "epoch": 2.4216300330674967,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004408308000521715,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20781175345182418,
      "num_tokens": 53994064.0,
      "step": 31125
    },
    {
      "entropy": 5.601336336135864,
      "epoch": 2.4220190624392144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044081206680863556,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2104591101408005,
      "num_tokens": 54003516.0,
      "step": 31130
    },
    {
      "entropy": 5.643300151824951,
      "epoch": 2.4224080918109316,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044079333104916554,
      "loss": 5.0207,
      "mean_token_accuracy": 0.20369343757629393,
      "num_tokens": 54012503.0,
      "step": 31135
    },
    {
      "entropy": 5.718322515487671,
      "epoch": 2.4227971211826493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004407745927740454,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2048995554447174,
      "num_tokens": 54021170.0,
      "step": 31140
    },
    {
      "entropy": 5.666590833663941,
      "epoch": 2.423186150554367,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004407558519835598,
      "loss": 5.027,
      "mean_token_accuracy": 0.20990532636642456,
      "num_tokens": 54030520.0,
      "step": 31145
    },
    {
      "entropy": 5.716016054153442,
      "epoch": 2.4235751799260843,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044073710867799295,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21078892946243286,
      "num_tokens": 54039625.0,
      "step": 31150
    },
    {
      "entropy": 5.60653133392334,
      "epoch": 2.423964209297802,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004407183628576293,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2164502814412117,
      "num_tokens": 54047756.0,
      "step": 31155
    },
    {
      "entropy": 5.636724233627319,
      "epoch": 2.4243532386695197,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004406996145227532,
      "loss": 5.0264,
      "mean_token_accuracy": 0.20057591199874877,
      "num_tokens": 54056051.0,
      "step": 31160
    },
    {
      "entropy": 5.73341212272644,
      "epoch": 2.424742268041237,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004406808636736492,
      "loss": 5.0097,
      "mean_token_accuracy": 0.2039689764380455,
      "num_tokens": 54065454.0,
      "step": 31165
    },
    {
      "entropy": 5.656882286071777,
      "epoch": 2.4251312974129546,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044066211031060183,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2052305519580841,
      "num_tokens": 54073501.0,
      "step": 31170
    },
    {
      "entropy": 5.658548355102539,
      "epoch": 2.4255203267846723,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004406433544338956,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20272246748209,
      "num_tokens": 54082314.0,
      "step": 31175
    },
    {
      "entropy": 5.594903087615966,
      "epoch": 2.4259093561563896,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004406245960438151,
      "loss": 4.9158,
      "mean_token_accuracy": 0.21184185147285461,
      "num_tokens": 54090590.0,
      "step": 31180
    },
    {
      "entropy": 5.686072492599488,
      "epoch": 2.4262983855281073,
      "grad_norm": 1.15625,
      "learning_rate": 0.000440605835140645,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20533467233181,
      "num_tokens": 54100366.0,
      "step": 31185
    },
    {
      "entropy": 5.629287910461426,
      "epoch": 2.426687414899825,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004405870717246699,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2146964594721794,
      "num_tokens": 54109007.0,
      "step": 31190
    },
    {
      "entropy": 5.645850419998169,
      "epoch": 2.4270764442715427,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044056830579617464,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2113846704363823,
      "num_tokens": 54117696.0,
      "step": 31195
    },
    {
      "entropy": 5.5990081310272215,
      "epoch": 2.42746547364326,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044054953735544376,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2097751021385193,
      "num_tokens": 54125955.0,
      "step": 31200
    },
    {
      "entropy": 5.643500423431396,
      "epoch": 2.4278545030149776,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004405307664027622,
      "loss": 4.9814,
      "mean_token_accuracy": 0.20560961216688156,
      "num_tokens": 54134311.0,
      "step": 31205
    },
    {
      "entropy": 5.64795331954956,
      "epoch": 2.4282435323866953,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004405119929384147,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20884929448366166,
      "num_tokens": 54144080.0,
      "step": 31210
    },
    {
      "entropy": 5.714404773712158,
      "epoch": 2.4286325617584126,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004404932169626862,
      "loss": 5.0872,
      "mean_token_accuracy": 0.19786328673362732,
      "num_tokens": 54152860.0,
      "step": 31215
    },
    {
      "entropy": 5.649265241622925,
      "epoch": 2.4290215911301303,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004404744384758615,
      "loss": 4.9324,
      "mean_token_accuracy": 0.20750968307256698,
      "num_tokens": 54162049.0,
      "step": 31220
    },
    {
      "entropy": 5.638836669921875,
      "epoch": 2.429410620501848,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004404556574782256,
      "loss": 4.903,
      "mean_token_accuracy": 0.20597613900899886,
      "num_tokens": 54170321.0,
      "step": 31225
    },
    {
      "entropy": 5.6488649368286135,
      "epoch": 2.4297996498735657,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044043687397006336,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2103401854634285,
      "num_tokens": 54179381.0,
      "step": 31230
    },
    {
      "entropy": 5.745428991317749,
      "epoch": 2.430188679245283,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044041808795166,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20150024741888045,
      "num_tokens": 54188416.0,
      "step": 31235
    },
    {
      "entropy": 5.650040149688721,
      "epoch": 2.4305777086170006,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004403992994233004,
      "loss": 4.912,
      "mean_token_accuracy": 0.21094109416007994,
      "num_tokens": 54196483.0,
      "step": 31240
    },
    {
      "entropy": 5.6612834453582765,
      "epoch": 2.4309667379887183,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044038050838526974,
      "loss": 4.9672,
      "mean_token_accuracy": 0.194662643969059,
      "num_tokens": 54205782.0,
      "step": 31245
    },
    {
      "entropy": 5.604939270019531,
      "epoch": 2.4313557673604356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000440361714837853,
      "loss": 4.9936,
      "mean_token_accuracy": 0.203917196393013,
      "num_tokens": 54214014.0,
      "step": 31250
    },
    {
      "entropy": 5.647617292404175,
      "epoch": 2.4317447967321533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004403429187813355,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21088347882032393,
      "num_tokens": 54222018.0,
      "step": 31255
    },
    {
      "entropy": 5.672516536712647,
      "epoch": 2.432133826103871,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004403241202160024,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20752330869436264,
      "num_tokens": 54231549.0,
      "step": 31260
    },
    {
      "entropy": 5.599980163574219,
      "epoch": 2.432522855475588,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004403053191421388,
      "loss": 4.9539,
      "mean_token_accuracy": 0.20914627909660338,
      "num_tokens": 54240605.0,
      "step": 31265
    },
    {
      "entropy": 5.673921918869018,
      "epoch": 2.432911884847306,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004402865155600302,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20804419815540315,
      "num_tokens": 54249201.0,
      "step": 31270
    },
    {
      "entropy": 5.6475982666015625,
      "epoch": 2.4333009142190236,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044026770946996183,
      "loss": 4.932,
      "mean_token_accuracy": 0.2089228630065918,
      "num_tokens": 54258039.0,
      "step": 31275
    },
    {
      "entropy": 5.707819271087646,
      "epoch": 2.433689943590741,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044024890087221896,
      "loss": 4.9769,
      "mean_token_accuracy": 0.20197538286447525,
      "num_tokens": 54266242.0,
      "step": 31280
    },
    {
      "entropy": 5.649300765991211,
      "epoch": 2.4340789729624586,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044023008976708705,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2095598727464676,
      "num_tokens": 54274838.0,
      "step": 31285
    },
    {
      "entropy": 5.634437084197998,
      "epoch": 2.4344680023341763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004402112761548515,
      "loss": 4.9178,
      "mean_token_accuracy": 0.20656898319721223,
      "num_tokens": 54283633.0,
      "step": 31290
    },
    {
      "entropy": 5.624357223510742,
      "epoch": 2.434857031705894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044019246003579776,
      "loss": 4.9778,
      "mean_token_accuracy": 0.20720132291316987,
      "num_tokens": 54291656.0,
      "step": 31295
    },
    {
      "entropy": 5.641462230682373,
      "epoch": 2.435246061077611,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004401736414102115,
      "loss": 4.9087,
      "mean_token_accuracy": 0.20748155415058137,
      "num_tokens": 54300353.0,
      "step": 31300
    },
    {
      "entropy": 5.707808780670166,
      "epoch": 2.435635090449329,
      "grad_norm": 1.1875,
      "learning_rate": 0.000440154820278378,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2048929288983345,
      "num_tokens": 54308771.0,
      "step": 31305
    },
    {
      "entropy": 5.682120227813721,
      "epoch": 2.4360241198210466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004401359966405831,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20163409560918807,
      "num_tokens": 54318069.0,
      "step": 31310
    },
    {
      "entropy": 5.626701736450196,
      "epoch": 2.436413149192764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044011717049711216,
      "loss": 4.9599,
      "mean_token_accuracy": 0.20719880014657974,
      "num_tokens": 54326363.0,
      "step": 31315
    },
    {
      "entropy": 5.662750864028931,
      "epoch": 2.4368021785644816,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044009834184825103,
      "loss": 5.031,
      "mean_token_accuracy": 0.20478837043046952,
      "num_tokens": 54334439.0,
      "step": 31320
    },
    {
      "entropy": 5.694104242324829,
      "epoch": 2.4371912079361993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004400795106942853,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20656394511461257,
      "num_tokens": 54343799.0,
      "step": 31325
    },
    {
      "entropy": 5.714408111572266,
      "epoch": 2.437580237307917,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044006067703550083,
      "loss": 5.0594,
      "mean_token_accuracy": 0.1996875137090683,
      "num_tokens": 54351382.0,
      "step": 31330
    },
    {
      "entropy": 5.729832124710083,
      "epoch": 2.437969266679634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004400418408721833,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20484249889850617,
      "num_tokens": 54360803.0,
      "step": 31335
    },
    {
      "entropy": 5.745359659194946,
      "epoch": 2.438358296051352,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044002300220461843,
      "loss": 5.0368,
      "mean_token_accuracy": 0.20131905525922775,
      "num_tokens": 54369540.0,
      "step": 31340
    },
    {
      "entropy": 5.709685945510865,
      "epoch": 2.4387473254230696,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004400041610330922,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19478348791599273,
      "num_tokens": 54378585.0,
      "step": 31345
    },
    {
      "entropy": 5.669918060302734,
      "epoch": 2.439136354794787,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004399853173578905,
      "loss": 4.913,
      "mean_token_accuracy": 0.21886730045080185,
      "num_tokens": 54387477.0,
      "step": 31350
    },
    {
      "entropy": 5.693781042098999,
      "epoch": 2.4395253841665046,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043996647117929916,
      "loss": 4.9648,
      "mean_token_accuracy": 0.21326095312833787,
      "num_tokens": 54396599.0,
      "step": 31355
    },
    {
      "entropy": 5.783183670043945,
      "epoch": 2.4399144135382222,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004399476224976043,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19120818078517915,
      "num_tokens": 54406302.0,
      "step": 31360
    },
    {
      "entropy": 5.695734453201294,
      "epoch": 2.44030344290994,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043992877131309165,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20827574282884598,
      "num_tokens": 54414467.0,
      "step": 31365
    },
    {
      "entropy": 5.627566194534301,
      "epoch": 2.440692472281657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004399099176260475,
      "loss": 4.9807,
      "mean_token_accuracy": 0.20642670542001723,
      "num_tokens": 54423699.0,
      "step": 31370
    },
    {
      "entropy": 5.636210489273071,
      "epoch": 2.441081501653375,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004398910614367579,
      "loss": 4.9827,
      "mean_token_accuracy": 0.19866613000631334,
      "num_tokens": 54432617.0,
      "step": 31375
    },
    {
      "entropy": 5.70959849357605,
      "epoch": 2.441470531025092,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043987220274550876,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20565340369939805,
      "num_tokens": 54441748.0,
      "step": 31380
    },
    {
      "entropy": 5.672737121582031,
      "epoch": 2.44185956039681,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043985334155258643,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2008816972374916,
      "num_tokens": 54450785.0,
      "step": 31385
    },
    {
      "entropy": 5.662129259109497,
      "epoch": 2.4422485897685275,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000439834477858277,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20017824620008468,
      "num_tokens": 54459967.0,
      "step": 31390
    },
    {
      "entropy": 5.749491453170776,
      "epoch": 2.4426376191402452,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043981561166286676,
      "loss": 5.0866,
      "mean_token_accuracy": 0.20291004180908204,
      "num_tokens": 54469066.0,
      "step": 31395
    },
    {
      "entropy": 5.7649617195129395,
      "epoch": 2.4430266485119625,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000439796742966642,
      "loss": 4.9813,
      "mean_token_accuracy": 0.20544808059930803,
      "num_tokens": 54477985.0,
      "step": 31400
    },
    {
      "entropy": 5.615476655960083,
      "epoch": 2.44341567788368,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043977787176988884,
      "loss": 4.8897,
      "mean_token_accuracy": 0.21326467394828796,
      "num_tokens": 54486223.0,
      "step": 31405
    },
    {
      "entropy": 5.644390821456909,
      "epoch": 2.443804707255398,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004397589980728938,
      "loss": 4.9206,
      "mean_token_accuracy": 0.20813342332839965,
      "num_tokens": 54494237.0,
      "step": 31410
    },
    {
      "entropy": 5.7613929271697994,
      "epoch": 2.444193736627115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043974012187594324,
      "loss": 5.037,
      "mean_token_accuracy": 0.20038892179727555,
      "num_tokens": 54503405.0,
      "step": 31415
    },
    {
      "entropy": 5.710390186309814,
      "epoch": 2.444582765998833,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004397212431793235,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2032597243785858,
      "num_tokens": 54511417.0,
      "step": 31420
    },
    {
      "entropy": 5.646000671386719,
      "epoch": 2.4449717953705505,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043970236198332107,
      "loss": 4.9951,
      "mean_token_accuracy": 0.203825244307518,
      "num_tokens": 54519548.0,
      "step": 31425
    },
    {
      "entropy": 5.570358180999756,
      "epoch": 2.4453608247422682,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043968347828822246,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21202671229839326,
      "num_tokens": 54528102.0,
      "step": 31430
    },
    {
      "entropy": 5.667199993133545,
      "epoch": 2.4457498541139855,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004396645920943142,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20727344900369643,
      "num_tokens": 54536619.0,
      "step": 31435
    },
    {
      "entropy": 5.649351930618286,
      "epoch": 2.446138883485703,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004396457034018828,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21250502169132232,
      "num_tokens": 54545091.0,
      "step": 31440
    },
    {
      "entropy": 5.681482362747192,
      "epoch": 2.446527912857421,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000439626812211215,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20347914695739747,
      "num_tokens": 54554055.0,
      "step": 31445
    },
    {
      "entropy": 5.700871372222901,
      "epoch": 2.446916942229138,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043960791852259735,
      "loss": 5.0197,
      "mean_token_accuracy": 0.19463416188955307,
      "num_tokens": 54562629.0,
      "step": 31450
    },
    {
      "entropy": 5.7141071319580075,
      "epoch": 2.447305971600856,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004395890223363165,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20914117991924286,
      "num_tokens": 54570929.0,
      "step": 31455
    },
    {
      "entropy": 5.598820304870605,
      "epoch": 2.4476950009725735,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043957012365265915,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21017364114522935,
      "num_tokens": 54579344.0,
      "step": 31460
    },
    {
      "entropy": 5.6737250804901125,
      "epoch": 2.4480840303442912,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043955122247191214,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2017744317650795,
      "num_tokens": 54588364.0,
      "step": 31465
    },
    {
      "entropy": 5.680770635604858,
      "epoch": 2.4484730597160085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004395323187943623,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1996506780385971,
      "num_tokens": 54597071.0,
      "step": 31470
    },
    {
      "entropy": 5.6994743824005125,
      "epoch": 2.448862089087726,
      "grad_norm": 1.25,
      "learning_rate": 0.0004395134126202964,
      "loss": 4.9981,
      "mean_token_accuracy": 0.201479409635067,
      "num_tokens": 54605625.0,
      "step": 31475
    },
    {
      "entropy": 5.691399240493775,
      "epoch": 2.449251118459444,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004394945039500012,
      "loss": 5.0248,
      "mean_token_accuracy": 0.19991101026535035,
      "num_tokens": 54614883.0,
      "step": 31480
    },
    {
      "entropy": 5.739169692993164,
      "epoch": 2.449640147831161,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043947559278376385,
      "loss": 5.0619,
      "mean_token_accuracy": 0.20442538261413573,
      "num_tokens": 54624689.0,
      "step": 31485
    },
    {
      "entropy": 5.6834924697875975,
      "epoch": 2.450029177202879,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043945667912187117,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2078772231936455,
      "num_tokens": 54633111.0,
      "step": 31490
    },
    {
      "entropy": 5.717809963226318,
      "epoch": 2.4504182065745965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004394377629646101,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2059326723217964,
      "num_tokens": 54642250.0,
      "step": 31495
    },
    {
      "entropy": 5.695753526687622,
      "epoch": 2.4508072359463138,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043941884431226763,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20664994418621063,
      "num_tokens": 54652025.0,
      "step": 31500
    },
    {
      "entropy": 5.714011907577515,
      "epoch": 2.4511962653180315,
      "grad_norm": 1.28125,
      "learning_rate": 0.000439399923165131,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2069253772497177,
      "num_tokens": 54659535.0,
      "step": 31505
    },
    {
      "entropy": 5.68516526222229,
      "epoch": 2.451585294689749,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004393809995234872,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20886108726263047,
      "num_tokens": 54667986.0,
      "step": 31510
    },
    {
      "entropy": 5.684633541107178,
      "epoch": 2.4519743240614664,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004393620733876233,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20059615075588227,
      "num_tokens": 54676867.0,
      "step": 31515
    },
    {
      "entropy": 5.630230808258057,
      "epoch": 2.452363353433184,
      "grad_norm": 1.25,
      "learning_rate": 0.0004393431447578267,
      "loss": 4.885,
      "mean_token_accuracy": 0.22034584879875183,
      "num_tokens": 54685458.0,
      "step": 31520
    },
    {
      "entropy": 5.594038009643555,
      "epoch": 2.452752382804902,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043932421363438427,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2083751529455185,
      "num_tokens": 54694453.0,
      "step": 31525
    },
    {
      "entropy": 5.641142320632935,
      "epoch": 2.4531414121766195,
      "grad_norm": 1.25,
      "learning_rate": 0.0004393052800175835,
      "loss": 4.9296,
      "mean_token_accuracy": 0.20773604065179824,
      "num_tokens": 54702241.0,
      "step": 31530
    },
    {
      "entropy": 5.7276038646698,
      "epoch": 2.4535304415483368,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043928634390771166,
      "loss": 4.9453,
      "mean_token_accuracy": 0.20741995573043823,
      "num_tokens": 54710823.0,
      "step": 31535
    },
    {
      "entropy": 5.598044776916504,
      "epoch": 2.4539194709200545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043926740530505603,
      "loss": 4.9491,
      "mean_token_accuracy": 0.20602016746997834,
      "num_tokens": 54718982.0,
      "step": 31540
    },
    {
      "entropy": 5.68038592338562,
      "epoch": 2.454308500291772,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000439248464209904,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2081098586320877,
      "num_tokens": 54728685.0,
      "step": 31545
    },
    {
      "entropy": 5.667522716522217,
      "epoch": 2.4546975296634894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000439229520622543,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20258067101240157,
      "num_tokens": 54737090.0,
      "step": 31550
    },
    {
      "entropy": 5.672182607650757,
      "epoch": 2.455086559035207,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004392105745432603,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2025071620941162,
      "num_tokens": 54746059.0,
      "step": 31555
    },
    {
      "entropy": 5.625269746780395,
      "epoch": 2.455475588406925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004391916259723436,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21703445613384248,
      "num_tokens": 54754577.0,
      "step": 31560
    },
    {
      "entropy": 5.686138677597046,
      "epoch": 2.4558646177786425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043917267491008023,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20145866721868516,
      "num_tokens": 54763065.0,
      "step": 31565
    },
    {
      "entropy": 5.729212284088135,
      "epoch": 2.4562536471503598,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004391537213567579,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20111447423696518,
      "num_tokens": 54771189.0,
      "step": 31570
    },
    {
      "entropy": 5.64577350616455,
      "epoch": 2.4566426765220775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004391347653126641,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21296558082103728,
      "num_tokens": 54779141.0,
      "step": 31575
    },
    {
      "entropy": 5.665112113952636,
      "epoch": 2.457031705893795,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043911580677808646,
      "loss": 4.962,
      "mean_token_accuracy": 0.2039460927248001,
      "num_tokens": 54787684.0,
      "step": 31580
    },
    {
      "entropy": 5.683018016815185,
      "epoch": 2.4574207352655124,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043909684575331264,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20838956981897355,
      "num_tokens": 54795831.0,
      "step": 31585
    },
    {
      "entropy": 5.657744884490967,
      "epoch": 2.45780976463723,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004390778822386304,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20842453688383103,
      "num_tokens": 54804500.0,
      "step": 31590
    },
    {
      "entropy": 5.716184806823731,
      "epoch": 2.458198794008948,
      "grad_norm": 1.25,
      "learning_rate": 0.00043905891623432754,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20390603691339493,
      "num_tokens": 54812802.0,
      "step": 31595
    },
    {
      "entropy": 5.730127000808716,
      "epoch": 2.458587823380665,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043903994774069165,
      "loss": 4.994,
      "mean_token_accuracy": 0.20891667157411575,
      "num_tokens": 54821699.0,
      "step": 31600
    },
    {
      "entropy": 5.7126106262207035,
      "epoch": 2.4589768527523828,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004390209767580107,
      "loss": 4.982,
      "mean_token_accuracy": 0.2106185659766197,
      "num_tokens": 54830808.0,
      "step": 31605
    },
    {
      "entropy": 5.61960735321045,
      "epoch": 2.4593658821241005,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004390020032865725,
      "loss": 4.9978,
      "mean_token_accuracy": 0.2042613595724106,
      "num_tokens": 54839226.0,
      "step": 31610
    },
    {
      "entropy": 5.764332294464111,
      "epoch": 2.4597549114958177,
      "grad_norm": 1.25,
      "learning_rate": 0.0004389830273266649,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20466790348291397,
      "num_tokens": 54847538.0,
      "step": 31615
    },
    {
      "entropy": 5.716181373596191,
      "epoch": 2.4601439408675354,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043896404887857596,
      "loss": 5.066,
      "mean_token_accuracy": 0.19339053630828856,
      "num_tokens": 54855697.0,
      "step": 31620
    },
    {
      "entropy": 5.6145366668701175,
      "epoch": 2.460532970239253,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004389450679425935,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22149752378463744,
      "num_tokens": 54865044.0,
      "step": 31625
    },
    {
      "entropy": 5.692435026168823,
      "epoch": 2.460921999610971,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004389260845190055,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21142982542514802,
      "num_tokens": 54874559.0,
      "step": 31630
    },
    {
      "entropy": 5.677493333816528,
      "epoch": 2.461311028982688,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043890709860810014,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21648336052894593,
      "num_tokens": 54882412.0,
      "step": 31635
    },
    {
      "entropy": 5.611321926116943,
      "epoch": 2.4617000583544058,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043888811021016545,
      "loss": 4.8608,
      "mean_token_accuracy": 0.2200280860066414,
      "num_tokens": 54890762.0,
      "step": 31640
    },
    {
      "entropy": 5.6379204273223875,
      "epoch": 2.4620890877261234,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043886911932548953,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21045771837234498,
      "num_tokens": 54899098.0,
      "step": 31645
    },
    {
      "entropy": 5.655600547790527,
      "epoch": 2.4624781170978407,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004388501259543605,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21452776938676835,
      "num_tokens": 54907471.0,
      "step": 31650
    },
    {
      "entropy": 5.672487068176269,
      "epoch": 2.4628671464695584,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004388311300970667,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21161340922117233,
      "num_tokens": 54915957.0,
      "step": 31655
    },
    {
      "entropy": 5.6851075172424315,
      "epoch": 2.463256175841276,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004388121317538962,
      "loss": 4.9243,
      "mean_token_accuracy": 0.20780953168869018,
      "num_tokens": 54924850.0,
      "step": 31660
    },
    {
      "entropy": 5.687220239639283,
      "epoch": 2.463645205212994,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004387931309251374,
      "loss": 5.0127,
      "mean_token_accuracy": 0.19894726127386092,
      "num_tokens": 54933839.0,
      "step": 31665
    },
    {
      "entropy": 5.713906621932983,
      "epoch": 2.464034234584711,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004387741276110785,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20014604479074477,
      "num_tokens": 54942152.0,
      "step": 31670
    },
    {
      "entropy": 5.702347946166992,
      "epoch": 2.4644232639564287,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043875512181200784,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2082337036728859,
      "num_tokens": 54951597.0,
      "step": 31675
    },
    {
      "entropy": 5.670708847045899,
      "epoch": 2.4648122933281464,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004387361135282139,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21602654904127122,
      "num_tokens": 54960507.0,
      "step": 31680
    },
    {
      "entropy": 5.7032946109771725,
      "epoch": 2.4652013226998637,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043871710275998507,
      "loss": 5.056,
      "mean_token_accuracy": 0.19846072942018508,
      "num_tokens": 54969897.0,
      "step": 31685
    },
    {
      "entropy": 5.735730981826782,
      "epoch": 2.4655903520715814,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004386980895076097,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20097761154174804,
      "num_tokens": 54978418.0,
      "step": 31690
    },
    {
      "entropy": 5.688509225845337,
      "epoch": 2.465979381443299,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043867907377137646,
      "loss": 4.8673,
      "mean_token_accuracy": 0.2146831378340721,
      "num_tokens": 54987011.0,
      "step": 31695
    },
    {
      "entropy": 5.635149621963501,
      "epoch": 2.4663684108150163,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043866005555157376,
      "loss": 4.9547,
      "mean_token_accuracy": 0.20898949056863786,
      "num_tokens": 54995402.0,
      "step": 31700
    },
    {
      "entropy": 5.605459976196289,
      "epoch": 2.466757440186734,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004386410348484903,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21309436708688737,
      "num_tokens": 55003842.0,
      "step": 31705
    },
    {
      "entropy": 5.681981468200684,
      "epoch": 2.4671464695584517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004386220116624145,
      "loss": 4.9936,
      "mean_token_accuracy": 0.20375392884016036,
      "num_tokens": 55012213.0,
      "step": 31710
    },
    {
      "entropy": 5.6612461566925045,
      "epoch": 2.467535498930169,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043860298599363506,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2041635125875473,
      "num_tokens": 55021247.0,
      "step": 31715
    },
    {
      "entropy": 5.6500404357910154,
      "epoch": 2.4679245283018867,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004385839578424408,
      "loss": 4.97,
      "mean_token_accuracy": 0.2060944050550461,
      "num_tokens": 55030336.0,
      "step": 31720
    },
    {
      "entropy": 5.639929103851318,
      "epoch": 2.4683135576736044,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043856492720912037,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20423865914344788,
      "num_tokens": 55038551.0,
      "step": 31725
    },
    {
      "entropy": 5.66362361907959,
      "epoch": 2.468702587045322,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004385458940939624,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20334839224815368,
      "num_tokens": 55047417.0,
      "step": 31730
    },
    {
      "entropy": 5.7419641494750975,
      "epoch": 2.4690916164170393,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004385268584972559,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19610781967639923,
      "num_tokens": 55055396.0,
      "step": 31735
    },
    {
      "entropy": 5.675347375869751,
      "epoch": 2.469480645788757,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004385078204192896,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2113894298672676,
      "num_tokens": 55063697.0,
      "step": 31740
    },
    {
      "entropy": 5.666456413269043,
      "epoch": 2.4698696751604747,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004384887798603523,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20795352905988693,
      "num_tokens": 55072697.0,
      "step": 31745
    },
    {
      "entropy": 5.59608907699585,
      "epoch": 2.470258704532192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000438469736820733,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21329332143068314,
      "num_tokens": 55081077.0,
      "step": 31750
    },
    {
      "entropy": 5.69738359451294,
      "epoch": 2.4706477339039097,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043845069130072067,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19986434727907182,
      "num_tokens": 55089915.0,
      "step": 31755
    },
    {
      "entropy": 5.720067119598388,
      "epoch": 2.4710367632756274,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004384316433006043,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19812064617872238,
      "num_tokens": 55099536.0,
      "step": 31760
    },
    {
      "entropy": 5.730847692489624,
      "epoch": 2.471425792647345,
      "grad_norm": 1.25,
      "learning_rate": 0.00043841259282067276,
      "loss": 5.0388,
      "mean_token_accuracy": 0.19697016030550002,
      "num_tokens": 55108350.0,
      "step": 31765
    },
    {
      "entropy": 5.786119604110718,
      "epoch": 2.4718148220190623,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043839353986121533,
      "loss": 5.0592,
      "mean_token_accuracy": 0.20507774800062178,
      "num_tokens": 55117134.0,
      "step": 31770
    },
    {
      "entropy": 5.710341167449951,
      "epoch": 2.47220385139078,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004383744844225209,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2017264649271965,
      "num_tokens": 55125821.0,
      "step": 31775
    },
    {
      "entropy": 5.707550477981568,
      "epoch": 2.4725928807624977,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043835542650487875,
      "loss": 4.9685,
      "mean_token_accuracy": 0.20483948141336442,
      "num_tokens": 55135373.0,
      "step": 31780
    },
    {
      "entropy": 5.694150733947754,
      "epoch": 2.472981910134215,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004383363661085781,
      "loss": 5.0143,
      "mean_token_accuracy": 0.19905594885349273,
      "num_tokens": 55144166.0,
      "step": 31785
    },
    {
      "entropy": 5.676385927200317,
      "epoch": 2.4733709395059327,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004383173032339079,
      "loss": 4.946,
      "mean_token_accuracy": 0.20544901341199875,
      "num_tokens": 55152983.0,
      "step": 31790
    },
    {
      "entropy": 5.664535570144653,
      "epoch": 2.4737599688776504,
      "grad_norm": 1.25,
      "learning_rate": 0.0004382982378811577,
      "loss": 4.8291,
      "mean_token_accuracy": 0.2129627525806427,
      "num_tokens": 55160954.0,
      "step": 31795
    },
    {
      "entropy": 5.64508056640625,
      "epoch": 2.474148998249368,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004382791700506166,
      "loss": 4.966,
      "mean_token_accuracy": 0.20942001789808273,
      "num_tokens": 55169393.0,
      "step": 31800
    },
    {
      "entropy": 5.63147006034851,
      "epoch": 2.4745380276210853,
      "grad_norm": 1.171875,
      "learning_rate": 0.000438260099742574,
      "loss": 5.0215,
      "mean_token_accuracy": 0.19730190932750702,
      "num_tokens": 55178301.0,
      "step": 31805
    },
    {
      "entropy": 5.6634947776794435,
      "epoch": 2.474927056992803,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004382410269573192,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2072368323802948,
      "num_tokens": 55186717.0,
      "step": 31810
    },
    {
      "entropy": 5.6114266872406,
      "epoch": 2.4753160863645203,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043822195169514167,
      "loss": 4.8261,
      "mean_token_accuracy": 0.2229962721467018,
      "num_tokens": 55195457.0,
      "step": 31815
    },
    {
      "entropy": 5.757020568847656,
      "epoch": 2.475705115736238,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043820287395633086,
      "loss": 5.0539,
      "mean_token_accuracy": 0.20305736511945724,
      "num_tokens": 55204623.0,
      "step": 31820
    },
    {
      "entropy": 5.654867315292359,
      "epoch": 2.4760941451079557,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004381837937411761,
      "loss": 5.0144,
      "mean_token_accuracy": 0.19713883846998215,
      "num_tokens": 55213396.0,
      "step": 31825
    },
    {
      "entropy": 5.671772384643555,
      "epoch": 2.4764831744796734,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043816471104996706,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20379563719034194,
      "num_tokens": 55222364.0,
      "step": 31830
    },
    {
      "entropy": 5.6942222118377686,
      "epoch": 2.4768722038513906,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004381456258829933,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21020497679710387,
      "num_tokens": 55230907.0,
      "step": 31835
    },
    {
      "entropy": 5.67466459274292,
      "epoch": 2.4772612332231083,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043812653824054426,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2102232500910759,
      "num_tokens": 55240047.0,
      "step": 31840
    },
    {
      "entropy": 5.667136096954346,
      "epoch": 2.477650262594826,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004381074481229097,
      "loss": 4.9281,
      "mean_token_accuracy": 0.20815111100673675,
      "num_tokens": 55248631.0,
      "step": 31845
    },
    {
      "entropy": 5.645478439331055,
      "epoch": 2.4780392919665433,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004380883555303791,
      "loss": 4.93,
      "mean_token_accuracy": 0.20750757157802582,
      "num_tokens": 55256785.0,
      "step": 31850
    },
    {
      "entropy": 5.624342727661133,
      "epoch": 2.478428321338261,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004380692604632424,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21617498248815536,
      "num_tokens": 55265904.0,
      "step": 31855
    },
    {
      "entropy": 5.601974201202393,
      "epoch": 2.4788173507099787,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043805016292178923,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2072703942656517,
      "num_tokens": 55274223.0,
      "step": 31860
    },
    {
      "entropy": 5.629825401306152,
      "epoch": 2.4792063800816964,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004380310629063094,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21248839795589447,
      "num_tokens": 55282667.0,
      "step": 31865
    },
    {
      "entropy": 5.6558802127838135,
      "epoch": 2.4795954094534136,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004380119604170925,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19753156155347823,
      "num_tokens": 55291207.0,
      "step": 31870
    },
    {
      "entropy": 5.6954515933990475,
      "epoch": 2.4799844388251313,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043799285545442876,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2024501919746399,
      "num_tokens": 55299828.0,
      "step": 31875
    },
    {
      "entropy": 5.772912406921387,
      "epoch": 2.480373468196849,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004379737480186078,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19498128443956375,
      "num_tokens": 55308056.0,
      "step": 31880
    },
    {
      "entropy": 5.731743907928466,
      "epoch": 2.4807624975685663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043795463810991956,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19635724872350693,
      "num_tokens": 55316888.0,
      "step": 31885
    },
    {
      "entropy": 5.672018194198609,
      "epoch": 2.481151526940284,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004379355257286541,
      "loss": 4.9552,
      "mean_token_accuracy": 0.20273196548223496,
      "num_tokens": 55325797.0,
      "step": 31890
    },
    {
      "entropy": 5.7102738380432125,
      "epoch": 2.4815405563120017,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043791641087510136,
      "loss": 4.9437,
      "mean_token_accuracy": 0.20762696862220764,
      "num_tokens": 55335071.0,
      "step": 31895
    },
    {
      "entropy": 5.752525854110718,
      "epoch": 2.4819295856837194,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004378972935495514,
      "loss": 5.0322,
      "mean_token_accuracy": 0.2034935846924782,
      "num_tokens": 55343718.0,
      "step": 31900
    },
    {
      "entropy": 5.656074571609497,
      "epoch": 2.4823186150554366,
      "grad_norm": 1.125,
      "learning_rate": 0.0004378781737522943,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20646874159574508,
      "num_tokens": 55353122.0,
      "step": 31905
    },
    {
      "entropy": 5.6390049934387205,
      "epoch": 2.4827076444271543,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043785905148362007,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20813973546028136,
      "num_tokens": 55362494.0,
      "step": 31910
    },
    {
      "entropy": 5.775249004364014,
      "epoch": 2.483096673798872,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043783992674381903,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20909199267625808,
      "num_tokens": 55371383.0,
      "step": 31915
    },
    {
      "entropy": 5.704949378967285,
      "epoch": 2.4834857031705893,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043782079953318126,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2141811117529869,
      "num_tokens": 55380086.0,
      "step": 31920
    },
    {
      "entropy": 5.586907196044922,
      "epoch": 2.483874732542307,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004378016698519971,
      "loss": 4.9311,
      "mean_token_accuracy": 0.20430454164743422,
      "num_tokens": 55388971.0,
      "step": 31925
    },
    {
      "entropy": 5.6551319599151615,
      "epoch": 2.4842637619140246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043778253770055666,
      "loss": 4.9261,
      "mean_token_accuracy": 0.20904773473739624,
      "num_tokens": 55396820.0,
      "step": 31930
    },
    {
      "entropy": 5.60729284286499,
      "epoch": 2.484652791285742,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043776340307915033,
      "loss": 4.8811,
      "mean_token_accuracy": 0.20773734152317047,
      "num_tokens": 55405653.0,
      "step": 31935
    },
    {
      "entropy": 5.677891397476197,
      "epoch": 2.4850418206574596,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004377442659880684,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2077256605029106,
      "num_tokens": 55414397.0,
      "step": 31940
    },
    {
      "entropy": 5.74730110168457,
      "epoch": 2.4854308500291773,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004377251264276013,
      "loss": 5.0364,
      "mean_token_accuracy": 0.20087112933397294,
      "num_tokens": 55422057.0,
      "step": 31945
    },
    {
      "entropy": 5.6124410152435305,
      "epoch": 2.4858198794008945,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043770598439803936,
      "loss": 4.894,
      "mean_token_accuracy": 0.21265334039926528,
      "num_tokens": 55430084.0,
      "step": 31950
    },
    {
      "entropy": 5.731271886825562,
      "epoch": 2.4862089087726122,
      "grad_norm": 1.375,
      "learning_rate": 0.00043768683989967304,
      "loss": 5.0498,
      "mean_token_accuracy": 0.20552590638399124,
      "num_tokens": 55438551.0,
      "step": 31955
    },
    {
      "entropy": 5.608238935470581,
      "epoch": 2.48659793814433,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043766769293279294,
      "loss": 4.7725,
      "mean_token_accuracy": 0.21722216308116912,
      "num_tokens": 55446862.0,
      "step": 31960
    },
    {
      "entropy": 5.644069194793701,
      "epoch": 2.4869869675160476,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043764854349768957,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21209183931350709,
      "num_tokens": 55455254.0,
      "step": 31965
    },
    {
      "entropy": 5.680535554885864,
      "epoch": 2.487375996887765,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043762939159465333,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20971533954143523,
      "num_tokens": 55463426.0,
      "step": 31970
    },
    {
      "entropy": 5.711111688613892,
      "epoch": 2.4877650262594826,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043761023722397495,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2080627664923668,
      "num_tokens": 55471543.0,
      "step": 31975
    },
    {
      "entropy": 5.7546099662780765,
      "epoch": 2.4881540556312003,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000437591080385945,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19794660955667495,
      "num_tokens": 55480118.0,
      "step": 31980
    },
    {
      "entropy": 5.644216823577881,
      "epoch": 2.4885430850029175,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004375719210808543,
      "loss": 4.852,
      "mean_token_accuracy": 0.21419955044984818,
      "num_tokens": 55488818.0,
      "step": 31985
    },
    {
      "entropy": 5.740829229354858,
      "epoch": 2.4889321143746352,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004375527593089934,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2065274953842163,
      "num_tokens": 55497576.0,
      "step": 31990
    },
    {
      "entropy": 5.767935514450073,
      "epoch": 2.489321143746353,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004375335950706531,
      "loss": 5.0826,
      "mean_token_accuracy": 0.2017228290438652,
      "num_tokens": 55505892.0,
      "step": 31995
    },
    {
      "entropy": 5.681882572174072,
      "epoch": 2.4897101731180706,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004375144283661242,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20213755667209626,
      "num_tokens": 55514595.0,
      "step": 32000
    },
    {
      "entropy": 5.66088171005249,
      "epoch": 2.490099202489788,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004374952591956975,
      "loss": 5.005,
      "mean_token_accuracy": 0.19875537902116774,
      "num_tokens": 55523681.0,
      "step": 32005
    },
    {
      "entropy": 5.684203815460205,
      "epoch": 2.4904882318615056,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043747608755966396,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21048945039510727,
      "num_tokens": 55532368.0,
      "step": 32010
    },
    {
      "entropy": 5.655999422073364,
      "epoch": 2.4908772612332233,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004374569134583143,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20949650853872298,
      "num_tokens": 55541873.0,
      "step": 32015
    },
    {
      "entropy": 5.695771932601929,
      "epoch": 2.4912662906049405,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004374377368919396,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20825198143720627,
      "num_tokens": 55551094.0,
      "step": 32020
    },
    {
      "entropy": 5.713011074066162,
      "epoch": 2.4916553199766582,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043741855786083085,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21307261735200883,
      "num_tokens": 55559378.0,
      "step": 32025
    },
    {
      "entropy": 5.692988538742066,
      "epoch": 2.492044349348376,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043739937636527894,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1969159349799156,
      "num_tokens": 55568134.0,
      "step": 32030
    },
    {
      "entropy": 5.7495428085327145,
      "epoch": 2.492433378720093,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043738019240557494,
      "loss": 5.025,
      "mean_token_accuracy": 0.19781984239816666,
      "num_tokens": 55576962.0,
      "step": 32035
    },
    {
      "entropy": 5.746601200103759,
      "epoch": 2.492822408091811,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043736100598201,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20618593990802764,
      "num_tokens": 55585944.0,
      "step": 32040
    },
    {
      "entropy": 5.657977342605591,
      "epoch": 2.4932114374635286,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004373418170948752,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2016979292035103,
      "num_tokens": 55594964.0,
      "step": 32045
    },
    {
      "entropy": 5.671987819671631,
      "epoch": 2.493600466835246,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004373226257444617,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20406342148780823,
      "num_tokens": 55604344.0,
      "step": 32050
    },
    {
      "entropy": 5.66226806640625,
      "epoch": 2.4939894962069635,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004373034319310608,
      "loss": 4.9294,
      "mean_token_accuracy": 0.20238882601261138,
      "num_tokens": 55612146.0,
      "step": 32055
    },
    {
      "entropy": 5.633600187301636,
      "epoch": 2.4943785255786812,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043728423565496367,
      "loss": 4.9587,
      "mean_token_accuracy": 0.19904157221317292,
      "num_tokens": 55621070.0,
      "step": 32060
    },
    {
      "entropy": 5.631651878356934,
      "epoch": 2.494767554950399,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004372650369164615,
      "loss": 4.944,
      "mean_token_accuracy": 0.21302301585674285,
      "num_tokens": 55629541.0,
      "step": 32065
    },
    {
      "entropy": 5.735749244689941,
      "epoch": 2.495156584322116,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043724583571584565,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21011031419038773,
      "num_tokens": 55637915.0,
      "step": 32070
    },
    {
      "entropy": 5.684532880783081,
      "epoch": 2.495545613693834,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043722663205340753,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2060002028942108,
      "num_tokens": 55645832.0,
      "step": 32075
    },
    {
      "entropy": 5.640752172470092,
      "epoch": 2.4959346430655516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004372074259294384,
      "loss": 4.8776,
      "mean_token_accuracy": 0.21371103078126907,
      "num_tokens": 55654035.0,
      "step": 32080
    },
    {
      "entropy": 5.649341630935669,
      "epoch": 2.496323672437269,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004371882173442299,
      "loss": 5.0314,
      "mean_token_accuracy": 0.19378727823495864,
      "num_tokens": 55663520.0,
      "step": 32085
    },
    {
      "entropy": 5.737103891372681,
      "epoch": 2.4967127018089865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043716900629807323,
      "loss": 5.021,
      "mean_token_accuracy": 0.2035091519355774,
      "num_tokens": 55672448.0,
      "step": 32090
    },
    {
      "entropy": 5.755094003677368,
      "epoch": 2.497101731180704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043714979279126006,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20535364300012587,
      "num_tokens": 55681389.0,
      "step": 32095
    },
    {
      "entropy": 5.7384827613830565,
      "epoch": 2.497490760552422,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004371305768240818,
      "loss": 5.048,
      "mean_token_accuracy": 0.20075265616178511,
      "num_tokens": 55690143.0,
      "step": 32100
    },
    {
      "entropy": 5.699243593215942,
      "epoch": 2.497879789924139,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004371113583968302,
      "loss": 4.9845,
      "mean_token_accuracy": 0.20367400199174882,
      "num_tokens": 55698097.0,
      "step": 32105
    },
    {
      "entropy": 5.661271381378174,
      "epoch": 2.498268819295857,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004370921375097967,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20679828971624375,
      "num_tokens": 55706570.0,
      "step": 32110
    },
    {
      "entropy": 5.686183166503906,
      "epoch": 2.4986578486675746,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043707291416327307,
      "loss": 4.8858,
      "mean_token_accuracy": 0.20662733167409897,
      "num_tokens": 55715926.0,
      "step": 32115
    },
    {
      "entropy": 5.648556137084961,
      "epoch": 2.499046878039292,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004370536883575509,
      "loss": 4.882,
      "mean_token_accuracy": 0.2083893224596977,
      "num_tokens": 55724313.0,
      "step": 32120
    },
    {
      "entropy": 5.66943621635437,
      "epoch": 2.4994359074110095,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004370344600929219,
      "loss": 5.031,
      "mean_token_accuracy": 0.20136660039424897,
      "num_tokens": 55732197.0,
      "step": 32125
    },
    {
      "entropy": 5.598745918273925,
      "epoch": 2.499824936782727,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004370152293696779,
      "loss": 4.882,
      "mean_token_accuracy": 0.21106997281312942,
      "num_tokens": 55740428.0,
      "step": 32130
    },
    {
      "entropy": 5.648098611831665,
      "epoch": 2.500213966154445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004369959961881107,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20415072590112687,
      "num_tokens": 55748832.0,
      "step": 32135
    },
    {
      "entropy": 5.725146150588989,
      "epoch": 2.500602995526162,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000436976760548512,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2077835977077484,
      "num_tokens": 55757717.0,
      "step": 32140
    },
    {
      "entropy": 5.625393867492676,
      "epoch": 2.50099202489788,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043695752245117393,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20914046913385392,
      "num_tokens": 55765763.0,
      "step": 32145
    },
    {
      "entropy": 5.7464649200439455,
      "epoch": 2.501381054269597,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004369382818963882,
      "loss": 5.0761,
      "mean_token_accuracy": 0.205593079328537,
      "num_tokens": 55774683.0,
      "step": 32150
    },
    {
      "entropy": 5.690613555908203,
      "epoch": 2.501770083641315,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004369190388844467,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20448861867189408,
      "num_tokens": 55782998.0,
      "step": 32155
    },
    {
      "entropy": 5.681671905517578,
      "epoch": 2.5021591130130325,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043689979341564154,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21488549709320068,
      "num_tokens": 55791411.0,
      "step": 32160
    },
    {
      "entropy": 5.651175880432129,
      "epoch": 2.50254814238475,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004368805454902648,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20857308357954024,
      "num_tokens": 55800169.0,
      "step": 32165
    },
    {
      "entropy": 5.646779727935791,
      "epoch": 2.5029371717564675,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043686129510860826,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20576000809669495,
      "num_tokens": 55809168.0,
      "step": 32170
    },
    {
      "entropy": 5.6551258087158205,
      "epoch": 2.503326201128185,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043684204227096435,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20671856999397278,
      "num_tokens": 55817824.0,
      "step": 32175
    },
    {
      "entropy": 5.670963907241822,
      "epoch": 2.503715230499903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000436822786977625,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2034686878323555,
      "num_tokens": 55825851.0,
      "step": 32180
    },
    {
      "entropy": 5.583332586288452,
      "epoch": 2.50410425987162,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043680352922888244,
      "loss": 4.8411,
      "mean_token_accuracy": 0.2108028545975685,
      "num_tokens": 55834862.0,
      "step": 32185
    },
    {
      "entropy": 5.655129384994507,
      "epoch": 2.504493289243338,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004367842690250288,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2119355618953705,
      "num_tokens": 55843267.0,
      "step": 32190
    },
    {
      "entropy": 5.60051908493042,
      "epoch": 2.5048823186150555,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004367650063663565,
      "loss": 4.9135,
      "mean_token_accuracy": 0.20874572098255156,
      "num_tokens": 55851458.0,
      "step": 32195
    },
    {
      "entropy": 5.7406857967376705,
      "epoch": 2.505271347986773,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004367457412531576,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1927231252193451,
      "num_tokens": 55860271.0,
      "step": 32200
    },
    {
      "entropy": 5.656378126144409,
      "epoch": 2.5056603773584905,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004367264736857245,
      "loss": 4.9159,
      "mean_token_accuracy": 0.20747559368610383,
      "num_tokens": 55868695.0,
      "step": 32205
    },
    {
      "entropy": 5.706352233886719,
      "epoch": 2.506049406730208,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004367072036643497,
      "loss": 4.9222,
      "mean_token_accuracy": 0.207609324157238,
      "num_tokens": 55877166.0,
      "step": 32210
    },
    {
      "entropy": 5.708856582641602,
      "epoch": 2.506438436101926,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043668793118932536,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20696360021829605,
      "num_tokens": 55886129.0,
      "step": 32215
    },
    {
      "entropy": 5.723817920684814,
      "epoch": 2.506827465473643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043666865626094405,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20683227628469467,
      "num_tokens": 55895090.0,
      "step": 32220
    },
    {
      "entropy": 5.776579809188843,
      "epoch": 2.507216494845361,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043664937887949825,
      "loss": 5.0311,
      "mean_token_accuracy": 0.202361561357975,
      "num_tokens": 55903231.0,
      "step": 32225
    },
    {
      "entropy": 5.640281772613525,
      "epoch": 2.5076055242170785,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004366300990452803,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21507373899221421,
      "num_tokens": 55912282.0,
      "step": 32230
    },
    {
      "entropy": 5.634460210800171,
      "epoch": 2.507994553588796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043661081675858295,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2084139883518219,
      "num_tokens": 55920871.0,
      "step": 32235
    },
    {
      "entropy": 5.752755355834961,
      "epoch": 2.5083835829605134,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043659153201969865,
      "loss": 5.0595,
      "mean_token_accuracy": 0.20084677189588546,
      "num_tokens": 55929787.0,
      "step": 32240
    },
    {
      "entropy": 5.796993684768677,
      "epoch": 2.508772612332231,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004365722448289201,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2032591298222542,
      "num_tokens": 55938747.0,
      "step": 32245
    },
    {
      "entropy": 5.760586738586426,
      "epoch": 2.5091616417039484,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004365529551865399,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20517097413539886,
      "num_tokens": 55947063.0,
      "step": 32250
    },
    {
      "entropy": 5.653596687316894,
      "epoch": 2.509550671075666,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043653366309285066,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20922656655311583,
      "num_tokens": 55955871.0,
      "step": 32255
    },
    {
      "entropy": 5.6612584590911865,
      "epoch": 2.509939700447384,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004365143685481452,
      "loss": 4.8564,
      "mean_token_accuracy": 0.21618293970823288,
      "num_tokens": 55965018.0,
      "step": 32260
    },
    {
      "entropy": 5.657132244110107,
      "epoch": 2.5103287298191015,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004364950715527164,
      "loss": 4.9132,
      "mean_token_accuracy": 0.21198441386222838,
      "num_tokens": 55973687.0,
      "step": 32265
    },
    {
      "entropy": 5.658427333831787,
      "epoch": 2.5107177591908187,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004364757721068568,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2087695762515068,
      "num_tokens": 55981783.0,
      "step": 32270
    },
    {
      "entropy": 5.726798152923584,
      "epoch": 2.5111067885625364,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043645647021085945,
      "loss": 5.1571,
      "mean_token_accuracy": 0.1963214322924614,
      "num_tokens": 55990266.0,
      "step": 32275
    },
    {
      "entropy": 5.685345983505249,
      "epoch": 2.511495817934254,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043643716586501706,
      "loss": 4.9528,
      "mean_token_accuracy": 0.20877471566200256,
      "num_tokens": 55998788.0,
      "step": 32280
    },
    {
      "entropy": 5.705431699752808,
      "epoch": 2.5118848473059714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043641785906962276,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20321722477674484,
      "num_tokens": 56007835.0,
      "step": 32285
    },
    {
      "entropy": 5.636080074310303,
      "epoch": 2.512273876677689,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004363985498249693,
      "loss": 4.8876,
      "mean_token_accuracy": 0.21303538084030152,
      "num_tokens": 56016008.0,
      "step": 32290
    },
    {
      "entropy": 5.654402208328247,
      "epoch": 2.512662906049407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043637923813134974,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21099064499139786,
      "num_tokens": 56025231.0,
      "step": 32295
    },
    {
      "entropy": 5.716175270080567,
      "epoch": 2.5130519354211245,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043635992398905713,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20109148919582367,
      "num_tokens": 56033471.0,
      "step": 32300
    },
    {
      "entropy": 5.69792742729187,
      "epoch": 2.5134409647928417,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004363406073983844,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2071823701262474,
      "num_tokens": 56041729.0,
      "step": 32305
    },
    {
      "entropy": 5.7276369571685795,
      "epoch": 2.5138299941645594,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043632128835962485,
      "loss": 5.0065,
      "mean_token_accuracy": 0.2059928610920906,
      "num_tokens": 56051055.0,
      "step": 32310
    },
    {
      "entropy": 5.737316274642945,
      "epoch": 2.514219023536277,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043630196687307156,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20673743188381194,
      "num_tokens": 56059629.0,
      "step": 32315
    },
    {
      "entropy": 5.6228334426879885,
      "epoch": 2.5146080529079944,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004362826429390176,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20814375281333924,
      "num_tokens": 56069334.0,
      "step": 32320
    },
    {
      "entropy": 5.714158201217652,
      "epoch": 2.514997082279712,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043626331655775624,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2045416310429573,
      "num_tokens": 56078113.0,
      "step": 32325
    },
    {
      "entropy": 5.661888170242309,
      "epoch": 2.51538611165143,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043624398772958075,
      "loss": 4.921,
      "mean_token_accuracy": 0.21139978021383285,
      "num_tokens": 56086378.0,
      "step": 32330
    },
    {
      "entropy": 5.700076389312744,
      "epoch": 2.5157751410231475,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004362246564547844,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2089908853173256,
      "num_tokens": 56095126.0,
      "step": 32335
    },
    {
      "entropy": 5.673971509933471,
      "epoch": 2.5161641703948647,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043620532273366044,
      "loss": 5.0322,
      "mean_token_accuracy": 0.20550710409879686,
      "num_tokens": 56105130.0,
      "step": 32340
    },
    {
      "entropy": 5.675560808181762,
      "epoch": 2.5165531997665824,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004361859865665024,
      "loss": 4.9589,
      "mean_token_accuracy": 0.209336294233799,
      "num_tokens": 56113925.0,
      "step": 32345
    },
    {
      "entropy": 5.6971235275268555,
      "epoch": 2.5169422291382997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043616664795360346,
      "loss": 4.938,
      "mean_token_accuracy": 0.20908759385347367,
      "num_tokens": 56121961.0,
      "step": 32350
    },
    {
      "entropy": 5.627387857437133,
      "epoch": 2.5173312585100174,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043614730689525716,
      "loss": 4.895,
      "mean_token_accuracy": 0.20576364248991014,
      "num_tokens": 56129809.0,
      "step": 32355
    },
    {
      "entropy": 5.641656875610352,
      "epoch": 2.517720287881735,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043612796339175707,
      "loss": 4.938,
      "mean_token_accuracy": 0.21156662404537202,
      "num_tokens": 56139118.0,
      "step": 32360
    },
    {
      "entropy": 5.711101531982422,
      "epoch": 2.5181093172534528,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043610861744339653,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20853212773799895,
      "num_tokens": 56148134.0,
      "step": 32365
    },
    {
      "entropy": 5.714322710037232,
      "epoch": 2.5184983466251705,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004360892690504692,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20284264385700226,
      "num_tokens": 56157207.0,
      "step": 32370
    },
    {
      "entropy": 5.626499605178833,
      "epoch": 2.5188873759968877,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004360699182132685,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21440672427415847,
      "num_tokens": 56165510.0,
      "step": 32375
    },
    {
      "entropy": 5.652453136444092,
      "epoch": 2.5192764053686054,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043605056493208825,
      "loss": 4.8664,
      "mean_token_accuracy": 0.213122221827507,
      "num_tokens": 56174151.0,
      "step": 32380
    },
    {
      "entropy": 5.689515781402588,
      "epoch": 2.5196654347403227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000436031209207222,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21207275092601777,
      "num_tokens": 56183834.0,
      "step": 32385
    },
    {
      "entropy": 5.722074127197265,
      "epoch": 2.5200544641120404,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043601185103896346,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2123647764325142,
      "num_tokens": 56192492.0,
      "step": 32390
    },
    {
      "entropy": 5.622499179840088,
      "epoch": 2.520443493483758,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043599249042760636,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2121364563703537,
      "num_tokens": 56201795.0,
      "step": 32395
    },
    {
      "entropy": 5.699419593811035,
      "epoch": 2.5208325228554758,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043597312737344446,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20700410157442092,
      "num_tokens": 56210586.0,
      "step": 32400
    },
    {
      "entropy": 5.743770503997803,
      "epoch": 2.521221552227193,
      "grad_norm": 1.25,
      "learning_rate": 0.00043595376187677157,
      "loss": 5.0538,
      "mean_token_accuracy": 0.19940631985664367,
      "num_tokens": 56219512.0,
      "step": 32405
    },
    {
      "entropy": 5.7590395450592045,
      "epoch": 2.5216105815989107,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043593439393788147,
      "loss": 4.9602,
      "mean_token_accuracy": 0.20781890600919722,
      "num_tokens": 56228709.0,
      "step": 32410
    },
    {
      "entropy": 5.638609838485718,
      "epoch": 2.5219996109706284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004359150235570682,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2033781334757805,
      "num_tokens": 56238071.0,
      "step": 32415
    },
    {
      "entropy": 5.6474395275115965,
      "epoch": 2.5223886403423457,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004358956507346255,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20622272789478302,
      "num_tokens": 56246818.0,
      "step": 32420
    },
    {
      "entropy": 5.819520139694214,
      "epoch": 2.5227776697140634,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004358762754708474,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20656449794769288,
      "num_tokens": 56255330.0,
      "step": 32425
    },
    {
      "entropy": 5.701710987091064,
      "epoch": 2.523166699085781,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043585689776602795,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20782554447650908,
      "num_tokens": 56264482.0,
      "step": 32430
    },
    {
      "entropy": 5.726718854904175,
      "epoch": 2.5235557284574988,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043583751762046104,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20012907534837723,
      "num_tokens": 56273230.0,
      "step": 32435
    },
    {
      "entropy": 5.682690954208374,
      "epoch": 2.523944757829216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004358181350344408,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21130677312612534,
      "num_tokens": 56281639.0,
      "step": 32440
    },
    {
      "entropy": 5.621569442749023,
      "epoch": 2.5243337872009337,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004357987500082613,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2043316885828972,
      "num_tokens": 56289680.0,
      "step": 32445
    },
    {
      "entropy": 5.635527753829956,
      "epoch": 2.524722816572651,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043577936254221675,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21374280452728273,
      "num_tokens": 56298091.0,
      "step": 32450
    },
    {
      "entropy": 5.7402314186096195,
      "epoch": 2.5251118459443687,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004357599726366013,
      "loss": 5.0298,
      "mean_token_accuracy": 0.19726824015378952,
      "num_tokens": 56307662.0,
      "step": 32455
    },
    {
      "entropy": 5.755006980895996,
      "epoch": 2.5255008753160864,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004357405802917091,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2072552040219307,
      "num_tokens": 56316147.0,
      "step": 32460
    },
    {
      "entropy": 5.654056167602539,
      "epoch": 2.525889904687804,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004357211855078345,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20124156326055526,
      "num_tokens": 56324270.0,
      "step": 32465
    },
    {
      "entropy": 5.696562385559082,
      "epoch": 2.5262789340595218,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043570178828527157,
      "loss": 5.0262,
      "mean_token_accuracy": 0.19444778859615325,
      "num_tokens": 56333209.0,
      "step": 32470
    },
    {
      "entropy": 5.695141315460205,
      "epoch": 2.526667963431239,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004356823886243149,
      "loss": 4.92,
      "mean_token_accuracy": 0.21222498416900634,
      "num_tokens": 56341784.0,
      "step": 32475
    },
    {
      "entropy": 5.714824151992798,
      "epoch": 2.5270569928029567,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043566298652525875,
      "loss": 4.9896,
      "mean_token_accuracy": 0.1996238797903061,
      "num_tokens": 56350607.0,
      "step": 32480
    },
    {
      "entropy": 5.6874607563018795,
      "epoch": 2.527446022174674,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043564358198839746,
      "loss": 5.006,
      "mean_token_accuracy": 0.20309099406003953,
      "num_tokens": 56359135.0,
      "step": 32485
    },
    {
      "entropy": 5.71419768333435,
      "epoch": 2.5278350515463917,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004356241750140256,
      "loss": 5.0156,
      "mean_token_accuracy": 0.207475246489048,
      "num_tokens": 56367524.0,
      "step": 32490
    },
    {
      "entropy": 5.666716527938843,
      "epoch": 2.5282240809181094,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043560476560243743,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2057400956749916,
      "num_tokens": 56376976.0,
      "step": 32495
    },
    {
      "entropy": 5.669465637207031,
      "epoch": 2.528613110289827,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004355853537539276,
      "loss": 5.0016,
      "mean_token_accuracy": 0.20162253230810165,
      "num_tokens": 56385576.0,
      "step": 32500
    },
    {
      "entropy": 5.760670804977417,
      "epoch": 2.5290021396615443,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004355659394687906,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21134088337421417,
      "num_tokens": 56395144.0,
      "step": 32505
    },
    {
      "entropy": 5.742590379714966,
      "epoch": 2.529391169033262,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004355465227473211,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20283859521150588,
      "num_tokens": 56403779.0,
      "step": 32510
    },
    {
      "entropy": 5.694804859161377,
      "epoch": 2.5297801984049797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004355271035898136,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19875340461730956,
      "num_tokens": 56413389.0,
      "step": 32515
    },
    {
      "entropy": 5.7027078628540036,
      "epoch": 2.530169227776697,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004355076819965628,
      "loss": 5.007,
      "mean_token_accuracy": 0.2044414237141609,
      "num_tokens": 56421475.0,
      "step": 32520
    },
    {
      "entropy": 5.731250762939453,
      "epoch": 2.5305582571484146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043548825796786353,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2083599954843521,
      "num_tokens": 56429590.0,
      "step": 32525
    },
    {
      "entropy": 5.640769052505493,
      "epoch": 2.5309472865201323,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043546883150401023,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21199076175689696,
      "num_tokens": 56438572.0,
      "step": 32530
    },
    {
      "entropy": 5.651857995986939,
      "epoch": 2.53133631589185,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043544940260529787,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2105289801955223,
      "num_tokens": 56446939.0,
      "step": 32535
    },
    {
      "entropy": 5.719775915145874,
      "epoch": 2.5317253452635673,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004354299712720213,
      "loss": 4.9146,
      "mean_token_accuracy": 0.20464773178100587,
      "num_tokens": 56456178.0,
      "step": 32540
    },
    {
      "entropy": 5.701127624511718,
      "epoch": 2.532114374635285,
      "grad_norm": 1.25,
      "learning_rate": 0.0004354105375044752,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2014423966407776,
      "num_tokens": 56464847.0,
      "step": 32545
    },
    {
      "entropy": 5.698019886016846,
      "epoch": 2.5325034040070027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043539110130295445,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2068748027086258,
      "num_tokens": 56473416.0,
      "step": 32550
    },
    {
      "entropy": 5.652522420883178,
      "epoch": 2.53289243337872,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043537166266775405,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20531934052705764,
      "num_tokens": 56481756.0,
      "step": 32555
    },
    {
      "entropy": 5.7069353580474855,
      "epoch": 2.5332814627504376,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000435352221599169,
      "loss": 5.0086,
      "mean_token_accuracy": 0.20639286935329437,
      "num_tokens": 56490384.0,
      "step": 32560
    },
    {
      "entropy": 5.7193603515625,
      "epoch": 2.5336704921221553,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004353327780974941,
      "loss": 4.9694,
      "mean_token_accuracy": 0.20970795303583145,
      "num_tokens": 56498858.0,
      "step": 32565
    },
    {
      "entropy": 5.717054939270019,
      "epoch": 2.534059521493873,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004353133321630246,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2046177104115486,
      "num_tokens": 56507182.0,
      "step": 32570
    },
    {
      "entropy": 5.655843496322632,
      "epoch": 2.5344485508655903,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004352938837960554,
      "loss": 4.9393,
      "mean_token_accuracy": 0.20523628890514373,
      "num_tokens": 56515912.0,
      "step": 32575
    },
    {
      "entropy": 5.662250518798828,
      "epoch": 2.534837580237308,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043527443299688166,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21039534062147142,
      "num_tokens": 56523888.0,
      "step": 32580
    },
    {
      "entropy": 5.764285850524902,
      "epoch": 2.5352266096090252,
      "grad_norm": 1.25,
      "learning_rate": 0.00043525497976579843,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20337315350770951,
      "num_tokens": 56533117.0,
      "step": 32585
    },
    {
      "entropy": 5.672673940658569,
      "epoch": 2.535615638980743,
      "grad_norm": 1.234375,
      "learning_rate": 0.000435235524103101,
      "loss": 4.9455,
      "mean_token_accuracy": 0.20398035198450087,
      "num_tokens": 56541633.0,
      "step": 32590
    },
    {
      "entropy": 5.716252565383911,
      "epoch": 2.5360046683524606,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004352160660090846,
      "loss": 4.964,
      "mean_token_accuracy": 0.20609863400459288,
      "num_tokens": 56551123.0,
      "step": 32595
    },
    {
      "entropy": 5.677213954925537,
      "epoch": 2.5363936977241783,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004351966054840443,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21039691865444182,
      "num_tokens": 56559096.0,
      "step": 32600
    },
    {
      "entropy": 5.637770748138427,
      "epoch": 2.5367827270958956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043517714252827556,
      "loss": 4.8885,
      "mean_token_accuracy": 0.21334285885095597,
      "num_tokens": 56567401.0,
      "step": 32605
    },
    {
      "entropy": 5.702492904663086,
      "epoch": 2.5371717564676133,
      "grad_norm": 1.25,
      "learning_rate": 0.0004351576771420737,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20664068758487703,
      "num_tokens": 56575456.0,
      "step": 32610
    },
    {
      "entropy": 5.742669725418091,
      "epoch": 2.537560785839331,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004351382093257339,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21570178121328354,
      "num_tokens": 56583810.0,
      "step": 32615
    },
    {
      "entropy": 5.663485383987426,
      "epoch": 2.5379498152110482,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004351187390795517,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20849667340517045,
      "num_tokens": 56592748.0,
      "step": 32620
    },
    {
      "entropy": 5.662202167510986,
      "epoch": 2.538338844582766,
      "grad_norm": 2.296875,
      "learning_rate": 0.00043509926640382244,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2230575293302536,
      "num_tokens": 56601461.0,
      "step": 32625
    },
    {
      "entropy": 5.644738578796387,
      "epoch": 2.5387278739544836,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004350797912988417,
      "loss": 4.94,
      "mean_token_accuracy": 0.20806770026683807,
      "num_tokens": 56610227.0,
      "step": 32630
    },
    {
      "entropy": 5.701452827453613,
      "epoch": 2.5391169033262013,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004350603137649049,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2069085046648979,
      "num_tokens": 56619236.0,
      "step": 32635
    },
    {
      "entropy": 5.689003038406372,
      "epoch": 2.5395059326979186,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004350408338023075,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21578714549541472,
      "num_tokens": 56628022.0,
      "step": 32640
    },
    {
      "entropy": 5.777026653289795,
      "epoch": 2.5398949620696363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004350213514113453,
      "loss": 5.048,
      "mean_token_accuracy": 0.20260293632745743,
      "num_tokens": 56637191.0,
      "step": 32645
    },
    {
      "entropy": 5.731728506088257,
      "epoch": 2.540283991441354,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043500186659231377,
      "loss": 4.9546,
      "mean_token_accuracy": 0.20533868819475173,
      "num_tokens": 56646395.0,
      "step": 32650
    },
    {
      "entropy": 5.647430086135865,
      "epoch": 2.5406730208130712,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004349823793455086,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21665948778390884,
      "num_tokens": 56655325.0,
      "step": 32655
    },
    {
      "entropy": 5.639145088195801,
      "epoch": 2.541062050184789,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043496288967122547,
      "loss": 4.9692,
      "mean_token_accuracy": 0.20423408448696137,
      "num_tokens": 56664470.0,
      "step": 32660
    },
    {
      "entropy": 5.747791528701782,
      "epoch": 2.5414510795565066,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043494339756976007,
      "loss": 4.9319,
      "mean_token_accuracy": 0.20249375700950623,
      "num_tokens": 56671883.0,
      "step": 32665
    },
    {
      "entropy": 5.749244356155396,
      "epoch": 2.5418401089282243,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004349239030414082,
      "loss": 5.0183,
      "mean_token_accuracy": 0.20322106331586837,
      "num_tokens": 56680877.0,
      "step": 32670
    },
    {
      "entropy": 5.686707925796509,
      "epoch": 2.5422291382999416,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043490440608646557,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2049104690551758,
      "num_tokens": 56689143.0,
      "step": 32675
    },
    {
      "entropy": 5.688883447647095,
      "epoch": 2.5426181676716593,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043488490670522817,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21230608522891997,
      "num_tokens": 56698696.0,
      "step": 32680
    },
    {
      "entropy": 5.6704846858978275,
      "epoch": 2.5430071970433765,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004348654048979918,
      "loss": 4.9525,
      "mean_token_accuracy": 0.20557556301355362,
      "num_tokens": 56707805.0,
      "step": 32685
    },
    {
      "entropy": 5.697279977798462,
      "epoch": 2.543396226415094,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004348459006650523,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2059640809893608,
      "num_tokens": 56716341.0,
      "step": 32690
    },
    {
      "entropy": 5.7409766674041744,
      "epoch": 2.543785255786812,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043482639400670576,
      "loss": 5.1121,
      "mean_token_accuracy": 0.19511549472808837,
      "num_tokens": 56725296.0,
      "step": 32695
    },
    {
      "entropy": 5.760735893249512,
      "epoch": 2.5441742851585296,
      "grad_norm": 1.203125,
      "learning_rate": 0.000434806884923248,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19755200296640396,
      "num_tokens": 56735118.0,
      "step": 32700
    },
    {
      "entropy": 5.721780824661255,
      "epoch": 2.5445633145302473,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004347873734149752,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21299345344305037,
      "num_tokens": 56743880.0,
      "step": 32705
    },
    {
      "entropy": 5.7009447574615475,
      "epoch": 2.5449523439019646,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043476785948218325,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2028905615210533,
      "num_tokens": 56751945.0,
      "step": 32710
    },
    {
      "entropy": 5.619579887390136,
      "epoch": 2.5453413732736823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043474834312516835,
      "loss": 4.9005,
      "mean_token_accuracy": 0.20717936456203462,
      "num_tokens": 56760645.0,
      "step": 32715
    },
    {
      "entropy": 5.762608289718628,
      "epoch": 2.5457304026453995,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004347288243442266,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20624287873506547,
      "num_tokens": 56769496.0,
      "step": 32720
    },
    {
      "entropy": 5.68577299118042,
      "epoch": 2.546119432017117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004347093031396543,
      "loss": 4.8918,
      "mean_token_accuracy": 0.20478617995977402,
      "num_tokens": 56778313.0,
      "step": 32725
    },
    {
      "entropy": 5.7061316013336185,
      "epoch": 2.546508461388835,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004346897795117474,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2017899841070175,
      "num_tokens": 56786210.0,
      "step": 32730
    },
    {
      "entropy": 5.691309309005737,
      "epoch": 2.5468974907605526,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004346702534608023,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20654103606939317,
      "num_tokens": 56795641.0,
      "step": 32735
    },
    {
      "entropy": 5.722550964355468,
      "epoch": 2.54728652013227,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004346507249871153,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19576369673013688,
      "num_tokens": 56804625.0,
      "step": 32740
    },
    {
      "entropy": 5.700703859329224,
      "epoch": 2.5476755495039876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004346311940909826,
      "loss": 4.9123,
      "mean_token_accuracy": 0.20989010035991668,
      "num_tokens": 56812835.0,
      "step": 32745
    },
    {
      "entropy": 5.689013433456421,
      "epoch": 2.5480645788757053,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043461166077270056,
      "loss": 4.9719,
      "mean_token_accuracy": 0.20878061801195144,
      "num_tokens": 56822005.0,
      "step": 32750
    },
    {
      "entropy": 5.697546815872192,
      "epoch": 2.5484536082474225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043459212503256566,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21193284541368484,
      "num_tokens": 56830490.0,
      "step": 32755
    },
    {
      "entropy": 5.735522079467773,
      "epoch": 2.54884263761914,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004345725868708743,
      "loss": 4.9823,
      "mean_token_accuracy": 0.20693827271461487,
      "num_tokens": 56839699.0,
      "step": 32760
    },
    {
      "entropy": 5.707082605361938,
      "epoch": 2.549231666990858,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043455304628792294,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2002649948000908,
      "num_tokens": 56848664.0,
      "step": 32765
    },
    {
      "entropy": 5.735581731796264,
      "epoch": 2.5496206963625756,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000434533503284008,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20198367685079574,
      "num_tokens": 56856567.0,
      "step": 32770
    },
    {
      "entropy": 5.799953413009644,
      "epoch": 2.550009725734293,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004345139578594261,
      "loss": 5.0604,
      "mean_token_accuracy": 0.19971122741699218,
      "num_tokens": 56864479.0,
      "step": 32775
    },
    {
      "entropy": 5.762111473083496,
      "epoch": 2.5503987551060106,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004344944100144738,
      "loss": 5.1018,
      "mean_token_accuracy": 0.19957056790590286,
      "num_tokens": 56872976.0,
      "step": 32780
    },
    {
      "entropy": 5.741881942749023,
      "epoch": 2.550787784477728,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043447485974944765,
      "loss": 4.9622,
      "mean_token_accuracy": 0.20680367350578308,
      "num_tokens": 56882146.0,
      "step": 32785
    },
    {
      "entropy": 5.699673795700074,
      "epoch": 2.5511768138494455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043445530706464437,
      "loss": 4.928,
      "mean_token_accuracy": 0.2110580861568451,
      "num_tokens": 56890267.0,
      "step": 32790
    },
    {
      "entropy": 5.609733295440674,
      "epoch": 2.551565843221163,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004344357519603606,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21562192738056182,
      "num_tokens": 56897887.0,
      "step": 32795
    },
    {
      "entropy": 5.678753471374511,
      "epoch": 2.551954872592881,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004344161944368931,
      "loss": 5.0536,
      "mean_token_accuracy": 0.19957194328308106,
      "num_tokens": 56906396.0,
      "step": 32800
    },
    {
      "entropy": 5.737152433395385,
      "epoch": 2.5523439019645986,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004343966344945385,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20492859780788422,
      "num_tokens": 56914552.0,
      "step": 32805
    },
    {
      "entropy": 5.657803535461426,
      "epoch": 2.552732931336316,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043437707213359375,
      "loss": 4.8858,
      "mean_token_accuracy": 0.20682284235954285,
      "num_tokens": 56921976.0,
      "step": 32810
    },
    {
      "entropy": 5.674916791915893,
      "epoch": 2.5531219607080335,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043435750735435554,
      "loss": 5.0288,
      "mean_token_accuracy": 0.19863365441560746,
      "num_tokens": 56930518.0,
      "step": 32815
    },
    {
      "entropy": 5.727136039733887,
      "epoch": 2.553510990079751,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043433794015712085,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20179325193166733,
      "num_tokens": 56939742.0,
      "step": 32820
    },
    {
      "entropy": 5.704491424560547,
      "epoch": 2.5539000194514685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004343183705421865,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20859729647636413,
      "num_tokens": 56947925.0,
      "step": 32825
    },
    {
      "entropy": 5.676928091049194,
      "epoch": 2.554289048823186,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004342987985098495,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20821321457624437,
      "num_tokens": 56956546.0,
      "step": 32830
    },
    {
      "entropy": 5.708922672271728,
      "epoch": 2.554678078194904,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043427922406040676,
      "loss": 5.0502,
      "mean_token_accuracy": 0.19950754642486573,
      "num_tokens": 56965323.0,
      "step": 32835
    },
    {
      "entropy": 5.775407505035401,
      "epoch": 2.555067107566621,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043425964719415525,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20044715702533722,
      "num_tokens": 56974566.0,
      "step": 32840
    },
    {
      "entropy": 5.671373128890991,
      "epoch": 2.555456136938339,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043424006791139213,
      "loss": 4.994,
      "mean_token_accuracy": 0.21012261956930162,
      "num_tokens": 56983565.0,
      "step": 32845
    },
    {
      "entropy": 5.631256484985352,
      "epoch": 2.5558451663100565,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043422048621241443,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21466976404190063,
      "num_tokens": 56992011.0,
      "step": 32850
    },
    {
      "entropy": 5.70928168296814,
      "epoch": 2.556234195681774,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043420090209751933,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2065294861793518,
      "num_tokens": 57001012.0,
      "step": 32855
    },
    {
      "entropy": 5.754943895339966,
      "epoch": 2.5566232250534915,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004341813155670039,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21120057702064515,
      "num_tokens": 57009572.0,
      "step": 32860
    },
    {
      "entropy": 5.725535202026367,
      "epoch": 2.557012254425209,
      "grad_norm": 1.25,
      "learning_rate": 0.0004341617266211654,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20633513629436492,
      "num_tokens": 57018173.0,
      "step": 32865
    },
    {
      "entropy": 5.614037418365479,
      "epoch": 2.557401283796927,
      "grad_norm": 1.140625,
      "learning_rate": 0.000434142135260301,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20994732081890105,
      "num_tokens": 57027565.0,
      "step": 32870
    },
    {
      "entropy": 5.733452415466308,
      "epoch": 2.557790313168644,
      "grad_norm": 1.28125,
      "learning_rate": 0.000434122541484708,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20279324501752855,
      "num_tokens": 57036660.0,
      "step": 32875
    },
    {
      "entropy": 5.733560180664062,
      "epoch": 2.558179342540362,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004341029452946837,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20620784312486648,
      "num_tokens": 57045461.0,
      "step": 32880
    },
    {
      "entropy": 5.733819437026978,
      "epoch": 2.558568371912079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004340833466905254,
      "loss": 5.021,
      "mean_token_accuracy": 0.20758386552333832,
      "num_tokens": 57053566.0,
      "step": 32885
    },
    {
      "entropy": 5.681113862991333,
      "epoch": 2.558957401283797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043406374567253055,
      "loss": 5.0221,
      "mean_token_accuracy": 0.19690076112747193,
      "num_tokens": 57061530.0,
      "step": 32890
    },
    {
      "entropy": 5.596763801574707,
      "epoch": 2.5593464306555145,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004340441422409965,
      "loss": 4.7902,
      "mean_token_accuracy": 0.2219374492764473,
      "num_tokens": 57069645.0,
      "step": 32895
    },
    {
      "entropy": 5.663645696640015,
      "epoch": 2.559735460027232,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340245363962209,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21029461771249772,
      "num_tokens": 57078998.0,
      "step": 32900
    },
    {
      "entropy": 5.719699048995972,
      "epoch": 2.56012448939895,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043400492813850083,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2050677388906479,
      "num_tokens": 57087363.0,
      "step": 32905
    },
    {
      "entropy": 5.732450151443482,
      "epoch": 2.560513518770667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043398531746813426,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20135474056005478,
      "num_tokens": 57096242.0,
      "step": 32910
    },
    {
      "entropy": 5.68996114730835,
      "epoch": 2.560902548142385,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043396570438541845,
      "loss": 4.9675,
      "mean_token_accuracy": 0.20162001699209214,
      "num_tokens": 57104791.0,
      "step": 32915
    },
    {
      "entropy": 5.713347291946411,
      "epoch": 2.561291577514102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004339460888906512,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2021369069814682,
      "num_tokens": 57113170.0,
      "step": 32920
    },
    {
      "entropy": 5.712580060958862,
      "epoch": 2.5616806068858198,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004339264709841299,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20526072680950164,
      "num_tokens": 57121746.0,
      "step": 32925
    },
    {
      "entropy": 5.724859285354614,
      "epoch": 2.5620696362575375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043390685066615244,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21671376675367354,
      "num_tokens": 57130347.0,
      "step": 32930
    },
    {
      "entropy": 5.683323574066162,
      "epoch": 2.562458665629255,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004338872279370164,
      "loss": 4.952,
      "mean_token_accuracy": 0.20928529798984527,
      "num_tokens": 57138370.0,
      "step": 32935
    },
    {
      "entropy": 5.721246576309204,
      "epoch": 2.5628476950009724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004338676027970196,
      "loss": 5.0429,
      "mean_token_accuracy": 0.2026049852371216,
      "num_tokens": 57147582.0,
      "step": 32940
    },
    {
      "entropy": 5.725758504867554,
      "epoch": 2.56323672437269,
      "grad_norm": 1.25,
      "learning_rate": 0.00043384797524645977,
      "loss": 4.973,
      "mean_token_accuracy": 0.20203132778406144,
      "num_tokens": 57156477.0,
      "step": 32945
    },
    {
      "entropy": 5.6618485927581785,
      "epoch": 2.563625753744408,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004338283452856348,
      "loss": 4.8953,
      "mean_token_accuracy": 0.20873745828866958,
      "num_tokens": 57164463.0,
      "step": 32950
    },
    {
      "entropy": 5.666740083694458,
      "epoch": 2.564014783116125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004338087129148425,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2036084160208702,
      "num_tokens": 57173640.0,
      "step": 32955
    },
    {
      "entropy": 5.723639822006225,
      "epoch": 2.5644038124878428,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043378907813438066,
      "loss": 5.0605,
      "mean_token_accuracy": 0.19265889078378678,
      "num_tokens": 57182239.0,
      "step": 32960
    },
    {
      "entropy": 5.7019875049591064,
      "epoch": 2.5647928418595605,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043376944094454734,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20237672030925752,
      "num_tokens": 57192411.0,
      "step": 32965
    },
    {
      "entropy": 5.79729437828064,
      "epoch": 2.565181871231278,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043374980134564044,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21248824894428253,
      "num_tokens": 57200888.0,
      "step": 32970
    },
    {
      "entropy": 5.722712993621826,
      "epoch": 2.5655709006029954,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043373015933795804,
      "loss": 4.999,
      "mean_token_accuracy": 0.2108399361371994,
      "num_tokens": 57209816.0,
      "step": 32975
    },
    {
      "entropy": 5.619099569320679,
      "epoch": 2.565959929974713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004337105149217981,
      "loss": 4.905,
      "mean_token_accuracy": 0.2130533903837204,
      "num_tokens": 57218190.0,
      "step": 32980
    },
    {
      "entropy": 5.648235940933228,
      "epoch": 2.566348959346431,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043369086809745875,
      "loss": 4.9796,
      "mean_token_accuracy": 0.20216379314661026,
      "num_tokens": 57226697.0,
      "step": 32985
    },
    {
      "entropy": 5.712060499191284,
      "epoch": 2.566737988718148,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043367121886523795,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20425665825605394,
      "num_tokens": 57235058.0,
      "step": 32990
    },
    {
      "entropy": 5.678001022338867,
      "epoch": 2.5671270180898658,
      "grad_norm": 1.171875,
      "learning_rate": 0.000433651567225434,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20356193482875823,
      "num_tokens": 57244344.0,
      "step": 32995
    },
    {
      "entropy": 5.737708473205567,
      "epoch": 2.5675160474615835,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004336319131783451,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20553993582725524,
      "num_tokens": 57253502.0,
      "step": 33000
    },
    {
      "epoch": 2.5675160474615835,
      "eval_entropy": 5.458543120632961,
      "eval_loss": 5.062242031097412,
      "eval_mean_token_accuracy": 0.21182624165029681,
      "eval_num_tokens": 57253502.0,
      "eval_runtime": 21.9341,
      "eval_samples_per_second": 1894.672,
      "eval_steps_per_second": 236.845,
      "step": 33000
    },
    {
      "entropy": 5.678699731826782,
      "epoch": 2.567905076833301,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043361225672426933,
      "loss": 4.927,
      "mean_token_accuracy": 0.211125884950161,
      "num_tokens": 57262552.0,
      "step": 33005
    },
    {
      "entropy": 5.7359199047088625,
      "epoch": 2.5682941062050184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004335925978635051,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20972345769405365,
      "num_tokens": 57271608.0,
      "step": 33010
    },
    {
      "entropy": 5.747376298904419,
      "epoch": 2.568683135576736,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043357293659635057,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2043384149670601,
      "num_tokens": 57280112.0,
      "step": 33015
    },
    {
      "entropy": 5.703431558609009,
      "epoch": 2.5690721649484534,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004335532729231041,
      "loss": 4.9561,
      "mean_token_accuracy": 0.19906243234872817,
      "num_tokens": 57289565.0,
      "step": 33020
    },
    {
      "entropy": 5.782798433303833,
      "epoch": 2.569461194320171,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043353360684406415,
      "loss": 5.08,
      "mean_token_accuracy": 0.19317682534456254,
      "num_tokens": 57298879.0,
      "step": 33025
    },
    {
      "entropy": 5.687327432632446,
      "epoch": 2.5698502236918888,
      "grad_norm": 1.21875,
      "learning_rate": 0.000433513938359529,
      "loss": 4.9491,
      "mean_token_accuracy": 0.20716072320938111,
      "num_tokens": 57307508.0,
      "step": 33030
    },
    {
      "entropy": 5.716709232330322,
      "epoch": 2.5702392530636065,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004334942674697971,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20679048746824263,
      "num_tokens": 57316187.0,
      "step": 33035
    },
    {
      "entropy": 5.689046382904053,
      "epoch": 2.5706282824353237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043347459417516696,
      "loss": 4.9026,
      "mean_token_accuracy": 0.21230379343032837,
      "num_tokens": 57325025.0,
      "step": 33040
    },
    {
      "entropy": 5.709733676910401,
      "epoch": 2.5710173118070414,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004334549184759371,
      "loss": 4.9723,
      "mean_token_accuracy": 0.20071024894714357,
      "num_tokens": 57333427.0,
      "step": 33045
    },
    {
      "entropy": 5.687884521484375,
      "epoch": 2.571406341178759,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004334352403724062,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21026103645563127,
      "num_tokens": 57342088.0,
      "step": 33050
    },
    {
      "entropy": 5.788322496414184,
      "epoch": 2.5717953705504764,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043341555986487254,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20250421166419982,
      "num_tokens": 57350075.0,
      "step": 33055
    },
    {
      "entropy": 5.691873407363891,
      "epoch": 2.572184399922194,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004333958769536349,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2098337396979332,
      "num_tokens": 57358517.0,
      "step": 33060
    },
    {
      "entropy": 5.680976438522339,
      "epoch": 2.5725734292939118,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004333761916389921,
      "loss": 5.0313,
      "mean_token_accuracy": 0.20166352540254592,
      "num_tokens": 57367980.0,
      "step": 33065
    },
    {
      "entropy": 5.774142694473267,
      "epoch": 2.5729624586656294,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004333565039212425,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2047487825155258,
      "num_tokens": 57376575.0,
      "step": 33070
    },
    {
      "entropy": 5.737471961975098,
      "epoch": 2.5733514880373467,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004333368138006851,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2064204216003418,
      "num_tokens": 57384950.0,
      "step": 33075
    },
    {
      "entropy": 5.711123847961426,
      "epoch": 2.5737405174090644,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004333171212776185,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2065430536866188,
      "num_tokens": 57393895.0,
      "step": 33080
    },
    {
      "entropy": 5.732526874542236,
      "epoch": 2.574129546780782,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004332974263523416,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2021734520792961,
      "num_tokens": 57403579.0,
      "step": 33085
    },
    {
      "entropy": 5.723679494857788,
      "epoch": 2.5745185761524993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043327772902515327,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20182851552963257,
      "num_tokens": 57411975.0,
      "step": 33090
    },
    {
      "entropy": 5.703841257095337,
      "epoch": 2.574907605524217,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004332580292963523,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2088599056005478,
      "num_tokens": 57419876.0,
      "step": 33095
    },
    {
      "entropy": 5.6517270565032955,
      "epoch": 2.5752966348959347,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004332383271662376,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2118717223405838,
      "num_tokens": 57428842.0,
      "step": 33100
    },
    {
      "entropy": 5.655807828903198,
      "epoch": 2.5756856642676524,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043321862263510816,
      "loss": 4.8904,
      "mean_token_accuracy": 0.21277089267969132,
      "num_tokens": 57437239.0,
      "step": 33105
    },
    {
      "entropy": 5.7198327541351315,
      "epoch": 2.5760746936393697,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004331989157032629,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20722661167383194,
      "num_tokens": 57446172.0,
      "step": 33110
    },
    {
      "entropy": 5.699471426010132,
      "epoch": 2.5764637230110874,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043317920637100097,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2135789304971695,
      "num_tokens": 57454907.0,
      "step": 33115
    },
    {
      "entropy": 5.628232955932617,
      "epoch": 2.5768527523828046,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004331594946386213,
      "loss": 4.9576,
      "mean_token_accuracy": 0.20341273993253708,
      "num_tokens": 57464309.0,
      "step": 33120
    },
    {
      "entropy": 5.702980327606201,
      "epoch": 2.5772417817545223,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043313978050642303,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21044119000434874,
      "num_tokens": 57472920.0,
      "step": 33125
    },
    {
      "entropy": 5.791906118392944,
      "epoch": 2.57763081112624,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004331200639747053,
      "loss": 5.0226,
      "mean_token_accuracy": 0.20145918428897858,
      "num_tokens": 57481608.0,
      "step": 33130
    },
    {
      "entropy": 5.67944302558899,
      "epoch": 2.5780198404979577,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004331003450437674,
      "loss": 4.8894,
      "mean_token_accuracy": 0.20856887102127075,
      "num_tokens": 57490373.0,
      "step": 33135
    },
    {
      "entropy": 5.664708471298217,
      "epoch": 2.5784088698696754,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004330806237139083,
      "loss": 4.9553,
      "mean_token_accuracy": 0.20704852491617204,
      "num_tokens": 57499224.0,
      "step": 33140
    },
    {
      "entropy": 5.672581577301026,
      "epoch": 2.5787978992413927,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004330608999854273,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20406946241855622,
      "num_tokens": 57508496.0,
      "step": 33145
    },
    {
      "entropy": 5.7370223045349125,
      "epoch": 2.5791869286131104,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004330411738586238,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2168922543525696,
      "num_tokens": 57517348.0,
      "step": 33150
    },
    {
      "entropy": 5.731233692169189,
      "epoch": 2.5795759579848276,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000433021445333797,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2071181669831276,
      "num_tokens": 57525511.0,
      "step": 33155
    },
    {
      "entropy": 5.734325408935547,
      "epoch": 2.5799649873565453,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043300171441124633,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20513100028038025,
      "num_tokens": 57534445.0,
      "step": 33160
    },
    {
      "entropy": 5.681735706329346,
      "epoch": 2.580354016728263,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004329819810912711,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2070796549320221,
      "num_tokens": 57542543.0,
      "step": 33165
    },
    {
      "entropy": 5.743970775604248,
      "epoch": 2.5807430460999807,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043296224537417075,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20299624353647233,
      "num_tokens": 57551407.0,
      "step": 33170
    },
    {
      "entropy": 5.675380754470825,
      "epoch": 2.581132075471698,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043294250726024473,
      "loss": 4.8817,
      "mean_token_accuracy": 0.21594064086675643,
      "num_tokens": 57559576.0,
      "step": 33175
    },
    {
      "entropy": 5.673430109024048,
      "epoch": 2.5815211048434157,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004329227667497926,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2143855422735214,
      "num_tokens": 57567893.0,
      "step": 33180
    },
    {
      "entropy": 5.696310329437256,
      "epoch": 2.5819101342151334,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043290302384311373,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21030088365077973,
      "num_tokens": 57576168.0,
      "step": 33185
    },
    {
      "entropy": 5.710881757736206,
      "epoch": 2.5822991635868506,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043288327854050794,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20549533665180206,
      "num_tokens": 57584540.0,
      "step": 33190
    },
    {
      "entropy": 5.710496234893799,
      "epoch": 2.5826881929585683,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043286353084227467,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2025822803378105,
      "num_tokens": 57592315.0,
      "step": 33195
    },
    {
      "entropy": 5.592957592010498,
      "epoch": 2.583077222330286,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043284378074871354,
      "loss": 4.8778,
      "mean_token_accuracy": 0.20903212130069732,
      "num_tokens": 57600362.0,
      "step": 33200
    },
    {
      "entropy": 5.701590967178345,
      "epoch": 2.5834662517020037,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004328240282601243,
      "loss": 4.9844,
      "mean_token_accuracy": 0.19968596398830413,
      "num_tokens": 57608852.0,
      "step": 33205
    },
    {
      "entropy": 5.688631963729859,
      "epoch": 2.583855281073721,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004328042733768066,
      "loss": 4.982,
      "mean_token_accuracy": 0.2107900321483612,
      "num_tokens": 57617081.0,
      "step": 33210
    },
    {
      "entropy": 5.7073366165161135,
      "epoch": 2.5842443104454387,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043278451609906027,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21303458958864213,
      "num_tokens": 57625378.0,
      "step": 33215
    },
    {
      "entropy": 5.6871226787567135,
      "epoch": 2.584633339817156,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043276475642718503,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2021857351064682,
      "num_tokens": 57634025.0,
      "step": 33220
    },
    {
      "entropy": 5.638347482681274,
      "epoch": 2.5850223691888736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004327449943614808,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21259480118751525,
      "num_tokens": 57642794.0,
      "step": 33225
    },
    {
      "entropy": 5.678907871246338,
      "epoch": 2.5854113985605913,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043272522990224727,
      "loss": 4.9623,
      "mean_token_accuracy": 0.20482416003942489,
      "num_tokens": 57651330.0,
      "step": 33230
    },
    {
      "entropy": 5.720505142211914,
      "epoch": 2.585800427932309,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004327054630497844,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20163153111934662,
      "num_tokens": 57660673.0,
      "step": 33235
    },
    {
      "entropy": 5.680821514129638,
      "epoch": 2.5861894573040267,
      "grad_norm": 1.25,
      "learning_rate": 0.00043268569380439223,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21334424912929534,
      "num_tokens": 57669349.0,
      "step": 33240
    },
    {
      "entropy": 5.7376665592193605,
      "epoch": 2.586578486675744,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004326659221663705,
      "loss": 5.0829,
      "mean_token_accuracy": 0.19897127896547318,
      "num_tokens": 57678904.0,
      "step": 33245
    },
    {
      "entropy": 5.75347375869751,
      "epoch": 2.5869675160474617,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004326461481360194,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20304772406816482,
      "num_tokens": 57686957.0,
      "step": 33250
    },
    {
      "entropy": 5.6729227066040036,
      "epoch": 2.587356545419179,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043262637171363906,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21375613063573837,
      "num_tokens": 57695201.0,
      "step": 33255
    },
    {
      "entropy": 5.750540733337402,
      "epoch": 2.5877455747908966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043260659289952926,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2031563937664032,
      "num_tokens": 57703638.0,
      "step": 33260
    },
    {
      "entropy": 5.748252725601196,
      "epoch": 2.5881346041626143,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043258681169399034,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20177083015441893,
      "num_tokens": 57713427.0,
      "step": 33265
    },
    {
      "entropy": 5.758066654205322,
      "epoch": 2.588523633534332,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043256702809732234,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20760197937488556,
      "num_tokens": 57721953.0,
      "step": 33270
    },
    {
      "entropy": 5.773918676376343,
      "epoch": 2.5889126629060493,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004325472421098255,
      "loss": 5.0367,
      "mean_token_accuracy": 0.19845963269472122,
      "num_tokens": 57730913.0,
      "step": 33275
    },
    {
      "entropy": 5.628821468353271,
      "epoch": 2.589301692277767,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043252745373180006,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2092765063047409,
      "num_tokens": 57739959.0,
      "step": 33280
    },
    {
      "entropy": 5.695522499084473,
      "epoch": 2.5896907216494847,
      "grad_norm": 1.25,
      "learning_rate": 0.0004325076629635462,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21130251437425612,
      "num_tokens": 57748226.0,
      "step": 33285
    },
    {
      "entropy": 5.659872817993164,
      "epoch": 2.590079751021202,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043248786980536424,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2050025463104248,
      "num_tokens": 57756521.0,
      "step": 33290
    },
    {
      "entropy": 5.737359952926636,
      "epoch": 2.5904687803929196,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004324680742575545,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20454122871160507,
      "num_tokens": 57765402.0,
      "step": 33295
    },
    {
      "entropy": 5.636382102966309,
      "epoch": 2.5908578097646373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043244827632041744,
      "loss": 4.9083,
      "mean_token_accuracy": 0.20721125453710557,
      "num_tokens": 57773645.0,
      "step": 33300
    },
    {
      "entropy": 5.749105024337768,
      "epoch": 2.591246839136355,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004324284759942534,
      "loss": 5.0899,
      "mean_token_accuracy": 0.19986154437065123,
      "num_tokens": 57781735.0,
      "step": 33305
    },
    {
      "entropy": 5.683712577819824,
      "epoch": 2.5916358685080723,
      "grad_norm": 1.25,
      "learning_rate": 0.0004324086732793627,
      "loss": 4.9596,
      "mean_token_accuracy": 0.20780885815620423,
      "num_tokens": 57790046.0,
      "step": 33310
    },
    {
      "entropy": 5.667308282852173,
      "epoch": 2.59202489787979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043238886817604605,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21017854362726213,
      "num_tokens": 57799069.0,
      "step": 33315
    },
    {
      "entropy": 5.735767030715943,
      "epoch": 2.5924139272515077,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004323690606846039,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20101184397935867,
      "num_tokens": 57808292.0,
      "step": 33320
    },
    {
      "entropy": 5.702739429473877,
      "epoch": 2.592802956623225,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043234925080533657,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19922055453062057,
      "num_tokens": 57816373.0,
      "step": 33325
    },
    {
      "entropy": 5.7166204929351805,
      "epoch": 2.5931919859949426,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004323294385385448,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2023086205124855,
      "num_tokens": 57825564.0,
      "step": 33330
    },
    {
      "entropy": 5.665809106826782,
      "epoch": 2.5935810153666603,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004323096238845293,
      "loss": 4.9934,
      "mean_token_accuracy": 0.1973897024989128,
      "num_tokens": 57834664.0,
      "step": 33335
    },
    {
      "entropy": 5.6826859474182125,
      "epoch": 2.593970044738378,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004322898068435906,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20228078961372375,
      "num_tokens": 57842910.0,
      "step": 33340
    },
    {
      "entropy": 5.689031314849854,
      "epoch": 2.5943590741100953,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004322699874160294,
      "loss": 4.8738,
      "mean_token_accuracy": 0.20522424280643464,
      "num_tokens": 57852038.0,
      "step": 33345
    },
    {
      "entropy": 5.75601167678833,
      "epoch": 2.594748103481813,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043225016560214654,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20295971482992173,
      "num_tokens": 57860578.0,
      "step": 33350
    },
    {
      "entropy": 5.719333791732788,
      "epoch": 2.59513713285353,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043223034140224266,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20018404126167297,
      "num_tokens": 57869124.0,
      "step": 33355
    },
    {
      "entropy": 5.690766048431397,
      "epoch": 2.595526162225248,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004322105148166187,
      "loss": 4.8617,
      "mean_token_accuracy": 0.21584379822015762,
      "num_tokens": 57877771.0,
      "step": 33360
    },
    {
      "entropy": 5.634674549102783,
      "epoch": 2.5959151915969656,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043219068584557526,
      "loss": 4.9055,
      "mean_token_accuracy": 0.20817106068134308,
      "num_tokens": 57885883.0,
      "step": 33365
    },
    {
      "entropy": 5.6868212699890135,
      "epoch": 2.5963042209686833,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004321708544894134,
      "loss": 4.94,
      "mean_token_accuracy": 0.20734603106975555,
      "num_tokens": 57894287.0,
      "step": 33370
    },
    {
      "entropy": 5.754468965530395,
      "epoch": 2.5966932503404005,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043215102074843403,
      "loss": 5.007,
      "mean_token_accuracy": 0.2059357762336731,
      "num_tokens": 57902046.0,
      "step": 33375
    },
    {
      "entropy": 5.7047460079193115,
      "epoch": 2.5970822797121182,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043213118462293796,
      "loss": 4.9544,
      "mean_token_accuracy": 0.20955901145935057,
      "num_tokens": 57910776.0,
      "step": 33380
    },
    {
      "entropy": 5.750837087631226,
      "epoch": 2.597471309083836,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004321113461132264,
      "loss": 5.1002,
      "mean_token_accuracy": 0.19460802674293518,
      "num_tokens": 57920242.0,
      "step": 33385
    },
    {
      "entropy": 5.686353969573974,
      "epoch": 2.597860338455553,
      "grad_norm": 1.25,
      "learning_rate": 0.00043209150521960016,
      "loss": 4.9501,
      "mean_token_accuracy": 0.20909306555986404,
      "num_tokens": 57928902.0,
      "step": 33390
    },
    {
      "entropy": 5.6704418659210205,
      "epoch": 2.598249367827271,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043207166194236037,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2170700505375862,
      "num_tokens": 57937660.0,
      "step": 33395
    },
    {
      "entropy": 5.679188251495361,
      "epoch": 2.5986383971989886,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004320518162818082,
      "loss": 4.9984,
      "mean_token_accuracy": 0.19788794219493866,
      "num_tokens": 57946483.0,
      "step": 33400
    },
    {
      "entropy": 5.6116539478302006,
      "epoch": 2.5990274265707063,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004320319682382445,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2063265785574913,
      "num_tokens": 57955963.0,
      "step": 33405
    },
    {
      "entropy": 5.683806324005127,
      "epoch": 2.5994164559424235,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043201211781197075,
      "loss": 4.9478,
      "mean_token_accuracy": 0.20308968275785447,
      "num_tokens": 57963750.0,
      "step": 33410
    },
    {
      "entropy": 5.714489126205445,
      "epoch": 2.5998054853141412,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043199226500328805,
      "loss": 4.978,
      "mean_token_accuracy": 0.20906989127397538,
      "num_tokens": 57972884.0,
      "step": 33415
    },
    {
      "entropy": 5.754586982727051,
      "epoch": 2.600194514685859,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004319724098124975,
      "loss": 5.0405,
      "mean_token_accuracy": 0.2036917746067047,
      "num_tokens": 57982143.0,
      "step": 33420
    },
    {
      "entropy": 5.732813024520874,
      "epoch": 2.600583544057576,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043195255223990055,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21198856085538864,
      "num_tokens": 57990954.0,
      "step": 33425
    },
    {
      "entropy": 5.74173583984375,
      "epoch": 2.600972573429294,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004319326922857984,
      "loss": 5.0284,
      "mean_token_accuracy": 0.19902270138263703,
      "num_tokens": 57998662.0,
      "step": 33430
    },
    {
      "entropy": 5.759254884719849,
      "epoch": 2.6013616028010116,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004319128299504925,
      "loss": 4.958,
      "mean_token_accuracy": 0.2119678348302841,
      "num_tokens": 58006956.0,
      "step": 33435
    },
    {
      "entropy": 5.703525257110596,
      "epoch": 2.6017506321727293,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043189296523428407,
      "loss": 4.9632,
      "mean_token_accuracy": 0.20530094504356383,
      "num_tokens": 58015390.0,
      "step": 33440
    },
    {
      "entropy": 5.6485944271087645,
      "epoch": 2.6021396615444465,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043187309813747464,
      "loss": 4.9215,
      "mean_token_accuracy": 0.20642916560173036,
      "num_tokens": 58024285.0,
      "step": 33445
    },
    {
      "entropy": 5.683557987213135,
      "epoch": 2.6025286909161642,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043185322866036565,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2070930927991867,
      "num_tokens": 58032411.0,
      "step": 33450
    },
    {
      "entropy": 5.694254636764526,
      "epoch": 2.6029177202878815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043183335680325866,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2118395522236824,
      "num_tokens": 58041815.0,
      "step": 33455
    },
    {
      "entropy": 5.7016747951507565,
      "epoch": 2.603306749659599,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043181348256645497,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2078417256474495,
      "num_tokens": 58050046.0,
      "step": 33460
    },
    {
      "entropy": 5.686977815628052,
      "epoch": 2.603695779031317,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043179360595025637,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2010343924164772,
      "num_tokens": 58058582.0,
      "step": 33465
    },
    {
      "entropy": 5.7467793941497805,
      "epoch": 2.6040848084030346,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004317737269549643,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20324262082576752,
      "num_tokens": 58067543.0,
      "step": 33470
    },
    {
      "entropy": 5.74483699798584,
      "epoch": 2.604473837774752,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004317538455808805,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21242719888687134,
      "num_tokens": 58076882.0,
      "step": 33475
    },
    {
      "entropy": 5.719075679779053,
      "epoch": 2.6048628671464695,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004317339618283065,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21096653044223784,
      "num_tokens": 58085775.0,
      "step": 33480
    },
    {
      "entropy": 5.705637693405151,
      "epoch": 2.6052518965181872,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004317140756975442,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21007921993732454,
      "num_tokens": 58094488.0,
      "step": 33485
    },
    {
      "entropy": 5.681598377227783,
      "epoch": 2.6056409258899045,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004316941871888951,
      "loss": 4.9455,
      "mean_token_accuracy": 0.20398151576519014,
      "num_tokens": 58103385.0,
      "step": 33490
    },
    {
      "entropy": 5.672256135940552,
      "epoch": 2.606029955261622,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043167429630266134,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21057169735431672,
      "num_tokens": 58111047.0,
      "step": 33495
    },
    {
      "entropy": 5.671245288848877,
      "epoch": 2.60641898463334,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004316544030391444,
      "loss": 4.9023,
      "mean_token_accuracy": 0.20500483065843583,
      "num_tokens": 58119676.0,
      "step": 33500
    },
    {
      "entropy": 5.607272672653198,
      "epoch": 2.6068080140050576,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004316345073986461,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20581505298614503,
      "num_tokens": 58128513.0,
      "step": 33505
    },
    {
      "entropy": 5.752149677276611,
      "epoch": 2.607197043376775,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004316146093814686,
      "loss": 5.0527,
      "mean_token_accuracy": 0.19899842888116837,
      "num_tokens": 58137859.0,
      "step": 33510
    },
    {
      "entropy": 5.791234254837036,
      "epoch": 2.6075860727484925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004315947089879137,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20264592617750168,
      "num_tokens": 58147314.0,
      "step": 33515
    },
    {
      "entropy": 5.664511585235596,
      "epoch": 2.60797510212021,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043157480621828327,
      "loss": 4.9706,
      "mean_token_accuracy": 0.20773538202047348,
      "num_tokens": 58156299.0,
      "step": 33520
    },
    {
      "entropy": 5.7481142520904545,
      "epoch": 2.6083641314919275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043155490107287935,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2047212690114975,
      "num_tokens": 58164799.0,
      "step": 33525
    },
    {
      "entropy": 5.75290207862854,
      "epoch": 2.608753160863645,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004315349935520041,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19438305348157883,
      "num_tokens": 58173901.0,
      "step": 33530
    },
    {
      "entropy": 5.658109855651856,
      "epoch": 2.609142190235363,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004315150836559594,
      "loss": 4.9201,
      "mean_token_accuracy": 0.20487455129623414,
      "num_tokens": 58183178.0,
      "step": 33535
    },
    {
      "entropy": 5.736638736724854,
      "epoch": 2.6095312196070806,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004314951713850474,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2029966190457344,
      "num_tokens": 58191556.0,
      "step": 33540
    },
    {
      "entropy": 5.67273097038269,
      "epoch": 2.609920248978798,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043147525673957026,
      "loss": 4.929,
      "mean_token_accuracy": 0.21210908591747285,
      "num_tokens": 58200720.0,
      "step": 33545
    },
    {
      "entropy": 5.6656747341156,
      "epoch": 2.6103092783505155,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043145533971983025,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2079576373100281,
      "num_tokens": 58209635.0,
      "step": 33550
    },
    {
      "entropy": 5.733290386199951,
      "epoch": 2.6106983077222328,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043143542032612935,
      "loss": 4.954,
      "mean_token_accuracy": 0.21250915229320527,
      "num_tokens": 58217926.0,
      "step": 33555
    },
    {
      "entropy": 5.678271389007568,
      "epoch": 2.6110873370939505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004314154985587701,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2145895317196846,
      "num_tokens": 58227086.0,
      "step": 33560
    },
    {
      "entropy": 5.696404838562012,
      "epoch": 2.611476366465668,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043139557441805457,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20539588034152984,
      "num_tokens": 58236110.0,
      "step": 33565
    },
    {
      "entropy": 5.670447063446045,
      "epoch": 2.611865395837386,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043137564790428503,
      "loss": 4.9004,
      "mean_token_accuracy": 0.2063648819923401,
      "num_tokens": 58244643.0,
      "step": 33570
    },
    {
      "entropy": 5.644930124282837,
      "epoch": 2.6122544252091036,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043135571901776397,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21398013681173325,
      "num_tokens": 58252700.0,
      "step": 33575
    },
    {
      "entropy": 5.692533731460571,
      "epoch": 2.612643454580821,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004313357877587937,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20700304806232453,
      "num_tokens": 58261274.0,
      "step": 33580
    },
    {
      "entropy": 5.76711802482605,
      "epoch": 2.6130324839525385,
      "grad_norm": 1.25,
      "learning_rate": 0.0004313158541276768,
      "loss": 5.0455,
      "mean_token_accuracy": 0.19952981173992157,
      "num_tokens": 58270957.0,
      "step": 33585
    },
    {
      "entropy": 5.719300222396851,
      "epoch": 2.6134215133242558,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004312959181247156,
      "loss": 4.9578,
      "mean_token_accuracy": 0.20843596011400223,
      "num_tokens": 58279454.0,
      "step": 33590
    },
    {
      "entropy": 5.705104827880859,
      "epoch": 2.6138105426959735,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004312759797502126,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20996173918247224,
      "num_tokens": 58289085.0,
      "step": 33595
    },
    {
      "entropy": 5.699274921417237,
      "epoch": 2.614199572067691,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043125603900447026,
      "loss": 5.0732,
      "mean_token_accuracy": 0.20572656244039536,
      "num_tokens": 58297901.0,
      "step": 33600
    },
    {
      "entropy": 5.730238533020019,
      "epoch": 2.614588601439409,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004312360958877913,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20761232674121857,
      "num_tokens": 58307045.0,
      "step": 33605
    },
    {
      "entropy": 5.755982208251953,
      "epoch": 2.614977630811126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004312161504004782,
      "loss": 5.0727,
      "mean_token_accuracy": 0.19465955197811127,
      "num_tokens": 58315767.0,
      "step": 33610
    },
    {
      "entropy": 5.766023921966553,
      "epoch": 2.615366660182844,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043119620254283375,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20591259598731995,
      "num_tokens": 58324405.0,
      "step": 33615
    },
    {
      "entropy": 5.670374727249145,
      "epoch": 2.6157556895545615,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043117625231516054,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21567057222127914,
      "num_tokens": 58332944.0,
      "step": 33620
    },
    {
      "entropy": 5.653215122222901,
      "epoch": 2.6161447189262788,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004311562997177612,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20514826625585555,
      "num_tokens": 58341569.0,
      "step": 33625
    },
    {
      "entropy": 5.677874994277954,
      "epoch": 2.6165337482979965,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004311363447509386,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20138383209705352,
      "num_tokens": 58350636.0,
      "step": 33630
    },
    {
      "entropy": 5.744486618041992,
      "epoch": 2.616922777669714,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004311163874149955,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2077670142054558,
      "num_tokens": 58359756.0,
      "step": 33635
    },
    {
      "entropy": 5.670916986465454,
      "epoch": 2.617311807041432,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043109642771023464,
      "loss": 4.8948,
      "mean_token_accuracy": 0.20582329332828522,
      "num_tokens": 58368086.0,
      "step": 33640
    },
    {
      "entropy": 5.727845144271851,
      "epoch": 2.617700836413149,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004310764656369591,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2002275764942169,
      "num_tokens": 58377597.0,
      "step": 33645
    },
    {
      "entropy": 5.704433441162109,
      "epoch": 2.618089865784867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004310565011954715,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2091353043913841,
      "num_tokens": 58385910.0,
      "step": 33650
    },
    {
      "entropy": 5.7768150806427006,
      "epoch": 2.618478895156584,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004310365343860748,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1979433074593544,
      "num_tokens": 58394879.0,
      "step": 33655
    },
    {
      "entropy": 5.658612442016602,
      "epoch": 2.6188679245283017,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043101656520907225,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2037658676505089,
      "num_tokens": 58403600.0,
      "step": 33660
    },
    {
      "entropy": 5.611598157882691,
      "epoch": 2.6192569539000194,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004309965936647665,
      "loss": 4.7898,
      "mean_token_accuracy": 0.22781807482242583,
      "num_tokens": 58412470.0,
      "step": 33665
    },
    {
      "entropy": 5.588254594802857,
      "epoch": 2.619645983271737,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004309766197534608,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2120668888092041,
      "num_tokens": 58421331.0,
      "step": 33670
    },
    {
      "entropy": 5.684003305435181,
      "epoch": 2.620035012643455,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043095664347545816,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2020118221640587,
      "num_tokens": 58429387.0,
      "step": 33675
    },
    {
      "entropy": 5.726310777664184,
      "epoch": 2.620424042015172,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004309366648310616,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20813866555690766,
      "num_tokens": 58438724.0,
      "step": 33680
    },
    {
      "entropy": 5.734375858306885,
      "epoch": 2.62081307138689,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043091668382057443,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20983607769012452,
      "num_tokens": 58448217.0,
      "step": 33685
    },
    {
      "entropy": 5.711432981491089,
      "epoch": 2.621202100758607,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043089670044429973,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21108884513378143,
      "num_tokens": 58457572.0,
      "step": 33690
    },
    {
      "entropy": 5.690261316299439,
      "epoch": 2.6215911301303247,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043087671470254076,
      "loss": 4.9463,
      "mean_token_accuracy": 0.20680274516344072,
      "num_tokens": 58466403.0,
      "step": 33695
    },
    {
      "entropy": 5.660318851470947,
      "epoch": 2.6219801595020424,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043085672659560077,
      "loss": 4.7873,
      "mean_token_accuracy": 0.21833567768335344,
      "num_tokens": 58475435.0,
      "step": 33700
    },
    {
      "entropy": 5.689179372787476,
      "epoch": 2.62236918887376,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043083673612378295,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20444071739912034,
      "num_tokens": 58483564.0,
      "step": 33705
    },
    {
      "entropy": 5.600613689422607,
      "epoch": 2.6227582182454774,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043081674328739066,
      "loss": 4.902,
      "mean_token_accuracy": 0.21259290874004363,
      "num_tokens": 58491991.0,
      "step": 33710
    },
    {
      "entropy": 5.660284423828125,
      "epoch": 2.623147247617195,
      "grad_norm": 1.25,
      "learning_rate": 0.00043079674808672744,
      "loss": 4.9853,
      "mean_token_accuracy": 0.20495769381523132,
      "num_tokens": 58500452.0,
      "step": 33715
    },
    {
      "entropy": 5.720640802383423,
      "epoch": 2.623536276988913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004307767505220964,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21203888207674026,
      "num_tokens": 58509124.0,
      "step": 33720
    },
    {
      "entropy": 5.699628162384033,
      "epoch": 2.62392530636063,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004307567505938012,
      "loss": 4.9607,
      "mean_token_accuracy": 0.20883955508470536,
      "num_tokens": 58516902.0,
      "step": 33725
    },
    {
      "entropy": 5.646574592590332,
      "epoch": 2.6243143357323477,
      "grad_norm": 1.375,
      "learning_rate": 0.0004307367483021452,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2091030552983284,
      "num_tokens": 58525401.0,
      "step": 33730
    },
    {
      "entropy": 5.787299871444702,
      "epoch": 2.6247033651040654,
      "grad_norm": 1.125,
      "learning_rate": 0.000430716743647432,
      "loss": 5.1049,
      "mean_token_accuracy": 0.20118573158979416,
      "num_tokens": 58534473.0,
      "step": 33735
    },
    {
      "entropy": 5.655651378631592,
      "epoch": 2.625092394475783,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000430696736629965,
      "loss": 4.7877,
      "mean_token_accuracy": 0.22510315775871276,
      "num_tokens": 58543137.0,
      "step": 33740
    },
    {
      "entropy": 5.634085607528687,
      "epoch": 2.6254814238475004,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004306767272500478,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21139079481363296,
      "num_tokens": 58551260.0,
      "step": 33745
    },
    {
      "entropy": 5.629272842407227,
      "epoch": 2.625870453219218,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043065671550798415,
      "loss": 4.9033,
      "mean_token_accuracy": 0.209025838971138,
      "num_tokens": 58559428.0,
      "step": 33750
    },
    {
      "entropy": 5.6439038753509525,
      "epoch": 2.626259482590936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043063670140407753,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2089892089366913,
      "num_tokens": 58568153.0,
      "step": 33755
    },
    {
      "entropy": 5.7283501625061035,
      "epoch": 2.626648511962653,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004306166849386317,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2047397866845131,
      "num_tokens": 58577432.0,
      "step": 33760
    },
    {
      "entropy": 5.739789247512817,
      "epoch": 2.6270375413343707,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043059666611195037,
      "loss": 5.0883,
      "mean_token_accuracy": 0.19985555857419968,
      "num_tokens": 58586988.0,
      "step": 33765
    },
    {
      "entropy": 5.743124294281006,
      "epoch": 2.6274265707060884,
      "grad_norm": 1.25,
      "learning_rate": 0.0004305766449243372,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22166897505521774,
      "num_tokens": 58595094.0,
      "step": 33770
    },
    {
      "entropy": 5.655226945877075,
      "epoch": 2.627815600077806,
      "grad_norm": 1.25,
      "learning_rate": 0.0004305566213760962,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21146383434534072,
      "num_tokens": 58603927.0,
      "step": 33775
    },
    {
      "entropy": 5.719522190093994,
      "epoch": 2.6282046294495234,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043053659546753094,
      "loss": 5.0671,
      "mean_token_accuracy": 0.19773264974355698,
      "num_tokens": 58612497.0,
      "step": 33780
    },
    {
      "entropy": 5.7501390933990475,
      "epoch": 2.628593658821241,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004305165671989455,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20600755661725997,
      "num_tokens": 58621814.0,
      "step": 33785
    },
    {
      "entropy": 5.766373348236084,
      "epoch": 2.6289826881929583,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004304965365706437,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20424779802560805,
      "num_tokens": 58630215.0,
      "step": 33790
    },
    {
      "entropy": 5.6639776706695555,
      "epoch": 2.629371717564676,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004304765035829294,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2028355360031128,
      "num_tokens": 58638710.0,
      "step": 33795
    },
    {
      "entropy": 5.686469793319702,
      "epoch": 2.6297607469363937,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043045646823610664,
      "loss": 4.9598,
      "mean_token_accuracy": 0.201075978577137,
      "num_tokens": 58647119.0,
      "step": 33800
    },
    {
      "entropy": 5.746336984634399,
      "epoch": 2.6301497763081114,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043043643053047935,
      "loss": 5.0267,
      "mean_token_accuracy": 0.1985120490193367,
      "num_tokens": 58655669.0,
      "step": 33805
    },
    {
      "entropy": 5.73438868522644,
      "epoch": 2.6305388056798287,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004304163904663517,
      "loss": 5.0183,
      "mean_token_accuracy": 0.19953179806470872,
      "num_tokens": 58664249.0,
      "step": 33810
    },
    {
      "entropy": 5.69872407913208,
      "epoch": 2.6309278350515464,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043039634804402767,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20161376148462296,
      "num_tokens": 58672823.0,
      "step": 33815
    },
    {
      "entropy": 5.713734197616577,
      "epoch": 2.631316864423264,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004303763032638114,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20238484740257262,
      "num_tokens": 58681430.0,
      "step": 33820
    },
    {
      "entropy": 5.687529611587524,
      "epoch": 2.6317058937949813,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004303562561260071,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21030967980623244,
      "num_tokens": 58690388.0,
      "step": 33825
    },
    {
      "entropy": 5.674468851089477,
      "epoch": 2.632094923166699,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043033620663091883,
      "loss": 4.959,
      "mean_token_accuracy": 0.21134776324033738,
      "num_tokens": 58698651.0,
      "step": 33830
    },
    {
      "entropy": 5.769495248794556,
      "epoch": 2.6324839525384167,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004303161547788509,
      "loss": 5.03,
      "mean_token_accuracy": 0.2049405574798584,
      "num_tokens": 58707951.0,
      "step": 33835
    },
    {
      "entropy": 5.70444884300232,
      "epoch": 2.6328729819101344,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004302961005701074,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2088685691356659,
      "num_tokens": 58716220.0,
      "step": 33840
    },
    {
      "entropy": 5.7517335414886475,
      "epoch": 2.6332620112818517,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043027604400499295,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20654278099536896,
      "num_tokens": 58725534.0,
      "step": 33845
    },
    {
      "entropy": 5.714947652816773,
      "epoch": 2.6336510406535694,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043025598508381155,
      "loss": 4.8576,
      "mean_token_accuracy": 0.21352887153625488,
      "num_tokens": 58733725.0,
      "step": 33850
    },
    {
      "entropy": 5.656706809997559,
      "epoch": 2.634040070025287,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004302359238068677,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2101040467619896,
      "num_tokens": 58742483.0,
      "step": 33855
    },
    {
      "entropy": 5.688205194473267,
      "epoch": 2.6344290993970043,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043021586017446585,
      "loss": 4.9616,
      "mean_token_accuracy": 0.20769460946321489,
      "num_tokens": 58750442.0,
      "step": 33860
    },
    {
      "entropy": 5.731773853302002,
      "epoch": 2.634818128768722,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043019579418691027,
      "loss": 5.0065,
      "mean_token_accuracy": 0.19845550060272216,
      "num_tokens": 58759640.0,
      "step": 33865
    },
    {
      "entropy": 5.671589183807373,
      "epoch": 2.6352071581404397,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004301757258445056,
      "loss": 4.8321,
      "mean_token_accuracy": 0.21180316656827927,
      "num_tokens": 58768241.0,
      "step": 33870
    },
    {
      "entropy": 5.693029546737671,
      "epoch": 2.6355961875121574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004301556551475563,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20384200364351274,
      "num_tokens": 58777125.0,
      "step": 33875
    },
    {
      "entropy": 5.726770401000977,
      "epoch": 2.6359852168838747,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004301355820963669,
      "loss": 5.0091,
      "mean_token_accuracy": 0.210397107899189,
      "num_tokens": 58785855.0,
      "step": 33880
    },
    {
      "entropy": 5.748571586608887,
      "epoch": 2.6363742462555924,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004301155066912419,
      "loss": 5.0416,
      "mean_token_accuracy": 0.20423906594514846,
      "num_tokens": 58795235.0,
      "step": 33885
    },
    {
      "entropy": 5.620795965194702,
      "epoch": 2.6367632756273096,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000430095428932486,
      "loss": 4.9055,
      "mean_token_accuracy": 0.20913298577070236,
      "num_tokens": 58803434.0,
      "step": 33890
    },
    {
      "entropy": 5.701079559326172,
      "epoch": 2.6371523049990273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043007534882040385,
      "loss": 5.0433,
      "mean_token_accuracy": 0.19801911413669587,
      "num_tokens": 58811995.0,
      "step": 33895
    },
    {
      "entropy": 5.7786987781524655,
      "epoch": 2.637541334370745,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004300552663553001,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20190469920635223,
      "num_tokens": 58820035.0,
      "step": 33900
    },
    {
      "entropy": 5.714387607574463,
      "epoch": 2.6379303637424627,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004300351815374795,
      "loss": 4.9201,
      "mean_token_accuracy": 0.20515310615301133,
      "num_tokens": 58829322.0,
      "step": 33905
    },
    {
      "entropy": 5.624718236923218,
      "epoch": 2.63831939311418,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004300150943672467,
      "loss": 5.0119,
      "mean_token_accuracy": 0.2068459451198578,
      "num_tokens": 58838227.0,
      "step": 33910
    },
    {
      "entropy": 5.734870576858521,
      "epoch": 2.6387084224858977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004299950048449067,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20370962023735045,
      "num_tokens": 58847118.0,
      "step": 33915
    },
    {
      "entropy": 5.67009129524231,
      "epoch": 2.6390974518576154,
      "grad_norm": 1.25,
      "learning_rate": 0.0004299749129707641,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2081275314092636,
      "num_tokens": 58855211.0,
      "step": 33920
    },
    {
      "entropy": 5.693898296356201,
      "epoch": 2.6394864812293326,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004299548187451239,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20446945279836654,
      "num_tokens": 58864412.0,
      "step": 33925
    },
    {
      "entropy": 5.674614906311035,
      "epoch": 2.6398755106010503,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042993472216829097,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20582833439111708,
      "num_tokens": 58873102.0,
      "step": 33930
    },
    {
      "entropy": 5.777045440673828,
      "epoch": 2.640264539972768,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042991462324057025,
      "loss": 5.0768,
      "mean_token_accuracy": 0.20361398607492448,
      "num_tokens": 58881974.0,
      "step": 33935
    },
    {
      "entropy": 5.701866626739502,
      "epoch": 2.6406535693444857,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004298945219622667,
      "loss": 4.992,
      "mean_token_accuracy": 0.21201558858156205,
      "num_tokens": 58890906.0,
      "step": 33940
    },
    {
      "entropy": 5.6590558052062985,
      "epoch": 2.641042598716203,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042987441833368525,
      "loss": 4.9716,
      "mean_token_accuracy": 0.20415595173835754,
      "num_tokens": 58900368.0,
      "step": 33945
    },
    {
      "entropy": 5.7435822010040285,
      "epoch": 2.6414316280879206,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042985431235513104,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19571671038866043,
      "num_tokens": 58910023.0,
      "step": 33950
    },
    {
      "entropy": 5.717715883255005,
      "epoch": 2.6418206574596383,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042983420402690917,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20852474272251129,
      "num_tokens": 58918451.0,
      "step": 33955
    },
    {
      "entropy": 5.629948568344116,
      "epoch": 2.6422096868313556,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042981409334932457,
      "loss": 4.9187,
      "mean_token_accuracy": 0.20955066084861756,
      "num_tokens": 58927017.0,
      "step": 33960
    },
    {
      "entropy": 5.6800707340240475,
      "epoch": 2.6425987162030733,
      "grad_norm": 1.25,
      "learning_rate": 0.0004297939803226826,
      "loss": 4.905,
      "mean_token_accuracy": 0.2099366456270218,
      "num_tokens": 58935438.0,
      "step": 33965
    },
    {
      "entropy": 5.688488292694092,
      "epoch": 2.642987745574791,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004297738649472884,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20595991909503936,
      "num_tokens": 58943806.0,
      "step": 33970
    },
    {
      "entropy": 5.759329032897949,
      "epoch": 2.6433767749465087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042975374722344713,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2057760939002037,
      "num_tokens": 58952997.0,
      "step": 33975
    },
    {
      "entropy": 5.655094146728516,
      "epoch": 2.643765804318226,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042973362715146397,
      "loss": 4.944,
      "mean_token_accuracy": 0.21307116448879243,
      "num_tokens": 58961820.0,
      "step": 33980
    },
    {
      "entropy": 5.756137752532959,
      "epoch": 2.6441548336899436,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004297135047316444,
      "loss": 5.0276,
      "mean_token_accuracy": 0.1944836527109146,
      "num_tokens": 58969856.0,
      "step": 33985
    },
    {
      "entropy": 5.650730228424072,
      "epoch": 2.644543863061661,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004296933799642936,
      "loss": 4.8743,
      "mean_token_accuracy": 0.21194353997707366,
      "num_tokens": 58978341.0,
      "step": 33990
    },
    {
      "entropy": 5.708328342437744,
      "epoch": 2.6449328924333786,
      "grad_norm": 1.234375,
      "learning_rate": 0.000429673252849717,
      "loss": 4.961,
      "mean_token_accuracy": 0.20882129073143005,
      "num_tokens": 58987468.0,
      "step": 33995
    },
    {
      "entropy": 5.6675660610198975,
      "epoch": 2.6453219218050963,
      "grad_norm": 1.25,
      "learning_rate": 0.00042965312338822005,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20905167758464813,
      "num_tokens": 58995824.0,
      "step": 34000
    },
    {
      "entropy": 5.643595409393311,
      "epoch": 2.645710951176814,
      "grad_norm": 1.28125,
      "learning_rate": 0.000429632991580108,
      "loss": 4.93,
      "mean_token_accuracy": 0.20923031866550446,
      "num_tokens": 59005209.0,
      "step": 34005
    },
    {
      "entropy": 5.674721527099609,
      "epoch": 2.6460999805485317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004296128574256864,
      "loss": 4.9957,
      "mean_token_accuracy": 0.20790353119373323,
      "num_tokens": 59013675.0,
      "step": 34010
    },
    {
      "entropy": 5.6995950698852536,
      "epoch": 2.646489009920249,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042959272092526086,
      "loss": 4.9594,
      "mean_token_accuracy": 0.2089621603488922,
      "num_tokens": 59022237.0,
      "step": 34015
    },
    {
      "entropy": 5.662182998657227,
      "epoch": 2.6468780392919666,
      "grad_norm": 1.25,
      "learning_rate": 0.00042957258207913687,
      "loss": 4.873,
      "mean_token_accuracy": 0.21584845036268235,
      "num_tokens": 59030442.0,
      "step": 34020
    },
    {
      "entropy": 5.659041070938111,
      "epoch": 2.647267068663684,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042955244088761985,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2088031530380249,
      "num_tokens": 59038160.0,
      "step": 34025
    },
    {
      "entropy": 5.7612823963165285,
      "epoch": 2.6476560980354016,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004295322973510156,
      "loss": 5.0586,
      "mean_token_accuracy": 0.2021506905555725,
      "num_tokens": 59047083.0,
      "step": 34030
    },
    {
      "entropy": 5.701892423629761,
      "epoch": 2.6480451274071193,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004295121514696297,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20804806649684907,
      "num_tokens": 59055161.0,
      "step": 34035
    },
    {
      "entropy": 5.756841325759888,
      "epoch": 2.648434156778837,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042949200324376784,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20500402599573136,
      "num_tokens": 59063870.0,
      "step": 34040
    },
    {
      "entropy": 5.614480018615723,
      "epoch": 2.6488231861505542,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004294718526737357,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21204107850790024,
      "num_tokens": 59072577.0,
      "step": 34045
    },
    {
      "entropy": 5.677929353713989,
      "epoch": 2.649212215522272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004294516997598391,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20653841346502305,
      "num_tokens": 59081764.0,
      "step": 34050
    },
    {
      "entropy": 5.740802335739136,
      "epoch": 2.6496012448939896,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004294315445023838,
      "loss": 4.9332,
      "mean_token_accuracy": 0.20827574282884598,
      "num_tokens": 59090494.0,
      "step": 34055
    },
    {
      "entropy": 5.712126922607422,
      "epoch": 2.649990274265707,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042941138690167556,
      "loss": 4.9311,
      "mean_token_accuracy": 0.209780690073967,
      "num_tokens": 59099357.0,
      "step": 34060
    },
    {
      "entropy": 5.690656995773315,
      "epoch": 2.6503793036374246,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004293912269580204,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21206434220075607,
      "num_tokens": 59108377.0,
      "step": 34065
    },
    {
      "entropy": 5.666376733779908,
      "epoch": 2.6507683330091423,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000429371064671724,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21006055623292924,
      "num_tokens": 59116621.0,
      "step": 34070
    },
    {
      "entropy": 5.730780506134034,
      "epoch": 2.65115736238086,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042935090004309247,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20594560354948044,
      "num_tokens": 59124886.0,
      "step": 34075
    },
    {
      "entropy": 5.708247232437134,
      "epoch": 2.6515463917525772,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042933073307243165,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2122181162238121,
      "num_tokens": 59133315.0,
      "step": 34080
    },
    {
      "entropy": 5.733900547027588,
      "epoch": 2.651935421124295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042931056376004767,
      "loss": 4.9445,
      "mean_token_accuracy": 0.214688478410244,
      "num_tokens": 59141906.0,
      "step": 34085
    },
    {
      "entropy": 5.752964639663697,
      "epoch": 2.652324450496012,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004292903921062465,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20351265966892243,
      "num_tokens": 59150072.0,
      "step": 34090
    },
    {
      "entropy": 5.698544263839722,
      "epoch": 2.65271347986773,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042927021811133416,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20696386247873305,
      "num_tokens": 59158798.0,
      "step": 34095
    },
    {
      "entropy": 5.651756477355957,
      "epoch": 2.6531025092394476,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004292500417756167,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21073096990585327,
      "num_tokens": 59167327.0,
      "step": 34100
    },
    {
      "entropy": 5.752778005599976,
      "epoch": 2.6534915386111653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042922986309940054,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2058023437857628,
      "num_tokens": 59176113.0,
      "step": 34105
    },
    {
      "entropy": 5.769353532791138,
      "epoch": 2.653880567982883,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042920968208299165,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2041608363389969,
      "num_tokens": 59184410.0,
      "step": 34110
    },
    {
      "entropy": 5.681175279617309,
      "epoch": 2.6542695973546,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004291894987266963,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2078913688659668,
      "num_tokens": 59193540.0,
      "step": 34115
    },
    {
      "entropy": 5.741752624511719,
      "epoch": 2.654658626726318,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042916931303082064,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20146892219781876,
      "num_tokens": 59202219.0,
      "step": 34120
    },
    {
      "entropy": 5.754435062408447,
      "epoch": 2.655047656098035,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042914912499567113,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20359433591365814,
      "num_tokens": 59211469.0,
      "step": 34125
    },
    {
      "entropy": 5.726271629333496,
      "epoch": 2.655436685469753,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042912893462155395,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21012914478778838,
      "num_tokens": 59220276.0,
      "step": 34130
    },
    {
      "entropy": 5.745905876159668,
      "epoch": 2.6558257148414706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042910874190877545,
      "loss": 4.9804,
      "mean_token_accuracy": 0.20637679398059844,
      "num_tokens": 59229412.0,
      "step": 34135
    },
    {
      "entropy": 5.751666498184204,
      "epoch": 2.6562147442131883,
      "grad_norm": 1.125,
      "learning_rate": 0.0004290885468576422,
      "loss": 4.8737,
      "mean_token_accuracy": 0.21003971695899964,
      "num_tokens": 59237442.0,
      "step": 34140
    },
    {
      "entropy": 5.72420744895935,
      "epoch": 2.6566037735849055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004290683494684604,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2018063485622406,
      "num_tokens": 59245964.0,
      "step": 34145
    },
    {
      "entropy": 5.714412879943848,
      "epoch": 2.656992802956623,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004290481497415367,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20915480703115463,
      "num_tokens": 59254456.0,
      "step": 34150
    },
    {
      "entropy": 5.641820478439331,
      "epoch": 2.657381832328341,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004290279476771775,
      "loss": 4.9379,
      "mean_token_accuracy": 0.20672687590122224,
      "num_tokens": 59262550.0,
      "step": 34155
    },
    {
      "entropy": 5.64005823135376,
      "epoch": 2.657770861700058,
      "grad_norm": 1.25,
      "learning_rate": 0.0004290077432756894,
      "loss": 4.9074,
      "mean_token_accuracy": 0.21478297114372252,
      "num_tokens": 59271877.0,
      "step": 34160
    },
    {
      "entropy": 5.729611587524414,
      "epoch": 2.658159891071776,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004289875365373788,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21334144920110704,
      "num_tokens": 59280991.0,
      "step": 34165
    },
    {
      "entropy": 5.71203351020813,
      "epoch": 2.6585489204434936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042896732746255256,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2146846041083336,
      "num_tokens": 59289508.0,
      "step": 34170
    },
    {
      "entropy": 5.642649745941162,
      "epoch": 2.6589379498152113,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042894711605151714,
      "loss": 4.8933,
      "mean_token_accuracy": 0.211956287920475,
      "num_tokens": 59298084.0,
      "step": 34175
    },
    {
      "entropy": 5.682379913330078,
      "epoch": 2.6593269791869285,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042892690230457924,
      "loss": 4.969,
      "mean_token_accuracy": 0.21003276258707046,
      "num_tokens": 59306846.0,
      "step": 34180
    },
    {
      "entropy": 5.677598667144776,
      "epoch": 2.659716008558646,
      "grad_norm": 1.25,
      "learning_rate": 0.00042890668622204566,
      "loss": 4.8789,
      "mean_token_accuracy": 0.20794605016708373,
      "num_tokens": 59315086.0,
      "step": 34185
    },
    {
      "entropy": 5.68616509437561,
      "epoch": 2.660105037930364,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000428886467804223,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20802532285451888,
      "num_tokens": 59323699.0,
      "step": 34190
    },
    {
      "entropy": 5.694668388366699,
      "epoch": 2.660494067302081,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042886624705141825,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20793934017419816,
      "num_tokens": 59332444.0,
      "step": 34195
    },
    {
      "entropy": 5.648365497589111,
      "epoch": 2.660883096673799,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042884602396393796,
      "loss": 4.9538,
      "mean_token_accuracy": 0.20524705350399017,
      "num_tokens": 59340980.0,
      "step": 34200
    },
    {
      "entropy": 5.727013778686524,
      "epoch": 2.6612721260455166,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004288257985420892,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21216763854026793,
      "num_tokens": 59348937.0,
      "step": 34205
    },
    {
      "entropy": 5.702333259582519,
      "epoch": 2.6616611554172342,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004288055707861788,
      "loss": 5.052,
      "mean_token_accuracy": 0.19866923242807388,
      "num_tokens": 59356936.0,
      "step": 34210
    },
    {
      "entropy": 5.741358995437622,
      "epoch": 2.6620501847889515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042878534069651364,
      "loss": 5.052,
      "mean_token_accuracy": 0.19417907148599625,
      "num_tokens": 59366052.0,
      "step": 34215
    },
    {
      "entropy": 5.687316513061523,
      "epoch": 2.662439214160669,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004287651082734007,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20594962537288666,
      "num_tokens": 59374607.0,
      "step": 34220
    },
    {
      "entropy": 5.68810453414917,
      "epoch": 2.6628282435323865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000428744873517147,
      "loss": 5.0529,
      "mean_token_accuracy": 0.19593643695116042,
      "num_tokens": 59383607.0,
      "step": 34225
    },
    {
      "entropy": 5.707661867141724,
      "epoch": 2.663217272904104,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004287246364280596,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2139095664024353,
      "num_tokens": 59391472.0,
      "step": 34230
    },
    {
      "entropy": 5.682311487197876,
      "epoch": 2.663606302275822,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004287043970064455,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21063620448112488,
      "num_tokens": 59400185.0,
      "step": 34235
    },
    {
      "entropy": 5.6669997215271,
      "epoch": 2.6639953316475395,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004286841552526118,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20499003678560257,
      "num_tokens": 59408188.0,
      "step": 34240
    },
    {
      "entropy": 5.7498384475708,
      "epoch": 2.664384361019257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042866391116686567,
      "loss": 5.0477,
      "mean_token_accuracy": 0.20273904651403427,
      "num_tokens": 59417302.0,
      "step": 34245
    },
    {
      "entropy": 5.761425495147705,
      "epoch": 2.6647733903909745,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004286436647495142,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20874613970518113,
      "num_tokens": 59425913.0,
      "step": 34250
    },
    {
      "entropy": 5.666897487640381,
      "epoch": 2.665162419762692,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042862341600086477,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2113860309123993,
      "num_tokens": 59434325.0,
      "step": 34255
    },
    {
      "entropy": 5.661088848114014,
      "epoch": 2.6655514491344094,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042860316492122447,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21041697561740874,
      "num_tokens": 59443280.0,
      "step": 34260
    },
    {
      "entropy": 5.755367374420166,
      "epoch": 2.665940478506127,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004285829115109006,
      "loss": 5.038,
      "mean_token_accuracy": 0.20381179749965667,
      "num_tokens": 59452498.0,
      "step": 34265
    },
    {
      "entropy": 5.799960088729859,
      "epoch": 2.666329507877845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004285626557702005,
      "loss": 5.091,
      "mean_token_accuracy": 0.19749799519777297,
      "num_tokens": 59461629.0,
      "step": 34270
    },
    {
      "entropy": 5.742047739028931,
      "epoch": 2.6667185372495625,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004285423976994316,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20312186628580092,
      "num_tokens": 59470835.0,
      "step": 34275
    },
    {
      "entropy": 5.674464464187622,
      "epoch": 2.66710756662128,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042852213729890117,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2139824390411377,
      "num_tokens": 59479102.0,
      "step": 34280
    },
    {
      "entropy": 5.798505115509033,
      "epoch": 2.6674965959929975,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004285018745689166,
      "loss": 5.0904,
      "mean_token_accuracy": 0.2040783166885376,
      "num_tokens": 59487622.0,
      "step": 34285
    },
    {
      "entropy": 5.692944669723511,
      "epoch": 2.667885625364715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004284816095097855,
      "loss": 4.9678,
      "mean_token_accuracy": 0.20580867379903794,
      "num_tokens": 59496038.0,
      "step": 34290
    },
    {
      "entropy": 5.7193889141082765,
      "epoch": 2.6682746547364324,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004284613421218152,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20729649513959886,
      "num_tokens": 59504214.0,
      "step": 34295
    },
    {
      "entropy": 5.6824049949646,
      "epoch": 2.66866368410815,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042844107240531324,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20677757412195205,
      "num_tokens": 59512478.0,
      "step": 34300
    },
    {
      "entropy": 5.695660638809204,
      "epoch": 2.669052713479868,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004284208003605873,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2104514554142952,
      "num_tokens": 59520528.0,
      "step": 34305
    },
    {
      "entropy": 5.661789226531982,
      "epoch": 2.6694417428515855,
      "grad_norm": 1.25,
      "learning_rate": 0.000428400525987945,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2099365249276161,
      "num_tokens": 59529011.0,
      "step": 34310
    },
    {
      "entropy": 5.618891906738281,
      "epoch": 2.669830772223303,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042838024928769375,
      "loss": 4.8466,
      "mean_token_accuracy": 0.2162046879529953,
      "num_tokens": 59537361.0,
      "step": 34315
    },
    {
      "entropy": 5.685580587387085,
      "epoch": 2.6702198015950205,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042835997026014145,
      "loss": 4.938,
      "mean_token_accuracy": 0.20677391737699508,
      "num_tokens": 59545154.0,
      "step": 34320
    },
    {
      "entropy": 5.639438247680664,
      "epoch": 2.6706088309667377,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004283396889055957,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2005898490548134,
      "num_tokens": 59553415.0,
      "step": 34325
    },
    {
      "entropy": 5.689557504653931,
      "epoch": 2.6709978603384554,
      "grad_norm": 1.25,
      "learning_rate": 0.0004283194052243641,
      "loss": 4.9459,
      "mean_token_accuracy": 0.20871423482894896,
      "num_tokens": 59561128.0,
      "step": 34330
    },
    {
      "entropy": 5.779829120635986,
      "epoch": 2.671386889710173,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042829911921675456,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2086014539003372,
      "num_tokens": 59570223.0,
      "step": 34335
    },
    {
      "entropy": 5.7249342918396,
      "epoch": 2.671775919081891,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000428278830883075,
      "loss": 4.9379,
      "mean_token_accuracy": 0.20448742359876632,
      "num_tokens": 59578643.0,
      "step": 34340
    },
    {
      "entropy": 5.693987083435059,
      "epoch": 2.6721649484536085,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042825854022363307,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2049323335289955,
      "num_tokens": 59587640.0,
      "step": 34345
    },
    {
      "entropy": 5.7125767230987545,
      "epoch": 2.6725539778253258,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042823824723873676,
      "loss": 5.011,
      "mean_token_accuracy": 0.2052415132522583,
      "num_tokens": 59596146.0,
      "step": 34350
    },
    {
      "entropy": 5.684187602996826,
      "epoch": 2.6729430071970435,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042821795192869407,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21603763103485107,
      "num_tokens": 59604774.0,
      "step": 34355
    },
    {
      "entropy": 5.715900039672851,
      "epoch": 2.6733320365687607,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004281976542938127,
      "loss": 4.9342,
      "mean_token_accuracy": 0.20937345921993256,
      "num_tokens": 59613241.0,
      "step": 34360
    },
    {
      "entropy": 5.6809711933135985,
      "epoch": 2.6737210659404784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042817735433440076,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20062430649995805,
      "num_tokens": 59621844.0,
      "step": 34365
    },
    {
      "entropy": 5.717845869064331,
      "epoch": 2.674110095312196,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004281570520507663,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2075674444437027,
      "num_tokens": 59630596.0,
      "step": 34370
    },
    {
      "entropy": 5.701227521896362,
      "epoch": 2.674499124683914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042813674744321737,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20935696512460708,
      "num_tokens": 59638830.0,
      "step": 34375
    },
    {
      "entropy": 5.646785163879395,
      "epoch": 2.674888154055631,
      "grad_norm": 1.3125,
      "learning_rate": 0.000428116440512062,
      "loss": 4.8799,
      "mean_token_accuracy": 0.20783390551805497,
      "num_tokens": 59647425.0,
      "step": 34380
    },
    {
      "entropy": 5.753915023803711,
      "epoch": 2.6752771834273488,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004280961312576084,
      "loss": 5.0107,
      "mean_token_accuracy": 0.20373741090297698,
      "num_tokens": 59655369.0,
      "step": 34385
    },
    {
      "entropy": 5.68843297958374,
      "epoch": 2.6756662127990665,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004280758196801646,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2136740803718567,
      "num_tokens": 59663843.0,
      "step": 34390
    },
    {
      "entropy": 5.687620639801025,
      "epoch": 2.6760552421707837,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042805550578003894,
      "loss": 4.8684,
      "mean_token_accuracy": 0.21488696932792664,
      "num_tokens": 59672115.0,
      "step": 34395
    },
    {
      "entropy": 5.663598251342774,
      "epoch": 2.6764442715425014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004280351895575396,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2170512318611145,
      "num_tokens": 59680884.0,
      "step": 34400
    },
    {
      "entropy": 5.6509974002838135,
      "epoch": 2.676833300914219,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004280148710129748,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21847243309020997,
      "num_tokens": 59689213.0,
      "step": 34405
    },
    {
      "entropy": 5.698428344726563,
      "epoch": 2.677222330285937,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042799455014665296,
      "loss": 4.9674,
      "mean_token_accuracy": 0.20648182928562164,
      "num_tokens": 59698508.0,
      "step": 34410
    },
    {
      "entropy": 5.738311386108398,
      "epoch": 2.677611359657654,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042797422695888226,
      "loss": 4.9902,
      "mean_token_accuracy": 0.20802370607852935,
      "num_tokens": 59706568.0,
      "step": 34415
    },
    {
      "entropy": 5.766266775131226,
      "epoch": 2.6780003890293718,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279539014499712,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20134620368480682,
      "num_tokens": 59715933.0,
      "step": 34420
    },
    {
      "entropy": 5.725120401382446,
      "epoch": 2.678389418401089,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004279335736202281,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20476993173360825,
      "num_tokens": 59724498.0,
      "step": 34425
    },
    {
      "entropy": 5.73232626914978,
      "epoch": 2.6787784477728067,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004279132434699615,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2089187815785408,
      "num_tokens": 59733489.0,
      "step": 34430
    },
    {
      "entropy": 5.7201393127441404,
      "epoch": 2.6791674771445244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004278929109994798,
      "loss": 4.9007,
      "mean_token_accuracy": 0.21600978821516037,
      "num_tokens": 59741915.0,
      "step": 34435
    },
    {
      "entropy": 5.755215406417847,
      "epoch": 2.679556506516242,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042787257620909144,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20344263762235643,
      "num_tokens": 59750058.0,
      "step": 34440
    },
    {
      "entropy": 5.795769166946411,
      "epoch": 2.67994553588796,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042785223909910515,
      "loss": 5.0304,
      "mean_token_accuracy": 0.201067653298378,
      "num_tokens": 59758849.0,
      "step": 34445
    },
    {
      "entropy": 5.783490324020386,
      "epoch": 2.680334565259677,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004278318996698294,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2044227570295334,
      "num_tokens": 59768088.0,
      "step": 34450
    },
    {
      "entropy": 5.734393835067749,
      "epoch": 2.6807235946313948,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004278115579215728,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21135684102773666,
      "num_tokens": 59776677.0,
      "step": 34455
    },
    {
      "entropy": 5.645063018798828,
      "epoch": 2.681112624003112,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004277912138546441,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20521750748157502,
      "num_tokens": 59785737.0,
      "step": 34460
    },
    {
      "entropy": 5.716525030136109,
      "epoch": 2.6815016533748297,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004277708674693519,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21365132927894592,
      "num_tokens": 59794075.0,
      "step": 34465
    },
    {
      "entropy": 5.795771932601928,
      "epoch": 2.6818906827465474,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042775051876600483,
      "loss": 4.9815,
      "mean_token_accuracy": 0.20844742208719252,
      "num_tokens": 59803150.0,
      "step": 34470
    },
    {
      "entropy": 5.721402835845947,
      "epoch": 2.682279712118265,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004277301677449119,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20386959314346315,
      "num_tokens": 59811423.0,
      "step": 34475
    },
    {
      "entropy": 5.76378870010376,
      "epoch": 2.6826687414899824,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004277098144063818,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2123885542154312,
      "num_tokens": 59820566.0,
      "step": 34480
    },
    {
      "entropy": 5.689324903488159,
      "epoch": 2.6830577708617,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042768945875072326,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21122681349515915,
      "num_tokens": 59829340.0,
      "step": 34485
    },
    {
      "entropy": 5.6613836765289305,
      "epoch": 2.6834468002334178,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004276691007782451,
      "loss": 4.926,
      "mean_token_accuracy": 0.20438716411590577,
      "num_tokens": 59837796.0,
      "step": 34490
    },
    {
      "entropy": 5.718779277801514,
      "epoch": 2.683835829605135,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004276487404892565,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20124615877866744,
      "num_tokens": 59846467.0,
      "step": 34495
    },
    {
      "entropy": 5.713428688049317,
      "epoch": 2.6842248589768527,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042762837788406615,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21138769686222075,
      "num_tokens": 59855127.0,
      "step": 34500
    },
    {
      "entropy": 5.754569339752197,
      "epoch": 2.6846138883485704,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004276080129629831,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21079697757959365,
      "num_tokens": 59863020.0,
      "step": 34505
    },
    {
      "entropy": 5.691145133972168,
      "epoch": 2.685002917720288,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042758764572631644,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20498876571655272,
      "num_tokens": 59872136.0,
      "step": 34510
    },
    {
      "entropy": 5.7243153095245365,
      "epoch": 2.6853919470920053,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00042756727617437504,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2030166357755661,
      "num_tokens": 59880955.0,
      "step": 34515
    },
    {
      "entropy": 5.675222682952881,
      "epoch": 2.685780976463723,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042754690430746805,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22288704961538314,
      "num_tokens": 59889260.0,
      "step": 34520
    },
    {
      "entropy": 5.559013366699219,
      "epoch": 2.6861700058354403,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004275265301259046,
      "loss": 4.8095,
      "mean_token_accuracy": 0.2229264885187149,
      "num_tokens": 59897808.0,
      "step": 34525
    },
    {
      "entropy": 5.6939507007598875,
      "epoch": 2.686559035207158,
      "grad_norm": 1.375,
      "learning_rate": 0.00042750615362999395,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21014904826879502,
      "num_tokens": 59906333.0,
      "step": 34530
    },
    {
      "entropy": 5.664986896514892,
      "epoch": 2.6869480645788757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042748577482004495,
      "loss": 4.9277,
      "mean_token_accuracy": 0.20926148146390916,
      "num_tokens": 59914812.0,
      "step": 34535
    },
    {
      "entropy": 5.700321197509766,
      "epoch": 2.6873370939505934,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042746539369636715,
      "loss": 4.9853,
      "mean_token_accuracy": 0.20648254752159118,
      "num_tokens": 59924235.0,
      "step": 34540
    },
    {
      "entropy": 5.765041780471802,
      "epoch": 2.687726123322311,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042744501025926966,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20751434415578843,
      "num_tokens": 59932643.0,
      "step": 34545
    },
    {
      "entropy": 5.683780193328857,
      "epoch": 2.6881151526940283,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004274246245090618,
      "loss": 4.8857,
      "mean_token_accuracy": 0.20909959971904754,
      "num_tokens": 59941380.0,
      "step": 34550
    },
    {
      "entropy": 5.6958390235900875,
      "epoch": 2.688504182065746,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004274042364460529,
      "loss": 4.9852,
      "mean_token_accuracy": 0.20385556071996688,
      "num_tokens": 59950167.0,
      "step": 34555
    },
    {
      "entropy": 5.67959623336792,
      "epoch": 2.6888932114374633,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042738384607055214,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20975787490606307,
      "num_tokens": 59959057.0,
      "step": 34560
    },
    {
      "entropy": 5.6308681011199955,
      "epoch": 2.689282240809181,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042736345338286917,
      "loss": 4.8289,
      "mean_token_accuracy": 0.21492743790149688,
      "num_tokens": 59967605.0,
      "step": 34565
    },
    {
      "entropy": 5.752499294281006,
      "epoch": 2.6896712701808987,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004273430583833133,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2016143783926964,
      "num_tokens": 59976208.0,
      "step": 34570
    },
    {
      "entropy": 5.667640590667725,
      "epoch": 2.6900602995526164,
      "grad_norm": 1.234375,
      "learning_rate": 0.000427322661072194,
      "loss": 4.9312,
      "mean_token_accuracy": 0.20445919930934905,
      "num_tokens": 59984431.0,
      "step": 34575
    },
    {
      "entropy": 5.648494291305542,
      "epoch": 2.6904493289243336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004273022614498208,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21284218579530717,
      "num_tokens": 59993414.0,
      "step": 34580
    },
    {
      "entropy": 5.691666889190674,
      "epoch": 2.6908383582960513,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004272818595165031,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2066776990890503,
      "num_tokens": 60002140.0,
      "step": 34585
    },
    {
      "entropy": 5.7191328525543215,
      "epoch": 2.691227387667769,
      "grad_norm": 1.25,
      "learning_rate": 0.0004272614552725506,
      "loss": 5.0463,
      "mean_token_accuracy": 0.19787559062242507,
      "num_tokens": 60009954.0,
      "step": 34590
    },
    {
      "entropy": 5.664065647125244,
      "epoch": 2.6916164170394863,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042724104871827293,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20897474139928818,
      "num_tokens": 60018754.0,
      "step": 34595
    },
    {
      "entropy": 5.698142147064209,
      "epoch": 2.692005446411204,
      "grad_norm": 1.25,
      "learning_rate": 0.0004272206398539795,
      "loss": 4.94,
      "mean_token_accuracy": 0.2032383933663368,
      "num_tokens": 60027795.0,
      "step": 34600
    },
    {
      "entropy": 5.697566509246826,
      "epoch": 2.6923944757829217,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042720022867998026,
      "loss": 4.9174,
      "mean_token_accuracy": 0.20928091406822205,
      "num_tokens": 60037764.0,
      "step": 34605
    },
    {
      "entropy": 5.590538597106933,
      "epoch": 2.6927835051546394,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004271798151965848,
      "loss": 4.7777,
      "mean_token_accuracy": 0.21928421407938004,
      "num_tokens": 60045325.0,
      "step": 34610
    },
    {
      "entropy": 5.694454193115234,
      "epoch": 2.6931725345263566,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004271593994041029,
      "loss": 4.982,
      "mean_token_accuracy": 0.206500281393528,
      "num_tokens": 60053536.0,
      "step": 34615
    },
    {
      "entropy": 5.744762706756592,
      "epoch": 2.6935615638980743,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004271389813028443,
      "loss": 5.0335,
      "mean_token_accuracy": 0.20599687248468398,
      "num_tokens": 60062003.0,
      "step": 34620
    },
    {
      "entropy": 5.714696884155273,
      "epoch": 2.693950593269792,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004271185608931187,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21346332877874374,
      "num_tokens": 60070633.0,
      "step": 34625
    },
    {
      "entropy": 5.718768548965454,
      "epoch": 2.6943396226415093,
      "grad_norm": 1.25,
      "learning_rate": 0.00042709813817523617,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21596226394176482,
      "num_tokens": 60079202.0,
      "step": 34630
    },
    {
      "entropy": 5.681667518615723,
      "epoch": 2.694728652013227,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042707771314950645,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22034886330366135,
      "num_tokens": 60087676.0,
      "step": 34635
    },
    {
      "entropy": 5.752464771270752,
      "epoch": 2.6951176813849447,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042705728581623944,
      "loss": 5.008,
      "mean_token_accuracy": 0.20564533025026321,
      "num_tokens": 60096968.0,
      "step": 34640
    },
    {
      "entropy": 5.727945280075073,
      "epoch": 2.6955067107566624,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004270368561757452,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21356792747974396,
      "num_tokens": 60105751.0,
      "step": 34645
    },
    {
      "entropy": 5.739959621429444,
      "epoch": 2.6958957401283796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004270164242283336,
      "loss": 5.0709,
      "mean_token_accuracy": 0.20289001613855362,
      "num_tokens": 60114957.0,
      "step": 34650
    },
    {
      "entropy": 5.774513626098633,
      "epoch": 2.6962847695000973,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004269959899743148,
      "loss": 5.0568,
      "mean_token_accuracy": 0.203794626891613,
      "num_tokens": 60123734.0,
      "step": 34655
    },
    {
      "entropy": 5.7639899253845215,
      "epoch": 2.6966737988718146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004269755534139987,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20050682723522187,
      "num_tokens": 60132178.0,
      "step": 34660
    },
    {
      "entropy": 5.777216100692749,
      "epoch": 2.6970628282435323,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042695511454769557,
      "loss": 5.0675,
      "mean_token_accuracy": 0.19938138276338577,
      "num_tokens": 60140436.0,
      "step": 34665
    },
    {
      "entropy": 5.748891162872314,
      "epoch": 2.69745185761525,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042693467337571534,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2088001400232315,
      "num_tokens": 60149154.0,
      "step": 34670
    },
    {
      "entropy": 5.715223979949951,
      "epoch": 2.6978408869869677,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004269142298983684,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20135734528303145,
      "num_tokens": 60158054.0,
      "step": 34675
    },
    {
      "entropy": 5.765547370910644,
      "epoch": 2.698229916358685,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042689378411596467,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20037845373153687,
      "num_tokens": 60166924.0,
      "step": 34680
    },
    {
      "entropy": 5.692225599288941,
      "epoch": 2.6986189457304026,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004268733360288146,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21295427531003952,
      "num_tokens": 60175223.0,
      "step": 34685
    },
    {
      "entropy": 5.736296939849853,
      "epoch": 2.6990079751021203,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042685288563722835,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20365485846996306,
      "num_tokens": 60184158.0,
      "step": 34690
    },
    {
      "entropy": 5.778026342391968,
      "epoch": 2.6993970044738376,
      "grad_norm": 1.25,
      "learning_rate": 0.0004268324329415163,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20334272235631942,
      "num_tokens": 60192816.0,
      "step": 34695
    },
    {
      "entropy": 5.772414684295654,
      "epoch": 2.6997860338455553,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004268119779419887,
      "loss": 4.9518,
      "mean_token_accuracy": 0.210853473842144,
      "num_tokens": 60200985.0,
      "step": 34700
    },
    {
      "entropy": 5.6309410572052006,
      "epoch": 2.700175063217273,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042679152063895603,
      "loss": 4.8727,
      "mean_token_accuracy": 0.20996491909027098,
      "num_tokens": 60208892.0,
      "step": 34705
    },
    {
      "entropy": 5.590454864501953,
      "epoch": 2.7005640925889907,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042677106103272867,
      "loss": 4.8288,
      "mean_token_accuracy": 0.22113702148199083,
      "num_tokens": 60216946.0,
      "step": 34710
    },
    {
      "entropy": 5.618599462509155,
      "epoch": 2.700953121960708,
      "grad_norm": 1.25,
      "learning_rate": 0.000426750599123617,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2169335663318634,
      "num_tokens": 60224973.0,
      "step": 34715
    },
    {
      "entropy": 5.699407243728638,
      "epoch": 2.7013421513324256,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042673013491193145,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2064083769917488,
      "num_tokens": 60232841.0,
      "step": 34720
    },
    {
      "entropy": 5.6886077404022215,
      "epoch": 2.7017311807041433,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042670966839798263,
      "loss": 4.9631,
      "mean_token_accuracy": 0.20922969430685043,
      "num_tokens": 60242162.0,
      "step": 34725
    },
    {
      "entropy": 5.7533472061157225,
      "epoch": 2.7021202100758606,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004266891995820811,
      "loss": 5.0683,
      "mean_token_accuracy": 0.195574551820755,
      "num_tokens": 60250924.0,
      "step": 34730
    },
    {
      "entropy": 5.788560676574707,
      "epoch": 2.7025092394475783,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042666872846453734,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20068838745355605,
      "num_tokens": 60260012.0,
      "step": 34735
    },
    {
      "entropy": 5.702512884140015,
      "epoch": 2.702898268819296,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042664825504566206,
      "loss": 4.8943,
      "mean_token_accuracy": 0.20823795050382615,
      "num_tokens": 60268172.0,
      "step": 34740
    },
    {
      "entropy": 5.621559906005859,
      "epoch": 2.7032872981910137,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000426627779325766,
      "loss": 4.8855,
      "mean_token_accuracy": 0.21660529673099518,
      "num_tokens": 60276753.0,
      "step": 34745
    },
    {
      "entropy": 5.641514921188355,
      "epoch": 2.703676327562731,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004266073013051595,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21847340166568757,
      "num_tokens": 60286388.0,
      "step": 34750
    },
    {
      "entropy": 5.669678497314453,
      "epoch": 2.7040653569344486,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004265868209841537,
      "loss": 4.8679,
      "mean_token_accuracy": 0.21334236562252046,
      "num_tokens": 60294672.0,
      "step": 34755
    },
    {
      "entropy": 5.719014310836792,
      "epoch": 2.704454386306166,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004265663383630591,
      "loss": 4.971,
      "mean_token_accuracy": 0.20974157899618148,
      "num_tokens": 60303293.0,
      "step": 34760
    },
    {
      "entropy": 5.683840942382813,
      "epoch": 2.7048434156778836,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042654585344218657,
      "loss": 4.9643,
      "mean_token_accuracy": 0.20707135647535324,
      "num_tokens": 60312123.0,
      "step": 34765
    },
    {
      "entropy": 5.6957923412323,
      "epoch": 2.7052324450496013,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004265253662218469,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21450020372867584,
      "num_tokens": 60320774.0,
      "step": 34770
    },
    {
      "entropy": 5.634758806228637,
      "epoch": 2.705621474421319,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042650487670235093,
      "loss": 4.8152,
      "mean_token_accuracy": 0.21598894745111466,
      "num_tokens": 60329536.0,
      "step": 34775
    },
    {
      "entropy": 5.705045890808106,
      "epoch": 2.7060105037930366,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004264843848840097,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20317461490631103,
      "num_tokens": 60338333.0,
      "step": 34780
    },
    {
      "entropy": 5.76559419631958,
      "epoch": 2.706399533164754,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000426463890767134,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20530707985162736,
      "num_tokens": 60347653.0,
      "step": 34785
    },
    {
      "entropy": 5.765258359909057,
      "epoch": 2.7067885625364716,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004264433943520349,
      "loss": 5.0329,
      "mean_token_accuracy": 0.20013868063688278,
      "num_tokens": 60355548.0,
      "step": 34790
    },
    {
      "entropy": 5.723854351043701,
      "epoch": 2.707177591908189,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042642289563902326,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2061668485403061,
      "num_tokens": 60364486.0,
      "step": 34795
    },
    {
      "entropy": 5.673714590072632,
      "epoch": 2.7075666212799065,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042640239462841023,
      "loss": 4.9835,
      "mean_token_accuracy": 0.20432132929563523,
      "num_tokens": 60372919.0,
      "step": 34800
    },
    {
      "entropy": 5.680074262619018,
      "epoch": 2.7079556506516242,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042638189132050675,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21110885590314865,
      "num_tokens": 60382721.0,
      "step": 34805
    },
    {
      "entropy": 5.71739239692688,
      "epoch": 2.708344680023342,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042636138571562415,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21098169684410095,
      "num_tokens": 60391478.0,
      "step": 34810
    },
    {
      "entropy": 5.745327806472778,
      "epoch": 2.708733709395059,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042634087781407337,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21791640520095826,
      "num_tokens": 60399724.0,
      "step": 34815
    },
    {
      "entropy": 5.737909030914307,
      "epoch": 2.709122738766777,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042632036761616564,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20079292804002763,
      "num_tokens": 60408505.0,
      "step": 34820
    },
    {
      "entropy": 5.729704666137695,
      "epoch": 2.7095117681384946,
      "grad_norm": 1.25,
      "learning_rate": 0.00042629985512221224,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20355501621961594,
      "num_tokens": 60417713.0,
      "step": 34825
    },
    {
      "entropy": 5.719590330123902,
      "epoch": 2.709900797510212,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004262793403325243,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21108058393001555,
      "num_tokens": 60426508.0,
      "step": 34830
    },
    {
      "entropy": 5.739744424819946,
      "epoch": 2.7102898268819295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042625882324741316,
      "loss": 4.981,
      "mean_token_accuracy": 0.20599597096443176,
      "num_tokens": 60435500.0,
      "step": 34835
    },
    {
      "entropy": 5.824837589263916,
      "epoch": 2.7106788562536472,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004262383038671902,
      "loss": 5.0877,
      "mean_token_accuracy": 0.19558206349611282,
      "num_tokens": 60444036.0,
      "step": 34840
    },
    {
      "entropy": 5.765301084518432,
      "epoch": 2.711067885625365,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004262177821921666,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2045507475733757,
      "num_tokens": 60453369.0,
      "step": 34845
    },
    {
      "entropy": 5.733113622665405,
      "epoch": 2.711456914997082,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042619725822265394,
      "loss": 5.0202,
      "mean_token_accuracy": 0.20741942524909973,
      "num_tokens": 60462232.0,
      "step": 34850
    },
    {
      "entropy": 5.741755437850952,
      "epoch": 2.7118459443688,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042617673195896345,
      "loss": 4.994,
      "mean_token_accuracy": 0.20558920353651047,
      "num_tokens": 60471337.0,
      "step": 34855
    },
    {
      "entropy": 5.823948764801026,
      "epoch": 2.712234973740517,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042615620340140674,
      "loss": 5.0881,
      "mean_token_accuracy": 0.19707985818386078,
      "num_tokens": 60480676.0,
      "step": 34860
    },
    {
      "entropy": 5.68254337310791,
      "epoch": 2.712624003112235,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042613567255029525,
      "loss": 5.0337,
      "mean_token_accuracy": 0.206841878592968,
      "num_tokens": 60489727.0,
      "step": 34865
    },
    {
      "entropy": 5.731939506530762,
      "epoch": 2.7130130324839525,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042611513940594037,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2060682937502861,
      "num_tokens": 60497859.0,
      "step": 34870
    },
    {
      "entropy": 5.708287000656128,
      "epoch": 2.7134020618556702,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004260946039686539,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20529041439294815,
      "num_tokens": 60506948.0,
      "step": 34875
    },
    {
      "entropy": 5.662458038330078,
      "epoch": 2.713791091227388,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042607406623874726,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20289556682109833,
      "num_tokens": 60515414.0,
      "step": 34880
    },
    {
      "entropy": 5.6291773319244385,
      "epoch": 2.714180120599105,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004260535262165322,
      "loss": 4.8369,
      "mean_token_accuracy": 0.21740337312221528,
      "num_tokens": 60523644.0,
      "step": 34885
    },
    {
      "entropy": 5.688690996170044,
      "epoch": 2.714569149970823,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042603298390232015,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20565861761569976,
      "num_tokens": 60531760.0,
      "step": 34890
    },
    {
      "entropy": 5.80579833984375,
      "epoch": 2.71495817934254,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042601243929642305,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20386364459991455,
      "num_tokens": 60540919.0,
      "step": 34895
    },
    {
      "entropy": 5.815211439132691,
      "epoch": 2.715347208714258,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042599189239915253,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20082654058933258,
      "num_tokens": 60549607.0,
      "step": 34900
    },
    {
      "entropy": 5.668854188919068,
      "epoch": 2.7157362380859755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004259713432108204,
      "loss": 4.8606,
      "mean_token_accuracy": 0.21700717657804489,
      "num_tokens": 60557926.0,
      "step": 34905
    },
    {
      "entropy": 5.616941356658936,
      "epoch": 2.7161252674576932,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004259507917317384,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20468156188726425,
      "num_tokens": 60567490.0,
      "step": 34910
    },
    {
      "entropy": 5.711208486557007,
      "epoch": 2.7165142968294105,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042593023796221843,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20440669655799865,
      "num_tokens": 60576279.0,
      "step": 34915
    },
    {
      "entropy": 5.768986415863037,
      "epoch": 2.716903326201128,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004259096819025723,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21294366717338561,
      "num_tokens": 60585856.0,
      "step": 34920
    },
    {
      "entropy": 5.74533486366272,
      "epoch": 2.717292355572846,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042588912355311196,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20219914019107818,
      "num_tokens": 60594139.0,
      "step": 34925
    },
    {
      "entropy": 5.717974281311035,
      "epoch": 2.717681384944563,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042586856291414925,
      "loss": 5.0559,
      "mean_token_accuracy": 0.1986674815416336,
      "num_tokens": 60602951.0,
      "step": 34930
    },
    {
      "entropy": 5.79165096282959,
      "epoch": 2.718070414316281,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042584799998599633,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20919229984283447,
      "num_tokens": 60611417.0,
      "step": 34935
    },
    {
      "entropy": 5.776333808898926,
      "epoch": 2.7184594436879985,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004258274347689651,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20613384395837783,
      "num_tokens": 60620419.0,
      "step": 34940
    },
    {
      "entropy": 5.7965678691864015,
      "epoch": 2.718848473059716,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042580686726336766,
      "loss": 5.1083,
      "mean_token_accuracy": 0.1974281683564186,
      "num_tokens": 60630029.0,
      "step": 34945
    },
    {
      "entropy": 5.612726163864136,
      "epoch": 2.7192375024314335,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042578629746951597,
      "loss": 4.8569,
      "mean_token_accuracy": 0.21830385327339172,
      "num_tokens": 60638331.0,
      "step": 34950
    },
    {
      "entropy": 5.811482667922974,
      "epoch": 2.719626531803151,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004257657253877222,
      "loss": 5.127,
      "mean_token_accuracy": 0.19877810627222062,
      "num_tokens": 60647645.0,
      "step": 34955
    },
    {
      "entropy": 5.7216551303863525,
      "epoch": 2.720015561174869,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042574515101829856,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2120066374540329,
      "num_tokens": 60656639.0,
      "step": 34960
    },
    {
      "entropy": 5.688572692871094,
      "epoch": 2.720404590546586,
      "grad_norm": 1.453125,
      "learning_rate": 0.00042572457436155717,
      "loss": 4.8295,
      "mean_token_accuracy": 0.2090499445796013,
      "num_tokens": 60665358.0,
      "step": 34965
    },
    {
      "entropy": 5.695967435836792,
      "epoch": 2.720793619918304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042570399541781023,
      "loss": 5.016,
      "mean_token_accuracy": 0.2105242446064949,
      "num_tokens": 60673998.0,
      "step": 34970
    },
    {
      "entropy": 5.7359638690948485,
      "epoch": 2.7211826492900215,
      "grad_norm": 1.25,
      "learning_rate": 0.00042568341418737016,
      "loss": 4.9638,
      "mean_token_accuracy": 0.1989905506372452,
      "num_tokens": 60682125.0,
      "step": 34975
    },
    {
      "entropy": 5.766455507278442,
      "epoch": 2.721571678661739,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000425662830670549,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20355624407529832,
      "num_tokens": 60690941.0,
      "step": 34980
    },
    {
      "entropy": 5.750086116790771,
      "epoch": 2.7219607080334565,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042564224486765925,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20628662407398224,
      "num_tokens": 60700353.0,
      "step": 34985
    },
    {
      "entropy": 5.6956874370574955,
      "epoch": 2.722349737405174,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004256216567790131,
      "loss": 4.921,
      "mean_token_accuracy": 0.21118637472391127,
      "num_tokens": 60709081.0,
      "step": 34990
    },
    {
      "entropy": 5.783931589126587,
      "epoch": 2.7227387667768914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004256010664049232,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20907147824764252,
      "num_tokens": 60717670.0,
      "step": 34995
    },
    {
      "entropy": 5.653836441040039,
      "epoch": 2.723127796148609,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004255804737457017,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21085964441299437,
      "num_tokens": 60725856.0,
      "step": 35000
    },
    {
      "entropy": 5.63736481666565,
      "epoch": 2.723516825520327,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042555987880166127,
      "loss": 4.928,
      "mean_token_accuracy": 0.20989930182695388,
      "num_tokens": 60734590.0,
      "step": 35005
    },
    {
      "entropy": 5.731264686584472,
      "epoch": 2.7239058548920445,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004255392815731142,
      "loss": 5.0071,
      "mean_token_accuracy": 0.2059839740395546,
      "num_tokens": 60743885.0,
      "step": 35010
    },
    {
      "entropy": 5.776842403411865,
      "epoch": 2.7242948842637618,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042551868206037324,
      "loss": 4.994,
      "mean_token_accuracy": 0.19641777575016023,
      "num_tokens": 60752212.0,
      "step": 35015
    },
    {
      "entropy": 5.721878433227539,
      "epoch": 2.7246839136354795,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042549808026375084,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20972406268119811,
      "num_tokens": 60760392.0,
      "step": 35020
    },
    {
      "entropy": 5.747177839279175,
      "epoch": 2.725072943007197,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042547747618355966,
      "loss": 5.0506,
      "mean_token_accuracy": 0.19544049203395844,
      "num_tokens": 60769650.0,
      "step": 35025
    },
    {
      "entropy": 5.75801944732666,
      "epoch": 2.7254619723789144,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004254568698201122,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2140042945742607,
      "num_tokens": 60778376.0,
      "step": 35030
    },
    {
      "entropy": 5.759399843215943,
      "epoch": 2.725851001750632,
      "grad_norm": 1.25,
      "learning_rate": 0.00042543626117372133,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20363978594541549,
      "num_tokens": 60786612.0,
      "step": 35035
    },
    {
      "entropy": 5.723813438415528,
      "epoch": 2.72624003112235,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042541565024469945,
      "loss": 5.0134,
      "mean_token_accuracy": 0.2095305249094963,
      "num_tokens": 60795054.0,
      "step": 35040
    },
    {
      "entropy": 5.718794584274292,
      "epoch": 2.7266290604940675,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042539503703335963,
      "loss": 4.9851,
      "mean_token_accuracy": 0.20630062371492386,
      "num_tokens": 60803343.0,
      "step": 35045
    },
    {
      "entropy": 5.688753080368042,
      "epoch": 2.7270180898657848,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042537442154001455,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21088998764753342,
      "num_tokens": 60811688.0,
      "step": 35050
    },
    {
      "entropy": 5.720266389846802,
      "epoch": 2.7274071192375025,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042535380376497684,
      "loss": 4.8935,
      "mean_token_accuracy": 0.21079623252153395,
      "num_tokens": 60820763.0,
      "step": 35055
    },
    {
      "entropy": 5.655383968353272,
      "epoch": 2.72779614860922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004253331837085595,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20842889249324797,
      "num_tokens": 60829512.0,
      "step": 35060
    },
    {
      "entropy": 5.649409532546997,
      "epoch": 2.7281851779809374,
      "grad_norm": 1.25,
      "learning_rate": 0.00042531256137107545,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21113201677799226,
      "num_tokens": 60838081.0,
      "step": 35065
    },
    {
      "entropy": 5.75114164352417,
      "epoch": 2.728574207352655,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004252919367528375,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2037801593542099,
      "num_tokens": 60847249.0,
      "step": 35070
    },
    {
      "entropy": 5.827528953552246,
      "epoch": 2.728963236724373,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042527130985415857,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20022119283676149,
      "num_tokens": 60855786.0,
      "step": 35075
    },
    {
      "entropy": 5.703303956985474,
      "epoch": 2.7293522660960905,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004252506806753517,
      "loss": 4.8634,
      "mean_token_accuracy": 0.21019433587789535,
      "num_tokens": 60864911.0,
      "step": 35080
    },
    {
      "entropy": 5.818068313598633,
      "epoch": 2.7297412954678077,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042523004921673,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2087898448109627,
      "num_tokens": 60872837.0,
      "step": 35085
    },
    {
      "entropy": 5.723009014129639,
      "epoch": 2.7301303248395254,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004252094154786063,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2039785251021385,
      "num_tokens": 60880470.0,
      "step": 35090
    },
    {
      "entropy": 5.659205484390259,
      "epoch": 2.7305193542112427,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004251887794612939,
      "loss": 4.9669,
      "mean_token_accuracy": 0.20906025320291519,
      "num_tokens": 60889763.0,
      "step": 35095
    },
    {
      "entropy": 5.732339763641358,
      "epoch": 2.7309083835829604,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004251681411651057,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21472941040992738,
      "num_tokens": 60898745.0,
      "step": 35100
    },
    {
      "entropy": 5.728853702545166,
      "epoch": 2.731297412954678,
      "grad_norm": 1.25,
      "learning_rate": 0.00042514750059035504,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20048222094774246,
      "num_tokens": 60907171.0,
      "step": 35105
    },
    {
      "entropy": 5.612569093704224,
      "epoch": 2.731686442326396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004251268577373551,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21187941282987593,
      "num_tokens": 60915837.0,
      "step": 35110
    },
    {
      "entropy": 5.70489501953125,
      "epoch": 2.732075471698113,
      "grad_norm": 1.234375,
      "learning_rate": 0.000425106212606419,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21067481935024263,
      "num_tokens": 60924705.0,
      "step": 35115
    },
    {
      "entropy": 5.6612220287323,
      "epoch": 2.7324645010698307,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042508556519786004,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21275871247053146,
      "num_tokens": 60934175.0,
      "step": 35120
    },
    {
      "entropy": 5.703955221176147,
      "epoch": 2.7328535304415484,
      "grad_norm": 1.25,
      "learning_rate": 0.0004250649155119915,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20019659847021104,
      "num_tokens": 60943099.0,
      "step": 35125
    },
    {
      "entropy": 5.757194566726684,
      "epoch": 2.7332425598132657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042504426354912666,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20433209985494613,
      "num_tokens": 60952346.0,
      "step": 35130
    },
    {
      "entropy": 5.71891598701477,
      "epoch": 2.7336315891849834,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042502360930957904,
      "loss": 4.9671,
      "mean_token_accuracy": 0.20493173152208327,
      "num_tokens": 60960285.0,
      "step": 35135
    },
    {
      "entropy": 5.780902862548828,
      "epoch": 2.734020618556701,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042500295279366196,
      "loss": 5.0663,
      "mean_token_accuracy": 0.19962110072374345,
      "num_tokens": 60969358.0,
      "step": 35140
    },
    {
      "entropy": 5.655981254577637,
      "epoch": 2.734409647928419,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042498229400168874,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2122568517923355,
      "num_tokens": 60977229.0,
      "step": 35145
    },
    {
      "entropy": 5.728889131546021,
      "epoch": 2.734798677300136,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000424961632933973,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2002171039581299,
      "num_tokens": 60986846.0,
      "step": 35150
    },
    {
      "entropy": 5.739418363571167,
      "epoch": 2.7351877066718537,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004249409695908281,
      "loss": 5.0082,
      "mean_token_accuracy": 0.1983056530356407,
      "num_tokens": 60995021.0,
      "step": 35155
    },
    {
      "entropy": 5.7087007522583,
      "epoch": 2.7355767360435714,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004249203039725677,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21663795709609984,
      "num_tokens": 61003219.0,
      "step": 35160
    },
    {
      "entropy": 5.664568758010864,
      "epoch": 2.7359657654152887,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042489963607950527,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20622797608375548,
      "num_tokens": 61011918.0,
      "step": 35165
    },
    {
      "entropy": 5.6633445739746096,
      "epoch": 2.7363547947870064,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042487896591195454,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1982610896229744,
      "num_tokens": 61021224.0,
      "step": 35170
    },
    {
      "entropy": 5.719748640060425,
      "epoch": 2.736743824158724,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042485829347022893,
      "loss": 4.9627,
      "mean_token_accuracy": 0.20518887042999268,
      "num_tokens": 61030660.0,
      "step": 35175
    },
    {
      "entropy": 5.750778293609619,
      "epoch": 2.737132853530442,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042483761875464226,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2122131645679474,
      "num_tokens": 61038542.0,
      "step": 35180
    },
    {
      "entropy": 5.697680044174194,
      "epoch": 2.737521882902159,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042481694176550825,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2069752186536789,
      "num_tokens": 61047175.0,
      "step": 35185
    },
    {
      "entropy": 5.692663669586182,
      "epoch": 2.7379109122738767,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042479626250314066,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2124638557434082,
      "num_tokens": 61056053.0,
      "step": 35190
    },
    {
      "entropy": 5.729515314102173,
      "epoch": 2.738299941645594,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042477558096785316,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2009968489408493,
      "num_tokens": 61064389.0,
      "step": 35195
    },
    {
      "entropy": 5.720431280136109,
      "epoch": 2.7386889710173117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042475489715995956,
      "loss": 4.918,
      "mean_token_accuracy": 0.21390856802463531,
      "num_tokens": 61072155.0,
      "step": 35200
    },
    {
      "entropy": 5.683716344833374,
      "epoch": 2.7390780003890294,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042473421107977385,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20617857724428176,
      "num_tokens": 61081644.0,
      "step": 35205
    },
    {
      "entropy": 5.796208381652832,
      "epoch": 2.739467029760747,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042471352272760974,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2070917844772339,
      "num_tokens": 61089392.0,
      "step": 35210
    },
    {
      "entropy": 5.762926340103149,
      "epoch": 2.7398560591324648,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004246928321037813,
      "loss": 5.0413,
      "mean_token_accuracy": 0.206692673265934,
      "num_tokens": 61098541.0,
      "step": 35215
    },
    {
      "entropy": 5.7560523509979244,
      "epoch": 2.740245088504182,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004246721392086023,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2055869072675705,
      "num_tokens": 61106759.0,
      "step": 35220
    },
    {
      "entropy": 5.736579942703247,
      "epoch": 2.7406341178758997,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042465144404238686,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21012219339609145,
      "num_tokens": 61115273.0,
      "step": 35225
    },
    {
      "entropy": 5.714414119720459,
      "epoch": 2.741023147247617,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042463074660544897,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20555173009634017,
      "num_tokens": 61124889.0,
      "step": 35230
    },
    {
      "entropy": 5.775128173828125,
      "epoch": 2.7414121766193347,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004246100468981026,
      "loss": 5.0613,
      "mean_token_accuracy": 0.19887160807847976,
      "num_tokens": 61134137.0,
      "step": 35235
    },
    {
      "entropy": 5.700501298904419,
      "epoch": 2.7418012059910524,
      "grad_norm": 1.234375,
      "learning_rate": 0.000424589344920662,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20897936969995498,
      "num_tokens": 61142974.0,
      "step": 35240
    },
    {
      "entropy": 5.741481637954712,
      "epoch": 2.74219023536277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004245686406734411,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20410853028297424,
      "num_tokens": 61151469.0,
      "step": 35245
    },
    {
      "entropy": 5.804116821289062,
      "epoch": 2.7425792647344873,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004245479341567541,
      "loss": 5.0693,
      "mean_token_accuracy": 0.19618377983570098,
      "num_tokens": 61160221.0,
      "step": 35250
    },
    {
      "entropy": 5.775238466262818,
      "epoch": 2.742968294106205,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004245272253709154,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20609673708677292,
      "num_tokens": 61168509.0,
      "step": 35255
    },
    {
      "entropy": 5.661110782623291,
      "epoch": 2.7433573234779227,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004245065143162389,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21041444987058638,
      "num_tokens": 61176774.0,
      "step": 35260
    },
    {
      "entropy": 5.6296485424041744,
      "epoch": 2.74374635284964,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042448580099303904,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2123877674341202,
      "num_tokens": 61185846.0,
      "step": 35265
    },
    {
      "entropy": 5.7228316307067875,
      "epoch": 2.7441353822213577,
      "grad_norm": 1.375,
      "learning_rate": 0.00042446508540163017,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2124114528298378,
      "num_tokens": 61194284.0,
      "step": 35270
    },
    {
      "entropy": 5.725844669342041,
      "epoch": 2.7445244115930754,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004244443675423264,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20283122211694718,
      "num_tokens": 61203119.0,
      "step": 35275
    },
    {
      "entropy": 5.713010597229004,
      "epoch": 2.744913440964793,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042442364741544227,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2141757935285568,
      "num_tokens": 61211866.0,
      "step": 35280
    },
    {
      "entropy": 5.717109203338623,
      "epoch": 2.7453024703365103,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042440292502129214,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20491950809955597,
      "num_tokens": 61220837.0,
      "step": 35285
    },
    {
      "entropy": 5.694370889663697,
      "epoch": 2.745691499708228,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004243822003601904,
      "loss": 4.962,
      "mean_token_accuracy": 0.20951741486787795,
      "num_tokens": 61229474.0,
      "step": 35290
    },
    {
      "entropy": 5.69015736579895,
      "epoch": 2.7460805290799453,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004243614734324515,
      "loss": 4.8676,
      "mean_token_accuracy": 0.211755108833313,
      "num_tokens": 61238271.0,
      "step": 35295
    },
    {
      "entropy": 5.733093929290772,
      "epoch": 2.746469558451663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004243407442383901,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20911089032888414,
      "num_tokens": 61246621.0,
      "step": 35300
    },
    {
      "entropy": 5.727267837524414,
      "epoch": 2.7468585878233807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042432001277832045,
      "loss": 5.0328,
      "mean_token_accuracy": 0.19624968320131303,
      "num_tokens": 61255945.0,
      "step": 35305
    },
    {
      "entropy": 5.69983229637146,
      "epoch": 2.7472476171950984,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004242992790525573,
      "loss": 4.8993,
      "mean_token_accuracy": 0.20778922140598297,
      "num_tokens": 61264432.0,
      "step": 35310
    },
    {
      "entropy": 5.725306177139283,
      "epoch": 2.747636646566816,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004242785430614154,
      "loss": 4.9255,
      "mean_token_accuracy": 0.20870424509048463,
      "num_tokens": 61272955.0,
      "step": 35315
    },
    {
      "entropy": 5.775846481323242,
      "epoch": 2.7480256759385333,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042425780480520906,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21643990874290467,
      "num_tokens": 61281224.0,
      "step": 35320
    },
    {
      "entropy": 5.657282018661499,
      "epoch": 2.748414705310251,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042423706428425305,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21579451709985734,
      "num_tokens": 61290149.0,
      "step": 35325
    },
    {
      "entropy": 5.688134384155274,
      "epoch": 2.7488037346819683,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004242163214988622,
      "loss": 4.982,
      "mean_token_accuracy": 0.19995140433311462,
      "num_tokens": 61299107.0,
      "step": 35330
    },
    {
      "entropy": 5.635414028167725,
      "epoch": 2.749192764053686,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004241955764493512,
      "loss": 4.869,
      "mean_token_accuracy": 0.21385025829076768,
      "num_tokens": 61308053.0,
      "step": 35335
    },
    {
      "entropy": 5.770064067840576,
      "epoch": 2.7495817934254037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042417482913603477,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20388067215681077,
      "num_tokens": 61317224.0,
      "step": 35340
    },
    {
      "entropy": 5.719659471511841,
      "epoch": 2.7499708227971214,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042415407955922774,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20206313729286193,
      "num_tokens": 61326073.0,
      "step": 35345
    },
    {
      "entropy": 5.712783575057983,
      "epoch": 2.7503598521688386,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042413332771924496,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2130627527832985,
      "num_tokens": 61334615.0,
      "step": 35350
    },
    {
      "entropy": 5.710699558258057,
      "epoch": 2.7507488815405563,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004241125736164013,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2085053250193596,
      "num_tokens": 61342875.0,
      "step": 35355
    },
    {
      "entropy": 5.688929271697998,
      "epoch": 2.751137910912274,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004240918172510117,
      "loss": 4.9132,
      "mean_token_accuracy": 0.2135462284088135,
      "num_tokens": 61351355.0,
      "step": 35360
    },
    {
      "entropy": 5.681088972091675,
      "epoch": 2.7515269402839913,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000424071058623391,
      "loss": 4.9325,
      "mean_token_accuracy": 0.20771088749170302,
      "num_tokens": 61359793.0,
      "step": 35365
    },
    {
      "entropy": 5.786489963531494,
      "epoch": 2.751915969655709,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004240502977338543,
      "loss": 5.1078,
      "mean_token_accuracy": 0.2048235386610031,
      "num_tokens": 61369302.0,
      "step": 35370
    },
    {
      "entropy": 5.820939588546753,
      "epoch": 2.7523049990274266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004240295345827166,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19990177154541017,
      "num_tokens": 61377753.0,
      "step": 35375
    },
    {
      "entropy": 5.73698320388794,
      "epoch": 2.7526940283991443,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004240087691702929,
      "loss": 4.997,
      "mean_token_accuracy": 0.20807427763938904,
      "num_tokens": 61386977.0,
      "step": 35380
    },
    {
      "entropy": 5.7078382015228275,
      "epoch": 2.7530830577708616,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004239880014968983,
      "loss": 5.0668,
      "mean_token_accuracy": 0.20457228720188142,
      "num_tokens": 61395756.0,
      "step": 35385
    },
    {
      "entropy": 5.6951357364654545,
      "epoch": 2.7534720871425793,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042396723156284786,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2087944209575653,
      "num_tokens": 61404622.0,
      "step": 35390
    },
    {
      "entropy": 5.751816177368164,
      "epoch": 2.753861116514297,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004239464593684568,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20451211780309678,
      "num_tokens": 61412685.0,
      "step": 35395
    },
    {
      "entropy": 5.648649644851685,
      "epoch": 2.7542501458860142,
      "grad_norm": 1.125,
      "learning_rate": 0.00042392568491404043,
      "loss": 4.8484,
      "mean_token_accuracy": 0.21240620613098143,
      "num_tokens": 61421069.0,
      "step": 35400
    },
    {
      "entropy": 5.629816675186158,
      "epoch": 2.754639175257732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042390490819991367,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21290070712566375,
      "num_tokens": 61430349.0,
      "step": 35405
    },
    {
      "entropy": 5.737220096588135,
      "epoch": 2.7550282046294496,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000423884129226392,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20736957341432571,
      "num_tokens": 61439125.0,
      "step": 35410
    },
    {
      "entropy": 5.630013513565063,
      "epoch": 2.7554172340011673,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004238633479937906,
      "loss": 4.886,
      "mean_token_accuracy": 0.21245695501565934,
      "num_tokens": 61447649.0,
      "step": 35415
    },
    {
      "entropy": 5.65569953918457,
      "epoch": 2.7558062633728846,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004238425645024249,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21050097048282623,
      "num_tokens": 61455990.0,
      "step": 35420
    },
    {
      "entropy": 5.712196159362793,
      "epoch": 2.7561952927446023,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004238217787526102,
      "loss": 5.0257,
      "mean_token_accuracy": 0.19362433552742003,
      "num_tokens": 61464793.0,
      "step": 35425
    },
    {
      "entropy": 5.6828032493591305,
      "epoch": 2.7565843221163195,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004238009907446618,
      "loss": 4.9677,
      "mean_token_accuracy": 0.20889329314231872,
      "num_tokens": 61474072.0,
      "step": 35430
    },
    {
      "entropy": 5.693464088439941,
      "epoch": 2.7569733514880372,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042378020047889527,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21002658754587172,
      "num_tokens": 61483518.0,
      "step": 35435
    },
    {
      "entropy": 5.7235311508178714,
      "epoch": 2.757362380859755,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042375940795562594,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20611643195152282,
      "num_tokens": 61491967.0,
      "step": 35440
    },
    {
      "entropy": 5.803478050231933,
      "epoch": 2.7577514102314726,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042373861317516944,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20099324882030487,
      "num_tokens": 61501184.0,
      "step": 35445
    },
    {
      "entropy": 5.796035718917847,
      "epoch": 2.75814043960319,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004237178161378412,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20037172138690948,
      "num_tokens": 61509716.0,
      "step": 35450
    },
    {
      "entropy": 5.732233142852783,
      "epoch": 2.7585294689749076,
      "grad_norm": 1.25,
      "learning_rate": 0.0004236970168439568,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20371731817722322,
      "num_tokens": 61518267.0,
      "step": 35455
    },
    {
      "entropy": 5.654461622238159,
      "epoch": 2.7589184983466253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042367621529383186,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21106842011213303,
      "num_tokens": 61527204.0,
      "step": 35460
    },
    {
      "entropy": 5.6929707527160645,
      "epoch": 2.7593075277183425,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042365541148778205,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2074694573879242,
      "num_tokens": 61535888.0,
      "step": 35465
    },
    {
      "entropy": 5.701891756057739,
      "epoch": 2.7596965570900602,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042363460542612285,
      "loss": 4.9625,
      "mean_token_accuracy": 0.20828531682491302,
      "num_tokens": 61544570.0,
      "step": 35470
    },
    {
      "entropy": 5.780369234085083,
      "epoch": 2.760085586461778,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004236137971091702,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20018284171819686,
      "num_tokens": 61553450.0,
      "step": 35475
    },
    {
      "entropy": 5.709358787536621,
      "epoch": 2.7604746158334956,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042359298653723963,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21125064939260482,
      "num_tokens": 61562063.0,
      "step": 35480
    },
    {
      "entropy": 5.760407876968384,
      "epoch": 2.760863645205213,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004235721737106471,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19726235121488572,
      "num_tokens": 61570962.0,
      "step": 35485
    },
    {
      "entropy": 5.729723215103149,
      "epoch": 2.7612526745769306,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004235513586297082,
      "loss": 4.9792,
      "mean_token_accuracy": 0.20766578018665313,
      "num_tokens": 61579070.0,
      "step": 35490
    },
    {
      "entropy": 5.749112558364868,
      "epoch": 2.7616417039486483,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042353054129473894,
      "loss": 4.8855,
      "mean_token_accuracy": 0.20797699838876724,
      "num_tokens": 61587573.0,
      "step": 35495
    },
    {
      "entropy": 5.667645025253296,
      "epoch": 2.7620307333203655,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004235097217060552,
      "loss": 4.9528,
      "mean_token_accuracy": 0.20631778091192246,
      "num_tokens": 61596242.0,
      "step": 35500
    },
    {
      "entropy": 5.622595405578613,
      "epoch": 2.7624197626920832,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004234888998639727,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2085738644003868,
      "num_tokens": 61604529.0,
      "step": 35505
    },
    {
      "entropy": 5.677980232238769,
      "epoch": 2.762808792063801,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004234680757688076,
      "loss": 4.9884,
      "mean_token_accuracy": 0.19754796326160431,
      "num_tokens": 61612600.0,
      "step": 35510
    },
    {
      "entropy": 5.761327028274536,
      "epoch": 2.7631978214355186,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042344724942087564,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2115863710641861,
      "num_tokens": 61621267.0,
      "step": 35515
    },
    {
      "entropy": 5.772805881500244,
      "epoch": 2.763586850807236,
      "grad_norm": 1.265625,
      "learning_rate": 0.000423426420820493,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20608966797590256,
      "num_tokens": 61630010.0,
      "step": 35520
    },
    {
      "entropy": 5.659929895401001,
      "epoch": 2.7639758801789536,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042340558996797564,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2056085631251335,
      "num_tokens": 61638118.0,
      "step": 35525
    },
    {
      "entropy": 5.716986083984375,
      "epoch": 2.764364909550671,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042338475686363964,
      "loss": 5.0057,
      "mean_token_accuracy": 0.208828803896904,
      "num_tokens": 61646895.0,
      "step": 35530
    },
    {
      "entropy": 5.736470317840576,
      "epoch": 2.7647539389223885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004233639215078012,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20760715305805205,
      "num_tokens": 61655285.0,
      "step": 35535
    },
    {
      "entropy": 5.779675054550171,
      "epoch": 2.765142968294106,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042334308390077637,
      "loss": 4.9438,
      "mean_token_accuracy": 0.20602003633975982,
      "num_tokens": 61663910.0,
      "step": 35540
    },
    {
      "entropy": 5.682765626907349,
      "epoch": 2.765531997665824,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004233222440428814,
      "loss": 5.003,
      "mean_token_accuracy": 0.20733555108308793,
      "num_tokens": 61673241.0,
      "step": 35545
    },
    {
      "entropy": 5.6576752185821535,
      "epoch": 2.765921027037541,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004233014019344324,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21079717427492142,
      "num_tokens": 61683171.0,
      "step": 35550
    },
    {
      "entropy": 5.722889804840088,
      "epoch": 2.766310056409259,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004232805575757457,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20943217873573303,
      "num_tokens": 61691100.0,
      "step": 35555
    },
    {
      "entropy": 5.722838687896728,
      "epoch": 2.7666990857809766,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004232597109671375,
      "loss": 5.011,
      "mean_token_accuracy": 0.205571249127388,
      "num_tokens": 61699658.0,
      "step": 35560
    },
    {
      "entropy": 5.675302743911743,
      "epoch": 2.767088115152694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004232388621089242,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20427283346652986,
      "num_tokens": 61708461.0,
      "step": 35565
    },
    {
      "entropy": 5.692220735549927,
      "epoch": 2.7674771445244115,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042321801100142216,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21660492122173308,
      "num_tokens": 61717114.0,
      "step": 35570
    },
    {
      "entropy": 5.6809436798095705,
      "epoch": 2.767866173896129,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042319715764494765,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20016030520200728,
      "num_tokens": 61725746.0,
      "step": 35575
    },
    {
      "entropy": 5.739760828018189,
      "epoch": 2.768255203267847,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042317630203981714,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2071981683373451,
      "num_tokens": 61733875.0,
      "step": 35580
    },
    {
      "entropy": 5.779660940170288,
      "epoch": 2.768644232639564,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042315544418634714,
      "loss": 5.068,
      "mean_token_accuracy": 0.2075582906603813,
      "num_tokens": 61742009.0,
      "step": 35585
    },
    {
      "entropy": 5.663666772842407,
      "epoch": 2.769033262011282,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042313458408485414,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21615182608366013,
      "num_tokens": 61750723.0,
      "step": 35590
    },
    {
      "entropy": 5.7175534725189205,
      "epoch": 2.7694222913829996,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042311372173565456,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1982097789645195,
      "num_tokens": 61760468.0,
      "step": 35595
    },
    {
      "entropy": 5.663492345809937,
      "epoch": 2.769811320754717,
      "grad_norm": 1.1875,
      "learning_rate": 0.000423092857139065,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21069636791944504,
      "num_tokens": 61769225.0,
      "step": 35600
    },
    {
      "entropy": 5.698661518096924,
      "epoch": 2.7702003501264345,
      "grad_norm": 1.21875,
      "learning_rate": 0.000423071990295402,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20075487792491914,
      "num_tokens": 61777585.0,
      "step": 35605
    },
    {
      "entropy": 5.693813514709473,
      "epoch": 2.770589379498152,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004230511212049822,
      "loss": 4.9709,
      "mean_token_accuracy": 0.20372925698757172,
      "num_tokens": 61785816.0,
      "step": 35610
    },
    {
      "entropy": 5.749909496307373,
      "epoch": 2.77097840886987,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004230302498681224,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20659043937921523,
      "num_tokens": 61794929.0,
      "step": 35615
    },
    {
      "entropy": 5.6977935314178465,
      "epoch": 2.771367438241587,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004230093762851392,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21018249839544295,
      "num_tokens": 61803607.0,
      "step": 35620
    },
    {
      "entropy": 5.657301187515259,
      "epoch": 2.771756467613305,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004229885004563491,
      "loss": 4.9397,
      "mean_token_accuracy": 0.20955567806959152,
      "num_tokens": 61812595.0,
      "step": 35625
    },
    {
      "entropy": 5.657483673095703,
      "epoch": 2.772145496985022,
      "grad_norm": 1.25,
      "learning_rate": 0.0004229676223820692,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21038243174552917,
      "num_tokens": 61820474.0,
      "step": 35630
    },
    {
      "entropy": 5.70485634803772,
      "epoch": 2.77253452635674,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042294674206261616,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2030025079846382,
      "num_tokens": 61829784.0,
      "step": 35635
    },
    {
      "entropy": 5.714876699447632,
      "epoch": 2.7729235557284575,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004229258594983067,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20952286720275878,
      "num_tokens": 61839316.0,
      "step": 35640
    },
    {
      "entropy": 5.697353458404541,
      "epoch": 2.773312585100175,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004229049746894578,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2109530121088028,
      "num_tokens": 61847705.0,
      "step": 35645
    },
    {
      "entropy": 5.711152839660644,
      "epoch": 2.773701614471893,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004228840876363864,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2050390660762787,
      "num_tokens": 61856028.0,
      "step": 35650
    },
    {
      "entropy": 5.803388404846191,
      "epoch": 2.77409064384361,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004228631983394093,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20413737446069719,
      "num_tokens": 61864836.0,
      "step": 35655
    },
    {
      "entropy": 5.68941102027893,
      "epoch": 2.774479673215328,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004228423067988434,
      "loss": 4.9534,
      "mean_token_accuracy": 0.20565738528966904,
      "num_tokens": 61873179.0,
      "step": 35660
    },
    {
      "entropy": 5.7224853515625,
      "epoch": 2.774868702587045,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042282141301500597,
      "loss": 5.044,
      "mean_token_accuracy": 0.20796093493700027,
      "num_tokens": 61881068.0,
      "step": 35665
    },
    {
      "entropy": 5.668208265304566,
      "epoch": 2.775257731958763,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004228005169882138,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2157965287566185,
      "num_tokens": 61889640.0,
      "step": 35670
    },
    {
      "entropy": 5.708583879470825,
      "epoch": 2.7756467613304805,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042277961871878413,
      "loss": 5.0447,
      "mean_token_accuracy": 0.19970574527978896,
      "num_tokens": 61898267.0,
      "step": 35675
    },
    {
      "entropy": 5.707902479171753,
      "epoch": 2.776035790702198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004227587182070338,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21368465274572374,
      "num_tokens": 61906166.0,
      "step": 35680
    },
    {
      "entropy": 5.64837007522583,
      "epoch": 2.7764248200739154,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004227378154532802,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21473677307367325,
      "num_tokens": 61915335.0,
      "step": 35685
    },
    {
      "entropy": 5.655986642837524,
      "epoch": 2.776813849445633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004227169104578404,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2053390085697174,
      "num_tokens": 61924969.0,
      "step": 35690
    },
    {
      "entropy": 5.758424377441406,
      "epoch": 2.777202878817351,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042269600322103164,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21231439411640168,
      "num_tokens": 61933817.0,
      "step": 35695
    },
    {
      "entropy": 5.74832673072815,
      "epoch": 2.777591908189068,
      "grad_norm": 1.34375,
      "learning_rate": 0.000422675093743171,
      "loss": 4.9901,
      "mean_token_accuracy": 0.1953140988945961,
      "num_tokens": 61941870.0,
      "step": 35700
    },
    {
      "entropy": 5.715137577056884,
      "epoch": 2.777980937560786,
      "grad_norm": 1.25,
      "learning_rate": 0.00042265418202457603,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2094820261001587,
      "num_tokens": 61950559.0,
      "step": 35705
    },
    {
      "entropy": 5.650943994522095,
      "epoch": 2.7783699669325035,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004226332680655637,
      "loss": 4.9542,
      "mean_token_accuracy": 0.20223014950752258,
      "num_tokens": 61960138.0,
      "step": 35710
    },
    {
      "entropy": 5.752165842056274,
      "epoch": 2.778758996304221,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004226123518664516,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2023313581943512,
      "num_tokens": 61968555.0,
      "step": 35715
    },
    {
      "entropy": 5.743624353408814,
      "epoch": 2.7791480256759384,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004225914334275569,
      "loss": 5.009,
      "mean_token_accuracy": 0.20128106325864792,
      "num_tokens": 61976572.0,
      "step": 35720
    },
    {
      "entropy": 5.657948112487793,
      "epoch": 2.779537055047656,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004225705127491972,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2100988283753395,
      "num_tokens": 61984479.0,
      "step": 35725
    },
    {
      "entropy": 5.729463958740235,
      "epoch": 2.7799260844193734,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004225495898316899,
      "loss": 5.022,
      "mean_token_accuracy": 0.20638817995786668,
      "num_tokens": 61993098.0,
      "step": 35730
    },
    {
      "entropy": 5.774903917312622,
      "epoch": 2.780315113791091,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00042252866467535233,
      "loss": 4.9735,
      "mean_token_accuracy": 0.20686670392751694,
      "num_tokens": 62001244.0,
      "step": 35735
    },
    {
      "entropy": 5.684518432617187,
      "epoch": 2.780704143162809,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042250773728050214,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2281483843922615,
      "num_tokens": 62010121.0,
      "step": 35740
    },
    {
      "entropy": 5.719108390808105,
      "epoch": 2.7810931725345265,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004224868076474568,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20883809477090837,
      "num_tokens": 62018668.0,
      "step": 35745
    },
    {
      "entropy": 5.748516178131103,
      "epoch": 2.781482201906244,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042246587577653394,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20620885342359543,
      "num_tokens": 62027198.0,
      "step": 35750
    },
    {
      "entropy": 5.839861679077148,
      "epoch": 2.7818712312779614,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004224449416680511,
      "loss": 5.0716,
      "mean_token_accuracy": 0.20140437930822372,
      "num_tokens": 62037359.0,
      "step": 35755
    },
    {
      "entropy": 5.723645544052124,
      "epoch": 2.782260260649679,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042242400532232593,
      "loss": 4.9379,
      "mean_token_accuracy": 0.20908773988485335,
      "num_tokens": 62046110.0,
      "step": 35760
    },
    {
      "entropy": 5.693311262130737,
      "epoch": 2.7826492900213964,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042240306673967614,
      "loss": 4.9137,
      "mean_token_accuracy": 0.212683741748333,
      "num_tokens": 62055265.0,
      "step": 35765
    },
    {
      "entropy": 5.706912231445313,
      "epoch": 2.783038319393114,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042238212592041945,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21394198387861252,
      "num_tokens": 62063406.0,
      "step": 35770
    },
    {
      "entropy": 5.6545721054077145,
      "epoch": 2.7834273487648318,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042236118286487364,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21695060431957244,
      "num_tokens": 62071998.0,
      "step": 35775
    },
    {
      "entropy": 5.6538702011108395,
      "epoch": 2.7838163781365495,
      "grad_norm": 1.25,
      "learning_rate": 0.00042234023757335633,
      "loss": 4.91,
      "mean_token_accuracy": 0.2060961663722992,
      "num_tokens": 62081007.0,
      "step": 35780
    },
    {
      "entropy": 5.787205791473388,
      "epoch": 2.7842054075082667,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042231929004618547,
      "loss": 4.999,
      "mean_token_accuracy": 0.20946573615074157,
      "num_tokens": 62089164.0,
      "step": 35785
    },
    {
      "entropy": 5.699540138244629,
      "epoch": 2.7845944368799844,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004222983402836789,
      "loss": 4.892,
      "mean_token_accuracy": 0.2104268491268158,
      "num_tokens": 62098300.0,
      "step": 35790
    },
    {
      "entropy": 5.799757480621338,
      "epoch": 2.784983466251702,
      "grad_norm": 1.421875,
      "learning_rate": 0.00042227738828615435,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2022045522928238,
      "num_tokens": 62106384.0,
      "step": 35795
    },
    {
      "entropy": 5.671920871734619,
      "epoch": 2.7853724956234194,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042225643405392993,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20571552217006683,
      "num_tokens": 62114988.0,
      "step": 35800
    },
    {
      "entropy": 5.740653562545776,
      "epoch": 2.785761524995137,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042223547758732356,
      "loss": 4.9827,
      "mean_token_accuracy": 0.20759759545326234,
      "num_tokens": 62123316.0,
      "step": 35805
    },
    {
      "entropy": 5.717961215972901,
      "epoch": 2.7861505543668548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042221451888665307,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21416011303663254,
      "num_tokens": 62132034.0,
      "step": 35810
    },
    {
      "entropy": 5.720687198638916,
      "epoch": 2.7865395837385725,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042219355795223657,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20303915143013002,
      "num_tokens": 62140671.0,
      "step": 35815
    },
    {
      "entropy": 5.787706518173218,
      "epoch": 2.7869286131102897,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042217259478439216,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20567448139190675,
      "num_tokens": 62149581.0,
      "step": 35820
    },
    {
      "entropy": 5.77656741142273,
      "epoch": 2.7873176424820074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004221516293834379,
      "loss": 4.9271,
      "mean_token_accuracy": 0.20878140777349471,
      "num_tokens": 62157618.0,
      "step": 35825
    },
    {
      "entropy": 5.738287925720215,
      "epoch": 2.787706671853725,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042213066174969177,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20463876128196717,
      "num_tokens": 62166508.0,
      "step": 35830
    },
    {
      "entropy": 5.669754219055176,
      "epoch": 2.7880957012254424,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042210969188347217,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20245258659124374,
      "num_tokens": 62174804.0,
      "step": 35835
    },
    {
      "entropy": 5.654208517074585,
      "epoch": 2.78848473059716,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004220887197850971,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21587610989809036,
      "num_tokens": 62183071.0,
      "step": 35840
    },
    {
      "entropy": 5.7013321876525875,
      "epoch": 2.7888737599688778,
      "grad_norm": 1.25,
      "learning_rate": 0.00042206774545488476,
      "loss": 4.8737,
      "mean_token_accuracy": 0.21504210829734802,
      "num_tokens": 62191595.0,
      "step": 35845
    },
    {
      "entropy": 5.748844242095947,
      "epoch": 2.7892627893405955,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004220467688931535,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21111344397068024,
      "num_tokens": 62200552.0,
      "step": 35850
    },
    {
      "entropy": 5.69690203666687,
      "epoch": 2.7896518187123127,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004220257901002216,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20231180489063263,
      "num_tokens": 62209492.0,
      "step": 35855
    },
    {
      "entropy": 5.720066785812378,
      "epoch": 2.7900408480840304,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004220048090764073,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20598178058862687,
      "num_tokens": 62218697.0,
      "step": 35860
    },
    {
      "entropy": 5.788217353820801,
      "epoch": 2.7904298774557477,
      "grad_norm": 1.25,
      "learning_rate": 0.00042198382582202905,
      "loss": 5.0089,
      "mean_token_accuracy": 0.20639549642801286,
      "num_tokens": 62227216.0,
      "step": 35865
    },
    {
      "entropy": 5.76427116394043,
      "epoch": 2.7908189068274654,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004219628403374052,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2036355823278427,
      "num_tokens": 62235782.0,
      "step": 35870
    },
    {
      "entropy": 5.690377855300904,
      "epoch": 2.791207936199183,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004219418526228541,
      "loss": 5.01,
      "mean_token_accuracy": 0.208826544880867,
      "num_tokens": 62244873.0,
      "step": 35875
    },
    {
      "entropy": 5.632970905303955,
      "epoch": 2.7915969655709008,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004219208626786943,
      "loss": 4.8512,
      "mean_token_accuracy": 0.21261836290359498,
      "num_tokens": 62253053.0,
      "step": 35880
    },
    {
      "entropy": 5.775230979919433,
      "epoch": 2.791985994942618,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004218998705052443,
      "loss": 5.0576,
      "mean_token_accuracy": 0.19887065440416335,
      "num_tokens": 62261107.0,
      "step": 35885
    },
    {
      "entropy": 5.751078319549561,
      "epoch": 2.7923750243143357,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042187887610282255,
      "loss": 5.0823,
      "mean_token_accuracy": 0.19353830963373184,
      "num_tokens": 62269722.0,
      "step": 35890
    },
    {
      "entropy": 5.7680480003356935,
      "epoch": 2.7927640536860534,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004218578794717476,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2026605099439621,
      "num_tokens": 62278396.0,
      "step": 35895
    },
    {
      "entropy": 5.759116315841675,
      "epoch": 2.7931530830577707,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042183688061233813,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21301019042730332,
      "num_tokens": 62286652.0,
      "step": 35900
    },
    {
      "entropy": 5.651119422912598,
      "epoch": 2.7935421124294884,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042181587952491267,
      "loss": 4.8258,
      "mean_token_accuracy": 0.22090516686439515,
      "num_tokens": 62294252.0,
      "step": 35905
    },
    {
      "entropy": 5.645628356933594,
      "epoch": 2.793931141801206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042179487620978997,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2145445540547371,
      "num_tokens": 62303222.0,
      "step": 35910
    },
    {
      "entropy": 5.763366746902466,
      "epoch": 2.7943201711729238,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004217738706672886,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21124972701072692,
      "num_tokens": 62311537.0,
      "step": 35915
    },
    {
      "entropy": 5.638851881027222,
      "epoch": 2.794709200544641,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042175286289772746,
      "loss": 4.8121,
      "mean_token_accuracy": 0.212005515396595,
      "num_tokens": 62320073.0,
      "step": 35920
    },
    {
      "entropy": 5.679330682754516,
      "epoch": 2.7950982299163587,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004217318529014251,
      "loss": 4.8991,
      "mean_token_accuracy": 0.213081257045269,
      "num_tokens": 62328253.0,
      "step": 35925
    },
    {
      "entropy": 5.745542716979981,
      "epoch": 2.7954872592880764,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004217108406787005,
      "loss": 5.0024,
      "mean_token_accuracy": 0.20184246748685836,
      "num_tokens": 62337127.0,
      "step": 35930
    },
    {
      "entropy": 5.764872026443482,
      "epoch": 2.7958762886597937,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004216898262298724,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2094544991850853,
      "num_tokens": 62345353.0,
      "step": 35935
    },
    {
      "entropy": 5.67405252456665,
      "epoch": 2.7962653180315113,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004216688095552596,
      "loss": 4.9036,
      "mean_token_accuracy": 0.20869094133377075,
      "num_tokens": 62354319.0,
      "step": 35940
    },
    {
      "entropy": 5.628344535827637,
      "epoch": 2.796654347403229,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004216477906551811,
      "loss": 4.8801,
      "mean_token_accuracy": 0.215924035012722,
      "num_tokens": 62363056.0,
      "step": 35945
    },
    {
      "entropy": 5.703462934494018,
      "epoch": 2.7970433767749467,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042162676952995585,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21067634969949722,
      "num_tokens": 62370999.0,
      "step": 35950
    },
    {
      "entropy": 5.74155559539795,
      "epoch": 2.797432406146664,
      "grad_norm": 1.375,
      "learning_rate": 0.00042160574617990267,
      "loss": 5.052,
      "mean_token_accuracy": 0.2084408476948738,
      "num_tokens": 62379021.0,
      "step": 35955
    },
    {
      "entropy": 5.708321475982666,
      "epoch": 2.7978214355183817,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042158472060534066,
      "loss": 4.9317,
      "mean_token_accuracy": 0.20339378118515014,
      "num_tokens": 62387223.0,
      "step": 35960
    },
    {
      "entropy": 5.725368881225586,
      "epoch": 2.798210464890099,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042156369280658886,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20852207243442536,
      "num_tokens": 62396702.0,
      "step": 35965
    },
    {
      "entropy": 5.778937292098999,
      "epoch": 2.7985994942618166,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004215426627839662,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19808615893125534,
      "num_tokens": 62405778.0,
      "step": 35970
    },
    {
      "entropy": 5.724852848052978,
      "epoch": 2.7989885236335343,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042152163053779196,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20039259791374206,
      "num_tokens": 62413882.0,
      "step": 35975
    },
    {
      "entropy": 5.725995206832886,
      "epoch": 2.799377553005252,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004215005960683852,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2115092784166336,
      "num_tokens": 62422142.0,
      "step": 35980
    },
    {
      "entropy": 5.626158285140991,
      "epoch": 2.7997665823769697,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004214795593760651,
      "loss": 4.8665,
      "mean_token_accuracy": 0.21968503743410112,
      "num_tokens": 62430902.0,
      "step": 35985
    },
    {
      "entropy": 5.7147032737731935,
      "epoch": 2.800155611748687,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004214585204611508,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20075087249279022,
      "num_tokens": 62440205.0,
      "step": 35990
    },
    {
      "entropy": 5.730164194107056,
      "epoch": 2.8005446411204047,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042143747932396153,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2176876559853554,
      "num_tokens": 62448730.0,
      "step": 35995
    },
    {
      "entropy": 5.714774942398071,
      "epoch": 2.800933670492122,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004214164359648166,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2043631926178932,
      "num_tokens": 62457323.0,
      "step": 36000
    },
    {
      "epoch": 2.800933670492122,
      "eval_entropy": 5.429411687199259,
      "eval_loss": 5.05598783493042,
      "eval_mean_token_accuracy": 0.21343226708450264,
      "eval_num_tokens": 62457323.0,
      "eval_runtime": 21.6577,
      "eval_samples_per_second": 1918.858,
      "eval_steps_per_second": 239.869,
      "step": 36000
    },
    {
      "entropy": 5.791463422775268,
      "epoch": 2.8013226998638396,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042139539038403535,
      "loss": 5.165,
      "mean_token_accuracy": 0.19221772253513336,
      "num_tokens": 62466328.0,
      "step": 36005
    },
    {
      "entropy": 5.803160095214844,
      "epoch": 2.8017117292355573,
      "grad_norm": 1.234375,
      "learning_rate": 0.000421374342581937,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20249293595552445,
      "num_tokens": 62475063.0,
      "step": 36010
    },
    {
      "entropy": 5.7295591831207275,
      "epoch": 2.802100758607275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042135329255884105,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21261500120162963,
      "num_tokens": 62483027.0,
      "step": 36015
    },
    {
      "entropy": 5.677012729644775,
      "epoch": 2.8024897879789923,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004213322403150668,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2160395175218582,
      "num_tokens": 62491762.0,
      "step": 36020
    },
    {
      "entropy": 5.730729246139527,
      "epoch": 2.80287881735071,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004213111858509337,
      "loss": 4.947,
      "mean_token_accuracy": 0.21364181041717528,
      "num_tokens": 62500340.0,
      "step": 36025
    },
    {
      "entropy": 5.767397356033325,
      "epoch": 2.8032678467224277,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004212901291667612,
      "loss": 4.9995,
      "mean_token_accuracy": 0.20482261925935746,
      "num_tokens": 62508580.0,
      "step": 36030
    },
    {
      "entropy": 5.642378950119019,
      "epoch": 2.803656876094145,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042126907026286887,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2144310876727104,
      "num_tokens": 62516790.0,
      "step": 36035
    },
    {
      "entropy": 5.7141297340393065,
      "epoch": 2.8040459054658626,
      "grad_norm": 1.375,
      "learning_rate": 0.00042124800913957624,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20203347951173783,
      "num_tokens": 62525525.0,
      "step": 36040
    },
    {
      "entropy": 5.7222494125366214,
      "epoch": 2.8044349348375803,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004212269457972028,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20563951283693313,
      "num_tokens": 62535948.0,
      "step": 36045
    },
    {
      "entropy": 5.6770265102386475,
      "epoch": 2.804823964209298,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004212058802360682,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2170703276991844,
      "num_tokens": 62544663.0,
      "step": 36050
    },
    {
      "entropy": 5.758439588546753,
      "epoch": 2.8052129935810153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004211848124564922,
      "loss": 5.0331,
      "mean_token_accuracy": 0.197563074529171,
      "num_tokens": 62553017.0,
      "step": 36055
    },
    {
      "entropy": 5.758427858352661,
      "epoch": 2.805602022952733,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004211637424587942,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22168685644865035,
      "num_tokens": 62561311.0,
      "step": 36060
    },
    {
      "entropy": 5.781798505783081,
      "epoch": 2.8059910523244502,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004211426702432941,
      "loss": 5.0574,
      "mean_token_accuracy": 0.20180109292268752,
      "num_tokens": 62569387.0,
      "step": 36065
    },
    {
      "entropy": 5.731270790100098,
      "epoch": 2.806380081696168,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004211215958103116,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20723959654569626,
      "num_tokens": 62579250.0,
      "step": 36070
    },
    {
      "entropy": 5.709207820892334,
      "epoch": 2.8067691110678856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042110051916016645,
      "loss": 4.9485,
      "mean_token_accuracy": 0.20611853450536727,
      "num_tokens": 62587848.0,
      "step": 36075
    },
    {
      "entropy": 5.713175868988037,
      "epoch": 2.8071581404396033,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004210794402931785,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21044569313526154,
      "num_tokens": 62596476.0,
      "step": 36080
    },
    {
      "entropy": 5.76529450416565,
      "epoch": 2.807547169811321,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004210583592096675,
      "loss": 5.0252,
      "mean_token_accuracy": 0.20681067556142807,
      "num_tokens": 62604785.0,
      "step": 36085
    },
    {
      "entropy": 5.798939323425293,
      "epoch": 2.8079361991830383,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004210372759099534,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20493322014808654,
      "num_tokens": 62612963.0,
      "step": 36090
    },
    {
      "entropy": 5.768968534469605,
      "epoch": 2.808325228554756,
      "grad_norm": 1.21875,
      "learning_rate": 0.000421016190394356,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2029451459646225,
      "num_tokens": 62621504.0,
      "step": 36095
    },
    {
      "entropy": 5.869975519180298,
      "epoch": 2.8087142579264732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042099510266319545,
      "loss": 5.1654,
      "mean_token_accuracy": 0.19004778265953065,
      "num_tokens": 62630475.0,
      "step": 36100
    },
    {
      "entropy": 5.712640905380249,
      "epoch": 2.809103287298191,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004209740127167915,
      "loss": 4.9202,
      "mean_token_accuracy": 0.20918382406234742,
      "num_tokens": 62639241.0,
      "step": 36105
    },
    {
      "entropy": 5.698992586135864,
      "epoch": 2.8094923166699086,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004209529205554643,
      "loss": 4.9185,
      "mean_token_accuracy": 0.20782988667488098,
      "num_tokens": 62648073.0,
      "step": 36110
    },
    {
      "entropy": 5.671531391143799,
      "epoch": 2.8098813460416263,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004209318261795339,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21369728296995164,
      "num_tokens": 62657455.0,
      "step": 36115
    },
    {
      "entropy": 5.728784275054932,
      "epoch": 2.8102703754133436,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004209107295893202,
      "loss": 4.9584,
      "mean_token_accuracy": 0.20220387279987334,
      "num_tokens": 62665795.0,
      "step": 36120
    },
    {
      "entropy": 5.7702690124511715,
      "epoch": 2.8106594047850613,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042088963078514344,
      "loss": 5.0101,
      "mean_token_accuracy": 0.20522227883338928,
      "num_tokens": 62674095.0,
      "step": 36125
    },
    {
      "entropy": 5.742775821685791,
      "epoch": 2.811048434156779,
      "grad_norm": 1.375,
      "learning_rate": 0.0004208685297673238,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20825452357530594,
      "num_tokens": 62683195.0,
      "step": 36130
    },
    {
      "entropy": 5.754679727554321,
      "epoch": 2.811437463528496,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004208474265361813,
      "loss": 5.0454,
      "mean_token_accuracy": 0.19705945551395415,
      "num_tokens": 62692135.0,
      "step": 36135
    },
    {
      "entropy": 5.735333490371704,
      "epoch": 2.811826492900214,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042082632109203627,
      "loss": 4.9302,
      "mean_token_accuracy": 0.21691171824932098,
      "num_tokens": 62700417.0,
      "step": 36140
    },
    {
      "entropy": 5.625975322723389,
      "epoch": 2.8122155222719316,
      "grad_norm": 1.234375,
      "learning_rate": 0.000420805213435209,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21859325021505355,
      "num_tokens": 62708995.0,
      "step": 36145
    },
    {
      "entropy": 5.696475839614868,
      "epoch": 2.8126045516436493,
      "grad_norm": 1.25,
      "learning_rate": 0.0004207841035660196,
      "loss": 5.02,
      "mean_token_accuracy": 0.1998225823044777,
      "num_tokens": 62718798.0,
      "step": 36150
    },
    {
      "entropy": 5.794991970062256,
      "epoch": 2.8129935810153666,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004207629914847885,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20139726251363754,
      "num_tokens": 62726916.0,
      "step": 36155
    },
    {
      "entropy": 5.784782934188843,
      "epoch": 2.8133826103870843,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000420741877191836,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20129092782735825,
      "num_tokens": 62735644.0,
      "step": 36160
    },
    {
      "entropy": 5.72823395729065,
      "epoch": 2.8137716397588015,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004207207606874825,
      "loss": 5.0134,
      "mean_token_accuracy": 0.20797159522771835,
      "num_tokens": 62743743.0,
      "step": 36165
    },
    {
      "entropy": 5.794711542129517,
      "epoch": 2.814160669130519,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004206996419720484,
      "loss": 5.112,
      "mean_token_accuracy": 0.19558885991573333,
      "num_tokens": 62753181.0,
      "step": 36170
    },
    {
      "entropy": 5.749361419677735,
      "epoch": 2.814549698502237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004206785210458541,
      "loss": 5.0607,
      "mean_token_accuracy": 0.2063676342368126,
      "num_tokens": 62763284.0,
      "step": 36175
    },
    {
      "entropy": 5.824263048171997,
      "epoch": 2.8149387278739546,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004206573979092202,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2060114935040474,
      "num_tokens": 62772051.0,
      "step": 36180
    },
    {
      "entropy": 5.833756780624389,
      "epoch": 2.8153277572456723,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042063627256246706,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2109333708882332,
      "num_tokens": 62781067.0,
      "step": 36185
    },
    {
      "entropy": 5.736289548873901,
      "epoch": 2.8157167866173896,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004206151450059153,
      "loss": 5.0706,
      "mean_token_accuracy": 0.20158449709415435,
      "num_tokens": 62789902.0,
      "step": 36190
    },
    {
      "entropy": 5.734937381744385,
      "epoch": 2.8161058159891073,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042059401523988546,
      "loss": 5.0163,
      "mean_token_accuracy": 0.20438360571861267,
      "num_tokens": 62799239.0,
      "step": 36195
    },
    {
      "entropy": 5.77956714630127,
      "epoch": 2.8164948453608245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004205728832646982,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20546602755784987,
      "num_tokens": 62808060.0,
      "step": 36200
    },
    {
      "entropy": 5.71492748260498,
      "epoch": 2.816883874732542,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042055174908067414,
      "loss": 4.8684,
      "mean_token_accuracy": 0.222498819231987,
      "num_tokens": 62816638.0,
      "step": 36205
    },
    {
      "entropy": 5.75117039680481,
      "epoch": 2.81727290410426,
      "grad_norm": 1.25,
      "learning_rate": 0.00042053061268813397,
      "loss": 4.9654,
      "mean_token_accuracy": 0.20772894620895385,
      "num_tokens": 62824895.0,
      "step": 36210
    },
    {
      "entropy": 5.730296993255616,
      "epoch": 2.8176619334759776,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042050947408739836,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20445550084114075,
      "num_tokens": 62834404.0,
      "step": 36215
    },
    {
      "entropy": 5.799823331832886,
      "epoch": 2.818050962847695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004204883332787881,
      "loss": 5.0055,
      "mean_token_accuracy": 0.20492597222328185,
      "num_tokens": 62842857.0,
      "step": 36220
    },
    {
      "entropy": 5.7177904605865475,
      "epoch": 2.8184399922194125,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042046719026262396,
      "loss": 4.9365,
      "mean_token_accuracy": 0.21294744610786437,
      "num_tokens": 62850845.0,
      "step": 36225
    },
    {
      "entropy": 5.773021268844604,
      "epoch": 2.8188290215911302,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042044604503922674,
      "loss": 5.08,
      "mean_token_accuracy": 0.20209589898586272,
      "num_tokens": 62859164.0,
      "step": 36230
    },
    {
      "entropy": 5.796916866302491,
      "epoch": 2.8192180509628475,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004204248976089173,
      "loss": 5.0269,
      "mean_token_accuracy": 0.2073966830968857,
      "num_tokens": 62868001.0,
      "step": 36235
    },
    {
      "entropy": 5.776257467269898,
      "epoch": 2.819607080334565,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004204037479720165,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20102550834417343,
      "num_tokens": 62877626.0,
      "step": 36240
    },
    {
      "entropy": 5.783353662490844,
      "epoch": 2.819996109706283,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042038259612884514,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21136777549982072,
      "num_tokens": 62886376.0,
      "step": 36245
    },
    {
      "entropy": 5.724270057678223,
      "epoch": 2.8203851390780006,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042036144207972445,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20730339288711547,
      "num_tokens": 62895001.0,
      "step": 36250
    },
    {
      "entropy": 5.683387088775635,
      "epoch": 2.820774168449718,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004203402858249752,
      "loss": 4.999,
      "mean_token_accuracy": 0.20336652249097825,
      "num_tokens": 62903425.0,
      "step": 36255
    },
    {
      "entropy": 5.746587467193604,
      "epoch": 2.8211631978214355,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004203191273649184,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21540874987840652,
      "num_tokens": 62911904.0,
      "step": 36260
    },
    {
      "entropy": 5.734385871887207,
      "epoch": 2.8215522271931532,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042029796669987513,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20774740725755692,
      "num_tokens": 62920576.0,
      "step": 36265
    },
    {
      "entropy": 5.72759370803833,
      "epoch": 2.8219412565648705,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004202768038301665,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21235346645116807,
      "num_tokens": 62929381.0,
      "step": 36270
    },
    {
      "entropy": 5.728389263153076,
      "epoch": 2.822330285936588,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042025563875611367,
      "loss": 4.9659,
      "mean_token_accuracy": 0.20625995546579362,
      "num_tokens": 62937903.0,
      "step": 36275
    },
    {
      "entropy": 5.71770830154419,
      "epoch": 2.822719315308306,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004202344714780376,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20706982910633087,
      "num_tokens": 62946630.0,
      "step": 36280
    },
    {
      "entropy": 5.759347295761108,
      "epoch": 2.8231083446800236,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042021330199625966,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21002335697412491,
      "num_tokens": 62954788.0,
      "step": 36285
    },
    {
      "entropy": 5.859901714324951,
      "epoch": 2.823497374051741,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042019213031110105,
      "loss": 5.1072,
      "mean_token_accuracy": 0.19550697207450868,
      "num_tokens": 62964091.0,
      "step": 36290
    },
    {
      "entropy": 5.770223808288574,
      "epoch": 2.8238864034234585,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004201709564228829,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20188231766223907,
      "num_tokens": 62972372.0,
      "step": 36295
    },
    {
      "entropy": 5.76106743812561,
      "epoch": 2.824275432795176,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042014978033192655,
      "loss": 5.0056,
      "mean_token_accuracy": 0.20466916412115096,
      "num_tokens": 62981065.0,
      "step": 36300
    },
    {
      "entropy": 5.801694440841675,
      "epoch": 2.8246644621668935,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042012860203855337,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20249920934438706,
      "num_tokens": 62989363.0,
      "step": 36305
    },
    {
      "entropy": 5.687857389450073,
      "epoch": 2.825053491538611,
      "grad_norm": 2.40625,
      "learning_rate": 0.00042010742154308453,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2207600951194763,
      "num_tokens": 62998454.0,
      "step": 36310
    },
    {
      "entropy": 5.810584354400635,
      "epoch": 2.825442520910329,
      "grad_norm": 1.25,
      "learning_rate": 0.00042008623884584167,
      "loss": 5.0968,
      "mean_token_accuracy": 0.20089921057224275,
      "num_tokens": 63007089.0,
      "step": 36315
    },
    {
      "entropy": 5.6966633796691895,
      "epoch": 2.825831550282046,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042006505394714597,
      "loss": 4.9673,
      "mean_token_accuracy": 0.20338371843099595,
      "num_tokens": 63016702.0,
      "step": 36320
    },
    {
      "entropy": 5.687336874008179,
      "epoch": 2.826220579653764,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042004386684731896,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20911576896905898,
      "num_tokens": 63024917.0,
      "step": 36325
    },
    {
      "entropy": 5.785512065887451,
      "epoch": 2.8266096090254815,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042002267754668216,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20397927463054658,
      "num_tokens": 63034527.0,
      "step": 36330
    },
    {
      "entropy": 5.659380483627319,
      "epoch": 2.826998638397199,
      "grad_norm": 1.15625,
      "learning_rate": 0.000420001486045557,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21329390704631807,
      "num_tokens": 63043214.0,
      "step": 36335
    },
    {
      "entropy": 5.735488557815552,
      "epoch": 2.8273876677689165,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004199802923442652,
      "loss": 4.9652,
      "mean_token_accuracy": 0.20599957555532455,
      "num_tokens": 63052494.0,
      "step": 36340
    },
    {
      "entropy": 5.794697856903076,
      "epoch": 2.827776697140634,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004199590964431282,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2032136246562004,
      "num_tokens": 63060543.0,
      "step": 36345
    },
    {
      "entropy": 5.773991823196411,
      "epoch": 2.828165726512352,
      "grad_norm": 1.25,
      "learning_rate": 0.0004199378983424676,
      "loss": 5.0519,
      "mean_token_accuracy": 0.2013360008597374,
      "num_tokens": 63069498.0,
      "step": 36350
    },
    {
      "entropy": 5.7104329586029055,
      "epoch": 2.828554755884069,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041991669804260505,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20889620780944823,
      "num_tokens": 63077887.0,
      "step": 36355
    },
    {
      "entropy": 5.761520767211914,
      "epoch": 2.828943785255787,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004198954955438623,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20222052931785583,
      "num_tokens": 63086131.0,
      "step": 36360
    },
    {
      "entropy": 5.7471414566040036,
      "epoch": 2.8293328146275045,
      "grad_norm": 1.3125,
      "learning_rate": 0.000419874290846561,
      "loss": 5.0145,
      "mean_token_accuracy": 0.20410784035921098,
      "num_tokens": 63094609.0,
      "step": 36365
    },
    {
      "entropy": 5.684627103805542,
      "epoch": 2.8297218439992218,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041985308395102303,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21429855674505233,
      "num_tokens": 63103953.0,
      "step": 36370
    },
    {
      "entropy": 5.805222797393799,
      "epoch": 2.8301108733709395,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041983187485756997,
      "loss": 5.0726,
      "mean_token_accuracy": 0.19919697642326356,
      "num_tokens": 63112450.0,
      "step": 36375
    },
    {
      "entropy": 5.746405601501465,
      "epoch": 2.830499902742657,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041981066356652373,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2076180547475815,
      "num_tokens": 63120475.0,
      "step": 36380
    },
    {
      "entropy": 5.734838628768921,
      "epoch": 2.830888932114375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041978945007820626,
      "loss": 5.013,
      "mean_token_accuracy": 0.20328641235828399,
      "num_tokens": 63128957.0,
      "step": 36385
    },
    {
      "entropy": 5.671457099914551,
      "epoch": 2.831277961486092,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041976823439293925,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2095578968524933,
      "num_tokens": 63138060.0,
      "step": 36390
    },
    {
      "entropy": 5.797523546218872,
      "epoch": 2.83166699085781,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004197470165110447,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19969479739665985,
      "num_tokens": 63146829.0,
      "step": 36395
    },
    {
      "entropy": 5.738124752044678,
      "epoch": 2.832056020229527,
      "grad_norm": 1.25,
      "learning_rate": 0.0004197257964328446,
      "loss": 4.9643,
      "mean_token_accuracy": 0.20790702998638153,
      "num_tokens": 63155456.0,
      "step": 36400
    },
    {
      "entropy": 5.750972318649292,
      "epoch": 2.8324450496012448,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004197045741586609,
      "loss": 5.0152,
      "mean_token_accuracy": 0.201012921333313,
      "num_tokens": 63164725.0,
      "step": 36405
    },
    {
      "entropy": 5.69929838180542,
      "epoch": 2.8328340789729625,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004196833496888156,
      "loss": 4.8887,
      "mean_token_accuracy": 0.218691910803318,
      "num_tokens": 63173538.0,
      "step": 36410
    },
    {
      "entropy": 5.764781904220581,
      "epoch": 2.83322310834468,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004196621230236308,
      "loss": 5.0704,
      "mean_token_accuracy": 0.20636032670736312,
      "num_tokens": 63183579.0,
      "step": 36415
    },
    {
      "entropy": 5.722826623916626,
      "epoch": 2.833612137716398,
      "grad_norm": 1.25,
      "learning_rate": 0.00041964089416342845,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20294328182935714,
      "num_tokens": 63192820.0,
      "step": 36420
    },
    {
      "entropy": 5.73057746887207,
      "epoch": 2.834001167088115,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004196196631085308,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20626851171255112,
      "num_tokens": 63201140.0,
      "step": 36425
    },
    {
      "entropy": 5.662766361236573,
      "epoch": 2.834390196459833,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041959842985925987,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2086496114730835,
      "num_tokens": 63210036.0,
      "step": 36430
    },
    {
      "entropy": 5.76242733001709,
      "epoch": 2.83477922583155,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000419577194415938,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20027022957801818,
      "num_tokens": 63219318.0,
      "step": 36435
    },
    {
      "entropy": 5.853606271743774,
      "epoch": 2.8351682552032678,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041955595677888735,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20167972296476364,
      "num_tokens": 63228343.0,
      "step": 36440
    },
    {
      "entropy": 5.7254638671875,
      "epoch": 2.8355572845749855,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004195347169484301,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21235917657613754,
      "num_tokens": 63237365.0,
      "step": 36445
    },
    {
      "entropy": 5.658693838119507,
      "epoch": 2.835946313946703,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004195134749248886,
      "loss": 4.8399,
      "mean_token_accuracy": 0.20872615575790404,
      "num_tokens": 63245706.0,
      "step": 36450
    },
    {
      "entropy": 5.730026388168335,
      "epoch": 2.8363353433184204,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004194922307085851,
      "loss": 4.982,
      "mean_token_accuracy": 0.20654136538505555,
      "num_tokens": 63254487.0,
      "step": 36455
    },
    {
      "entropy": 5.760679006576538,
      "epoch": 2.836724372690138,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004194709842998419,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2080386370420456,
      "num_tokens": 63263196.0,
      "step": 36460
    },
    {
      "entropy": 5.744549036026001,
      "epoch": 2.837113402061856,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041944973569898157,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21010482013225557,
      "num_tokens": 63271812.0,
      "step": 36465
    },
    {
      "entropy": 5.680595779418946,
      "epoch": 2.837502431433573,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004194284849063265,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2081234112381935,
      "num_tokens": 63280423.0,
      "step": 36470
    },
    {
      "entropy": 5.695953369140625,
      "epoch": 2.8378914608052908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004194072319221989,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20743531435728074,
      "num_tokens": 63289565.0,
      "step": 36475
    },
    {
      "entropy": 5.777735948562622,
      "epoch": 2.8382804901770085,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004193859767469214,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2080550014972687,
      "num_tokens": 63298542.0,
      "step": 36480
    },
    {
      "entropy": 5.75899658203125,
      "epoch": 2.838669519548726,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041936471938081667,
      "loss": 5.037,
      "mean_token_accuracy": 0.20003886818885802,
      "num_tokens": 63307541.0,
      "step": 36485
    },
    {
      "entropy": 5.690267181396484,
      "epoch": 2.8390585489204434,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041934345982420695,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20870667248964309,
      "num_tokens": 63316470.0,
      "step": 36490
    },
    {
      "entropy": 5.74273362159729,
      "epoch": 2.839447578292161,
      "grad_norm": 1.21875,
      "learning_rate": 0.000419322198077415,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21208949089050294,
      "num_tokens": 63325458.0,
      "step": 36495
    },
    {
      "entropy": 5.657659482955933,
      "epoch": 2.8398366076638784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041930093414076344,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21417147070169448,
      "num_tokens": 63334287.0,
      "step": 36500
    },
    {
      "entropy": 5.693455219268799,
      "epoch": 2.840225637035596,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041927966801457486,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21681063622236252,
      "num_tokens": 63342980.0,
      "step": 36505
    },
    {
      "entropy": 5.7105707168579105,
      "epoch": 2.8406146664073137,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000419258399699172,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21078926473855972,
      "num_tokens": 63351479.0,
      "step": 36510
    },
    {
      "entropy": 5.708147573471069,
      "epoch": 2.8410036957790314,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004192371291948775,
      "loss": 4.9534,
      "mean_token_accuracy": 0.20328305214643477,
      "num_tokens": 63360010.0,
      "step": 36515
    },
    {
      "entropy": 5.74080605506897,
      "epoch": 2.841392725150749,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041921585650201413,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19890889823436736,
      "num_tokens": 63369522.0,
      "step": 36520
    },
    {
      "entropy": 5.8027712345123295,
      "epoch": 2.8417817545224664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004191945816209047,
      "loss": 5.0468,
      "mean_token_accuracy": 0.20152828395366668,
      "num_tokens": 63379365.0,
      "step": 36525
    },
    {
      "entropy": 5.754071569442749,
      "epoch": 2.842170783894184,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041917330455187195,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21570038050413132,
      "num_tokens": 63387837.0,
      "step": 36530
    },
    {
      "entropy": 5.652082347869873,
      "epoch": 2.8425598132659013,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041915202529523894,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21188283264636992,
      "num_tokens": 63396583.0,
      "step": 36535
    },
    {
      "entropy": 5.7183265209198,
      "epoch": 2.842948842637619,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004191307438513283,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2099013641476631,
      "num_tokens": 63405542.0,
      "step": 36540
    },
    {
      "entropy": 5.784355211257934,
      "epoch": 2.8433378720093367,
      "grad_norm": 1.3125,
      "learning_rate": 0.000419109460220463,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20915253460407257,
      "num_tokens": 63414482.0,
      "step": 36545
    },
    {
      "entropy": 5.6849198818206785,
      "epoch": 2.8437269013810544,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041908817440296605,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21275498420000077,
      "num_tokens": 63423928.0,
      "step": 36550
    },
    {
      "entropy": 5.705389785766601,
      "epoch": 2.8441159307527717,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041906688639916035,
      "loss": 4.9821,
      "mean_token_accuracy": 0.20450742691755294,
      "num_tokens": 63432186.0,
      "step": 36555
    },
    {
      "entropy": 5.640087652206421,
      "epoch": 2.8445049601244894,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004190455962093689,
      "loss": 4.8586,
      "mean_token_accuracy": 0.22283801734447478,
      "num_tokens": 63439630.0,
      "step": 36560
    },
    {
      "entropy": 5.6778799533844,
      "epoch": 2.844893989496207,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041902430383391494,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21052723228931428,
      "num_tokens": 63448121.0,
      "step": 36565
    },
    {
      "entropy": 5.6989398956298825,
      "epoch": 2.8452830188679243,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004190030092731214,
      "loss": 4.8142,
      "mean_token_accuracy": 0.2197665184736252,
      "num_tokens": 63455937.0,
      "step": 36570
    },
    {
      "entropy": 5.677034711837768,
      "epoch": 2.845672048239642,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004189817125273113,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20383487939834594,
      "num_tokens": 63464851.0,
      "step": 36575
    },
    {
      "entropy": 5.704415225982666,
      "epoch": 2.8460610776113597,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041896041359680797,
      "loss": 4.9104,
      "mean_token_accuracy": 0.21304958760738374,
      "num_tokens": 63473535.0,
      "step": 36580
    },
    {
      "entropy": 5.760348224639893,
      "epoch": 2.8464501069830774,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041893911248193437,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20645393133163453,
      "num_tokens": 63482755.0,
      "step": 36585
    },
    {
      "entropy": 5.704601144790649,
      "epoch": 2.8468391363547947,
      "grad_norm": 1.21875,
      "learning_rate": 0.000418917809183014,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21195828467607497,
      "num_tokens": 63491481.0,
      "step": 36590
    },
    {
      "entropy": 5.754874897003174,
      "epoch": 2.8472281657265124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041889650370036986,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20547961741685866,
      "num_tokens": 63500876.0,
      "step": 36595
    },
    {
      "entropy": 5.743144273757935,
      "epoch": 2.84761719509823,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004188751960343253,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21147698312997817,
      "num_tokens": 63508119.0,
      "step": 36600
    },
    {
      "entropy": 5.787603521347046,
      "epoch": 2.8480062244699473,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004188538861852037,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20377288907766342,
      "num_tokens": 63516078.0,
      "step": 36605
    },
    {
      "entropy": 5.736608552932739,
      "epoch": 2.848395253841665,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004188325741533283,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20324206799268724,
      "num_tokens": 63524687.0,
      "step": 36610
    },
    {
      "entropy": 5.728704929351807,
      "epoch": 2.8487842832133827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004188112599390225,
      "loss": 5.0595,
      "mean_token_accuracy": 0.19933152496814727,
      "num_tokens": 63534496.0,
      "step": 36615
    },
    {
      "entropy": 5.761440134048462,
      "epoch": 2.8491733125851004,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004187899435426098,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21075938194990157,
      "num_tokens": 63542890.0,
      "step": 36620
    },
    {
      "entropy": 5.856949710845948,
      "epoch": 2.8495623419568177,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041876862496441347,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20125951766967773,
      "num_tokens": 63550942.0,
      "step": 36625
    },
    {
      "entropy": 5.76570234298706,
      "epoch": 2.8499513713285354,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041874730420475716,
      "loss": 5.0691,
      "mean_token_accuracy": 0.19709621369838715,
      "num_tokens": 63559077.0,
      "step": 36630
    },
    {
      "entropy": 5.701242113113404,
      "epoch": 2.8503404007002526,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041872598126396434,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2118488684296608,
      "num_tokens": 63567837.0,
      "step": 36635
    },
    {
      "entropy": 5.715847158432007,
      "epoch": 2.8507294300719703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041870465614235843,
      "loss": 4.8609,
      "mean_token_accuracy": 0.20970326513051987,
      "num_tokens": 63576688.0,
      "step": 36640
    },
    {
      "entropy": 5.702781295776367,
      "epoch": 2.851118459443688,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041868332884026317,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2198580250144005,
      "num_tokens": 63585027.0,
      "step": 36645
    },
    {
      "entropy": 5.733483409881591,
      "epoch": 2.8515074888154057,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004186619993580021,
      "loss": 5.0045,
      "mean_token_accuracy": 0.20151478052139282,
      "num_tokens": 63594560.0,
      "step": 36650
    },
    {
      "entropy": 5.725176858901977,
      "epoch": 2.851896518187123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041864066769589875,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20282082259655,
      "num_tokens": 63603301.0,
      "step": 36655
    },
    {
      "entropy": 5.761955881118775,
      "epoch": 2.8522855475588407,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041861933385427697,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20372612476348878,
      "num_tokens": 63612310.0,
      "step": 36660
    },
    {
      "entropy": 5.776137018203736,
      "epoch": 2.8526745769305584,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004185979978334604,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20261527895927428,
      "num_tokens": 63620736.0,
      "step": 36665
    },
    {
      "entropy": 5.7731468200683596,
      "epoch": 2.8530636063022756,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004185766596337727,
      "loss": 5.0193,
      "mean_token_accuracy": 0.19952816069126128,
      "num_tokens": 63629182.0,
      "step": 36670
    },
    {
      "entropy": 5.712402439117431,
      "epoch": 2.8534526356739933,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041855531925553773,
      "loss": 4.875,
      "mean_token_accuracy": 0.2137434408068657,
      "num_tokens": 63638197.0,
      "step": 36675
    },
    {
      "entropy": 5.7052281379699705,
      "epoch": 2.853841665045711,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004185339766990794,
      "loss": 5.0477,
      "mean_token_accuracy": 0.20401182025671005,
      "num_tokens": 63647461.0,
      "step": 36680
    },
    {
      "entropy": 5.689260625839234,
      "epoch": 2.8542306944174287,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004185126319647213,
      "loss": 4.9935,
      "mean_token_accuracy": 0.2039366379380226,
      "num_tokens": 63655402.0,
      "step": 36685
    },
    {
      "entropy": 5.74499626159668,
      "epoch": 2.854619723789146,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004184912850527875,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21402277499437333,
      "num_tokens": 63663650.0,
      "step": 36690
    },
    {
      "entropy": 5.73839430809021,
      "epoch": 2.8550087531608637,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004184699359636018,
      "loss": 4.9412,
      "mean_token_accuracy": 0.20440115928649902,
      "num_tokens": 63672403.0,
      "step": 36695
    },
    {
      "entropy": 5.697465991973877,
      "epoch": 2.8553977825325814,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004184485846974882,
      "loss": 4.879,
      "mean_token_accuracy": 0.21389620304107665,
      "num_tokens": 63681669.0,
      "step": 36700
    },
    {
      "entropy": 5.725019121170044,
      "epoch": 2.8557868119042986,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004184272312547706,
      "loss": 4.9337,
      "mean_token_accuracy": 0.20902230441570283,
      "num_tokens": 63690190.0,
      "step": 36705
    },
    {
      "entropy": 5.722491216659546,
      "epoch": 2.8561758412760163,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041840587563577315,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2070932000875473,
      "num_tokens": 63699389.0,
      "step": 36710
    },
    {
      "entropy": 5.758575105667115,
      "epoch": 2.856564870647734,
      "grad_norm": 1.25,
      "learning_rate": 0.0004183845178408197,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20925637781620027,
      "num_tokens": 63708314.0,
      "step": 36715
    },
    {
      "entropy": 5.669419670104981,
      "epoch": 2.8569539000194517,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041836315787023456,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2185925707221031,
      "num_tokens": 63716732.0,
      "step": 36720
    },
    {
      "entropy": 5.627235031127929,
      "epoch": 2.857342929391169,
      "grad_norm": 1.25,
      "learning_rate": 0.00041834179572434153,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2085905358195305,
      "num_tokens": 63725143.0,
      "step": 36725
    },
    {
      "entropy": 5.666985368728637,
      "epoch": 2.8577319587628867,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041832043140346495,
      "loss": 4.8924,
      "mean_token_accuracy": 0.21844515204429626,
      "num_tokens": 63732534.0,
      "step": 36730
    },
    {
      "entropy": 5.779353380203247,
      "epoch": 2.858120988134604,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004182990649079289,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20122582018375396,
      "num_tokens": 63740540.0,
      "step": 36735
    },
    {
      "entropy": 5.726066732406617,
      "epoch": 2.8585100175063216,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004182776962380577,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1952139765024185,
      "num_tokens": 63749364.0,
      "step": 36740
    },
    {
      "entropy": 5.734862899780273,
      "epoch": 2.8588990468780393,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041825632539417546,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20269803553819657,
      "num_tokens": 63758309.0,
      "step": 36745
    },
    {
      "entropy": 5.755665588378906,
      "epoch": 2.859288076249757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004182349523766065,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2120609849691391,
      "num_tokens": 63766882.0,
      "step": 36750
    },
    {
      "entropy": 5.683825206756592,
      "epoch": 2.8596771056214743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041821357718567514,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21510414630174637,
      "num_tokens": 63775108.0,
      "step": 36755
    },
    {
      "entropy": 5.6861590385437015,
      "epoch": 2.860066134993192,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004181921998217057,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20744120180606843,
      "num_tokens": 63783674.0,
      "step": 36760
    },
    {
      "entropy": 5.729680061340332,
      "epoch": 2.8604551643649097,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004181708202850225,
      "loss": 5.085,
      "mean_token_accuracy": 0.20357842296361922,
      "num_tokens": 63793068.0,
      "step": 36765
    },
    {
      "entropy": 5.675493001937866,
      "epoch": 2.860844193736627,
      "grad_norm": 1.25,
      "learning_rate": 0.00041814943857595,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21113720387220383,
      "num_tokens": 63801845.0,
      "step": 36770
    },
    {
      "entropy": 5.7200864315032955,
      "epoch": 2.8612332231083446,
      "grad_norm": 1.25,
      "learning_rate": 0.0004181280546948127,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20181134343147278,
      "num_tokens": 63810107.0,
      "step": 36775
    },
    {
      "entropy": 5.670366382598877,
      "epoch": 2.8616222524800623,
      "grad_norm": 1.25,
      "learning_rate": 0.0004181066686419349,
      "loss": 4.8998,
      "mean_token_accuracy": 0.20974615812301636,
      "num_tokens": 63818420.0,
      "step": 36780
    },
    {
      "entropy": 5.713176679611206,
      "epoch": 2.86201128185178,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041808528041764115,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2074020028114319,
      "num_tokens": 63826947.0,
      "step": 36785
    },
    {
      "entropy": 5.67804799079895,
      "epoch": 2.8624003112234973,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004180638900222561,
      "loss": 4.934,
      "mean_token_accuracy": 0.21596689373254777,
      "num_tokens": 63835483.0,
      "step": 36790
    },
    {
      "entropy": 5.639019918441773,
      "epoch": 2.862789340595215,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004180424974561042,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2096621200442314,
      "num_tokens": 63843983.0,
      "step": 36795
    },
    {
      "entropy": 5.73206901550293,
      "epoch": 2.8631783699669326,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041802110271951,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21251205056905748,
      "num_tokens": 63852478.0,
      "step": 36800
    },
    {
      "entropy": 5.758202600479126,
      "epoch": 2.86356739933865,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004179997058127983,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20636558532714844,
      "num_tokens": 63860684.0,
      "step": 36805
    },
    {
      "entropy": 5.729201030731201,
      "epoch": 2.8639564287103676,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004179783067362936,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2049630403518677,
      "num_tokens": 63869033.0,
      "step": 36810
    },
    {
      "entropy": 5.751136302947998,
      "epoch": 2.8643454580820853,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004179569054903207,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2036151483654976,
      "num_tokens": 63877645.0,
      "step": 36815
    },
    {
      "entropy": 5.741723918914795,
      "epoch": 2.864734487453803,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041793550207520436,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21337577253580092,
      "num_tokens": 63886168.0,
      "step": 36820
    },
    {
      "entropy": 5.705454874038696,
      "epoch": 2.8651235168255202,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004179140964912692,
      "loss": 4.9876,
      "mean_token_accuracy": 0.20757579058408737,
      "num_tokens": 63895110.0,
      "step": 36825
    },
    {
      "entropy": 5.746557235717773,
      "epoch": 2.865512546197238,
      "grad_norm": 1.25,
      "learning_rate": 0.0004178926887388401,
      "loss": 4.9314,
      "mean_token_accuracy": 0.20682869404554366,
      "num_tokens": 63903002.0,
      "step": 36830
    },
    {
      "entropy": 5.698002052307129,
      "epoch": 2.865901575568955,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041787127881824193,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2142242893576622,
      "num_tokens": 63911011.0,
      "step": 36835
    },
    {
      "entropy": 5.70913987159729,
      "epoch": 2.866290604940673,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041784986672979945,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20738430172204972,
      "num_tokens": 63919667.0,
      "step": 36840
    },
    {
      "entropy": 5.733752107620239,
      "epoch": 2.8666796343123906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041782845247383767,
      "loss": 4.959,
      "mean_token_accuracy": 0.2086784988641739,
      "num_tokens": 63929445.0,
      "step": 36845
    },
    {
      "entropy": 5.738183450698853,
      "epoch": 2.8670686636841083,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041780703605068145,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2168579801917076,
      "num_tokens": 63937730.0,
      "step": 36850
    },
    {
      "entropy": 5.706082773208618,
      "epoch": 2.867457693055826,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004177856174606558,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21206885129213332,
      "num_tokens": 63945877.0,
      "step": 36855
    },
    {
      "entropy": 5.743387174606323,
      "epoch": 2.8678467224275432,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004177641967040856,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2078859806060791,
      "num_tokens": 63953861.0,
      "step": 36860
    },
    {
      "entropy": 5.726610088348389,
      "epoch": 2.868235751799261,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041774277378129595,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2078307017683983,
      "num_tokens": 63962622.0,
      "step": 36865
    },
    {
      "entropy": 5.778795194625855,
      "epoch": 2.868624781170978,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041772134869261186,
      "loss": 5.0226,
      "mean_token_accuracy": 0.2049886092543602,
      "num_tokens": 63971100.0,
      "step": 36870
    },
    {
      "entropy": 5.644616508483887,
      "epoch": 2.869013810542696,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004176999214383585,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2086598828434944,
      "num_tokens": 63979763.0,
      "step": 36875
    },
    {
      "entropy": 5.682045221328735,
      "epoch": 2.8694028399144136,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041767849201886103,
      "loss": 4.9199,
      "mean_token_accuracy": 0.21428375393152238,
      "num_tokens": 63988383.0,
      "step": 36880
    },
    {
      "entropy": 5.6884088039398195,
      "epoch": 2.8697918692861313,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004176570604344445,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20352015793323516,
      "num_tokens": 63996578.0,
      "step": 36885
    },
    {
      "entropy": 5.671725559234619,
      "epoch": 2.8701808986578485,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041763562668543415,
      "loss": 4.8808,
      "mean_token_accuracy": 0.21336551904678344,
      "num_tokens": 64004755.0,
      "step": 36890
    },
    {
      "entropy": 5.677460145950318,
      "epoch": 2.8705699280295662,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004176141907721552,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20429760813713074,
      "num_tokens": 64013937.0,
      "step": 36895
    },
    {
      "entropy": 5.802621793746948,
      "epoch": 2.870958957401284,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004175927526949329,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20770859569311143,
      "num_tokens": 64022464.0,
      "step": 36900
    },
    {
      "entropy": 5.817018508911133,
      "epoch": 2.871347986773001,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004175713124540925,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20416144877672196,
      "num_tokens": 64031023.0,
      "step": 36905
    },
    {
      "entropy": 5.814526224136353,
      "epoch": 2.871737016144719,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041754987004995935,
      "loss": 5.0224,
      "mean_token_accuracy": 0.20754458457231523,
      "num_tokens": 64039740.0,
      "step": 36910
    },
    {
      "entropy": 5.7939453125,
      "epoch": 2.8721260455164366,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041752842548285887,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20747753381729125,
      "num_tokens": 64048638.0,
      "step": 36915
    },
    {
      "entropy": 5.7803596496582035,
      "epoch": 2.8725150748881543,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004175069787531163,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20558423846960067,
      "num_tokens": 64058068.0,
      "step": 36920
    },
    {
      "entropy": 5.723847818374634,
      "epoch": 2.8729041042598715,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041748552986105726,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2158821403980255,
      "num_tokens": 64066881.0,
      "step": 36925
    },
    {
      "entropy": 5.829465436935425,
      "epoch": 2.8732931336315892,
      "grad_norm": 1.28125,
      "learning_rate": 0.000417464078807007,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19796052724123,
      "num_tokens": 64076020.0,
      "step": 36930
    },
    {
      "entropy": 5.722281551361084,
      "epoch": 2.8736821630033065,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004174426255912912,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21289380490779877,
      "num_tokens": 64083940.0,
      "step": 36935
    },
    {
      "entropy": 5.731930494308472,
      "epoch": 2.874071192375024,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004174211702142352,
      "loss": 5.0137,
      "mean_token_accuracy": 0.208070370554924,
      "num_tokens": 64091385.0,
      "step": 36940
    },
    {
      "entropy": 5.665528917312622,
      "epoch": 2.874460221746742,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041739971267616466,
      "loss": 4.9295,
      "mean_token_accuracy": 0.20756968706846238,
      "num_tokens": 64099797.0,
      "step": 36945
    },
    {
      "entropy": 5.797987031936645,
      "epoch": 2.8748492511184596,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004173782529774051,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20333696603775026,
      "num_tokens": 64107790.0,
      "step": 36950
    },
    {
      "entropy": 5.754706525802613,
      "epoch": 2.8752382804901773,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041735679111828223,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21378969252109528,
      "num_tokens": 64115090.0,
      "step": 36955
    },
    {
      "entropy": 5.654418325424194,
      "epoch": 2.8756273098618945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041733532709912157,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2107856512069702,
      "num_tokens": 64124039.0,
      "step": 36960
    },
    {
      "entropy": 5.71700758934021,
      "epoch": 2.876016339233612,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041731386092024893,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20919185280799865,
      "num_tokens": 64132343.0,
      "step": 36965
    },
    {
      "entropy": 5.7032561779022215,
      "epoch": 2.8764053686053295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041729239258198996,
      "loss": 4.8725,
      "mean_token_accuracy": 0.21915805339813232,
      "num_tokens": 64140742.0,
      "step": 36970
    },
    {
      "entropy": 5.771155309677124,
      "epoch": 2.876794397977047,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041727092208467036,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21050880402326583,
      "num_tokens": 64148822.0,
      "step": 36975
    },
    {
      "entropy": 5.705517673492432,
      "epoch": 2.877183427348765,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041724944942861603,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21073877364397048,
      "num_tokens": 64157724.0,
      "step": 36980
    },
    {
      "entropy": 5.708099842071533,
      "epoch": 2.8775724567204826,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041722797461415267,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20981355905532836,
      "num_tokens": 64165895.0,
      "step": 36985
    },
    {
      "entropy": 5.744191312789917,
      "epoch": 2.8779614860922,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004172064976416062,
      "loss": 4.9189,
      "mean_token_accuracy": 0.2090353786945343,
      "num_tokens": 64174015.0,
      "step": 36990
    },
    {
      "entropy": 5.660231018066407,
      "epoch": 2.8783505154639175,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004171850185113024,
      "loss": 4.8677,
      "mean_token_accuracy": 0.21418857872486113,
      "num_tokens": 64182560.0,
      "step": 36995
    },
    {
      "entropy": 5.73729100227356,
      "epoch": 2.878739544835635,
      "grad_norm": 1.25,
      "learning_rate": 0.00041716353722356735,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20412824898958207,
      "num_tokens": 64190999.0,
      "step": 37000
    },
    {
      "entropy": 5.7428975105285645,
      "epoch": 2.8791285742073525,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004171420537787269,
      "loss": 5.0051,
      "mean_token_accuracy": 0.19977980405092238,
      "num_tokens": 64200073.0,
      "step": 37005
    },
    {
      "entropy": 5.804224061965942,
      "epoch": 2.87951760357907,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041712056817710697,
      "loss": 5.0151,
      "mean_token_accuracy": 0.21008194386959075,
      "num_tokens": 64208079.0,
      "step": 37010
    },
    {
      "entropy": 5.725548791885376,
      "epoch": 2.879906632950788,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004170990804190337,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20500696897506715,
      "num_tokens": 64216321.0,
      "step": 37015
    },
    {
      "entropy": 5.817561388015747,
      "epoch": 2.8802956623225056,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041707759050483304,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20466277450323106,
      "num_tokens": 64224804.0,
      "step": 37020
    },
    {
      "entropy": 5.801752805709839,
      "epoch": 2.880684691694223,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000417056098434831,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20028270334005355,
      "num_tokens": 64234022.0,
      "step": 37025
    },
    {
      "entropy": 5.742588806152344,
      "epoch": 2.8810737210659405,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004170346042093538,
      "loss": 4.9361,
      "mean_token_accuracy": 0.20720379054546356,
      "num_tokens": 64241913.0,
      "step": 37030
    },
    {
      "entropy": 5.765771102905274,
      "epoch": 2.881462750437658,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004170131078287276,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21175179183483123,
      "num_tokens": 64249494.0,
      "step": 37035
    },
    {
      "entropy": 5.709988403320312,
      "epoch": 2.8818517798093755,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004169916092932785,
      "loss": 4.9926,
      "mean_token_accuracy": 0.20719997137784957,
      "num_tokens": 64257540.0,
      "step": 37040
    },
    {
      "entropy": 5.702971124649048,
      "epoch": 2.882240809181093,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004169701086033327,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2179891437292099,
      "num_tokens": 64266583.0,
      "step": 37045
    },
    {
      "entropy": 5.710389232635498,
      "epoch": 2.882629838552811,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041694860575921654,
      "loss": 4.946,
      "mean_token_accuracy": 0.2072391539812088,
      "num_tokens": 64275313.0,
      "step": 37050
    },
    {
      "entropy": 5.80691237449646,
      "epoch": 2.8830188679245285,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041692710076125615,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20386543571949006,
      "num_tokens": 64284160.0,
      "step": 37055
    },
    {
      "entropy": 5.688087701797485,
      "epoch": 2.883407897296246,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000416905593609778,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2115530326962471,
      "num_tokens": 64293010.0,
      "step": 37060
    },
    {
      "entropy": 5.726702356338501,
      "epoch": 2.8837969266679635,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041688408430510827,
      "loss": 5.019,
      "mean_token_accuracy": 0.19847943186759948,
      "num_tokens": 64301728.0,
      "step": 37065
    },
    {
      "entropy": 5.748627710342407,
      "epoch": 2.8841859560396808,
      "grad_norm": 1.25,
      "learning_rate": 0.0004168625728475734,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21035687923431395,
      "num_tokens": 64310056.0,
      "step": 37070
    },
    {
      "entropy": 5.746639060974121,
      "epoch": 2.8845749854113985,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041684105923749977,
      "loss": 4.93,
      "mean_token_accuracy": 0.21742263287305832,
      "num_tokens": 64320016.0,
      "step": 37075
    },
    {
      "entropy": 5.71370677947998,
      "epoch": 2.884964014783116,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004168195434752139,
      "loss": 4.886,
      "mean_token_accuracy": 0.20994980335235597,
      "num_tokens": 64327763.0,
      "step": 37080
    },
    {
      "entropy": 5.686677932739258,
      "epoch": 2.885353044154834,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041679802556104207,
      "loss": 4.8897,
      "mean_token_accuracy": 0.21090674996376038,
      "num_tokens": 64336190.0,
      "step": 37085
    },
    {
      "entropy": 5.672700071334839,
      "epoch": 2.885742073526551,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041677650549531097,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2125129446387291,
      "num_tokens": 64345329.0,
      "step": 37090
    },
    {
      "entropy": 5.75876088142395,
      "epoch": 2.886131102898269,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041675498327834706,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20023034065961837,
      "num_tokens": 64353272.0,
      "step": 37095
    },
    {
      "entropy": 5.760937929153442,
      "epoch": 2.8865201322699865,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041673345891047686,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19787791818380357,
      "num_tokens": 64361866.0,
      "step": 37100
    },
    {
      "entropy": 5.771781349182129,
      "epoch": 2.8869091616417037,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004167119323920271,
      "loss": 5.0706,
      "mean_token_accuracy": 0.19937632232904434,
      "num_tokens": 64371294.0,
      "step": 37105
    },
    {
      "entropy": 5.7344911098480225,
      "epoch": 2.8872981910134214,
      "grad_norm": 1.25,
      "learning_rate": 0.0004166904037233243,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21333439350128175,
      "num_tokens": 64379166.0,
      "step": 37110
    },
    {
      "entropy": 5.711657857894897,
      "epoch": 2.887687220385139,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041666887290469506,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20718881934881211,
      "num_tokens": 64389092.0,
      "step": 37115
    },
    {
      "entropy": 5.703644847869873,
      "epoch": 2.888076249756857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004166473399364663,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22060548812150954,
      "num_tokens": 64397423.0,
      "step": 37120
    },
    {
      "entropy": 5.764362335205078,
      "epoch": 2.888465279128574,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041662580481896446,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20971526801586152,
      "num_tokens": 64406291.0,
      "step": 37125
    },
    {
      "entropy": 5.703642320632935,
      "epoch": 2.888854308500292,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004166042675525165,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20622216314077377,
      "num_tokens": 64414737.0,
      "step": 37130
    },
    {
      "entropy": 5.721449518203736,
      "epoch": 2.8892433378720095,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041658272813744924,
      "loss": 4.9739,
      "mean_token_accuracy": 0.20735889077186584,
      "num_tokens": 64423737.0,
      "step": 37135
    },
    {
      "entropy": 5.786657905578613,
      "epoch": 2.8896323672437267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004165611865740894,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2021647334098816,
      "num_tokens": 64432295.0,
      "step": 37140
    },
    {
      "entropy": 5.851624059677124,
      "epoch": 2.8900213966154444,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004165396428627638,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20253732353448867,
      "num_tokens": 64441090.0,
      "step": 37145
    },
    {
      "entropy": 5.813965129852295,
      "epoch": 2.890410425987162,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041651809700379947,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2010728284716606,
      "num_tokens": 64450186.0,
      "step": 37150
    },
    {
      "entropy": 5.71786904335022,
      "epoch": 2.89079945535888,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041649654899752326,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21107727140188218,
      "num_tokens": 64458519.0,
      "step": 37155
    },
    {
      "entropy": 5.656841897964478,
      "epoch": 2.891188484730597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004164749988442621,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2085729017853737,
      "num_tokens": 64467941.0,
      "step": 37160
    },
    {
      "entropy": 5.7249034404754635,
      "epoch": 2.891577514102315,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000416453446544343,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2036920368671417,
      "num_tokens": 64476811.0,
      "step": 37165
    },
    {
      "entropy": 5.759035205841064,
      "epoch": 2.891966543474032,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004164318920980931,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21390917748212815,
      "num_tokens": 64485905.0,
      "step": 37170
    },
    {
      "entropy": 5.7777036190032955,
      "epoch": 2.8923555728457497,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004164103355058393,
      "loss": 5.0998,
      "mean_token_accuracy": 0.19965046942234038,
      "num_tokens": 64495157.0,
      "step": 37175
    },
    {
      "entropy": 5.721639823913574,
      "epoch": 2.8927446022174674,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004163887767679087,
      "loss": 4.912,
      "mean_token_accuracy": 0.21985400915145875,
      "num_tokens": 64503598.0,
      "step": 37180
    },
    {
      "entropy": 5.772041988372803,
      "epoch": 2.893133631589185,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004163672158846285,
      "loss": 4.9398,
      "mean_token_accuracy": 0.20617058277130126,
      "num_tokens": 64511753.0,
      "step": 37185
    },
    {
      "entropy": 5.784866380691528,
      "epoch": 2.8935226609609024,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041634565285632573,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20023715645074844,
      "num_tokens": 64521011.0,
      "step": 37190
    },
    {
      "entropy": 5.756584596633911,
      "epoch": 2.89391169033262,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041632408768332776,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2042479172348976,
      "num_tokens": 64529716.0,
      "step": 37195
    },
    {
      "entropy": 5.755605554580688,
      "epoch": 2.8943007197043378,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041630252036596165,
      "loss": 4.9946,
      "mean_token_accuracy": 0.20901148468255998,
      "num_tokens": 64538673.0,
      "step": 37200
    },
    {
      "entropy": 5.694317817687988,
      "epoch": 2.894689749076055,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004162809509045547,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2105732575058937,
      "num_tokens": 64546399.0,
      "step": 37205
    },
    {
      "entropy": 5.749301195144653,
      "epoch": 2.8950787784477727,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041625937929943424,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20357396006584166,
      "num_tokens": 64554412.0,
      "step": 37210
    },
    {
      "entropy": 5.7358551025390625,
      "epoch": 2.8954678078194904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041623780555092747,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2126569226384163,
      "num_tokens": 64562630.0,
      "step": 37215
    },
    {
      "entropy": 5.7939635753631595,
      "epoch": 2.895856837191208,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004162162296593618,
      "loss": 5.1112,
      "mean_token_accuracy": 0.1941658228635788,
      "num_tokens": 64572492.0,
      "step": 37220
    },
    {
      "entropy": 5.73642897605896,
      "epoch": 2.8962458665629254,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041619465162506466,
      "loss": 4.9693,
      "mean_token_accuracy": 0.211011666059494,
      "num_tokens": 64580964.0,
      "step": 37225
    },
    {
      "entropy": 5.7402819156646725,
      "epoch": 2.896634895934643,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004161730714483633,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2097105413675308,
      "num_tokens": 64589513.0,
      "step": 37230
    },
    {
      "entropy": 5.718275165557861,
      "epoch": 2.8970239253063608,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004161514891295854,
      "loss": 4.9943,
      "mean_token_accuracy": 0.20621182918548583,
      "num_tokens": 64597462.0,
      "step": 37235
    },
    {
      "entropy": 5.7098325252532955,
      "epoch": 2.897412954678078,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041612990466905825,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2118037074804306,
      "num_tokens": 64604908.0,
      "step": 37240
    },
    {
      "entropy": 5.711767482757568,
      "epoch": 2.8978019840497957,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041610831806710954,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21341390311717987,
      "num_tokens": 64613983.0,
      "step": 37245
    },
    {
      "entropy": 5.668642330169678,
      "epoch": 2.8981910134215134,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041608672932406666,
      "loss": 4.883,
      "mean_token_accuracy": 0.21553361117839814,
      "num_tokens": 64622620.0,
      "step": 37250
    },
    {
      "entropy": 5.756299734115601,
      "epoch": 2.898580042793231,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041606513844025716,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2071797162294388,
      "num_tokens": 64631204.0,
      "step": 37255
    },
    {
      "entropy": 5.72422423362732,
      "epoch": 2.8989690721649484,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004160435454160087,
      "loss": 4.9451,
      "mean_token_accuracy": 0.20564515590667726,
      "num_tokens": 64639904.0,
      "step": 37260
    },
    {
      "entropy": 5.768924236297607,
      "epoch": 2.899358101536666,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000416021950251649,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20210411846637727,
      "num_tokens": 64649011.0,
      "step": 37265
    },
    {
      "entropy": 5.719493865966797,
      "epoch": 2.8997471309083833,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004160003529475057,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20151309072971343,
      "num_tokens": 64658174.0,
      "step": 37270
    },
    {
      "entropy": 5.722173118591309,
      "epoch": 2.900136160280101,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004159787535039064,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20730003118515014,
      "num_tokens": 64667301.0,
      "step": 37275
    },
    {
      "entropy": 5.793070363998413,
      "epoch": 2.9005251896518187,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004159571519211789,
      "loss": 5.057,
      "mean_token_accuracy": 0.20257613360881804,
      "num_tokens": 64676087.0,
      "step": 37280
    },
    {
      "entropy": 5.77973575592041,
      "epoch": 2.9009142190235364,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000415935548199651,
      "loss": 5.098,
      "mean_token_accuracy": 0.20026143491268159,
      "num_tokens": 64684990.0,
      "step": 37285
    },
    {
      "entropy": 5.889983367919922,
      "epoch": 2.901303248395254,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004159139423396504,
      "loss": 5.1194,
      "mean_token_accuracy": 0.1953114628791809,
      "num_tokens": 64693291.0,
      "step": 37290
    },
    {
      "entropy": 5.810815382003784,
      "epoch": 2.9016922777669714,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004158923343415051,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2033628612756729,
      "num_tokens": 64701922.0,
      "step": 37295
    },
    {
      "entropy": 5.744422340393067,
      "epoch": 2.902081307138689,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004158707242055429,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20656122863292695,
      "num_tokens": 64710782.0,
      "step": 37300
    },
    {
      "entropy": 5.817006635665893,
      "epoch": 2.9024703365104063,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041584911193209167,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2097572147846222,
      "num_tokens": 64719285.0,
      "step": 37305
    },
    {
      "entropy": 5.752814102172851,
      "epoch": 2.902859365882124,
      "grad_norm": 1.25,
      "learning_rate": 0.0004158274975214794,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2083319365978241,
      "num_tokens": 64727683.0,
      "step": 37310
    },
    {
      "entropy": 5.710377359390259,
      "epoch": 2.9032483952538417,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041580588097403384,
      "loss": 4.9469,
      "mean_token_accuracy": 0.20366083830595016,
      "num_tokens": 64736812.0,
      "step": 37315
    },
    {
      "entropy": 5.746654224395752,
      "epoch": 2.9036374246255594,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041578426229008325,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20517215132713318,
      "num_tokens": 64745858.0,
      "step": 37320
    },
    {
      "entropy": 5.787350845336914,
      "epoch": 2.9040264539972767,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041576264146995557,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19533346146345137,
      "num_tokens": 64754851.0,
      "step": 37325
    },
    {
      "entropy": 5.764447212219238,
      "epoch": 2.9044154833689944,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041574101851397883,
      "loss": 5.058,
      "mean_token_accuracy": 0.19854519963264466,
      "num_tokens": 64764028.0,
      "step": 37330
    },
    {
      "entropy": 5.718879508972168,
      "epoch": 2.904804512740712,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004157193934224811,
      "loss": 4.8766,
      "mean_token_accuracy": 0.2128916084766388,
      "num_tokens": 64773269.0,
      "step": 37335
    },
    {
      "entropy": 5.736349058151245,
      "epoch": 2.9051935421124293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004156977661957906,
      "loss": 4.943,
      "mean_token_accuracy": 0.2111997589468956,
      "num_tokens": 64781867.0,
      "step": 37340
    },
    {
      "entropy": 5.732470941543579,
      "epoch": 2.905582571484147,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004156761368342355,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2104508176445961,
      "num_tokens": 64791109.0,
      "step": 37345
    },
    {
      "entropy": 5.717786979675293,
      "epoch": 2.9059716008558647,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041565450533814384,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22144932746887208,
      "num_tokens": 64799122.0,
      "step": 37350
    },
    {
      "entropy": 5.725673198699951,
      "epoch": 2.9063606302275824,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041563287170784396,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2024384170770645,
      "num_tokens": 64806936.0,
      "step": 37355
    },
    {
      "entropy": 5.69240026473999,
      "epoch": 2.9067496595992997,
      "grad_norm": 1.25,
      "learning_rate": 0.0004156112359436641,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20318218916654587,
      "num_tokens": 64815633.0,
      "step": 37360
    },
    {
      "entropy": 5.699302911758423,
      "epoch": 2.9071386889710173,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041558959804593255,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21420224606990815,
      "num_tokens": 64824064.0,
      "step": 37365
    },
    {
      "entropy": 5.679358196258545,
      "epoch": 2.9075277183427346,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041556795801497766,
      "loss": 4.9449,
      "mean_token_accuracy": 0.210280342400074,
      "num_tokens": 64831970.0,
      "step": 37370
    },
    {
      "entropy": 5.693839168548584,
      "epoch": 2.9079167477144523,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004155463158511277,
      "loss": 4.951,
      "mean_token_accuracy": 0.20558794289827348,
      "num_tokens": 64840524.0,
      "step": 37375
    },
    {
      "entropy": 5.810556697845459,
      "epoch": 2.90830577708617,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041552467155471105,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19923983365297318,
      "num_tokens": 64849039.0,
      "step": 37380
    },
    {
      "entropy": 5.783484172821045,
      "epoch": 2.9086948064578877,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004155030251260563,
      "loss": 4.987,
      "mean_token_accuracy": 0.20251857340335847,
      "num_tokens": 64858172.0,
      "step": 37385
    },
    {
      "entropy": 5.711177396774292,
      "epoch": 2.9090838358296054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041548137656549167,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21165110170841217,
      "num_tokens": 64866108.0,
      "step": 37390
    },
    {
      "entropy": 5.763345050811767,
      "epoch": 2.9094728652013226,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004154597258733457,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2029396876692772,
      "num_tokens": 64874451.0,
      "step": 37395
    },
    {
      "entropy": 5.7915647506713865,
      "epoch": 2.9098618945730403,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004154380730499472,
      "loss": 5.0079,
      "mean_token_accuracy": 0.209265798330307,
      "num_tokens": 64882354.0,
      "step": 37400
    },
    {
      "entropy": 5.760558986663819,
      "epoch": 2.9102509239447576,
      "grad_norm": 1.25,
      "learning_rate": 0.00041541641809562425,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2126854658126831,
      "num_tokens": 64890716.0,
      "step": 37405
    },
    {
      "entropy": 5.628662061691284,
      "epoch": 2.9106399533164753,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004153947610107057,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2148878365755081,
      "num_tokens": 64899629.0,
      "step": 37410
    },
    {
      "entropy": 5.756655597686768,
      "epoch": 2.911028982688193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041537310179552017,
      "loss": 4.9762,
      "mean_token_accuracy": 0.1973027065396309,
      "num_tokens": 64908539.0,
      "step": 37415
    },
    {
      "entropy": 5.7430816173553465,
      "epoch": 2.9114180120599107,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004153514404503962,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2137985572218895,
      "num_tokens": 64917751.0,
      "step": 37420
    },
    {
      "entropy": 5.717655897140503,
      "epoch": 2.911807041431628,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004153297769756625,
      "loss": 5.0085,
      "mean_token_accuracy": 0.2057817056775093,
      "num_tokens": 64925591.0,
      "step": 37425
    },
    {
      "entropy": 5.765898847579956,
      "epoch": 2.9121960708033456,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004153081113716478,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20899815559387208,
      "num_tokens": 64934512.0,
      "step": 37430
    },
    {
      "entropy": 5.711842012405396,
      "epoch": 2.9125851001750633,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041528644363868087,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21274943351745607,
      "num_tokens": 64942682.0,
      "step": 37435
    },
    {
      "entropy": 5.7020327091217045,
      "epoch": 2.9129741295467806,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004152647737770904,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20841173678636551,
      "num_tokens": 64950961.0,
      "step": 37440
    },
    {
      "entropy": 5.737736415863037,
      "epoch": 2.9133631589184983,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004152431017872053,
      "loss": 5.0088,
      "mean_token_accuracy": 0.20696523189544677,
      "num_tokens": 64959392.0,
      "step": 37445
    },
    {
      "entropy": 5.736891603469848,
      "epoch": 2.913752188290216,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041522142766935426,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21029538065195083,
      "num_tokens": 64967786.0,
      "step": 37450
    },
    {
      "entropy": 5.792594385147095,
      "epoch": 2.9141412176619337,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041519975142386627,
      "loss": 4.9659,
      "mean_token_accuracy": 0.2075514629483223,
      "num_tokens": 64975987.0,
      "step": 37455
    },
    {
      "entropy": 5.722790670394898,
      "epoch": 2.914530247033651,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004151780730510702,
      "loss": 4.9334,
      "mean_token_accuracy": 0.2087882548570633,
      "num_tokens": 64984134.0,
      "step": 37460
    },
    {
      "entropy": 5.696947526931763,
      "epoch": 2.9149192764053686,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041515639255129504,
      "loss": 4.9208,
      "mean_token_accuracy": 0.20955941826105118,
      "num_tokens": 64992346.0,
      "step": 37465
    },
    {
      "entropy": 5.765525770187378,
      "epoch": 2.9153083057770863,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004151347099248697,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21498361229896545,
      "num_tokens": 65001029.0,
      "step": 37470
    },
    {
      "entropy": 5.761052417755127,
      "epoch": 2.9156973351488036,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004151130251721231,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20264137983322145,
      "num_tokens": 65008932.0,
      "step": 37475
    },
    {
      "entropy": 5.7972806930542,
      "epoch": 2.9160863645205213,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041509133829338454,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21167746484279631,
      "num_tokens": 65017774.0,
      "step": 37480
    },
    {
      "entropy": 5.804692983627319,
      "epoch": 2.916475393892239,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004150696492889827,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20427973866462706,
      "num_tokens": 65026641.0,
      "step": 37485
    },
    {
      "entropy": 5.742819976806641,
      "epoch": 2.9168644232639567,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041504795815924697,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20865918397903443,
      "num_tokens": 65035123.0,
      "step": 37490
    },
    {
      "entropy": 5.7964515209198,
      "epoch": 2.917253452635674,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041502626490450635,
      "loss": 5.0434,
      "mean_token_accuracy": 0.2015484243631363,
      "num_tokens": 65043553.0,
      "step": 37495
    },
    {
      "entropy": 5.746234893798828,
      "epoch": 2.9176424820073916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004150045695250901,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20630771964788436,
      "num_tokens": 65052267.0,
      "step": 37500
    },
    {
      "entropy": 5.7275035858154295,
      "epoch": 2.918031511379109,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004149828720213272,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21291425228118896,
      "num_tokens": 65060667.0,
      "step": 37505
    },
    {
      "entropy": 5.670481014251709,
      "epoch": 2.9184205407508266,
      "grad_norm": 1.25,
      "learning_rate": 0.0004149611723935472,
      "loss": 4.8784,
      "mean_token_accuracy": 0.21938779354095458,
      "num_tokens": 65069194.0,
      "step": 37510
    },
    {
      "entropy": 5.746001195907593,
      "epoch": 2.9188095701225443,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004149394706420791,
      "loss": 4.9647,
      "mean_token_accuracy": 0.20571626126766204,
      "num_tokens": 65077873.0,
      "step": 37515
    },
    {
      "entropy": 5.721691083908081,
      "epoch": 2.919198599494262,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041491776676725223,
      "loss": 4.964,
      "mean_token_accuracy": 0.2102587938308716,
      "num_tokens": 65087503.0,
      "step": 37520
    },
    {
      "entropy": 5.755048942565918,
      "epoch": 2.9195876288659792,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004148960607693961,
      "loss": 5.0249,
      "mean_token_accuracy": 0.19768452644348145,
      "num_tokens": 65096520.0,
      "step": 37525
    },
    {
      "entropy": 5.7073980331420895,
      "epoch": 2.919976658237697,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041487435264883974,
      "loss": 4.9281,
      "mean_token_accuracy": 0.21284224689006806,
      "num_tokens": 65105323.0,
      "step": 37530
    },
    {
      "entropy": 5.844447946548462,
      "epoch": 2.9203656876094146,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004148526424059128,
      "loss": 5.1139,
      "mean_token_accuracy": 0.19298322200775148,
      "num_tokens": 65114412.0,
      "step": 37535
    },
    {
      "entropy": 5.7610949039459225,
      "epoch": 2.920754716981132,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004148309300409446,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2147040292620659,
      "num_tokens": 65122285.0,
      "step": 37540
    },
    {
      "entropy": 5.705696487426758,
      "epoch": 2.9211437463528496,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004148092155542646,
      "loss": 4.9997,
      "mean_token_accuracy": 0.20700423866510392,
      "num_tokens": 65131812.0,
      "step": 37545
    },
    {
      "entropy": 5.756667184829712,
      "epoch": 2.9215327757245673,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004147874989462023,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21198780089616776,
      "num_tokens": 65140187.0,
      "step": 37550
    },
    {
      "entropy": 5.749834871292114,
      "epoch": 2.921921805096285,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004147657802170871,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2109106034040451,
      "num_tokens": 65149510.0,
      "step": 37555
    },
    {
      "entropy": 5.7667945384979244,
      "epoch": 2.922310834468002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041474405936724873,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2069396808743477,
      "num_tokens": 65157889.0,
      "step": 37560
    },
    {
      "entropy": 5.726014423370361,
      "epoch": 2.92269986383972,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004147223363970166,
      "loss": 4.9,
      "mean_token_accuracy": 0.21077440232038497,
      "num_tokens": 65166233.0,
      "step": 37565
    },
    {
      "entropy": 5.804691314697266,
      "epoch": 2.9230888932114376,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004147006113067205,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2060772731900215,
      "num_tokens": 65174093.0,
      "step": 37570
    },
    {
      "entropy": 5.805622673034668,
      "epoch": 2.923477922583155,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004146788840966901,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21100467145442964,
      "num_tokens": 65183059.0,
      "step": 37575
    },
    {
      "entropy": 5.7412818431854244,
      "epoch": 2.9238669519548726,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004146571547672548,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21339647471904755,
      "num_tokens": 65191412.0,
      "step": 37580
    },
    {
      "entropy": 5.733124160766602,
      "epoch": 2.9242559813265903,
      "grad_norm": 1.375,
      "learning_rate": 0.00041463542331874443,
      "loss": 5.0111,
      "mean_token_accuracy": 0.2081962689757347,
      "num_tokens": 65200401.0,
      "step": 37585
    },
    {
      "entropy": 5.710060405731201,
      "epoch": 2.924645010698308,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041461368975148876,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20566997528076172,
      "num_tokens": 65208987.0,
      "step": 37590
    },
    {
      "entropy": 5.767377090454102,
      "epoch": 2.925034040070025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041459195406581763,
      "loss": 4.961,
      "mean_token_accuracy": 0.21061062961816787,
      "num_tokens": 65217449.0,
      "step": 37595
    },
    {
      "entropy": 5.761460828781128,
      "epoch": 2.925423069441743,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004145702162620608,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2089443564414978,
      "num_tokens": 65225922.0,
      "step": 37600
    },
    {
      "entropy": 5.706326341629028,
      "epoch": 2.92581209881346,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041454847634054805,
      "loss": 4.9398,
      "mean_token_accuracy": 0.209811007976532,
      "num_tokens": 65233777.0,
      "step": 37605
    },
    {
      "entropy": 5.693035745620728,
      "epoch": 2.926201128185178,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041452673430160923,
      "loss": 4.9396,
      "mean_token_accuracy": 0.20394111424684525,
      "num_tokens": 65242140.0,
      "step": 37610
    },
    {
      "entropy": 5.783765268325806,
      "epoch": 2.9265901575568956,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004145049901455744,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20083435177803038,
      "num_tokens": 65251166.0,
      "step": 37615
    },
    {
      "entropy": 5.7911255836486815,
      "epoch": 2.9269791869286133,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041448324387277337,
      "loss": 4.9126,
      "mean_token_accuracy": 0.20130646973848343,
      "num_tokens": 65260099.0,
      "step": 37620
    },
    {
      "entropy": 5.790359830856323,
      "epoch": 2.927368216300331,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041446149548353606,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20328816622495652,
      "num_tokens": 65270450.0,
      "step": 37625
    },
    {
      "entropy": 5.735617208480835,
      "epoch": 2.927757245672048,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004144397449781926,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20121398121118544,
      "num_tokens": 65279298.0,
      "step": 37630
    },
    {
      "entropy": 5.736426067352295,
      "epoch": 2.928146275043766,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004144179923570729,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2144026577472687,
      "num_tokens": 65287015.0,
      "step": 37635
    },
    {
      "entropy": 5.7527649879455565,
      "epoch": 2.928535304415483,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004143962376205071,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21055516004562377,
      "num_tokens": 65295300.0,
      "step": 37640
    },
    {
      "entropy": 5.746245956420898,
      "epoch": 2.928924333787201,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041437448076882526,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21496892720460892,
      "num_tokens": 65304040.0,
      "step": 37645
    },
    {
      "entropy": 5.7900513172149655,
      "epoch": 2.9293133631589185,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041435272180235757,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2090310588479042,
      "num_tokens": 65311519.0,
      "step": 37650
    },
    {
      "entropy": 5.716901779174805,
      "epoch": 2.9297023925306362,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000414330960721434,
      "loss": 4.8858,
      "mean_token_accuracy": 0.2211376830935478,
      "num_tokens": 65319864.0,
      "step": 37655
    },
    {
      "entropy": 5.7015259742736815,
      "epoch": 2.9300914219023535,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041430919752638497,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2035437747836113,
      "num_tokens": 65328884.0,
      "step": 37660
    },
    {
      "entropy": 5.726526832580566,
      "epoch": 2.930480451274071,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004142874322175406,
      "loss": 4.9402,
      "mean_token_accuracy": 0.20868986397981643,
      "num_tokens": 65337507.0,
      "step": 37665
    },
    {
      "entropy": 5.7860719680786135,
      "epoch": 2.930869480645789,
      "grad_norm": 1.25,
      "learning_rate": 0.00041426566479523116,
      "loss": 4.934,
      "mean_token_accuracy": 0.20682337880134583,
      "num_tokens": 65345564.0,
      "step": 37670
    },
    {
      "entropy": 5.749407577514648,
      "epoch": 2.931258510017506,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041424389525978683,
      "loss": 5.042,
      "mean_token_accuracy": 0.2043719246983528,
      "num_tokens": 65354736.0,
      "step": 37675
    },
    {
      "entropy": 5.749319934844971,
      "epoch": 2.931647539389224,
      "grad_norm": 1.25,
      "learning_rate": 0.0004142221236115381,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20675768256187438,
      "num_tokens": 65363551.0,
      "step": 37680
    },
    {
      "entropy": 5.835529756546021,
      "epoch": 2.9320365687609415,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004142003498508152,
      "loss": 5.1094,
      "mean_token_accuracy": 0.20293712317943574,
      "num_tokens": 65372394.0,
      "step": 37685
    },
    {
      "entropy": 5.852826690673828,
      "epoch": 2.9324255981326592,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004141785739779486,
      "loss": 5.1173,
      "mean_token_accuracy": 0.19799267798662185,
      "num_tokens": 65381732.0,
      "step": 37690
    },
    {
      "entropy": 5.711467647552491,
      "epoch": 2.9328146275043765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041415679599326853,
      "loss": 5.0492,
      "mean_token_accuracy": 0.19774440973997115,
      "num_tokens": 65390718.0,
      "step": 37695
    },
    {
      "entropy": 5.723369026184082,
      "epoch": 2.933203656876094,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041413501589710567,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21028241217136384,
      "num_tokens": 65399340.0,
      "step": 37700
    },
    {
      "entropy": 5.746189212799072,
      "epoch": 2.9335926862478114,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004141132336897904,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20540821254253389,
      "num_tokens": 65408911.0,
      "step": 37705
    },
    {
      "entropy": 5.73016095161438,
      "epoch": 2.933981715619529,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004140914493716533,
      "loss": 5.0081,
      "mean_token_accuracy": 0.20626921057701111,
      "num_tokens": 65418496.0,
      "step": 37710
    },
    {
      "entropy": 5.71728024482727,
      "epoch": 2.934370744991247,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004140696629430248,
      "loss": 4.9029,
      "mean_token_accuracy": 0.21261151134967804,
      "num_tokens": 65426791.0,
      "step": 37715
    },
    {
      "entropy": 5.854611253738403,
      "epoch": 2.9347597743629645,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041404787440423534,
      "loss": 5.067,
      "mean_token_accuracy": 0.20267720222473146,
      "num_tokens": 65435513.0,
      "step": 37720
    },
    {
      "entropy": 5.782723808288575,
      "epoch": 2.9351488037346822,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041402608375561595,
      "loss": 5.031,
      "mean_token_accuracy": 0.20125727504491805,
      "num_tokens": 65444132.0,
      "step": 37725
    },
    {
      "entropy": 5.744702339172363,
      "epoch": 2.9355378331063995,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041400429099749687,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21486452519893645,
      "num_tokens": 65452313.0,
      "step": 37730
    },
    {
      "entropy": 5.765815258026123,
      "epoch": 2.935926862478117,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000413982496130209,
      "loss": 4.9975,
      "mean_token_accuracy": 0.20691528171300888,
      "num_tokens": 65460724.0,
      "step": 37735
    },
    {
      "entropy": 5.736054420471191,
      "epoch": 2.9363158918498344,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041396069915408294,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21830601543188094,
      "num_tokens": 65468902.0,
      "step": 37740
    },
    {
      "entropy": 5.749852323532105,
      "epoch": 2.936704921221552,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041393890006944945,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21082276552915574,
      "num_tokens": 65477614.0,
      "step": 37745
    },
    {
      "entropy": 5.746755647659302,
      "epoch": 2.93709395059327,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004139170988766393,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20665867626667023,
      "num_tokens": 65486369.0,
      "step": 37750
    },
    {
      "entropy": 5.809875011444092,
      "epoch": 2.9374829799649875,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004138952955759833,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20460011959075927,
      "num_tokens": 65495235.0,
      "step": 37755
    },
    {
      "entropy": 5.745532989501953,
      "epoch": 2.937872009336705,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041387349016781225,
      "loss": 4.981,
      "mean_token_accuracy": 0.21481469124555588,
      "num_tokens": 65504499.0,
      "step": 37760
    },
    {
      "entropy": 5.789349555969238,
      "epoch": 2.9382610387084225,
      "grad_norm": 1.203125,
      "learning_rate": 0.000413851682652457,
      "loss": 5.0835,
      "mean_token_accuracy": 0.1930072009563446,
      "num_tokens": 65512612.0,
      "step": 37765
    },
    {
      "entropy": 5.679971551895141,
      "epoch": 2.93865006808014,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004138298730302485,
      "loss": 4.8064,
      "mean_token_accuracy": 0.22338423430919646,
      "num_tokens": 65521129.0,
      "step": 37770
    },
    {
      "entropy": 5.696767711639405,
      "epoch": 2.9390390974518574,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004138080613015176,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20711466670036316,
      "num_tokens": 65529547.0,
      "step": 37775
    },
    {
      "entropy": 5.679032468795777,
      "epoch": 2.939428126823575,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041378624746659536,
      "loss": 4.9806,
      "mean_token_accuracy": 0.208369180560112,
      "num_tokens": 65538106.0,
      "step": 37780
    },
    {
      "entropy": 5.763957357406616,
      "epoch": 2.939817156195293,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041376443152581265,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2051646366715431,
      "num_tokens": 65546718.0,
      "step": 37785
    },
    {
      "entropy": 5.754372310638428,
      "epoch": 2.9402061855670105,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041374261347950064,
      "loss": 4.9729,
      "mean_token_accuracy": 0.20824459046125413,
      "num_tokens": 65555484.0,
      "step": 37790
    },
    {
      "entropy": 5.646736478805542,
      "epoch": 2.9405952149387278,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004137207933279903,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21545591354370117,
      "num_tokens": 65563285.0,
      "step": 37795
    },
    {
      "entropy": 5.632595872879028,
      "epoch": 2.9409842443104455,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041369897107161266,
      "loss": 4.8061,
      "mean_token_accuracy": 0.22062060683965684,
      "num_tokens": 65571702.0,
      "step": 37800
    },
    {
      "entropy": 5.701446962356568,
      "epoch": 2.941373273682163,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041367714671069897,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21719106435775756,
      "num_tokens": 65580380.0,
      "step": 37805
    },
    {
      "entropy": 5.71339111328125,
      "epoch": 2.9417623030538804,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041365532024558024,
      "loss": 5.0342,
      "mean_token_accuracy": 0.19979556351900102,
      "num_tokens": 65589218.0,
      "step": 37810
    },
    {
      "entropy": 5.731223392486572,
      "epoch": 2.942151332425598,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004136334916765877,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2182898238301277,
      "num_tokens": 65598020.0,
      "step": 37815
    },
    {
      "entropy": 5.808727169036866,
      "epoch": 2.942540361797316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041361166100405264,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20625956505537033,
      "num_tokens": 65606591.0,
      "step": 37820
    },
    {
      "entropy": 5.780256223678589,
      "epoch": 2.9429293911690335,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004135898282283062,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2016656830906868,
      "num_tokens": 65614597.0,
      "step": 37825
    },
    {
      "entropy": 5.695263767242432,
      "epoch": 2.9433184205407508,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004135679933496797,
      "loss": 4.9346,
      "mean_token_accuracy": 0.20670516043901443,
      "num_tokens": 65623092.0,
      "step": 37830
    },
    {
      "entropy": 5.789718103408814,
      "epoch": 2.9437074499124685,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004135461563685045,
      "loss": 5.0186,
      "mean_token_accuracy": 0.19989261478185655,
      "num_tokens": 65632623.0,
      "step": 37835
    },
    {
      "entropy": 5.800095653533935,
      "epoch": 2.9440964792841857,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041352431728511194,
      "loss": 4.9861,
      "mean_token_accuracy": 0.20626098215579985,
      "num_tokens": 65641049.0,
      "step": 37840
    },
    {
      "entropy": 5.792394161224365,
      "epoch": 2.9444855086559034,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004135024760998333,
      "loss": 5.0549,
      "mean_token_accuracy": 0.20023055970668793,
      "num_tokens": 65649928.0,
      "step": 37845
    },
    {
      "entropy": 5.783151388168335,
      "epoch": 2.944874538027621,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041348063281299995,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2002826750278473,
      "num_tokens": 65658597.0,
      "step": 37850
    },
    {
      "entropy": 5.725397109985352,
      "epoch": 2.945263567399339,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041345878742494346,
      "loss": 4.9636,
      "mean_token_accuracy": 0.207400144636631,
      "num_tokens": 65667532.0,
      "step": 37855
    },
    {
      "entropy": 5.702459907531738,
      "epoch": 2.945652596771056,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041343693993599535,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20669802129268647,
      "num_tokens": 65676209.0,
      "step": 37860
    },
    {
      "entropy": 5.757857036590576,
      "epoch": 2.9460416261427738,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004134150903464869,
      "loss": 4.9923,
      "mean_token_accuracy": 0.20346804857254028,
      "num_tokens": 65684372.0,
      "step": 37865
    },
    {
      "entropy": 5.73091607093811,
      "epoch": 2.9464306555144915,
      "grad_norm": 1.25,
      "learning_rate": 0.0004133932386567498,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21479010730981826,
      "num_tokens": 65692386.0,
      "step": 37870
    },
    {
      "entropy": 5.752978181838989,
      "epoch": 2.9468196848862087,
      "grad_norm": 1.25,
      "learning_rate": 0.0004133713848671156,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2064836710691452,
      "num_tokens": 65700707.0,
      "step": 37875
    },
    {
      "entropy": 5.787127494812012,
      "epoch": 2.9472087142579264,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004133495289779159,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20602697432041167,
      "num_tokens": 65710166.0,
      "step": 37880
    },
    {
      "entropy": 5.73648157119751,
      "epoch": 2.947597743629644,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004133276709894823,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21024233400821685,
      "num_tokens": 65718900.0,
      "step": 37885
    },
    {
      "entropy": 5.789086151123047,
      "epoch": 2.947986773001362,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004133058109021464,
      "loss": 5.0493,
      "mean_token_accuracy": 0.20787284821271895,
      "num_tokens": 65727882.0,
      "step": 37890
    },
    {
      "entropy": 5.710661458969116,
      "epoch": 2.948375802373079,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041328394871624003,
      "loss": 4.8575,
      "mean_token_accuracy": 0.212934772670269,
      "num_tokens": 65736681.0,
      "step": 37895
    },
    {
      "entropy": 5.805316829681397,
      "epoch": 2.9487648317447968,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004132620844320948,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20582326352596284,
      "num_tokens": 65745002.0,
      "step": 37900
    },
    {
      "entropy": 5.776008653640747,
      "epoch": 2.9491538611165145,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004132402180500425,
      "loss": 5.03,
      "mean_token_accuracy": 0.20382364243268966,
      "num_tokens": 65753918.0,
      "step": 37905
    },
    {
      "entropy": 5.7881636142730715,
      "epoch": 2.9495428904882317,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004132183495704149,
      "loss": 5.0042,
      "mean_token_accuracy": 0.20378240197896957,
      "num_tokens": 65762659.0,
      "step": 37910
    },
    {
      "entropy": 5.7585508823394775,
      "epoch": 2.9499319198599494,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004131964789935439,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20291766077280043,
      "num_tokens": 65771464.0,
      "step": 37915
    },
    {
      "entropy": 5.709188461303711,
      "epoch": 2.950320949231667,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004131746063197612,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21291617006063462,
      "num_tokens": 65779628.0,
      "step": 37920
    },
    {
      "entropy": 5.745810508728027,
      "epoch": 2.950709978603385,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004131527315493989,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20372420251369477,
      "num_tokens": 65788210.0,
      "step": 37925
    },
    {
      "entropy": 5.774693965911865,
      "epoch": 2.951099007975102,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041313085468278873,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2062986209988594,
      "num_tokens": 65796295.0,
      "step": 37930
    },
    {
      "entropy": 5.780889272689819,
      "epoch": 2.9514880373468197,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004131089757202627,
      "loss": 5.0184,
      "mean_token_accuracy": 0.20742080807685853,
      "num_tokens": 65805670.0,
      "step": 37935
    },
    {
      "entropy": 5.749173402786255,
      "epoch": 2.951877066718537,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041308709466215276,
      "loss": 4.9503,
      "mean_token_accuracy": 0.20759862959384917,
      "num_tokens": 65814835.0,
      "step": 37940
    },
    {
      "entropy": 5.726490116119384,
      "epoch": 2.9522660960902547,
      "grad_norm": 1.21875,
      "learning_rate": 0.000413065211508791,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21502220928668975,
      "num_tokens": 65823405.0,
      "step": 37945
    },
    {
      "entropy": 5.846900892257691,
      "epoch": 2.9526551254619724,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041304332626050937,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2012569412589073,
      "num_tokens": 65832127.0,
      "step": 37950
    },
    {
      "entropy": 5.782841920852661,
      "epoch": 2.95304415483369,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041302143891764,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20871592611074447,
      "num_tokens": 65840690.0,
      "step": 37955
    },
    {
      "entropy": 5.732952451705932,
      "epoch": 2.9534331842054073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000412999549480515,
      "loss": 4.968,
      "mean_token_accuracy": 0.20852576792240143,
      "num_tokens": 65849252.0,
      "step": 37960
    },
    {
      "entropy": 5.762992668151855,
      "epoch": 2.953822213577125,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004129776579494665,
      "loss": 4.8944,
      "mean_token_accuracy": 0.21150533407926558,
      "num_tokens": 65857872.0,
      "step": 37965
    },
    {
      "entropy": 5.7629297256469725,
      "epoch": 2.9542112429488427,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004129557643248265,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2112600699067116,
      "num_tokens": 65866378.0,
      "step": 37970
    },
    {
      "entropy": 5.707042026519775,
      "epoch": 2.95460027232056,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004129338686069276,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2069742888212204,
      "num_tokens": 65875578.0,
      "step": 37975
    },
    {
      "entropy": 5.768731880187988,
      "epoch": 2.9549893016922777,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004129119707961016,
      "loss": 5.028,
      "mean_token_accuracy": 0.19839782863855362,
      "num_tokens": 65884794.0,
      "step": 37980
    },
    {
      "entropy": 5.7582066535949705,
      "epoch": 2.9553783310639954,
      "grad_norm": 1.328125,
      "learning_rate": 0.000412890070892681,
      "loss": 4.92,
      "mean_token_accuracy": 0.2103092908859253,
      "num_tokens": 65893239.0,
      "step": 37985
    },
    {
      "entropy": 5.811394929885864,
      "epoch": 2.955767360435713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000412868168896998,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2042123109102249,
      "num_tokens": 65902420.0,
      "step": 37990
    },
    {
      "entropy": 5.758172035217285,
      "epoch": 2.9561563898074303,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004128462648093851,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20114406645298005,
      "num_tokens": 65911238.0,
      "step": 37995
    },
    {
      "entropy": 5.7628344058990475,
      "epoch": 2.956545419179148,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004128243586301745,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21276420056819917,
      "num_tokens": 65920366.0,
      "step": 38000
    },
    {
      "entropy": 5.684243249893188,
      "epoch": 2.9569344485508657,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004128024503596986,
      "loss": 4.9601,
      "mean_token_accuracy": 0.20868742018938063,
      "num_tokens": 65929753.0,
      "step": 38005
    },
    {
      "entropy": 5.783647251129151,
      "epoch": 2.957323477922583,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041278053999828986,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2072565257549286,
      "num_tokens": 65938387.0,
      "step": 38010
    },
    {
      "entropy": 5.8046647071838375,
      "epoch": 2.9577125072943007,
      "grad_norm": 1.25,
      "learning_rate": 0.0004127586275462806,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21438101530075074,
      "num_tokens": 65946522.0,
      "step": 38015
    },
    {
      "entropy": 5.718218278884888,
      "epoch": 2.9581015366660184,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041273671300400363,
      "loss": 4.8579,
      "mean_token_accuracy": 0.21954718381166458,
      "num_tokens": 65954387.0,
      "step": 38020
    },
    {
      "entropy": 5.683928394317627,
      "epoch": 2.958490566037736,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041271479637179116,
      "loss": 4.9729,
      "mean_token_accuracy": 0.20769625902175903,
      "num_tokens": 65962103.0,
      "step": 38025
    },
    {
      "entropy": 5.742477083206177,
      "epoch": 2.9588795954094533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004126928776499759,
      "loss": 5.0451,
      "mean_token_accuracy": 0.19920091927051545,
      "num_tokens": 65971045.0,
      "step": 38030
    },
    {
      "entropy": 5.813068103790283,
      "epoch": 2.959268624781171,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004126709568388903,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20825294703245162,
      "num_tokens": 65979927.0,
      "step": 38035
    },
    {
      "entropy": 5.816789865493774,
      "epoch": 2.9596576541528883,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004126490339388672,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2035070091485977,
      "num_tokens": 65988955.0,
      "step": 38040
    },
    {
      "entropy": 5.806856679916382,
      "epoch": 2.960046683524606,
      "grad_norm": 1.3125,
      "learning_rate": 0.000412627108950239,
      "loss": 5.084,
      "mean_token_accuracy": 0.19557463377714157,
      "num_tokens": 65997731.0,
      "step": 38045
    },
    {
      "entropy": 5.764254188537597,
      "epoch": 2.9604357128963237,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004126051818733385,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21491411477327346,
      "num_tokens": 66006425.0,
      "step": 38050
    },
    {
      "entropy": 5.756955909729004,
      "epoch": 2.9608247422680414,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004125832527084984,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20675763487815857,
      "num_tokens": 66014849.0,
      "step": 38055
    },
    {
      "entropy": 5.698051977157593,
      "epoch": 2.961213771639759,
      "grad_norm": 1.125,
      "learning_rate": 0.00041256132145605143,
      "loss": 4.9856,
      "mean_token_accuracy": 0.20604703575372696,
      "num_tokens": 66024036.0,
      "step": 38060
    },
    {
      "entropy": 5.816127157211303,
      "epoch": 2.9616028010114763,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004125393881163304,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2077329471707344,
      "num_tokens": 66032583.0,
      "step": 38065
    },
    {
      "entropy": 5.830139398574829,
      "epoch": 2.961991830383194,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041251745268966797,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1952809676527977,
      "num_tokens": 66041369.0,
      "step": 38070
    },
    {
      "entropy": 5.720626974105835,
      "epoch": 2.9623808597549113,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004124955151763971,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20653744041919708,
      "num_tokens": 66049608.0,
      "step": 38075
    },
    {
      "entropy": 5.747635841369629,
      "epoch": 2.962769889126629,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004124735755768506,
      "loss": 5.071,
      "mean_token_accuracy": 0.20562667548656463,
      "num_tokens": 66057612.0,
      "step": 38080
    },
    {
      "entropy": 5.747008991241455,
      "epoch": 2.9631589184983467,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004124516338913615,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21280138790607453,
      "num_tokens": 66065687.0,
      "step": 38085
    },
    {
      "entropy": 5.7730419635772705,
      "epoch": 2.9635479478700644,
      "grad_norm": 1.25,
      "learning_rate": 0.0004124296901202626,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21259047985076904,
      "num_tokens": 66074117.0,
      "step": 38090
    },
    {
      "entropy": 5.7326483726501465,
      "epoch": 2.9639369772417816,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041240774426388685,
      "loss": 4.902,
      "mean_token_accuracy": 0.2179764449596405,
      "num_tokens": 66083022.0,
      "step": 38095
    },
    {
      "entropy": 5.731894874572754,
      "epoch": 2.9643260066134993,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004123857963225673,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22280168384313584,
      "num_tokens": 66091453.0,
      "step": 38100
    },
    {
      "entropy": 5.714373826980591,
      "epoch": 2.964715035985217,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000412363846296637,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21985155194997788,
      "num_tokens": 66099279.0,
      "step": 38105
    },
    {
      "entropy": 5.748297500610351,
      "epoch": 2.9651040653569343,
      "grad_norm": 1.25,
      "learning_rate": 0.000412341894186429,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2076827511191368,
      "num_tokens": 66107224.0,
      "step": 38110
    },
    {
      "entropy": 5.765652322769165,
      "epoch": 2.965493094728652,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004123199399922763,
      "loss": 4.9568,
      "mean_token_accuracy": 0.20541047900915146,
      "num_tokens": 66115719.0,
      "step": 38115
    },
    {
      "entropy": 5.833788061141968,
      "epoch": 2.9658821241003697,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041229798371451203,
      "loss": 4.9704,
      "mean_token_accuracy": 0.20798690617084503,
      "num_tokens": 66123744.0,
      "step": 38120
    },
    {
      "entropy": 5.693296670913696,
      "epoch": 2.9662711534720874,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004122760253534695,
      "loss": 4.8704,
      "mean_token_accuracy": 0.21448576152324678,
      "num_tokens": 66131929.0,
      "step": 38125
    },
    {
      "entropy": 5.849803447723389,
      "epoch": 2.9666601828438046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004122540649094817,
      "loss": 5.1211,
      "mean_token_accuracy": 0.2034429654479027,
      "num_tokens": 66141185.0,
      "step": 38130
    },
    {
      "entropy": 5.719653654098511,
      "epoch": 2.9670492122155223,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041223210238288196,
      "loss": 4.875,
      "mean_token_accuracy": 0.21359510272741317,
      "num_tokens": 66150139.0,
      "step": 38135
    },
    {
      "entropy": 5.68509955406189,
      "epoch": 2.9674382415872396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004122101377740035,
      "loss": 4.8809,
      "mean_token_accuracy": 0.2157800778746605,
      "num_tokens": 66158650.0,
      "step": 38140
    },
    {
      "entropy": 5.719266366958618,
      "epoch": 2.9678272709589573,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004121881710831796,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20459336936473846,
      "num_tokens": 66166683.0,
      "step": 38145
    },
    {
      "entropy": 5.818162155151367,
      "epoch": 2.968216300330675,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004121662023107435,
      "loss": 5.0389,
      "mean_token_accuracy": 0.19992193430662156,
      "num_tokens": 66175100.0,
      "step": 38150
    },
    {
      "entropy": 5.783654689788818,
      "epoch": 2.9686053297023927,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004121442314570286,
      "loss": 4.963,
      "mean_token_accuracy": 0.20806353986263276,
      "num_tokens": 66183764.0,
      "step": 38155
    },
    {
      "entropy": 5.785940217971802,
      "epoch": 2.9689943590741104,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041212225852236834,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2015123963356018,
      "num_tokens": 66192177.0,
      "step": 38160
    },
    {
      "entropy": 5.77645092010498,
      "epoch": 2.9693833884458276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004121002835070961,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21003472656011582,
      "num_tokens": 66200685.0,
      "step": 38165
    },
    {
      "entropy": 5.813252544403076,
      "epoch": 2.9697724178175453,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004120783064115452,
      "loss": 5.028,
      "mean_token_accuracy": 0.20306553244590758,
      "num_tokens": 66209558.0,
      "step": 38170
    },
    {
      "entropy": 5.789865398406983,
      "epoch": 2.9701614471892626,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004120563272360492,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20890484303236007,
      "num_tokens": 66218352.0,
      "step": 38175
    },
    {
      "entropy": 5.830019950866699,
      "epoch": 2.9705504765609803,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004120343459809416,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20792610198259354,
      "num_tokens": 66226771.0,
      "step": 38180
    },
    {
      "entropy": 5.725120162963867,
      "epoch": 2.970939505932698,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004120123626465559,
      "loss": 4.8806,
      "mean_token_accuracy": 0.21828744262456895,
      "num_tokens": 66234950.0,
      "step": 38185
    },
    {
      "entropy": 5.689775991439819,
      "epoch": 2.9713285353044157,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041199037723322566,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20742257088422775,
      "num_tokens": 66243619.0,
      "step": 38190
    },
    {
      "entropy": 5.716256761550904,
      "epoch": 2.971717564676133,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041196838974128446,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20376623868942262,
      "num_tokens": 66252411.0,
      "step": 38195
    },
    {
      "entropy": 5.801603841781616,
      "epoch": 2.9721065940478506,
      "grad_norm": 1.25,
      "learning_rate": 0.000411946400171066,
      "loss": 4.9287,
      "mean_token_accuracy": 0.20895910859107972,
      "num_tokens": 66260660.0,
      "step": 38200
    },
    {
      "entropy": 5.761641025543213,
      "epoch": 2.9724956234195683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041192440852290385,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20128608494997025,
      "num_tokens": 66270072.0,
      "step": 38205
    },
    {
      "entropy": 5.76371169090271,
      "epoch": 2.9728846527912856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004119024147971318,
      "loss": 4.9787,
      "mean_token_accuracy": 0.21514695137739182,
      "num_tokens": 66278971.0,
      "step": 38210
    },
    {
      "entropy": 5.717041921615601,
      "epoch": 2.9732736821630033,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041188041899408345,
      "loss": 4.8183,
      "mean_token_accuracy": 0.21387315392494202,
      "num_tokens": 66287118.0,
      "step": 38215
    },
    {
      "entropy": 5.714354705810547,
      "epoch": 2.973662711534721,
      "grad_norm": 1.25,
      "learning_rate": 0.0004118584211140926,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2153848022222519,
      "num_tokens": 66296041.0,
      "step": 38220
    },
    {
      "entropy": 5.812448358535766,
      "epoch": 2.9740517409064386,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041183642115749316,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21117728054523469,
      "num_tokens": 66304478.0,
      "step": 38225
    },
    {
      "entropy": 5.749567651748658,
      "epoch": 2.974440770278156,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041181441912461873,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2033123105764389,
      "num_tokens": 66313175.0,
      "step": 38230
    },
    {
      "entropy": 5.796424102783203,
      "epoch": 2.9748297996498736,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004117924150158032,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20990918278694154,
      "num_tokens": 66322102.0,
      "step": 38235
    },
    {
      "entropy": 5.750087738037109,
      "epoch": 2.9752188290215913,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041177040883138064,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2057881996035576,
      "num_tokens": 66330930.0,
      "step": 38240
    },
    {
      "entropy": 5.734534645080567,
      "epoch": 2.9756078583933085,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041174840057168474,
      "loss": 5.0656,
      "mean_token_accuracy": 0.20691406726837158,
      "num_tokens": 66339161.0,
      "step": 38245
    },
    {
      "entropy": 5.777521228790283,
      "epoch": 2.9759968877650262,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004117263902370495,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2043324500322342,
      "num_tokens": 66347498.0,
      "step": 38250
    },
    {
      "entropy": 5.7785851001739506,
      "epoch": 2.976385917136744,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041170437782780896,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2072884812951088,
      "num_tokens": 66356848.0,
      "step": 38255
    },
    {
      "entropy": 5.7322922706604,
      "epoch": 2.9767749465084616,
      "grad_norm": 1.375,
      "learning_rate": 0.000411682363344297,
      "loss": 4.9691,
      "mean_token_accuracy": 0.20564228892326356,
      "num_tokens": 66365104.0,
      "step": 38260
    },
    {
      "entropy": 5.742816638946533,
      "epoch": 2.977163975880179,
      "grad_norm": 1.25,
      "learning_rate": 0.00041166034678684777,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21110543757677078,
      "num_tokens": 66373984.0,
      "step": 38265
    },
    {
      "entropy": 5.761313772201538,
      "epoch": 2.9775530052518966,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004116383281557953,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20297882407903672,
      "num_tokens": 66383187.0,
      "step": 38270
    },
    {
      "entropy": 5.749477815628052,
      "epoch": 2.977942034623614,
      "grad_norm": 1.25,
      "learning_rate": 0.00041161630745147374,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2054951921105385,
      "num_tokens": 66392134.0,
      "step": 38275
    },
    {
      "entropy": 5.7860331535339355,
      "epoch": 2.9783310639953315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004115942846742171,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20382155925035478,
      "num_tokens": 66400782.0,
      "step": 38280
    },
    {
      "entropy": 5.743750047683716,
      "epoch": 2.9787200933670492,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004115722598243596,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21231495589017868,
      "num_tokens": 66409185.0,
      "step": 38285
    },
    {
      "entropy": 5.710330152511597,
      "epoch": 2.979109122738767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041155023290223537,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20607083886861802,
      "num_tokens": 66417841.0,
      "step": 38290
    },
    {
      "entropy": 5.719877433776856,
      "epoch": 2.979498152110484,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004115282039081788,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20610310435295104,
      "num_tokens": 66426882.0,
      "step": 38295
    },
    {
      "entropy": 5.779862880706787,
      "epoch": 2.979887181482202,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041150617284252405,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21625621765851974,
      "num_tokens": 66434943.0,
      "step": 38300
    },
    {
      "entropy": 5.750674390792847,
      "epoch": 2.9802762108539196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004114841397056053,
      "loss": 5.0535,
      "mean_token_accuracy": 0.204359333217144,
      "num_tokens": 66444260.0,
      "step": 38305
    },
    {
      "entropy": 5.760622024536133,
      "epoch": 2.980665240225637,
      "grad_norm": 1.125,
      "learning_rate": 0.00041146210449775704,
      "loss": 4.886,
      "mean_token_accuracy": 0.2185773506760597,
      "num_tokens": 66453603.0,
      "step": 38310
    },
    {
      "entropy": 5.744404077529907,
      "epoch": 2.9810542695973545,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004114400672193136,
      "loss": 5.0182,
      "mean_token_accuracy": 0.20099300146102905,
      "num_tokens": 66462613.0,
      "step": 38315
    },
    {
      "entropy": 5.712764644622803,
      "epoch": 2.9814432989690722,
      "grad_norm": 1.375,
      "learning_rate": 0.00041141802787060923,
      "loss": 4.8678,
      "mean_token_accuracy": 0.21373123675584793,
      "num_tokens": 66470405.0,
      "step": 38320
    },
    {
      "entropy": 5.681039810180664,
      "epoch": 2.98183232834079,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004113959864519784,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21051537096500397,
      "num_tokens": 66479298.0,
      "step": 38325
    },
    {
      "entropy": 5.668348550796509,
      "epoch": 2.982221357712507,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004113739429637556,
      "loss": 4.8647,
      "mean_token_accuracy": 0.21617803275585173,
      "num_tokens": 66487939.0,
      "step": 38330
    },
    {
      "entropy": 5.642125415802002,
      "epoch": 2.982610387084225,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004113518974062753,
      "loss": 4.8524,
      "mean_token_accuracy": 0.21952768862247468,
      "num_tokens": 66496864.0,
      "step": 38335
    },
    {
      "entropy": 5.743410253524781,
      "epoch": 2.9829994164559426,
      "grad_norm": 1.25,
      "learning_rate": 0.000411329849779872,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20322478264570237,
      "num_tokens": 66505512.0,
      "step": 38340
    },
    {
      "entropy": 5.714822483062744,
      "epoch": 2.98338844582766,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041130780008488015,
      "loss": 4.901,
      "mean_token_accuracy": 0.21103117763996124,
      "num_tokens": 66513972.0,
      "step": 38345
    },
    {
      "entropy": 5.776326084136963,
      "epoch": 2.9837774751993775,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004112857483216344,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2057521626353264,
      "num_tokens": 66523032.0,
      "step": 38350
    },
    {
      "entropy": 5.768940258026123,
      "epoch": 2.9841665045710952,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041126369449046935,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21044726073741912,
      "num_tokens": 66531693.0,
      "step": 38355
    },
    {
      "entropy": 5.771009826660157,
      "epoch": 2.984555533942813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041124163859171967,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20415421575307846,
      "num_tokens": 66540834.0,
      "step": 38360
    },
    {
      "entropy": 5.8024513721466064,
      "epoch": 2.98494456331453,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004112195806257199,
      "loss": 4.978,
      "mean_token_accuracy": 0.21123577058315277,
      "num_tokens": 66549189.0,
      "step": 38365
    },
    {
      "entropy": 5.741639852523804,
      "epoch": 2.985333592686248,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004111975205928049,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21631405651569366,
      "num_tokens": 66558169.0,
      "step": 38370
    },
    {
      "entropy": 5.804769659042359,
      "epoch": 2.985722622057965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041117545849330924,
      "loss": 5.0569,
      "mean_token_accuracy": 0.19624873399734497,
      "num_tokens": 66567036.0,
      "step": 38375
    },
    {
      "entropy": 5.803778457641601,
      "epoch": 2.986111651429683,
      "grad_norm": 1.25,
      "learning_rate": 0.00041115339432756776,
      "loss": 5.0537,
      "mean_token_accuracy": 0.19827789813280106,
      "num_tokens": 66575990.0,
      "step": 38380
    },
    {
      "entropy": 5.8171892166137695,
      "epoch": 2.9865006808014005,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004111313280959152,
      "loss": 5.0158,
      "mean_token_accuracy": 0.2033352717757225,
      "num_tokens": 66584675.0,
      "step": 38385
    },
    {
      "entropy": 5.766924571990967,
      "epoch": 2.986889710173118,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041110925979868647,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2044031873345375,
      "num_tokens": 66593938.0,
      "step": 38390
    },
    {
      "entropy": 5.772219467163086,
      "epoch": 2.9872787395448355,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004110871894362163,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21301252394914627,
      "num_tokens": 66602043.0,
      "step": 38395
    },
    {
      "entropy": 5.797375917434692,
      "epoch": 2.987667768916553,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041106511700883975,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2050095722079277,
      "num_tokens": 66610376.0,
      "step": 38400
    },
    {
      "entropy": 5.739901971817017,
      "epoch": 2.988056798288271,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041104304251689147,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2035106137394905,
      "num_tokens": 66619253.0,
      "step": 38405
    },
    {
      "entropy": 5.813145399093628,
      "epoch": 2.988445827659988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041102096596070666,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20244687497615815,
      "num_tokens": 66627983.0,
      "step": 38410
    },
    {
      "entropy": 5.757772302627563,
      "epoch": 2.988834857031706,
      "grad_norm": 1.375,
      "learning_rate": 0.00041099888734062013,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21422244012355804,
      "num_tokens": 66635629.0,
      "step": 38415
    },
    {
      "entropy": 5.813073825836182,
      "epoch": 2.9892238864034235,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000410976806656967,
      "loss": 5.1356,
      "mean_token_accuracy": 0.19308022856712342,
      "num_tokens": 66643960.0,
      "step": 38420
    },
    {
      "entropy": 5.7791276454925535,
      "epoch": 2.989612915775141,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004109547239100822,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20786335170269013,
      "num_tokens": 66652252.0,
      "step": 38425
    },
    {
      "entropy": 5.716545104980469,
      "epoch": 2.9900019451468585,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004109326391003009,
      "loss": 4.9578,
      "mean_token_accuracy": 0.20923302620649337,
      "num_tokens": 66660773.0,
      "step": 38430
    },
    {
      "entropy": 5.786676692962646,
      "epoch": 2.990390974518576,
      "grad_norm": 1.25,
      "learning_rate": 0.00041091055222795814,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20908840596675873,
      "num_tokens": 66668681.0,
      "step": 38435
    },
    {
      "entropy": 5.722526264190674,
      "epoch": 2.990780003890294,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004108884632933891,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2074008494615555,
      "num_tokens": 66678971.0,
      "step": 38440
    },
    {
      "entropy": 5.814666414260865,
      "epoch": 2.991169033262011,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041086637229692893,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20274681597948074,
      "num_tokens": 66687217.0,
      "step": 38445
    },
    {
      "entropy": 5.755166864395141,
      "epoch": 2.991558062633729,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041084427923891275,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2109092652797699,
      "num_tokens": 66696020.0,
      "step": 38450
    },
    {
      "entropy": 5.784342670440674,
      "epoch": 2.9919470920054465,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041082218411967586,
      "loss": 5.0289,
      "mean_token_accuracy": 0.20709141194820405,
      "num_tokens": 66704652.0,
      "step": 38455
    },
    {
      "entropy": 5.765918159484864,
      "epoch": 2.992336121377164,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041080008693955355,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21026127338409423,
      "num_tokens": 66713803.0,
      "step": 38460
    },
    {
      "entropy": 5.732566833496094,
      "epoch": 2.9927251507488815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004107779876988811,
      "loss": 4.8838,
      "mean_token_accuracy": 0.21865480542182922,
      "num_tokens": 66722102.0,
      "step": 38465
    },
    {
      "entropy": 5.733516454696655,
      "epoch": 2.993114180120599,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041075588639799375,
      "loss": 4.963,
      "mean_token_accuracy": 0.21237993240356445,
      "num_tokens": 66729951.0,
      "step": 38470
    },
    {
      "entropy": 5.697514390945434,
      "epoch": 2.9935032094923164,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004107337830372269,
      "loss": 4.9372,
      "mean_token_accuracy": 0.20656849890947343,
      "num_tokens": 66738737.0,
      "step": 38475
    },
    {
      "entropy": 5.769795513153076,
      "epoch": 2.993892238864034,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000410711677616916,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20565010309219361,
      "num_tokens": 66747436.0,
      "step": 38480
    },
    {
      "entropy": 5.805146980285644,
      "epoch": 2.994281268235752,
      "grad_norm": 1.25,
      "learning_rate": 0.0004106895701373963,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2061385065317154,
      "num_tokens": 66756274.0,
      "step": 38485
    },
    {
      "entropy": 5.71694049835205,
      "epoch": 2.9946702976074695,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041066746059900336,
      "loss": 4.915,
      "mean_token_accuracy": 0.21304097920656204,
      "num_tokens": 66765002.0,
      "step": 38490
    },
    {
      "entropy": 5.711454439163208,
      "epoch": 2.995059326979187,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004106453490020727,
      "loss": 4.9663,
      "mean_token_accuracy": 0.20740969330072404,
      "num_tokens": 66773478.0,
      "step": 38495
    },
    {
      "entropy": 5.763480615615845,
      "epoch": 2.9954483563509045,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041062323534693973,
      "loss": 5.0176,
      "mean_token_accuracy": 0.20974729657173158,
      "num_tokens": 66781694.0,
      "step": 38500
    },
    {
      "entropy": 5.7504345893859865,
      "epoch": 2.995837385722622,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041060111963394004,
      "loss": 4.9875,
      "mean_token_accuracy": 0.20956162959337235,
      "num_tokens": 66790513.0,
      "step": 38505
    },
    {
      "entropy": 5.720385456085205,
      "epoch": 2.9962264150943394,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041057900186340917,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21445202231407165,
      "num_tokens": 66798661.0,
      "step": 38510
    },
    {
      "entropy": 5.744590139389038,
      "epoch": 2.996615444466057,
      "grad_norm": 1.25,
      "learning_rate": 0.00041055688203568274,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20719115436077118,
      "num_tokens": 66807516.0,
      "step": 38515
    },
    {
      "entropy": 5.707797813415527,
      "epoch": 2.997004473837775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004105347601510964,
      "loss": 4.8012,
      "mean_token_accuracy": 0.22146872878074647,
      "num_tokens": 66816753.0,
      "step": 38520
    },
    {
      "entropy": 5.746677684783935,
      "epoch": 2.9973935032094925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004105126362099859,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21039930284023284,
      "num_tokens": 66825313.0,
      "step": 38525
    },
    {
      "entropy": 5.774389600753784,
      "epoch": 2.9977825325812097,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004104905102126867,
      "loss": 5.0607,
      "mean_token_accuracy": 0.19798608422279357,
      "num_tokens": 66833899.0,
      "step": 38530
    },
    {
      "entropy": 5.7406891822814945,
      "epoch": 2.9981715619529274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004104683821595347,
      "loss": 4.965,
      "mean_token_accuracy": 0.20784751027822496,
      "num_tokens": 66842515.0,
      "step": 38535
    },
    {
      "entropy": 5.7687653541564945,
      "epoch": 2.998560591324645,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041044625205086566,
      "loss": 4.9371,
      "mean_token_accuracy": 0.20992920100688933,
      "num_tokens": 66851112.0,
      "step": 38540
    },
    {
      "entropy": 5.69510407447815,
      "epoch": 2.9989496206963624,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041042411988701535,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21230171918869017,
      "num_tokens": 66860163.0,
      "step": 38545
    },
    {
      "entropy": 5.75424108505249,
      "epoch": 2.99933865006808,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041040198566831944,
      "loss": 4.992,
      "mean_token_accuracy": 0.2104624703526497,
      "num_tokens": 66868532.0,
      "step": 38550
    },
    {
      "entropy": 5.69028058052063,
      "epoch": 2.999727679439798,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004103798493951141,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2060410648584366,
      "num_tokens": 66877524.0,
      "step": 38555
    },
    {
      "entropy": 5.710945765177409,
      "epoch": 3.0000778058743434,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041035771106773483,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21177851491504246,
      "num_tokens": 66884807.0,
      "step": 38560
    },
    {
      "entropy": 5.791142559051513,
      "epoch": 3.000466835246061,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004103355706865179,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21445183157920839,
      "num_tokens": 66894200.0,
      "step": 38565
    },
    {
      "entropy": 5.686918258666992,
      "epoch": 3.0008558646177788,
      "grad_norm": 1.171875,
      "learning_rate": 0.000410313428251799,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21561605483293533,
      "num_tokens": 66903302.0,
      "step": 38570
    },
    {
      "entropy": 5.7380280017852785,
      "epoch": 3.001244893989496,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041029128376391424,
      "loss": 4.9301,
      "mean_token_accuracy": 0.20791093111038209,
      "num_tokens": 66911441.0,
      "step": 38575
    },
    {
      "entropy": 5.720616292953491,
      "epoch": 3.0016339233612137,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041026913722319954,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2156571090221405,
      "num_tokens": 66920167.0,
      "step": 38580
    },
    {
      "entropy": 5.7749182224273685,
      "epoch": 3.0020229527329314,
      "grad_norm": 1.203125,
      "learning_rate": 0.000410246988629991,
      "loss": 4.9806,
      "mean_token_accuracy": 0.20120503902435302,
      "num_tokens": 66928839.0,
      "step": 38585
    },
    {
      "entropy": 5.754401540756225,
      "epoch": 3.002411982104649,
      "grad_norm": 1.25,
      "learning_rate": 0.00041022483798462466,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2106250911951065,
      "num_tokens": 66937779.0,
      "step": 38590
    },
    {
      "entropy": 5.764439344406128,
      "epoch": 3.0028010114763664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004102026852874367,
      "loss": 4.9333,
      "mean_token_accuracy": 0.213492488861084,
      "num_tokens": 66946059.0,
      "step": 38595
    },
    {
      "entropy": 5.786048173904419,
      "epoch": 3.003190040848084,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041018053053876316,
      "loss": 4.9617,
      "mean_token_accuracy": 0.20981108546257018,
      "num_tokens": 66955201.0,
      "step": 38600
    },
    {
      "entropy": 5.797392463684082,
      "epoch": 3.0035790702198017,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004101583737389402,
      "loss": 4.896,
      "mean_token_accuracy": 0.21738944351673126,
      "num_tokens": 66963529.0,
      "step": 38605
    },
    {
      "entropy": 5.820156097412109,
      "epoch": 3.003968099591519,
      "grad_norm": 1.25,
      "learning_rate": 0.00041013621488830406,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1968762382864952,
      "num_tokens": 66972224.0,
      "step": 38610
    },
    {
      "entropy": 5.825885152816772,
      "epoch": 3.0043571289632367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041011405398719096,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20220792293548584,
      "num_tokens": 66981036.0,
      "step": 38615
    },
    {
      "entropy": 5.800108289718628,
      "epoch": 3.0047461583349544,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041009189103593716,
      "loss": 4.9448,
      "mean_token_accuracy": 0.20585397332906724,
      "num_tokens": 66989511.0,
      "step": 38620
    },
    {
      "entropy": 5.782093048095703,
      "epoch": 3.0051351877066717,
      "grad_norm": 1.25,
      "learning_rate": 0.0004100697260348789,
      "loss": 4.9217,
      "mean_token_accuracy": 0.20810072124004364,
      "num_tokens": 66998521.0,
      "step": 38625
    },
    {
      "entropy": 5.716224718093872,
      "epoch": 3.0055242170783893,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041004755898435254,
      "loss": 4.8177,
      "mean_token_accuracy": 0.22549835741519927,
      "num_tokens": 67007173.0,
      "step": 38630
    },
    {
      "entropy": 5.71312403678894,
      "epoch": 3.005913246450107,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004100253898846945,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20299778431653975,
      "num_tokens": 67015671.0,
      "step": 38635
    },
    {
      "entropy": 5.72544641494751,
      "epoch": 3.0063022758218247,
      "grad_norm": 1.34375,
      "learning_rate": 0.000410003218736241,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21171251982450484,
      "num_tokens": 67023501.0,
      "step": 38640
    },
    {
      "entropy": 5.8085489749908445,
      "epoch": 3.006691305193542,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040998104553932856,
      "loss": 5.031,
      "mean_token_accuracy": 0.20339032411575317,
      "num_tokens": 67032409.0,
      "step": 38645
    },
    {
      "entropy": 5.84272141456604,
      "epoch": 3.0070803345652597,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004099588702942936,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20583265274763107,
      "num_tokens": 67041128.0,
      "step": 38650
    },
    {
      "entropy": 5.770159339904785,
      "epoch": 3.0074693639369774,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040993669300147257,
      "loss": 4.951,
      "mean_token_accuracy": 0.21565534323453903,
      "num_tokens": 67050000.0,
      "step": 38655
    },
    {
      "entropy": 5.763939285278321,
      "epoch": 3.0078583933086946,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040991451366120205,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20793238431215286,
      "num_tokens": 67059064.0,
      "step": 38660
    },
    {
      "entropy": 5.752371501922608,
      "epoch": 3.0082474226804123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004098923322738185,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21327605843544006,
      "num_tokens": 67067710.0,
      "step": 38665
    },
    {
      "entropy": 5.740877294540406,
      "epoch": 3.00863645205213,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040987014883965843,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21167029440402985,
      "num_tokens": 67075709.0,
      "step": 38670
    },
    {
      "entropy": 5.824728918075562,
      "epoch": 3.0090254814238473,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040984796335905867,
      "loss": 4.9979,
      "mean_token_accuracy": 0.20370995998382568,
      "num_tokens": 67084509.0,
      "step": 38675
    },
    {
      "entropy": 5.788042640686035,
      "epoch": 3.009414510795565,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004098257758323556,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2113847702741623,
      "num_tokens": 67092826.0,
      "step": 38680
    },
    {
      "entropy": 5.756755399703979,
      "epoch": 3.0098035401672827,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004098035862598859,
      "loss": 4.9216,
      "mean_token_accuracy": 0.210196353495121,
      "num_tokens": 67101379.0,
      "step": 38685
    },
    {
      "entropy": 5.796127796173096,
      "epoch": 3.0101925695390004,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004097813946419865,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2040318638086319,
      "num_tokens": 67110550.0,
      "step": 38690
    },
    {
      "entropy": 5.735618782043457,
      "epoch": 3.0105815989107176,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004097592009789939,
      "loss": 4.8391,
      "mean_token_accuracy": 0.22018862068653106,
      "num_tokens": 67119183.0,
      "step": 38695
    },
    {
      "entropy": 5.815178442001343,
      "epoch": 3.0109706282824353,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040973700527124494,
      "loss": 4.9564,
      "mean_token_accuracy": 0.20765087008476257,
      "num_tokens": 67127765.0,
      "step": 38700
    },
    {
      "entropy": 5.847450208663941,
      "epoch": 3.011359657654153,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040971480751907633,
      "loss": 5.0975,
      "mean_token_accuracy": 0.20357577949762345,
      "num_tokens": 67136395.0,
      "step": 38705
    },
    {
      "entropy": 5.769359302520752,
      "epoch": 3.0117486870258703,
      "grad_norm": 1.375,
      "learning_rate": 0.00040969260772282494,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21423709094524385,
      "num_tokens": 67144458.0,
      "step": 38710
    },
    {
      "entropy": 5.804913568496704,
      "epoch": 3.012137716397588,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004096704058828277,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20918814241886138,
      "num_tokens": 67153363.0,
      "step": 38715
    },
    {
      "entropy": 5.781003999710083,
      "epoch": 3.0125267457693057,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040964820199942135,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21344236582517623,
      "num_tokens": 67162231.0,
      "step": 38720
    },
    {
      "entropy": 5.741076421737671,
      "epoch": 3.012915775141023,
      "grad_norm": 1.25,
      "learning_rate": 0.0004096259960729428,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21578937768936157,
      "num_tokens": 67170726.0,
      "step": 38725
    },
    {
      "entropy": 5.777537107467651,
      "epoch": 3.0133048045127406,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040960378810372906,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20782314836978913,
      "num_tokens": 67179617.0,
      "step": 38730
    },
    {
      "entropy": 5.746387720108032,
      "epoch": 3.0136938338844583,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095815780921171,
      "loss": 4.931,
      "mean_token_accuracy": 0.2181040182709694,
      "num_tokens": 67187990.0,
      "step": 38735
    },
    {
      "entropy": 5.805026531219482,
      "epoch": 3.014082863256176,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004095593660384439,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21347876638174057,
      "num_tokens": 67196549.0,
      "step": 38740
    },
    {
      "entropy": 5.810295677185058,
      "epoch": 3.0144718926278933,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004095371519430465,
      "loss": 4.9648,
      "mean_token_accuracy": 0.20977624654769897,
      "num_tokens": 67205206.0,
      "step": 38745
    },
    {
      "entropy": 5.82351393699646,
      "epoch": 3.014860921999611,
      "grad_norm": 1.25,
      "learning_rate": 0.00040951493580626195,
      "loss": 5.1021,
      "mean_token_accuracy": 0.19830751270055771,
      "num_tokens": 67213582.0,
      "step": 38750
    },
    {
      "entropy": 5.7222741603851315,
      "epoch": 3.0152499513713287,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004094927176284273,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21655819267034532,
      "num_tokens": 67222300.0,
      "step": 38755
    },
    {
      "entropy": 5.736434507369995,
      "epoch": 3.015638980743046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040947049740987977,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2174398720264435,
      "num_tokens": 67230794.0,
      "step": 38760
    },
    {
      "entropy": 5.782102012634278,
      "epoch": 3.0160280101147636,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004094482751509565,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20948062241077423,
      "num_tokens": 67239134.0,
      "step": 38765
    },
    {
      "entropy": 5.779600763320923,
      "epoch": 3.0164170394864813,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004094260508519947,
      "loss": 4.9529,
      "mean_token_accuracy": 0.20633677691221236,
      "num_tokens": 67247929.0,
      "step": 38770
    },
    {
      "entropy": 5.835150527954101,
      "epoch": 3.0168060688581986,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040940382451333144,
      "loss": 5.0377,
      "mean_token_accuracy": 0.19676806926727294,
      "num_tokens": 67255649.0,
      "step": 38775
    },
    {
      "entropy": 5.749681997299194,
      "epoch": 3.0171950982299163,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040938159613530405,
      "loss": 4.8784,
      "mean_token_accuracy": 0.21579378992319107,
      "num_tokens": 67264004.0,
      "step": 38780
    },
    {
      "entropy": 5.729515218734742,
      "epoch": 3.017584127601634,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004093593657182499,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2167251303792,
      "num_tokens": 67273081.0,
      "step": 38785
    },
    {
      "entropy": 5.780004978179932,
      "epoch": 3.0179731569733517,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004093371332625062,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21318065822124482,
      "num_tokens": 67282414.0,
      "step": 38790
    },
    {
      "entropy": 5.810172986984253,
      "epoch": 3.018362186345069,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004093148987684103,
      "loss": 4.9406,
      "mean_token_accuracy": 0.20979322195053102,
      "num_tokens": 67291630.0,
      "step": 38795
    },
    {
      "entropy": 5.781027126312256,
      "epoch": 3.0187512157167866,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040929266223629966,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21050905287265778,
      "num_tokens": 67300429.0,
      "step": 38800
    },
    {
      "entropy": 5.771558094024658,
      "epoch": 3.0191402450885043,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040927042366651164,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22085565477609634,
      "num_tokens": 67309374.0,
      "step": 38805
    },
    {
      "entropy": 5.7766424179077145,
      "epoch": 3.0195292744602216,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040924818305938354,
      "loss": 4.8696,
      "mean_token_accuracy": 0.21970000118017197,
      "num_tokens": 67317706.0,
      "step": 38810
    },
    {
      "entropy": 5.692120313644409,
      "epoch": 3.0199183038319393,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040922594041525303,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21240868866443635,
      "num_tokens": 67326754.0,
      "step": 38815
    },
    {
      "entropy": 5.815887069702148,
      "epoch": 3.020307333203657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004092036957344574,
      "loss": 4.9862,
      "mean_token_accuracy": 0.20628483593463898,
      "num_tokens": 67336138.0,
      "step": 38820
    },
    {
      "entropy": 5.873916149139404,
      "epoch": 3.0206963625753747,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040918144901733443,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20660205334424972,
      "num_tokens": 67344712.0,
      "step": 38825
    },
    {
      "entropy": 5.753109359741211,
      "epoch": 3.021085391947092,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004091592002642215,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2241586908698082,
      "num_tokens": 67352662.0,
      "step": 38830
    },
    {
      "entropy": 5.70209002494812,
      "epoch": 3.0214744213188096,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004091369494754562,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2190609261393547,
      "num_tokens": 67361489.0,
      "step": 38835
    },
    {
      "entropy": 5.7553094863891605,
      "epoch": 3.0218634506905273,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004091146966513763,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2175418958067894,
      "num_tokens": 67369592.0,
      "step": 38840
    },
    {
      "entropy": 5.7961039543151855,
      "epoch": 3.0222524800622446,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040909244179231926,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21575529426336287,
      "num_tokens": 67378761.0,
      "step": 38845
    },
    {
      "entropy": 5.7453117847442625,
      "epoch": 3.0226415094339623,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004090701848986229,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2171661913394928,
      "num_tokens": 67387574.0,
      "step": 38850
    },
    {
      "entropy": 5.7421118259429935,
      "epoch": 3.02303053880568,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004090479259706247,
      "loss": 4.8517,
      "mean_token_accuracy": 0.22144483625888825,
      "num_tokens": 67395665.0,
      "step": 38855
    },
    {
      "entropy": 5.819369459152222,
      "epoch": 3.023419568177397,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004090256650086628,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21103609651327132,
      "num_tokens": 67403480.0,
      "step": 38860
    },
    {
      "entropy": 5.781932306289673,
      "epoch": 3.023808597549115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040900340201307475,
      "loss": 4.9684,
      "mean_token_accuracy": 0.20906000286340715,
      "num_tokens": 67411876.0,
      "step": 38865
    },
    {
      "entropy": 5.762634515762329,
      "epoch": 3.0241976269208326,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040898113698419823,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2084784358739853,
      "num_tokens": 67420267.0,
      "step": 38870
    },
    {
      "entropy": 5.753719615936279,
      "epoch": 3.0245866562925503,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004089588699223713,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2112307146191597,
      "num_tokens": 67428707.0,
      "step": 38875
    },
    {
      "entropy": 5.790412855148316,
      "epoch": 3.0249756856642676,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040893660082793167,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20740681886672974,
      "num_tokens": 67437061.0,
      "step": 38880
    },
    {
      "entropy": 5.782514524459839,
      "epoch": 3.0253647150359853,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00040891432970121736,
      "loss": 4.967,
      "mean_token_accuracy": 0.20520188361406327,
      "num_tokens": 67445270.0,
      "step": 38885
    },
    {
      "entropy": 5.739250564575196,
      "epoch": 3.025753744407703,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040889205654256615,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20400436818599701,
      "num_tokens": 67453960.0,
      "step": 38890
    },
    {
      "entropy": 5.835293865203857,
      "epoch": 3.02614277377942,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040886978135231625,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2024360179901123,
      "num_tokens": 67462124.0,
      "step": 38895
    },
    {
      "entropy": 5.746919250488281,
      "epoch": 3.026531803151138,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040884750413080535,
      "loss": 4.8568,
      "mean_token_accuracy": 0.21170284152030944,
      "num_tokens": 67470762.0,
      "step": 38900
    },
    {
      "entropy": 5.7531877040863035,
      "epoch": 3.0269208325228556,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040882522487837176,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21013229340314865,
      "num_tokens": 67479363.0,
      "step": 38905
    },
    {
      "entropy": 5.752018785476684,
      "epoch": 3.027309861894573,
      "grad_norm": 1.25,
      "learning_rate": 0.00040880294359535327,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21446655243635177,
      "num_tokens": 67487803.0,
      "step": 38910
    },
    {
      "entropy": 5.794632196426392,
      "epoch": 3.0276988912662905,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040878066028208804,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2022012948989868,
      "num_tokens": 67496524.0,
      "step": 38915
    },
    {
      "entropy": 5.712293338775635,
      "epoch": 3.0280879206380082,
      "grad_norm": 1.25,
      "learning_rate": 0.00040875837493891434,
      "loss": 4.9233,
      "mean_token_accuracy": 0.20832296162843705,
      "num_tokens": 67505327.0,
      "step": 38920
    },
    {
      "entropy": 5.712574481964111,
      "epoch": 3.028476950009726,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004087360875661701,
      "loss": 4.8732,
      "mean_token_accuracy": 0.216759617626667,
      "num_tokens": 67514216.0,
      "step": 38925
    },
    {
      "entropy": 5.749611568450928,
      "epoch": 3.028865979381443,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040871379816419376,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21068477183580397,
      "num_tokens": 67522957.0,
      "step": 38930
    },
    {
      "entropy": 5.728671646118164,
      "epoch": 3.029255008753161,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004086915067333232,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21332569420337677,
      "num_tokens": 67532013.0,
      "step": 38935
    },
    {
      "entropy": 5.649725866317749,
      "epoch": 3.0296440381248786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004086692132738968,
      "loss": 4.8191,
      "mean_token_accuracy": 0.21994998008012773,
      "num_tokens": 67540683.0,
      "step": 38940
    },
    {
      "entropy": 5.86691689491272,
      "epoch": 3.030033067496596,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000408646917786253,
      "loss": 5.0591,
      "mean_token_accuracy": 0.20109203904867173,
      "num_tokens": 67549697.0,
      "step": 38945
    },
    {
      "entropy": 5.739890289306641,
      "epoch": 3.0304220968683135,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004086246202707298,
      "loss": 4.871,
      "mean_token_accuracy": 0.21212777942419053,
      "num_tokens": 67557335.0,
      "step": 38950
    },
    {
      "entropy": 5.70845947265625,
      "epoch": 3.0308111262400312,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040860232072766577,
      "loss": 4.9246,
      "mean_token_accuracy": 0.2063554972410202,
      "num_tokens": 67565542.0,
      "step": 38955
    },
    {
      "entropy": 5.7423272132873535,
      "epoch": 3.0312001556117485,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004085800191573991,
      "loss": 4.8836,
      "mean_token_accuracy": 0.2126782938838005,
      "num_tokens": 67573413.0,
      "step": 38960
    },
    {
      "entropy": 5.718665933609008,
      "epoch": 3.031589184983466,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004085577155602683,
      "loss": 4.7558,
      "mean_token_accuracy": 0.22742581963539124,
      "num_tokens": 67581912.0,
      "step": 38965
    },
    {
      "entropy": 5.743589782714844,
      "epoch": 3.031978214355184,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040853540993661175,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21669421195983887,
      "num_tokens": 67591344.0,
      "step": 38970
    },
    {
      "entropy": 5.779635334014893,
      "epoch": 3.0323672437269016,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040851310228676787,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2045698031783104,
      "num_tokens": 67599680.0,
      "step": 38975
    },
    {
      "entropy": 5.785592937469483,
      "epoch": 3.032756273098619,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004084907926110752,
      "loss": 4.9215,
      "mean_token_accuracy": 0.20775238126516343,
      "num_tokens": 67608420.0,
      "step": 38980
    },
    {
      "entropy": 5.756100225448608,
      "epoch": 3.0331453024703365,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004084684809098722,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20475578159093857,
      "num_tokens": 67617174.0,
      "step": 38985
    },
    {
      "entropy": 5.787242794036866,
      "epoch": 3.0335343318420542,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040844616718349743,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2028190091252327,
      "num_tokens": 67625273.0,
      "step": 38990
    },
    {
      "entropy": 5.828755283355713,
      "epoch": 3.0339233612137715,
      "grad_norm": 1.25,
      "learning_rate": 0.00040842385143228946,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20483621656894685,
      "num_tokens": 67633982.0,
      "step": 38995
    },
    {
      "entropy": 5.816717910766601,
      "epoch": 3.034312390585489,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040840153365658693,
      "loss": 4.979,
      "mean_token_accuracy": 0.20942473262548447,
      "num_tokens": 67642739.0,
      "step": 39000
    },
    {
      "epoch": 3.034312390585489,
      "eval_entropy": 5.502825894966162,
      "eval_loss": 5.049113750457764,
      "eval_mean_token_accuracy": 0.21488728179480734,
      "eval_num_tokens": 67642739.0,
      "eval_runtime": 21.6709,
      "eval_samples_per_second": 1917.688,
      "eval_steps_per_second": 239.723,
      "step": 39000
    },
    {
      "entropy": 5.760551691055298,
      "epoch": 3.034701419957207,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040837921385672855,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2050407648086548,
      "num_tokens": 67652025.0,
      "step": 39005
    },
    {
      "entropy": 5.814124059677124,
      "epoch": 3.035090449328924,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040835689203305275,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2020278975367546,
      "num_tokens": 67660521.0,
      "step": 39010
    },
    {
      "entropy": 5.786734199523925,
      "epoch": 3.035479478700642,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040833456818589845,
      "loss": 4.9353,
      "mean_token_accuracy": 0.20891972333192826,
      "num_tokens": 67668590.0,
      "step": 39015
    },
    {
      "entropy": 5.720639610290528,
      "epoch": 3.0358685080723595,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040831224231560424,
      "loss": 4.904,
      "mean_token_accuracy": 0.211442469060421,
      "num_tokens": 67676815.0,
      "step": 39020
    },
    {
      "entropy": 5.724521923065185,
      "epoch": 3.0362575374440772,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000408289914422509,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21161848306655884,
      "num_tokens": 67685417.0,
      "step": 39025
    },
    {
      "entropy": 5.698367500305176,
      "epoch": 3.0366465668157945,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004082675845069515,
      "loss": 4.8637,
      "mean_token_accuracy": 0.21829329878091813,
      "num_tokens": 67693758.0,
      "step": 39030
    },
    {
      "entropy": 5.797330713272094,
      "epoch": 3.037035596187512,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004082452525692704,
      "loss": 4.9764,
      "mean_token_accuracy": 0.20921367257833481,
      "num_tokens": 67701762.0,
      "step": 39035
    },
    {
      "entropy": 5.715843391418457,
      "epoch": 3.03742462555923,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040822291860980484,
      "loss": 4.8741,
      "mean_token_accuracy": 0.2157423198223114,
      "num_tokens": 67710786.0,
      "step": 39040
    },
    {
      "entropy": 5.773695564270019,
      "epoch": 3.037813654930947,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040820058262889343,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20783009082078935,
      "num_tokens": 67718902.0,
      "step": 39045
    },
    {
      "entropy": 5.786623859405518,
      "epoch": 3.038202684302665,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040817824462687516,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21336370706558228,
      "num_tokens": 67727194.0,
      "step": 39050
    },
    {
      "entropy": 5.804684638977051,
      "epoch": 3.0385917136743825,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040815590460408906,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21190279722213745,
      "num_tokens": 67735463.0,
      "step": 39055
    },
    {
      "entropy": 5.715198373794555,
      "epoch": 3.0389807430460998,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040813356256087395,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21835399568080902,
      "num_tokens": 67744505.0,
      "step": 39060
    },
    {
      "entropy": 5.734085321426392,
      "epoch": 3.0393697724178175,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004081112184975691,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20090840011835098,
      "num_tokens": 67753709.0,
      "step": 39065
    },
    {
      "entropy": 5.7642467498779295,
      "epoch": 3.039758801789535,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004080888724145132,
      "loss": 4.9461,
      "mean_token_accuracy": 0.20791601538658142,
      "num_tokens": 67761659.0,
      "step": 39070
    },
    {
      "entropy": 5.7474609375,
      "epoch": 3.040147831161253,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004080665243120456,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2097438782453537,
      "num_tokens": 67769644.0,
      "step": 39075
    },
    {
      "entropy": 5.761591815948487,
      "epoch": 3.04053686053297,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004080441741905053,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20149799585342407,
      "num_tokens": 67779208.0,
      "step": 39080
    },
    {
      "entropy": 5.816317176818847,
      "epoch": 3.040925889904688,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004080218220502314,
      "loss": 5.013,
      "mean_token_accuracy": 0.21056080013513565,
      "num_tokens": 67788542.0,
      "step": 39085
    },
    {
      "entropy": 5.768927907943725,
      "epoch": 3.0413149192764055,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040799946789156304,
      "loss": 4.9581,
      "mean_token_accuracy": 0.20755518227815628,
      "num_tokens": 67796863.0,
      "step": 39090
    },
    {
      "entropy": 5.77177586555481,
      "epoch": 3.0417039486481228,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040797711171483954,
      "loss": 4.988,
      "mean_token_accuracy": 0.2144743651151657,
      "num_tokens": 67805860.0,
      "step": 39095
    },
    {
      "entropy": 5.726226186752319,
      "epoch": 3.0420929780198405,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004079547535203999,
      "loss": 4.8772,
      "mean_token_accuracy": 0.21508204191923141,
      "num_tokens": 67814110.0,
      "step": 39100
    },
    {
      "entropy": 5.769148826599121,
      "epoch": 3.042482007391558,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004079323933085836,
      "loss": 4.9477,
      "mean_token_accuracy": 0.20873139202594757,
      "num_tokens": 67822460.0,
      "step": 39105
    },
    {
      "entropy": 5.816865062713623,
      "epoch": 3.0428710367632754,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040791003107972976,
      "loss": 5.0464,
      "mean_token_accuracy": 0.2106510728597641,
      "num_tokens": 67830833.0,
      "step": 39110
    },
    {
      "entropy": 5.741622066497802,
      "epoch": 3.043260066134993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004078876668341778,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21480145901441575,
      "num_tokens": 67839009.0,
      "step": 39115
    },
    {
      "entropy": 5.748734474182129,
      "epoch": 3.043649095506711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004078653005722669,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2128509223461151,
      "num_tokens": 67847589.0,
      "step": 39120
    },
    {
      "entropy": 5.72815899848938,
      "epoch": 3.0440381248784285,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004078429322943366,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2067059949040413,
      "num_tokens": 67856516.0,
      "step": 39125
    },
    {
      "entropy": 5.745333528518676,
      "epoch": 3.0444271542501458,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040782056200072627,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2099539279937744,
      "num_tokens": 67864877.0,
      "step": 39130
    },
    {
      "entropy": 5.730934095382691,
      "epoch": 3.0448161836218635,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040779818969177527,
      "loss": 4.9225,
      "mean_token_accuracy": 0.20815183818340302,
      "num_tokens": 67873437.0,
      "step": 39135
    },
    {
      "entropy": 5.812160301208496,
      "epoch": 3.045205212993581,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040777581536782315,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21979974061250687,
      "num_tokens": 67882222.0,
      "step": 39140
    },
    {
      "entropy": 5.725315904617309,
      "epoch": 3.0455942423652984,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004077534390292093,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21306637674570084,
      "num_tokens": 67890886.0,
      "step": 39145
    },
    {
      "entropy": 5.764017820358276,
      "epoch": 3.045983271737016,
      "grad_norm": 1.25,
      "learning_rate": 0.00040773106067627335,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21176708489656448,
      "num_tokens": 67899239.0,
      "step": 39150
    },
    {
      "entropy": 5.785585546493531,
      "epoch": 3.046372301108734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040770868030935486,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2149684578180313,
      "num_tokens": 67907679.0,
      "step": 39155
    },
    {
      "entropy": 5.771771049499511,
      "epoch": 3.046761330480451,
      "grad_norm": 1.375,
      "learning_rate": 0.00040768629792879344,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2076895222067833,
      "num_tokens": 67916437.0,
      "step": 39160
    },
    {
      "entropy": 5.752957105636597,
      "epoch": 3.0471503598521688,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004076639135349284,
      "loss": 4.964,
      "mean_token_accuracy": 0.2052147164940834,
      "num_tokens": 67924961.0,
      "step": 39165
    },
    {
      "entropy": 5.7660839557647705,
      "epoch": 3.0475393892238865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040764152712809983,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2262889638543129,
      "num_tokens": 67933854.0,
      "step": 39170
    },
    {
      "entropy": 5.732032632827758,
      "epoch": 3.047928418595604,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004076191387086472,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2152099698781967,
      "num_tokens": 67942904.0,
      "step": 39175
    },
    {
      "entropy": 5.761520338058472,
      "epoch": 3.0483174479673214,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040759674827691014,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2117391973733902,
      "num_tokens": 67951861.0,
      "step": 39180
    },
    {
      "entropy": 5.761766386032105,
      "epoch": 3.048706477339039,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004075743558332285,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21096070408821105,
      "num_tokens": 67960644.0,
      "step": 39185
    },
    {
      "entropy": 5.815056562423706,
      "epoch": 3.049095506710757,
      "grad_norm": 1.234375,
      "learning_rate": 0.000407551961377942,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20504214018583297,
      "num_tokens": 67969290.0,
      "step": 39190
    },
    {
      "entropy": 5.6987669467926025,
      "epoch": 3.049484536082474,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004075295649113905,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2137141615152359,
      "num_tokens": 67978305.0,
      "step": 39195
    },
    {
      "entropy": 5.801343536376953,
      "epoch": 3.0498735654541917,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040750716643391384,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2107463374733925,
      "num_tokens": 67986861.0,
      "step": 39200
    },
    {
      "entropy": 5.805468416213989,
      "epoch": 3.0502625948259094,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004074847659458518,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21327392607927323,
      "num_tokens": 67996003.0,
      "step": 39205
    },
    {
      "entropy": 5.704204988479614,
      "epoch": 3.050651624197627,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040746236344754423,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2238614171743393,
      "num_tokens": 68004398.0,
      "step": 39210
    },
    {
      "entropy": 5.712455177307129,
      "epoch": 3.0510406535693444,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004074399589393312,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21066251546144485,
      "num_tokens": 68013386.0,
      "step": 39215
    },
    {
      "entropy": 5.7345024108886715,
      "epoch": 3.051429682941062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004074175524215526,
      "loss": 4.9274,
      "mean_token_accuracy": 0.21313873827457427,
      "num_tokens": 68022541.0,
      "step": 39220
    },
    {
      "entropy": 5.7668561935424805,
      "epoch": 3.05181871231278,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004073951438945484,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21823297441005707,
      "num_tokens": 68031690.0,
      "step": 39225
    },
    {
      "entropy": 5.80838532447815,
      "epoch": 3.052207741684497,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040737273335865866,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20372162163257598,
      "num_tokens": 68040863.0,
      "step": 39230
    },
    {
      "entropy": 5.8584551334381105,
      "epoch": 3.0525967710562147,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040735032081422334,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20700934529304504,
      "num_tokens": 68049066.0,
      "step": 39235
    },
    {
      "entropy": 5.832453966140747,
      "epoch": 3.0529858004279324,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040732790626158257,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21329862624406815,
      "num_tokens": 68057791.0,
      "step": 39240
    },
    {
      "entropy": 5.7638932228088375,
      "epoch": 3.0533748297996497,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040730548970107646,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2072589322924614,
      "num_tokens": 68066587.0,
      "step": 39245
    },
    {
      "entropy": 5.793990182876587,
      "epoch": 3.0537638591713674,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004072830711330451,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21428415626287461,
      "num_tokens": 68074952.0,
      "step": 39250
    },
    {
      "entropy": 5.7308127880096436,
      "epoch": 3.054152888543085,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004072606505578287,
      "loss": 4.963,
      "mean_token_accuracy": 0.2153785228729248,
      "num_tokens": 68084091.0,
      "step": 39255
    },
    {
      "entropy": 5.7604108333587645,
      "epoch": 3.054541917914803,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040723822797576737,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21375963389873504,
      "num_tokens": 68091968.0,
      "step": 39260
    },
    {
      "entropy": 5.751478052139282,
      "epoch": 3.05493094728652,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004072158033872015,
      "loss": 4.9955,
      "mean_token_accuracy": 0.1976011410355568,
      "num_tokens": 68100925.0,
      "step": 39265
    },
    {
      "entropy": 5.768335294723511,
      "epoch": 3.0553199766582377,
      "grad_norm": 1.25,
      "learning_rate": 0.0004071933767924713,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21290189623832703,
      "num_tokens": 68109518.0,
      "step": 39270
    },
    {
      "entropy": 5.7704774856567385,
      "epoch": 3.0557090060299554,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040717094819191685,
      "loss": 4.9836,
      "mean_token_accuracy": 0.20619669407606125,
      "num_tokens": 68118261.0,
      "step": 39275
    },
    {
      "entropy": 5.761379766464233,
      "epoch": 3.0560980354016727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040714851758587875,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21849191784858704,
      "num_tokens": 68126919.0,
      "step": 39280
    },
    {
      "entropy": 5.765096235275268,
      "epoch": 3.0564870647733904,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040712608497469715,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21525786221027374,
      "num_tokens": 68135157.0,
      "step": 39285
    },
    {
      "entropy": 5.82550277709961,
      "epoch": 3.056876094145108,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004071036503587126,
      "loss": 5.1522,
      "mean_token_accuracy": 0.19806270599365233,
      "num_tokens": 68145043.0,
      "step": 39290
    },
    {
      "entropy": 5.763582324981689,
      "epoch": 3.0572651235168253,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004070812137382654,
      "loss": 4.8751,
      "mean_token_accuracy": 0.21207141876220703,
      "num_tokens": 68154637.0,
      "step": 39295
    },
    {
      "entropy": 5.7778510570526125,
      "epoch": 3.057654152888543,
      "grad_norm": 1.28125,
      "learning_rate": 0.000407058775113696,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21008999794721603,
      "num_tokens": 68163511.0,
      "step": 39300
    },
    {
      "entropy": 5.750403881072998,
      "epoch": 3.0580431822602607,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004070363344853449,
      "loss": 4.9375,
      "mean_token_accuracy": 0.20594925284385682,
      "num_tokens": 68172108.0,
      "step": 39305
    },
    {
      "entropy": 5.717688846588135,
      "epoch": 3.0584322116319784,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040701389185355256,
      "loss": 4.8598,
      "mean_token_accuracy": 0.21840772330760955,
      "num_tokens": 68180937.0,
      "step": 39310
    },
    {
      "entropy": 5.717485618591309,
      "epoch": 3.0588212410036957,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004069914472186595,
      "loss": 4.8042,
      "mean_token_accuracy": 0.22489549964666367,
      "num_tokens": 68189588.0,
      "step": 39315
    },
    {
      "entropy": 5.732551908493042,
      "epoch": 3.0592102703754134,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004069690005810063,
      "loss": 4.8282,
      "mean_token_accuracy": 0.22416229099035262,
      "num_tokens": 68198051.0,
      "step": 39320
    },
    {
      "entropy": 5.691645717620849,
      "epoch": 3.059599299747131,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040694655194093365,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21503009647130966,
      "num_tokens": 68205967.0,
      "step": 39325
    },
    {
      "entropy": 5.777645063400269,
      "epoch": 3.0599883291188483,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040692410129878205,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2034858375787735,
      "num_tokens": 68213773.0,
      "step": 39330
    },
    {
      "entropy": 5.719740581512451,
      "epoch": 3.060377358490566,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004069016486548922,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21305294930934907,
      "num_tokens": 68222361.0,
      "step": 39335
    },
    {
      "entropy": 5.776933908462524,
      "epoch": 3.0607663878622837,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004068791940096048,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2090112641453743,
      "num_tokens": 68230797.0,
      "step": 39340
    },
    {
      "entropy": 5.771874284744262,
      "epoch": 3.061155417234001,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040685673736326044,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21527090817689895,
      "num_tokens": 68239147.0,
      "step": 39345
    },
    {
      "entropy": 5.723786163330078,
      "epoch": 3.0615444466057187,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004068342787162001,
      "loss": 4.8822,
      "mean_token_accuracy": 0.213077574968338,
      "num_tokens": 68247988.0,
      "step": 39350
    },
    {
      "entropy": 5.7855321884155275,
      "epoch": 3.0619334759774364,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040681181806876435,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20819674730300902,
      "num_tokens": 68257024.0,
      "step": 39355
    },
    {
      "entropy": 5.789840221405029,
      "epoch": 3.062322505349154,
      "grad_norm": 1.40625,
      "learning_rate": 0.000406789355421294,
      "loss": 4.9234,
      "mean_token_accuracy": 0.20716043710708618,
      "num_tokens": 68265875.0,
      "step": 39360
    },
    {
      "entropy": 5.734454679489136,
      "epoch": 3.0627115347208713,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004067668907741301,
      "loss": 4.98,
      "mean_token_accuracy": 0.20469937175512315,
      "num_tokens": 68274104.0,
      "step": 39365
    },
    {
      "entropy": 5.748278427124023,
      "epoch": 3.063100564092589,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004067444241276132,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21483258455991744,
      "num_tokens": 68282831.0,
      "step": 39370
    },
    {
      "entropy": 5.786056184768677,
      "epoch": 3.0634895934643067,
      "grad_norm": 1.4375,
      "learning_rate": 0.00040672195548208443,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20488976091146469,
      "num_tokens": 68291062.0,
      "step": 39375
    },
    {
      "entropy": 5.780897188186645,
      "epoch": 3.063878622836024,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004066994848378846,
      "loss": 4.9343,
      "mean_token_accuracy": 0.20997628718614578,
      "num_tokens": 68300490.0,
      "step": 39380
    },
    {
      "entropy": 5.713163757324219,
      "epoch": 3.0642676522077417,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040667701219535473,
      "loss": 4.8246,
      "mean_token_accuracy": 0.21923534870147704,
      "num_tokens": 68309451.0,
      "step": 39385
    },
    {
      "entropy": 5.774997568130493,
      "epoch": 3.0646566815794594,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004066545375548358,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21101301163434982,
      "num_tokens": 68317822.0,
      "step": 39390
    },
    {
      "entropy": 5.769029951095581,
      "epoch": 3.0650457109511766,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004066320609166689,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20862487256526946,
      "num_tokens": 68326732.0,
      "step": 39395
    },
    {
      "entropy": 5.864382553100586,
      "epoch": 3.0654347403228943,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004066095822811949,
      "loss": 5.0808,
      "mean_token_accuracy": 0.19868389517068863,
      "num_tokens": 68335082.0,
      "step": 39400
    },
    {
      "entropy": 5.782797908782959,
      "epoch": 3.065823769694612,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000406587101648755,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21629954725503922,
      "num_tokens": 68344331.0,
      "step": 39405
    },
    {
      "entropy": 5.734326553344727,
      "epoch": 3.0662127990663297,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004065646190196902,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20621538013219834,
      "num_tokens": 68353204.0,
      "step": 39410
    },
    {
      "entropy": 5.76788215637207,
      "epoch": 3.066601828438047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004065421343943417,
      "loss": 4.9611,
      "mean_token_accuracy": 0.20831728130578994,
      "num_tokens": 68362125.0,
      "step": 39415
    },
    {
      "entropy": 5.755284929275513,
      "epoch": 3.0669908578097647,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040651964777305074,
      "loss": 4.7834,
      "mean_token_accuracy": 0.21968006193637848,
      "num_tokens": 68370916.0,
      "step": 39420
    },
    {
      "entropy": 5.753920269012451,
      "epoch": 3.0673798871814824,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004064971591561585,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21481657028198242,
      "num_tokens": 68379955.0,
      "step": 39425
    },
    {
      "entropy": 5.67555341720581,
      "epoch": 3.0677689165531996,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040647466854400606,
      "loss": 4.8259,
      "mean_token_accuracy": 0.21728423684835435,
      "num_tokens": 68388698.0,
      "step": 39430
    },
    {
      "entropy": 5.776620817184448,
      "epoch": 3.0681579459249173,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004064521759369349,
      "loss": 4.9991,
      "mean_token_accuracy": 0.20366449654102325,
      "num_tokens": 68398304.0,
      "step": 39435
    },
    {
      "entropy": 5.69122462272644,
      "epoch": 3.068546975296635,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040642968133528613,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22800374031066895,
      "num_tokens": 68406605.0,
      "step": 39440
    },
    {
      "entropy": 5.789189863204956,
      "epoch": 3.0689360046683523,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004064071847394012,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21518452763557433,
      "num_tokens": 68415621.0,
      "step": 39445
    },
    {
      "entropy": 5.7972753047943115,
      "epoch": 3.06932503404007,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004063846861496213,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21743236631155013,
      "num_tokens": 68424571.0,
      "step": 39450
    },
    {
      "entropy": 5.711740207672119,
      "epoch": 3.0697140634117877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000406362185566288,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21436226963996888,
      "num_tokens": 68433584.0,
      "step": 39455
    },
    {
      "entropy": 5.702483177185059,
      "epoch": 3.0701030927835053,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004063396829897426,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21232127100229264,
      "num_tokens": 68441801.0,
      "step": 39460
    },
    {
      "entropy": 5.716058015823364,
      "epoch": 3.0704921221552226,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004063171784203265,
      "loss": 4.8526,
      "mean_token_accuracy": 0.21208134442567825,
      "num_tokens": 68449838.0,
      "step": 39465
    },
    {
      "entropy": 5.680799055099487,
      "epoch": 3.0708811515269403,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004062946718583813,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2202177032828331,
      "num_tokens": 68458335.0,
      "step": 39470
    },
    {
      "entropy": 5.65205659866333,
      "epoch": 3.071270180898658,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040627216330424844,
      "loss": 4.898,
      "mean_token_accuracy": 0.22518338561058043,
      "num_tokens": 68466851.0,
      "step": 39475
    },
    {
      "entropy": 5.680207872390747,
      "epoch": 3.0716592102703753,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004062496527582694,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2032962843775749,
      "num_tokens": 68475220.0,
      "step": 39480
    },
    {
      "entropy": 5.759002017974853,
      "epoch": 3.072048239642093,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004062271402207858,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2181516781449318,
      "num_tokens": 68484190.0,
      "step": 39485
    },
    {
      "entropy": 5.839335012435913,
      "epoch": 3.0724372690138106,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004062046256921391,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20833667367696762,
      "num_tokens": 68493102.0,
      "step": 39490
    },
    {
      "entropy": 5.7938965320587155,
      "epoch": 3.072826298385528,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040618210917267117,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21431828737258912,
      "num_tokens": 68501650.0,
      "step": 39495
    },
    {
      "entropy": 5.710832405090332,
      "epoch": 3.0732153277572456,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040615959066272355,
      "loss": 4.8148,
      "mean_token_accuracy": 0.22192294001579285,
      "num_tokens": 68509487.0,
      "step": 39500
    },
    {
      "entropy": 5.734316492080689,
      "epoch": 3.0736043571289633,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004061370701626379,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21724549382925035,
      "num_tokens": 68517644.0,
      "step": 39505
    },
    {
      "entropy": 5.765217208862305,
      "epoch": 3.073993386500681,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004061145476727559,
      "loss": 4.9154,
      "mean_token_accuracy": 0.21897059977054595,
      "num_tokens": 68526217.0,
      "step": 39510
    },
    {
      "entropy": 5.7855064392089846,
      "epoch": 3.0743824158723982,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004060920231934193,
      "loss": 4.8944,
      "mean_token_accuracy": 0.21419646888971328,
      "num_tokens": 68534778.0,
      "step": 39515
    },
    {
      "entropy": 5.819291400909424,
      "epoch": 3.074771445244116,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004060694967249699,
      "loss": 4.9952,
      "mean_token_accuracy": 0.20562713146209716,
      "num_tokens": 68543043.0,
      "step": 39520
    },
    {
      "entropy": 5.76445894241333,
      "epoch": 3.0751604746158336,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004060469682677496,
      "loss": 4.9429,
      "mean_token_accuracy": 0.20815117955207824,
      "num_tokens": 68551391.0,
      "step": 39525
    },
    {
      "entropy": 5.7462221622467045,
      "epoch": 3.075549503987551,
      "grad_norm": 1.375,
      "learning_rate": 0.00040602443782210003,
      "loss": 4.9241,
      "mean_token_accuracy": 0.212738299369812,
      "num_tokens": 68560251.0,
      "step": 39530
    },
    {
      "entropy": 5.761011791229248,
      "epoch": 3.0759385333592686,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004060019053883632,
      "loss": 4.9133,
      "mean_token_accuracy": 0.20744238942861556,
      "num_tokens": 68568974.0,
      "step": 39535
    },
    {
      "entropy": 5.714836597442627,
      "epoch": 3.0763275627309863,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000405979370966881,
      "loss": 4.8007,
      "mean_token_accuracy": 0.21903826743364335,
      "num_tokens": 68577280.0,
      "step": 39540
    },
    {
      "entropy": 5.735162353515625,
      "epoch": 3.0767165921027035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004059568345579953,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20670411586761475,
      "num_tokens": 68586102.0,
      "step": 39545
    },
    {
      "entropy": 5.692688417434693,
      "epoch": 3.0771056214744212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040593429616204806,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21331096738576888,
      "num_tokens": 68594420.0,
      "step": 39550
    },
    {
      "entropy": 5.865401697158814,
      "epoch": 3.077494650846139,
      "grad_norm": 1.25,
      "learning_rate": 0.00040591175577938125,
      "loss": 5.0346,
      "mean_token_accuracy": 0.20152995437383653,
      "num_tokens": 68603760.0,
      "step": 39555
    },
    {
      "entropy": 5.85372838973999,
      "epoch": 3.0778836802178566,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040588921341033697,
      "loss": 4.9996,
      "mean_token_accuracy": 0.20572633892297745,
      "num_tokens": 68612039.0,
      "step": 39560
    },
    {
      "entropy": 5.782596683502197,
      "epoch": 3.078272709589574,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004058666690552572,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20787141472101212,
      "num_tokens": 68620452.0,
      "step": 39565
    },
    {
      "entropy": 5.805132341384888,
      "epoch": 3.0786617389612916,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040584412271448407,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2142111137509346,
      "num_tokens": 68628526.0,
      "step": 39570
    },
    {
      "entropy": 5.775401639938354,
      "epoch": 3.0790507683330093,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004058215743883595,
      "loss": 4.8885,
      "mean_token_accuracy": 0.21455609947443008,
      "num_tokens": 68636342.0,
      "step": 39575
    },
    {
      "entropy": 5.731519889831543,
      "epoch": 3.0794397977047265,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004057990240772259,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22391967326402665,
      "num_tokens": 68645333.0,
      "step": 39580
    },
    {
      "entropy": 5.801797533035279,
      "epoch": 3.0798288270764442,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040577647178142523,
      "loss": 4.9638,
      "mean_token_accuracy": 0.21127763390541077,
      "num_tokens": 68653114.0,
      "step": 39585
    },
    {
      "entropy": 5.767221832275391,
      "epoch": 3.080217856448162,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004057539175012998,
      "loss": 4.9374,
      "mean_token_accuracy": 0.210533107817173,
      "num_tokens": 68661789.0,
      "step": 39590
    },
    {
      "entropy": 5.766764163970947,
      "epoch": 3.080606885819879,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040573136123719174,
      "loss": 4.9321,
      "mean_token_accuracy": 0.21174464523792266,
      "num_tokens": 68670481.0,
      "step": 39595
    },
    {
      "entropy": 5.819529008865357,
      "epoch": 3.080995915191597,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040570880298944334,
      "loss": 4.972,
      "mean_token_accuracy": 0.2124832347035408,
      "num_tokens": 68679270.0,
      "step": 39600
    },
    {
      "entropy": 5.801679372787476,
      "epoch": 3.0813849445633146,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004056862427583969,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20382715463638307,
      "num_tokens": 68688301.0,
      "step": 39605
    },
    {
      "entropy": 5.729549789428711,
      "epoch": 3.0817739739350323,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004056636805443948,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2161804050207138,
      "num_tokens": 68696382.0,
      "step": 39610
    },
    {
      "entropy": 5.699559593200684,
      "epoch": 3.0821630033067495,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040564111634777927,
      "loss": 4.8229,
      "mean_token_accuracy": 0.22478315234184265,
      "num_tokens": 68705552.0,
      "step": 39615
    },
    {
      "entropy": 5.814018774032593,
      "epoch": 3.0825520326784672,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040561855016889276,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20430726557970047,
      "num_tokens": 68714125.0,
      "step": 39620
    },
    {
      "entropy": 5.74969449043274,
      "epoch": 3.082941062050185,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040559598200807763,
      "loss": 4.8223,
      "mean_token_accuracy": 0.21761805266141893,
      "num_tokens": 68722119.0,
      "step": 39625
    },
    {
      "entropy": 5.803784656524658,
      "epoch": 3.083330091421902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004055734118656764,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21198455542325972,
      "num_tokens": 68730745.0,
      "step": 39630
    },
    {
      "entropy": 5.823920869827271,
      "epoch": 3.08371912079362,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004055508397420314,
      "loss": 5.0454,
      "mean_token_accuracy": 0.20396374613046647,
      "num_tokens": 68739747.0,
      "step": 39635
    },
    {
      "entropy": 5.7832857131958,
      "epoch": 3.0841081501653376,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040552826563748526,
      "loss": 4.868,
      "mean_token_accuracy": 0.21842247992753983,
      "num_tokens": 68748703.0,
      "step": 39640
    },
    {
      "entropy": 5.777443742752075,
      "epoch": 3.0844971795370553,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004055056895523804,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22148414701223373,
      "num_tokens": 68757412.0,
      "step": 39645
    },
    {
      "entropy": 5.737195587158203,
      "epoch": 3.0848862089087725,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040548311148705943,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20679867416620254,
      "num_tokens": 68765680.0,
      "step": 39650
    },
    {
      "entropy": 5.808868169784546,
      "epoch": 3.08527523828049,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004054605314418649,
      "loss": 4.98,
      "mean_token_accuracy": 0.2045988067984581,
      "num_tokens": 68773931.0,
      "step": 39655
    },
    {
      "entropy": 5.801799392700195,
      "epoch": 3.085664267652208,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040543794941713944,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2091333508491516,
      "num_tokens": 68781854.0,
      "step": 39660
    },
    {
      "entropy": 5.735847854614258,
      "epoch": 3.086053297023925,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004054153654132258,
      "loss": 4.9571,
      "mean_token_accuracy": 0.20761697739362717,
      "num_tokens": 68790384.0,
      "step": 39665
    },
    {
      "entropy": 5.7697385311126705,
      "epoch": 3.086442326395643,
      "grad_norm": 1.25,
      "learning_rate": 0.0004053927794304665,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20599237233400344,
      "num_tokens": 68798899.0,
      "step": 39670
    },
    {
      "entropy": 5.768698740005493,
      "epoch": 3.0868313557673606,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004053701914692043,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21541758328676225,
      "num_tokens": 68807567.0,
      "step": 39675
    },
    {
      "entropy": 5.772577381134033,
      "epoch": 3.087220385139078,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000405347601529782,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2147700846195221,
      "num_tokens": 68816039.0,
      "step": 39680
    },
    {
      "entropy": 5.732383060455322,
      "epoch": 3.0876094145107955,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004053250096125423,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21486999839544296,
      "num_tokens": 68825185.0,
      "step": 39685
    },
    {
      "entropy": 5.7795319080352785,
      "epoch": 3.087998443882513,
      "grad_norm": 1.375,
      "learning_rate": 0.0004053024157178281,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20260193347930908,
      "num_tokens": 68833757.0,
      "step": 39690
    },
    {
      "entropy": 5.726060438156128,
      "epoch": 3.088387473254231,
      "grad_norm": 1.375,
      "learning_rate": 0.000405279819845982,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22234226763248444,
      "num_tokens": 68843155.0,
      "step": 39695
    },
    {
      "entropy": 5.785486030578613,
      "epoch": 3.088776502625948,
      "grad_norm": 1.25,
      "learning_rate": 0.00040525722199734706,
      "loss": 4.9069,
      "mean_token_accuracy": 0.21966517716646194,
      "num_tokens": 68852183.0,
      "step": 39700
    },
    {
      "entropy": 5.679362869262695,
      "epoch": 3.089165531997666,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004052346221722661,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22701964080333709,
      "num_tokens": 68861003.0,
      "step": 39705
    },
    {
      "entropy": 5.752509498596192,
      "epoch": 3.0895545613693836,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040521202037108205,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21492766737937927,
      "num_tokens": 68869657.0,
      "step": 39710
    },
    {
      "entropy": 5.724474096298218,
      "epoch": 3.089943590741101,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004051894165941378,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22146259993314743,
      "num_tokens": 68878747.0,
      "step": 39715
    },
    {
      "entropy": 5.74120945930481,
      "epoch": 3.0903326201128185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040516681084177636,
      "loss": 4.8208,
      "mean_token_accuracy": 0.22383075803518296,
      "num_tokens": 68887460.0,
      "step": 39720
    },
    {
      "entropy": 5.787602186203003,
      "epoch": 3.090721649484536,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040514420311434085,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2022153690457344,
      "num_tokens": 68896129.0,
      "step": 39725
    },
    {
      "entropy": 5.810035943984985,
      "epoch": 3.0911106788562535,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040512159341217415,
      "loss": 4.907,
      "mean_token_accuracy": 0.21366390734910964,
      "num_tokens": 68904235.0,
      "step": 39730
    },
    {
      "entropy": 5.7596197605133055,
      "epoch": 3.091499708227971,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040509898173561934,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21626016497612,
      "num_tokens": 68912161.0,
      "step": 39735
    },
    {
      "entropy": 5.6932618618011475,
      "epoch": 3.091888737599689,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004050763680850196,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2093872532248497,
      "num_tokens": 68920316.0,
      "step": 39740
    },
    {
      "entropy": 5.652742910385132,
      "epoch": 3.0922777669714065,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004050537524607179,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2189144864678383,
      "num_tokens": 68929374.0,
      "step": 39745
    },
    {
      "entropy": 5.6921648502349855,
      "epoch": 3.092666796343124,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040503113486305766,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2253242090344429,
      "num_tokens": 68936997.0,
      "step": 39750
    },
    {
      "entropy": 5.731439161300659,
      "epoch": 3.0930558257148415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040500851529238184,
      "loss": 4.8369,
      "mean_token_accuracy": 0.2232477769255638,
      "num_tokens": 68945773.0,
      "step": 39755
    },
    {
      "entropy": 5.747562026977539,
      "epoch": 3.093444855086559,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004049858937490337,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21028582453727723,
      "num_tokens": 68954226.0,
      "step": 39760
    },
    {
      "entropy": 5.7482147216796875,
      "epoch": 3.0938338844582765,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004049632702333566,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21002336889505385,
      "num_tokens": 68962510.0,
      "step": 39765
    },
    {
      "entropy": 5.823799753189087,
      "epoch": 3.094222913829994,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040494064474569375,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20402777940034866,
      "num_tokens": 68971531.0,
      "step": 39770
    },
    {
      "entropy": 5.821825456619263,
      "epoch": 3.094611943201712,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040491801728638835,
      "loss": 5.0062,
      "mean_token_accuracy": 0.21250158846378325,
      "num_tokens": 68980217.0,
      "step": 39775
    },
    {
      "entropy": 5.756355237960816,
      "epoch": 3.095000972573429,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040489538785578385,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20372197180986404,
      "num_tokens": 68989243.0,
      "step": 39780
    },
    {
      "entropy": 5.798519897460937,
      "epoch": 3.095390001945147,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040487275645422356,
      "loss": 4.954,
      "mean_token_accuracy": 0.20905472487211227,
      "num_tokens": 68997102.0,
      "step": 39785
    },
    {
      "entropy": 5.771003293991089,
      "epoch": 3.0957790313168645,
      "grad_norm": 1.390625,
      "learning_rate": 0.000404850123082051,
      "loss": 4.9637,
      "mean_token_accuracy": 0.20962756723165513,
      "num_tokens": 69006230.0,
      "step": 39790
    },
    {
      "entropy": 5.850861501693726,
      "epoch": 3.096168060688582,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004048274877396094,
      "loss": 5.009,
      "mean_token_accuracy": 0.2072341784834862,
      "num_tokens": 69015303.0,
      "step": 39795
    },
    {
      "entropy": 5.835467147827148,
      "epoch": 3.0965570900602994,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004048048504272424,
      "loss": 4.8409,
      "mean_token_accuracy": 0.21624854058027268,
      "num_tokens": 69023434.0,
      "step": 39800
    },
    {
      "entropy": 5.837031364440918,
      "epoch": 3.096946119432017,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004047822111452933,
      "loss": 5.0601,
      "mean_token_accuracy": 0.19899210929870606,
      "num_tokens": 69031694.0,
      "step": 39805
    },
    {
      "entropy": 5.725315713882447,
      "epoch": 3.097335148803735,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040475956989410584,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21067645996809006,
      "num_tokens": 69040191.0,
      "step": 39810
    },
    {
      "entropy": 5.750631952285767,
      "epoch": 3.097724178175452,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004047369266740233,
      "loss": 5.0059,
      "mean_token_accuracy": 0.20805785655975342,
      "num_tokens": 69049277.0,
      "step": 39815
    },
    {
      "entropy": 5.76277117729187,
      "epoch": 3.09811320754717,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040471428148538944,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2096048802137375,
      "num_tokens": 69057483.0,
      "step": 39820
    },
    {
      "entropy": 5.751685047149659,
      "epoch": 3.0985022369188875,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004046916343285479,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2135115757584572,
      "num_tokens": 69066055.0,
      "step": 39825
    },
    {
      "entropy": 5.778376770019531,
      "epoch": 3.0988912662906047,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040466898520384213,
      "loss": 5.0135,
      "mean_token_accuracy": 0.20746229887008666,
      "num_tokens": 69074774.0,
      "step": 39830
    },
    {
      "entropy": 5.7578864097595215,
      "epoch": 3.0992802956623224,
      "grad_norm": 1.296875,
      "learning_rate": 0.000404646334111616,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20425669401884078,
      "num_tokens": 69083307.0,
      "step": 39835
    },
    {
      "entropy": 5.76624493598938,
      "epoch": 3.09966932503404,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004046236810522131,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2219396010041237,
      "num_tokens": 69091528.0,
      "step": 39840
    },
    {
      "entropy": 5.683811473846435,
      "epoch": 3.100058354405758,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004046010260259771,
      "loss": 4.875,
      "mean_token_accuracy": 0.21084967404603958,
      "num_tokens": 69100562.0,
      "step": 39845
    },
    {
      "entropy": 5.807650423049926,
      "epoch": 3.100447383777475,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040457836903325177,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20812997221946716,
      "num_tokens": 69108629.0,
      "step": 39850
    },
    {
      "entropy": 5.7626721382141115,
      "epoch": 3.100836413149193,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040455571007438104,
      "loss": 4.8129,
      "mean_token_accuracy": 0.21863297373056412,
      "num_tokens": 69117694.0,
      "step": 39855
    },
    {
      "entropy": 5.773023557662964,
      "epoch": 3.1012254425209105,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004045330491497085,
      "loss": 4.9976,
      "mean_token_accuracy": 0.19849037230014802,
      "num_tokens": 69126318.0,
      "step": 39860
    },
    {
      "entropy": 5.758947277069092,
      "epoch": 3.1016144718926277,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040451038625957827,
      "loss": 4.9191,
      "mean_token_accuracy": 0.20600378662347793,
      "num_tokens": 69135041.0,
      "step": 39865
    },
    {
      "entropy": 5.7290044784545895,
      "epoch": 3.1020035012643454,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004044877214043339,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2137392982840538,
      "num_tokens": 69143906.0,
      "step": 39870
    },
    {
      "entropy": 5.735186529159546,
      "epoch": 3.102392530636063,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004044650545843196,
      "loss": 4.8138,
      "mean_token_accuracy": 0.2151723623275757,
      "num_tokens": 69152204.0,
      "step": 39875
    },
    {
      "entropy": 5.711070394515991,
      "epoch": 3.1027815600077804,
      "grad_norm": 1.359375,
      "learning_rate": 0.000404442385799879,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21677784621715546,
      "num_tokens": 69161097.0,
      "step": 39880
    },
    {
      "entropy": 5.711878776550293,
      "epoch": 3.103170589379498,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004044197150513564,
      "loss": 4.913,
      "mean_token_accuracy": 0.20440458357334138,
      "num_tokens": 69169545.0,
      "step": 39885
    },
    {
      "entropy": 5.780318641662598,
      "epoch": 3.1035596187512158,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004043970423390955,
      "loss": 5.0165,
      "mean_token_accuracy": 0.19606248438358306,
      "num_tokens": 69178188.0,
      "step": 39890
    },
    {
      "entropy": 5.816859388351441,
      "epoch": 3.1039486481229335,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004043743676634405,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21090450137853622,
      "num_tokens": 69187361.0,
      "step": 39895
    },
    {
      "entropy": 5.7644054889678955,
      "epoch": 3.1043376774946507,
      "grad_norm": 1.25,
      "learning_rate": 0.0004043516910247353,
      "loss": 4.8545,
      "mean_token_accuracy": 0.22059257924556733,
      "num_tokens": 69196115.0,
      "step": 39900
    },
    {
      "entropy": 5.68728609085083,
      "epoch": 3.1047267068663684,
      "grad_norm": 1.25,
      "learning_rate": 0.0004043290124233242,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22214387059211732,
      "num_tokens": 69204937.0,
      "step": 39905
    },
    {
      "entropy": 5.683556747436524,
      "epoch": 3.105115736238086,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004043063318595511,
      "loss": 4.8697,
      "mean_token_accuracy": 0.21619418859481812,
      "num_tokens": 69213067.0,
      "step": 39910
    },
    {
      "entropy": 5.722344779968262,
      "epoch": 3.1055047656098034,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004042836493337603,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20884758234024048,
      "num_tokens": 69221069.0,
      "step": 39915
    },
    {
      "entropy": 5.711345052719116,
      "epoch": 3.105893794981521,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004042609648462959,
      "loss": 4.8807,
      "mean_token_accuracy": 0.21115671396255492,
      "num_tokens": 69229269.0,
      "step": 39920
    },
    {
      "entropy": 5.777154111862183,
      "epoch": 3.1062828243532388,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000404238278397502,
      "loss": 4.9712,
      "mean_token_accuracy": 0.20809506326913835,
      "num_tokens": 69237782.0,
      "step": 39925
    },
    {
      "entropy": 5.76377501487732,
      "epoch": 3.106671853724956,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040421558998772293,
      "loss": 4.8774,
      "mean_token_accuracy": 0.21217077374458312,
      "num_tokens": 69245673.0,
      "step": 39930
    },
    {
      "entropy": 5.817785310745239,
      "epoch": 3.1070608830966737,
      "grad_norm": 1.1875,
      "learning_rate": 0.000404192899617303,
      "loss": 5.0355,
      "mean_token_accuracy": 0.2041858285665512,
      "num_tokens": 69254161.0,
      "step": 39935
    },
    {
      "entropy": 5.731550788879394,
      "epoch": 3.1074499124683914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004041702072865864,
      "loss": 4.9016,
      "mean_token_accuracy": 0.20450989454984664,
      "num_tokens": 69263053.0,
      "step": 39940
    },
    {
      "entropy": 5.8307147979736325,
      "epoch": 3.107838941840109,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004041475129959176,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20861730426549913,
      "num_tokens": 69271016.0,
      "step": 39945
    },
    {
      "entropy": 5.8108643054962155,
      "epoch": 3.1082279712118264,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004041248167456408,
      "loss": 4.9082,
      "mean_token_accuracy": 0.20639184713363648,
      "num_tokens": 69279238.0,
      "step": 39950
    },
    {
      "entropy": 5.778442239761352,
      "epoch": 3.108617000583544,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004041021185361004,
      "loss": 5.02,
      "mean_token_accuracy": 0.2051006719470024,
      "num_tokens": 69288599.0,
      "step": 39955
    },
    {
      "entropy": 5.730287647247314,
      "epoch": 3.1090060299552618,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004040794183676409,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21562488824129106,
      "num_tokens": 69297255.0,
      "step": 39960
    },
    {
      "entropy": 5.70685043334961,
      "epoch": 3.109395059326979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004040567162406066,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22491381466388702,
      "num_tokens": 69306252.0,
      "step": 39965
    },
    {
      "entropy": 5.712982654571533,
      "epoch": 3.1097840886986967,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004040340121553421,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2170789659023285,
      "num_tokens": 69314513.0,
      "step": 39970
    },
    {
      "entropy": 5.734579133987427,
      "epoch": 3.1101731180704144,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004040113061121918,
      "loss": 4.9798,
      "mean_token_accuracy": 0.20679136216640473,
      "num_tokens": 69323537.0,
      "step": 39975
    },
    {
      "entropy": 5.769730186462402,
      "epoch": 3.110562147442132,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040398859811150036,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21523208916187286,
      "num_tokens": 69331677.0,
      "step": 39980
    },
    {
      "entropy": 5.777107810974121,
      "epoch": 3.1109511768138494,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004039658881536122,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2144148275256157,
      "num_tokens": 69340317.0,
      "step": 39985
    },
    {
      "entropy": 5.751975774765015,
      "epoch": 3.111340206185567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004039431762388719,
      "loss": 4.8826,
      "mean_token_accuracy": 0.21805571466684343,
      "num_tokens": 69349412.0,
      "step": 39990
    },
    {
      "entropy": 5.777456331253052,
      "epoch": 3.1117292355572848,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004039204623676242,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20605419874191283,
      "num_tokens": 69357717.0,
      "step": 39995
    },
    {
      "entropy": 5.728319311141968,
      "epoch": 3.112118264929002,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040389774654021367,
      "loss": 4.9376,
      "mean_token_accuracy": 0.21150212585926056,
      "num_tokens": 69365858.0,
      "step": 40000
    },
    {
      "entropy": 5.826995992660523,
      "epoch": 3.1125072943007197,
      "grad_norm": 1.5625,
      "learning_rate": 0.000403875028756985,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20935767590999604,
      "num_tokens": 69374535.0,
      "step": 40005
    },
    {
      "entropy": 5.808182907104492,
      "epoch": 3.1128963236724374,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004038523090182829,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20947306603193283,
      "num_tokens": 69383051.0,
      "step": 40010
    },
    {
      "entropy": 5.838730335235596,
      "epoch": 3.1132853530441547,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000403829587324452,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19739585667848586,
      "num_tokens": 69390876.0,
      "step": 40015
    },
    {
      "entropy": 5.850389289855957,
      "epoch": 3.1136743824158724,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004038068636758373,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21325714588165284,
      "num_tokens": 69399887.0,
      "step": 40020
    },
    {
      "entropy": 5.710236215591431,
      "epoch": 3.11406341178759,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004037841380727834,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20272131115198136,
      "num_tokens": 69407859.0,
      "step": 40025
    },
    {
      "entropy": 5.7235105514526365,
      "epoch": 3.1144524411593073,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004037614105156352,
      "loss": 4.9444,
      "mean_token_accuracy": 0.20799121558666228,
      "num_tokens": 69416157.0,
      "step": 40030
    },
    {
      "entropy": 5.883564233779907,
      "epoch": 3.114841470531025,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004037386810047376,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20723860263824462,
      "num_tokens": 69423784.0,
      "step": 40035
    },
    {
      "entropy": 5.745909070968628,
      "epoch": 3.1152304999027427,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004037159495404354,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21909126490354539,
      "num_tokens": 69432358.0,
      "step": 40040
    },
    {
      "entropy": 5.744107818603515,
      "epoch": 3.1156195292744604,
      "grad_norm": 1.375,
      "learning_rate": 0.00040369321612307343,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21131176948547364,
      "num_tokens": 69440173.0,
      "step": 40045
    },
    {
      "entropy": 5.855760192871093,
      "epoch": 3.1160085586461777,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040367048075299684,
      "loss": 5.1037,
      "mean_token_accuracy": 0.19977132678031922,
      "num_tokens": 69449302.0,
      "step": 40050
    },
    {
      "entropy": 5.849479484558105,
      "epoch": 3.1163975880178953,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004036477434305505,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21182374358177186,
      "num_tokens": 69458284.0,
      "step": 40055
    },
    {
      "entropy": 5.7925519943237305,
      "epoch": 3.116786617389613,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004036250041560794,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21010270118713378,
      "num_tokens": 69466960.0,
      "step": 40060
    },
    {
      "entropy": 5.688681364059448,
      "epoch": 3.1171756467613303,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004036022629299286,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22311098426580428,
      "num_tokens": 69476096.0,
      "step": 40065
    },
    {
      "entropy": 5.748182344436645,
      "epoch": 3.117564676133048,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004035795197524431,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2214929446578026,
      "num_tokens": 69484754.0,
      "step": 40070
    },
    {
      "entropy": 5.773387002944946,
      "epoch": 3.1179537055047657,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004035567746239681,
      "loss": 4.9465,
      "mean_token_accuracy": 0.20690796077251433,
      "num_tokens": 69494149.0,
      "step": 40075
    },
    {
      "entropy": 5.796993970870972,
      "epoch": 3.1183427348764834,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040353402754484866,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21049665212631224,
      "num_tokens": 69502833.0,
      "step": 40080
    },
    {
      "entropy": 5.751757526397705,
      "epoch": 3.1187317642482006,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004035112785154299,
      "loss": 4.8804,
      "mean_token_accuracy": 0.21231265962123871,
      "num_tokens": 69510995.0,
      "step": 40085
    },
    {
      "entropy": 5.7674877643585205,
      "epoch": 3.1191207936199183,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040348852753605707,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21192436814308166,
      "num_tokens": 69519126.0,
      "step": 40090
    },
    {
      "entropy": 5.772055625915527,
      "epoch": 3.119509822991636,
      "grad_norm": 1.25,
      "learning_rate": 0.00040346577460707535,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21388096809387208,
      "num_tokens": 69527621.0,
      "step": 40095
    },
    {
      "entropy": 5.794259834289551,
      "epoch": 3.1198988523633533,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040344301972882997,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21082728654146193,
      "num_tokens": 69536549.0,
      "step": 40100
    },
    {
      "entropy": 5.803535842895508,
      "epoch": 3.120287881735071,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004034202629016662,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2132548451423645,
      "num_tokens": 69546148.0,
      "step": 40105
    },
    {
      "entropy": 5.67745189666748,
      "epoch": 3.1206769111067887,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040339750412592937,
      "loss": 4.985,
      "mean_token_accuracy": 0.20287290066480637,
      "num_tokens": 69555875.0,
      "step": 40110
    },
    {
      "entropy": 5.784997606277466,
      "epoch": 3.121065940478506,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004033747434019647,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20895988643169403,
      "num_tokens": 69564561.0,
      "step": 40115
    },
    {
      "entropy": 5.777000141143799,
      "epoch": 3.1214549698502236,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040335198073011765,
      "loss": 4.9734,
      "mean_token_accuracy": 0.20545304715633392,
      "num_tokens": 69573122.0,
      "step": 40120
    },
    {
      "entropy": 5.822106981277466,
      "epoch": 3.1218439992219413,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004033292161107336,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2115643173456192,
      "num_tokens": 69582229.0,
      "step": 40125
    },
    {
      "entropy": 5.747476959228516,
      "epoch": 3.122233028593659,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040330644954415795,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2123159259557724,
      "num_tokens": 69591055.0,
      "step": 40130
    },
    {
      "entropy": 5.784395599365235,
      "epoch": 3.1226220579653763,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004032836810307362,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21085905730724336,
      "num_tokens": 69599342.0,
      "step": 40135
    },
    {
      "entropy": 5.82467827796936,
      "epoch": 3.123011087337094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004032609105708137,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20202109068632126,
      "num_tokens": 69609011.0,
      "step": 40140
    },
    {
      "entropy": 5.774012851715088,
      "epoch": 3.1234001167088117,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040323813816473604,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21122611314058304,
      "num_tokens": 69618167.0,
      "step": 40145
    },
    {
      "entropy": 5.851735830307007,
      "epoch": 3.123789146080529,
      "grad_norm": 1.25,
      "learning_rate": 0.0004032153638128488,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21043165773153305,
      "num_tokens": 69627109.0,
      "step": 40150
    },
    {
      "entropy": 5.688071632385254,
      "epoch": 3.1241781754522466,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040319258751549743,
      "loss": 4.7898,
      "mean_token_accuracy": 0.2140660509467125,
      "num_tokens": 69635806.0,
      "step": 40155
    },
    {
      "entropy": 5.687273120880127,
      "epoch": 3.1245672048239643,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040316980927302765,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21587437987327576,
      "num_tokens": 69644437.0,
      "step": 40160
    },
    {
      "entropy": 5.8111467361450195,
      "epoch": 3.1249562341956816,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040314702908578496,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20002802908420564,
      "num_tokens": 69653747.0,
      "step": 40165
    },
    {
      "entropy": 5.777410840988159,
      "epoch": 3.1253452635673993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004031242469541151,
      "loss": 4.8733,
      "mean_token_accuracy": 0.20874358862638473,
      "num_tokens": 69662291.0,
      "step": 40170
    },
    {
      "entropy": 5.7479095458984375,
      "epoch": 3.125734292939117,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040310146287836373,
      "loss": 4.8919,
      "mean_token_accuracy": 0.21335157752037048,
      "num_tokens": 69670680.0,
      "step": 40175
    },
    {
      "entropy": 5.694083547592163,
      "epoch": 3.1261233223108347,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040307867685887653,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2189897119998932,
      "num_tokens": 69679351.0,
      "step": 40180
    },
    {
      "entropy": 5.724062299728393,
      "epoch": 3.126512351682552,
      "grad_norm": 1.375,
      "learning_rate": 0.00040305588889599927,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21566104143857956,
      "num_tokens": 69687678.0,
      "step": 40185
    },
    {
      "entropy": 5.770704507827759,
      "epoch": 3.1269013810542696,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004030330989900777,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21533187478780746,
      "num_tokens": 69696940.0,
      "step": 40190
    },
    {
      "entropy": 5.695532131195068,
      "epoch": 3.1272904104259873,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040301030714145764,
      "loss": 4.904,
      "mean_token_accuracy": 0.20965037047863005,
      "num_tokens": 69706220.0,
      "step": 40195
    },
    {
      "entropy": 5.803440570831299,
      "epoch": 3.1276794397977046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000402987513350485,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2105756014585495,
      "num_tokens": 69715163.0,
      "step": 40200
    },
    {
      "entropy": 5.788340663909912,
      "epoch": 3.1280684691694223,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004029647176175055,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2197499841451645,
      "num_tokens": 69723677.0,
      "step": 40205
    },
    {
      "entropy": 5.695401573181153,
      "epoch": 3.12845749854114,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040294191994286513,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21331340670585633,
      "num_tokens": 69732483.0,
      "step": 40210
    },
    {
      "entropy": 5.6678542137146,
      "epoch": 3.1288465279128572,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040291912032690963,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2228936403989792,
      "num_tokens": 69740844.0,
      "step": 40215
    },
    {
      "entropy": 5.743181467056274,
      "epoch": 3.129235557284575,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004028963187699852,
      "loss": 4.907,
      "mean_token_accuracy": 0.21609598845243455,
      "num_tokens": 69749426.0,
      "step": 40220
    },
    {
      "entropy": 5.790025854110718,
      "epoch": 3.1296245866562926,
      "grad_norm": 1.25,
      "learning_rate": 0.0004028735152724377,
      "loss": 5.0022,
      "mean_token_accuracy": 0.1997210755944252,
      "num_tokens": 69758581.0,
      "step": 40225
    },
    {
      "entropy": 5.76169753074646,
      "epoch": 3.1300136160280103,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040285070983461313,
      "loss": 5.0197,
      "mean_token_accuracy": 0.20568263232707978,
      "num_tokens": 69767444.0,
      "step": 40230
    },
    {
      "entropy": 5.740632581710815,
      "epoch": 3.1304026453997276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040282790245685753,
      "loss": 4.8957,
      "mean_token_accuracy": 0.20835064649581908,
      "num_tokens": 69776726.0,
      "step": 40235
    },
    {
      "entropy": 5.795553493499756,
      "epoch": 3.1307916747714453,
      "grad_norm": 1.34375,
      "learning_rate": 0.000402805093139517,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2087833270430565,
      "num_tokens": 69785341.0,
      "step": 40240
    },
    {
      "entropy": 5.679021787643433,
      "epoch": 3.131180704143163,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004027822818829375,
      "loss": 4.8477,
      "mean_token_accuracy": 0.21551962643861772,
      "num_tokens": 69793515.0,
      "step": 40245
    },
    {
      "entropy": 5.727423810958863,
      "epoch": 3.13156973351488,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004027594686874653,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2085217460989952,
      "num_tokens": 69801861.0,
      "step": 40250
    },
    {
      "entropy": 5.740774488449096,
      "epoch": 3.131958762886598,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040273665355344655,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22003409117460251,
      "num_tokens": 69811662.0,
      "step": 40255
    },
    {
      "entropy": 5.718981409072876,
      "epoch": 3.1323477922583156,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004027138364812274,
      "loss": 4.8568,
      "mean_token_accuracy": 0.21883145570755005,
      "num_tokens": 69820107.0,
      "step": 40260
    },
    {
      "entropy": 5.844197463989258,
      "epoch": 3.132736821630033,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004026910174711541,
      "loss": 5.0335,
      "mean_token_accuracy": 0.19858429133892058,
      "num_tokens": 69829279.0,
      "step": 40265
    },
    {
      "entropy": 5.8277181625366214,
      "epoch": 3.1331258510017506,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040266819652357266,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2079933390021324,
      "num_tokens": 69837649.0,
      "step": 40270
    },
    {
      "entropy": 5.789802122116089,
      "epoch": 3.1335148803734683,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040264537363882976,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21123535186052322,
      "num_tokens": 69845557.0,
      "step": 40275
    },
    {
      "entropy": 5.73231725692749,
      "epoch": 3.133903909745186,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040262254881727135,
      "loss": 4.8784,
      "mean_token_accuracy": 0.2229197785258293,
      "num_tokens": 69855854.0,
      "step": 40280
    },
    {
      "entropy": 5.776639842987061,
      "epoch": 3.134292939116903,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040259972205924395,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21475390195846558,
      "num_tokens": 69864420.0,
      "step": 40285
    },
    {
      "entropy": 5.791183042526245,
      "epoch": 3.134681968488621,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040257689336509385,
      "loss": 4.944,
      "mean_token_accuracy": 0.21120575964450836,
      "num_tokens": 69873399.0,
      "step": 40290
    },
    {
      "entropy": 5.80324969291687,
      "epoch": 3.1350709978603386,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004025540627351675,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20177256762981416,
      "num_tokens": 69882908.0,
      "step": 40295
    },
    {
      "entropy": 5.760149192810059,
      "epoch": 3.135460027232056,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004025312301698112,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20189143121242523,
      "num_tokens": 69892603.0,
      "step": 40300
    },
    {
      "entropy": 5.7924097061157225,
      "epoch": 3.1358490566037736,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004025083956693715,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21391555070877075,
      "num_tokens": 69901008.0,
      "step": 40305
    },
    {
      "entropy": 5.902125453948974,
      "epoch": 3.1362380859754913,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004024855592341948,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20328429043293,
      "num_tokens": 69909726.0,
      "step": 40310
    },
    {
      "entropy": 5.713063144683838,
      "epoch": 3.136627115347209,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004024627208646277,
      "loss": 4.9114,
      "mean_token_accuracy": 0.219138403236866,
      "num_tokens": 69919073.0,
      "step": 40315
    },
    {
      "entropy": 5.7250877857208256,
      "epoch": 3.137016144718926,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00040243988056101667,
      "loss": 4.9272,
      "mean_token_accuracy": 0.20709774494171143,
      "num_tokens": 69927627.0,
      "step": 40320
    },
    {
      "entropy": 5.770431661605835,
      "epoch": 3.137405174090644,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004024170383237083,
      "loss": 4.947,
      "mean_token_accuracy": 0.21090989112854003,
      "num_tokens": 69935842.0,
      "step": 40325
    },
    {
      "entropy": 5.783553695678711,
      "epoch": 3.1377942034623616,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004023941941530492,
      "loss": 4.9756,
      "mean_token_accuracy": 0.2062254801392555,
      "num_tokens": 69944530.0,
      "step": 40330
    },
    {
      "entropy": 5.6843461990356445,
      "epoch": 3.138183232834079,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000402371348049386,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2142156332731247,
      "num_tokens": 69952908.0,
      "step": 40335
    },
    {
      "entropy": 5.850854110717774,
      "epoch": 3.1385722622057965,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004023485000130653,
      "loss": 5.049,
      "mean_token_accuracy": 0.1994370773434639,
      "num_tokens": 69960905.0,
      "step": 40340
    },
    {
      "entropy": 5.793711423873901,
      "epoch": 3.1389612915775142,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040232565004443373,
      "loss": 4.9558,
      "mean_token_accuracy": 0.2108398362994194,
      "num_tokens": 69969648.0,
      "step": 40345
    },
    {
      "entropy": 5.717907238006592,
      "epoch": 3.1393503209492315,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004023027981438382,
      "loss": 4.8545,
      "mean_token_accuracy": 0.21686673611402513,
      "num_tokens": 69978006.0,
      "step": 40350
    },
    {
      "entropy": 5.739992141723633,
      "epoch": 3.139739350320949,
      "grad_norm": 1.25,
      "learning_rate": 0.00040227994431162517,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21215471476316453,
      "num_tokens": 69986862.0,
      "step": 40355
    },
    {
      "entropy": 5.756698513031006,
      "epoch": 3.140128379692667,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040225708854814174,
      "loss": 4.9512,
      "mean_token_accuracy": 0.20926265716552733,
      "num_tokens": 69996066.0,
      "step": 40360
    },
    {
      "entropy": 5.769553184509277,
      "epoch": 3.140517409064384,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004022342308537345,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2225148320198059,
      "num_tokens": 70004394.0,
      "step": 40365
    },
    {
      "entropy": 5.784255886077881,
      "epoch": 3.140906438436102,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004022113712287503,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20053702145814895,
      "num_tokens": 70014298.0,
      "step": 40370
    },
    {
      "entropy": 5.803610134124756,
      "epoch": 3.1412954678078195,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000402188509673536,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2066754400730133,
      "num_tokens": 70022006.0,
      "step": 40375
    },
    {
      "entropy": 5.851156902313233,
      "epoch": 3.1416844971795372,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040216564618843854,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2005156770348549,
      "num_tokens": 70031100.0,
      "step": 40380
    },
    {
      "entropy": 5.85844407081604,
      "epoch": 3.1420735265512545,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004021427807738048,
      "loss": 4.9992,
      "mean_token_accuracy": 0.20618006139993666,
      "num_tokens": 70039830.0,
      "step": 40385
    },
    {
      "entropy": 5.679624891281128,
      "epoch": 3.142462555922972,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004021199134299817,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21236151307821274,
      "num_tokens": 70048187.0,
      "step": 40390
    },
    {
      "entropy": 5.828287172317505,
      "epoch": 3.14285158529469,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040209704415731623,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20655807256698608,
      "num_tokens": 70056432.0,
      "step": 40395
    },
    {
      "entropy": 5.8425600051879885,
      "epoch": 3.143240614666407,
      "grad_norm": 1.375,
      "learning_rate": 0.00040207417295615546,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20687542110681534,
      "num_tokens": 70065720.0,
      "step": 40400
    },
    {
      "entropy": 5.813409948348999,
      "epoch": 3.143629644038125,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040205129982684635,
      "loss": 4.9132,
      "mean_token_accuracy": 0.21038702130317688,
      "num_tokens": 70074030.0,
      "step": 40405
    },
    {
      "entropy": 5.788703870773316,
      "epoch": 3.1440186734098425,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040202842476973606,
      "loss": 4.93,
      "mean_token_accuracy": 0.2102228358387947,
      "num_tokens": 70082703.0,
      "step": 40410
    },
    {
      "entropy": 5.700043392181397,
      "epoch": 3.1444077027815602,
      "grad_norm": 1.359375,
      "learning_rate": 0.00040200554778517154,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22261289656162261,
      "num_tokens": 70091221.0,
      "step": 40415
    },
    {
      "entropy": 5.763620471954345,
      "epoch": 3.1447967321532775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040198266887349994,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20276841819286345,
      "num_tokens": 70100197.0,
      "step": 40420
    },
    {
      "entropy": 5.789965057373047,
      "epoch": 3.145185761524995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040195978803506853,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2111675336956978,
      "num_tokens": 70108759.0,
      "step": 40425
    },
    {
      "entropy": 5.81311993598938,
      "epoch": 3.145574790896713,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040193690527022443,
      "loss": 4.9649,
      "mean_token_accuracy": 0.20730699151754378,
      "num_tokens": 70117881.0,
      "step": 40430
    },
    {
      "entropy": 5.765024662017822,
      "epoch": 3.14596382026843,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040191402057931474,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21044069826602935,
      "num_tokens": 70127700.0,
      "step": 40435
    },
    {
      "entropy": 5.75165696144104,
      "epoch": 3.146352849640148,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004018911339626869,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2049643650650978,
      "num_tokens": 70136234.0,
      "step": 40440
    },
    {
      "entropy": 5.74533257484436,
      "epoch": 3.1467418790118655,
      "grad_norm": 1.265625,
      "learning_rate": 0.000401868245420688,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21151627004146575,
      "num_tokens": 70144940.0,
      "step": 40445
    },
    {
      "entropy": 5.772161912918091,
      "epoch": 3.147130908383583,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040184535495366543,
      "loss": 4.9118,
      "mean_token_accuracy": 0.21127618998289108,
      "num_tokens": 70154628.0,
      "step": 40450
    },
    {
      "entropy": 5.812704563140869,
      "epoch": 3.1475199377553005,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004018224625619664,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2111609622836113,
      "num_tokens": 70163207.0,
      "step": 40455
    },
    {
      "entropy": 5.735611534118652,
      "epoch": 3.147908967127018,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040179956824593847,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2170082852244377,
      "num_tokens": 70172103.0,
      "step": 40460
    },
    {
      "entropy": 5.87116265296936,
      "epoch": 3.1482979964987354,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040177667200592894,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20277344435453415,
      "num_tokens": 70180646.0,
      "step": 40465
    },
    {
      "entropy": 5.784483098983765,
      "epoch": 3.148687025870453,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004017537738422852,
      "loss": 4.911,
      "mean_token_accuracy": 0.21171634942293166,
      "num_tokens": 70188614.0,
      "step": 40470
    },
    {
      "entropy": 5.737367296218872,
      "epoch": 3.149076055242171,
      "grad_norm": 1.25,
      "learning_rate": 0.00040173087375535457,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21053340584039687,
      "num_tokens": 70196959.0,
      "step": 40475
    },
    {
      "entropy": 5.767529439926148,
      "epoch": 3.1494650846138885,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040170797174548476,
      "loss": 4.9843,
      "mean_token_accuracy": 0.208919021487236,
      "num_tokens": 70205763.0,
      "step": 40480
    },
    {
      "entropy": 5.7736616134643555,
      "epoch": 3.1498541139856058,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040168506781302303,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2041158750653267,
      "num_tokens": 70214586.0,
      "step": 40485
    },
    {
      "entropy": 5.759898519515991,
      "epoch": 3.1502431433573235,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004016621619583171,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2169592037796974,
      "num_tokens": 70223334.0,
      "step": 40490
    },
    {
      "entropy": 5.735036182403564,
      "epoch": 3.150632172729041,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040163925418171454,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21174300462007523,
      "num_tokens": 70232395.0,
      "step": 40495
    },
    {
      "entropy": 5.791286039352417,
      "epoch": 3.1510212021007584,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004016163444835628,
      "loss": 4.9683,
      "mean_token_accuracy": 0.20765730887651443,
      "num_tokens": 70241658.0,
      "step": 40500
    },
    {
      "entropy": 5.706120109558105,
      "epoch": 3.151410231472476,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004015934328642096,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21409650892019272,
      "num_tokens": 70250468.0,
      "step": 40505
    },
    {
      "entropy": 5.732951688766479,
      "epoch": 3.151799260844194,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040157051932400244,
      "loss": 4.9007,
      "mean_token_accuracy": 0.20881550759077072,
      "num_tokens": 70259440.0,
      "step": 40510
    },
    {
      "entropy": 5.794725465774536,
      "epoch": 3.1521882902159115,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004015476038632892,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21785697489976882,
      "num_tokens": 70268848.0,
      "step": 40515
    },
    {
      "entropy": 5.869899797439575,
      "epoch": 3.1525773195876288,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004015246864824175,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20374394953250885,
      "num_tokens": 70278470.0,
      "step": 40520
    },
    {
      "entropy": 5.787311649322509,
      "epoch": 3.1529663489593465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040150176718173504,
      "loss": 4.9877,
      "mean_token_accuracy": 0.20901406705379486,
      "num_tokens": 70288712.0,
      "step": 40525
    },
    {
      "entropy": 5.808365726470948,
      "epoch": 3.153355378331064,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004014788459615896,
      "loss": 4.9682,
      "mean_token_accuracy": 0.20850373059511185,
      "num_tokens": 70297848.0,
      "step": 40530
    },
    {
      "entropy": 5.82039852142334,
      "epoch": 3.1537444077027814,
      "grad_norm": 1.25,
      "learning_rate": 0.0004014559228223291,
      "loss": 4.9537,
      "mean_token_accuracy": 0.20202360302209854,
      "num_tokens": 70306313.0,
      "step": 40535
    },
    {
      "entropy": 5.744287300109863,
      "epoch": 3.154133437074499,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040143299776430115,
      "loss": 4.8944,
      "mean_token_accuracy": 0.21507159769535064,
      "num_tokens": 70315002.0,
      "step": 40540
    },
    {
      "entropy": 5.745880508422852,
      "epoch": 3.154522466446217,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004014100707878536,
      "loss": 4.929,
      "mean_token_accuracy": 0.21719609946012497,
      "num_tokens": 70323878.0,
      "step": 40545
    },
    {
      "entropy": 5.778616189956665,
      "epoch": 3.154911495817934,
      "grad_norm": 1.25,
      "learning_rate": 0.00040138714189333455,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2140916809439659,
      "num_tokens": 70332592.0,
      "step": 40550
    },
    {
      "entropy": 5.767165279388427,
      "epoch": 3.1553005251896518,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040136421108109176,
      "loss": 4.8963,
      "mean_token_accuracy": 0.20752975791692735,
      "num_tokens": 70340057.0,
      "step": 40555
    },
    {
      "entropy": 5.743940353393555,
      "epoch": 3.1556895545613695,
      "grad_norm": 1.25,
      "learning_rate": 0.0004013412783514732,
      "loss": 4.947,
      "mean_token_accuracy": 0.20901575386524202,
      "num_tokens": 70348501.0,
      "step": 40560
    },
    {
      "entropy": 5.781447839736939,
      "epoch": 3.156078583933087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004013183437048268,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2125971570611,
      "num_tokens": 70358110.0,
      "step": 40565
    },
    {
      "entropy": 5.800413608551025,
      "epoch": 3.1564676133048044,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040129540714150066,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20313268154859543,
      "num_tokens": 70366683.0,
      "step": 40570
    },
    {
      "entropy": 5.829809951782226,
      "epoch": 3.156856642676522,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004012724686618427,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2037683606147766,
      "num_tokens": 70375418.0,
      "step": 40575
    },
    {
      "entropy": 5.821298885345459,
      "epoch": 3.15724567204824,
      "grad_norm": 1.328125,
      "learning_rate": 0.000401249528266201,
      "loss": 4.9799,
      "mean_token_accuracy": 0.20957402437925338,
      "num_tokens": 70384168.0,
      "step": 40580
    },
    {
      "entropy": 5.841653299331665,
      "epoch": 3.157634701419957,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040122658595492366,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2039206475019455,
      "num_tokens": 70392304.0,
      "step": 40585
    },
    {
      "entropy": 5.7030939102172855,
      "epoch": 3.1580237307916748,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004012036417283588,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21099811941385269,
      "num_tokens": 70401186.0,
      "step": 40590
    },
    {
      "entropy": 5.680010175704956,
      "epoch": 3.1584127601633925,
      "grad_norm": 1.25,
      "learning_rate": 0.00040118069558685456,
      "loss": 4.7865,
      "mean_token_accuracy": 0.223158797621727,
      "num_tokens": 70409304.0,
      "step": 40595
    },
    {
      "entropy": 5.786808872222901,
      "epoch": 3.1588017895351097,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040115774753075905,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2181577280163765,
      "num_tokens": 70417206.0,
      "step": 40600
    },
    {
      "entropy": 5.714131927490234,
      "epoch": 3.1591908189068274,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004011347975604206,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21172747611999512,
      "num_tokens": 70425573.0,
      "step": 40605
    },
    {
      "entropy": 5.761755800247192,
      "epoch": 3.159579848278545,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004011118456761873,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21218641102313995,
      "num_tokens": 70434415.0,
      "step": 40610
    },
    {
      "entropy": 5.82722430229187,
      "epoch": 3.159968877650263,
      "grad_norm": 1.25,
      "learning_rate": 0.0004010888918784075,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2189266249537468,
      "num_tokens": 70443033.0,
      "step": 40615
    },
    {
      "entropy": 5.762420606613159,
      "epoch": 3.16035790702198,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004010659361674294,
      "loss": 4.942,
      "mean_token_accuracy": 0.21072487831115722,
      "num_tokens": 70451740.0,
      "step": 40620
    },
    {
      "entropy": 5.838847351074219,
      "epoch": 3.1607469363936977,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040104297854360145,
      "loss": 5.022,
      "mean_token_accuracy": 0.20983533263206483,
      "num_tokens": 70459893.0,
      "step": 40625
    },
    {
      "entropy": 5.815574836730957,
      "epoch": 3.1611359657654154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004010200190072719,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21154980063438417,
      "num_tokens": 70468852.0,
      "step": 40630
    },
    {
      "entropy": 5.797278022766113,
      "epoch": 3.1615249951371327,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004009970575587891,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20951933711767196,
      "num_tokens": 70477418.0,
      "step": 40635
    },
    {
      "entropy": 5.836968851089478,
      "epoch": 3.1619140245088504,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004009740941985016,
      "loss": 5.0137,
      "mean_token_accuracy": 0.20803424566984177,
      "num_tokens": 70486368.0,
      "step": 40640
    },
    {
      "entropy": 5.7701764583587645,
      "epoch": 3.162303053880568,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004009511289267576,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20609866082668304,
      "num_tokens": 70495097.0,
      "step": 40645
    },
    {
      "entropy": 5.773818111419677,
      "epoch": 3.1626920832522853,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004009281617439058,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21256339848041533,
      "num_tokens": 70503684.0,
      "step": 40650
    },
    {
      "entropy": 5.859715986251831,
      "epoch": 3.163081112624003,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004009051926502945,
      "loss": 4.9724,
      "mean_token_accuracy": 0.20651036947965623,
      "num_tokens": 70511917.0,
      "step": 40655
    },
    {
      "entropy": 5.706477928161621,
      "epoch": 3.1634701419957207,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040088222164627233,
      "loss": 4.8733,
      "mean_token_accuracy": 0.21568609029054642,
      "num_tokens": 70521074.0,
      "step": 40660
    },
    {
      "entropy": 5.883972930908203,
      "epoch": 3.1638591713674384,
      "grad_norm": 1.359375,
      "learning_rate": 0.00040085924873218783,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20678069442510605,
      "num_tokens": 70530261.0,
      "step": 40665
    },
    {
      "entropy": 5.774124050140381,
      "epoch": 3.1642482007391557,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004008362739083895,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21366457343101503,
      "num_tokens": 70539125.0,
      "step": 40670
    },
    {
      "entropy": 5.766443824768066,
      "epoch": 3.1646372301108734,
      "grad_norm": 1.28125,
      "learning_rate": 0.000400813297175226,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2173941671848297,
      "num_tokens": 70547420.0,
      "step": 40675
    },
    {
      "entropy": 5.72861967086792,
      "epoch": 3.165026259482591,
      "grad_norm": 1.375,
      "learning_rate": 0.000400790318533046,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21090103387832643,
      "num_tokens": 70556307.0,
      "step": 40680
    },
    {
      "entropy": 5.746558952331543,
      "epoch": 3.1654152888543083,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040076733798219817,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21188623011112212,
      "num_tokens": 70565275.0,
      "step": 40685
    },
    {
      "entropy": 5.924573564529419,
      "epoch": 3.165804318226026,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004007443555230312,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20019012540578843,
      "num_tokens": 70574038.0,
      "step": 40690
    },
    {
      "entropy": 5.834783601760864,
      "epoch": 3.1661933475977437,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040072137115589366,
      "loss": 4.9962,
      "mean_token_accuracy": 0.20957378298044205,
      "num_tokens": 70582118.0,
      "step": 40695
    },
    {
      "entropy": 5.746265983581543,
      "epoch": 3.166582376969461,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004006983848811345,
      "loss": 4.8774,
      "mean_token_accuracy": 0.21890666484832763,
      "num_tokens": 70590569.0,
      "step": 40700
    },
    {
      "entropy": 5.714414501190186,
      "epoch": 3.1669714063411787,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040067539669910235,
      "loss": 4.943,
      "mean_token_accuracy": 0.21356316655874252,
      "num_tokens": 70599518.0,
      "step": 40705
    },
    {
      "entropy": 5.686338996887207,
      "epoch": 3.1673604357128964,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004006524066101461,
      "loss": 4.8605,
      "mean_token_accuracy": 0.21602620482444762,
      "num_tokens": 70607638.0,
      "step": 40710
    },
    {
      "entropy": 5.773846817016602,
      "epoch": 3.167749465084614,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004006294146146146,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20743745863437651,
      "num_tokens": 70616260.0,
      "step": 40715
    },
    {
      "entropy": 5.831448078155518,
      "epoch": 3.1681384944563313,
      "grad_norm": 1.25,
      "learning_rate": 0.0004006064207128567,
      "loss": 4.9366,
      "mean_token_accuracy": 0.20551228672266006,
      "num_tokens": 70624654.0,
      "step": 40720
    },
    {
      "entropy": 5.821302461624145,
      "epoch": 3.168527523828049,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040058342490522134,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20598124116659164,
      "num_tokens": 70633141.0,
      "step": 40725
    },
    {
      "entropy": 5.767334985733032,
      "epoch": 3.1689165531997667,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004005604271920573,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2146398216485977,
      "num_tokens": 70641701.0,
      "step": 40730
    },
    {
      "entropy": 5.775250101089478,
      "epoch": 3.169305582571484,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040053742757371365,
      "loss": 4.9341,
      "mean_token_accuracy": 0.20857871621847152,
      "num_tokens": 70650566.0,
      "step": 40735
    },
    {
      "entropy": 5.761979579925537,
      "epoch": 3.1696946119432017,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004005144260505394,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21092050820589064,
      "num_tokens": 70660127.0,
      "step": 40740
    },
    {
      "entropy": 5.839926862716675,
      "epoch": 3.1700836413149194,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004004914226228834,
      "loss": 5.0167,
      "mean_token_accuracy": 0.20587404519319535,
      "num_tokens": 70669324.0,
      "step": 40745
    },
    {
      "entropy": 5.804857778549194,
      "epoch": 3.170472670686637,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004004684172910949,
      "loss": 4.957,
      "mean_token_accuracy": 0.2088740274310112,
      "num_tokens": 70678590.0,
      "step": 40750
    },
    {
      "entropy": 5.822592401504517,
      "epoch": 3.1708617000583543,
      "grad_norm": 1.25,
      "learning_rate": 0.0004004454100555229,
      "loss": 4.985,
      "mean_token_accuracy": 0.20957597345113754,
      "num_tokens": 70687348.0,
      "step": 40755
    },
    {
      "entropy": 5.735289144515991,
      "epoch": 3.171250729430072,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040042240091651645,
      "loss": 4.8156,
      "mean_token_accuracy": 0.22941847741603852,
      "num_tokens": 70695066.0,
      "step": 40760
    },
    {
      "entropy": 5.760897207260132,
      "epoch": 3.1716397588017897,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040039938987442464,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21118153035640716,
      "num_tokens": 70703026.0,
      "step": 40765
    },
    {
      "entropy": 5.79872579574585,
      "epoch": 3.172028788173507,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040037637692959667,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20884792059659957,
      "num_tokens": 70712046.0,
      "step": 40770
    },
    {
      "entropy": 5.791887044906616,
      "epoch": 3.1724178175452247,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040035336208238176,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2116324558854103,
      "num_tokens": 70720860.0,
      "step": 40775
    },
    {
      "entropy": 5.775736665725708,
      "epoch": 3.1728068469169424,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040033034533312913,
      "loss": 4.8298,
      "mean_token_accuracy": 0.21768372356891633,
      "num_tokens": 70729805.0,
      "step": 40780
    },
    {
      "entropy": 5.766955280303955,
      "epoch": 3.1731958762886596,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000400307326682188,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20681175738573074,
      "num_tokens": 70738960.0,
      "step": 40785
    },
    {
      "entropy": 5.811338138580322,
      "epoch": 3.1735849056603773,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040028430612990755,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21198581606149675,
      "num_tokens": 70747715.0,
      "step": 40790
    },
    {
      "entropy": 5.816431856155395,
      "epoch": 3.173973935032095,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040026128367663726,
      "loss": 4.9253,
      "mean_token_accuracy": 0.20980679243803024,
      "num_tokens": 70756573.0,
      "step": 40795
    },
    {
      "entropy": 5.849886560440064,
      "epoch": 3.1743629644038123,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040023825932272635,
      "loss": 5.1007,
      "mean_token_accuracy": 0.19771260470151902,
      "num_tokens": 70766161.0,
      "step": 40800
    },
    {
      "entropy": 5.798144292831421,
      "epoch": 3.17475199377553,
      "grad_norm": 1.25,
      "learning_rate": 0.00040021523306852416,
      "loss": 4.958,
      "mean_token_accuracy": 0.2085500717163086,
      "num_tokens": 70775252.0,
      "step": 40805
    },
    {
      "entropy": 5.800560998916626,
      "epoch": 3.1751410231472477,
      "grad_norm": 1.3125,
      "learning_rate": 0.00040019220491438007,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20991758704185487,
      "num_tokens": 70783429.0,
      "step": 40810
    },
    {
      "entropy": 5.741214942932129,
      "epoch": 3.1755300525189654,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004001691748606436,
      "loss": 4.917,
      "mean_token_accuracy": 0.21538470983505248,
      "num_tokens": 70792008.0,
      "step": 40815
    },
    {
      "entropy": 5.751021814346314,
      "epoch": 3.1759190818906826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004001461429076641,
      "loss": 4.8905,
      "mean_token_accuracy": 0.21819397509098054,
      "num_tokens": 70800190.0,
      "step": 40820
    },
    {
      "entropy": 5.692185926437378,
      "epoch": 3.1763081112624003,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004001231090557912,
      "loss": 4.8147,
      "mean_token_accuracy": 0.22185861468315124,
      "num_tokens": 70808148.0,
      "step": 40825
    },
    {
      "entropy": 5.826910495758057,
      "epoch": 3.176697140634118,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040010007330537405,
      "loss": 5.1216,
      "mean_token_accuracy": 0.19750313758850097,
      "num_tokens": 70817848.0,
      "step": 40830
    },
    {
      "entropy": 5.803548812866211,
      "epoch": 3.1770861700058353,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040007703565676263,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20751919001340866,
      "num_tokens": 70826737.0,
      "step": 40835
    },
    {
      "entropy": 5.801681280136108,
      "epoch": 3.177475199377553,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040005399611030615,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21708332002162933,
      "num_tokens": 70835478.0,
      "step": 40840
    },
    {
      "entropy": 5.81149377822876,
      "epoch": 3.1778642287492707,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004000309546663543,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21244437992572784,
      "num_tokens": 70844752.0,
      "step": 40845
    },
    {
      "entropy": 5.712144136428833,
      "epoch": 3.1782532581209884,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040000791132525676,
      "loss": 4.913,
      "mean_token_accuracy": 0.20900585651397705,
      "num_tokens": 70852634.0,
      "step": 40850
    },
    {
      "entropy": 5.804977226257324,
      "epoch": 3.1786422874927056,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039998486608736305,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21162612736225128,
      "num_tokens": 70860225.0,
      "step": 40855
    },
    {
      "entropy": 5.784864377975464,
      "epoch": 3.1790313168644233,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003999618189530231,
      "loss": 4.9473,
      "mean_token_accuracy": 0.20207290053367616,
      "num_tokens": 70868320.0,
      "step": 40860
    },
    {
      "entropy": 5.7427009582519535,
      "epoch": 3.179420346236141,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003999387699225863,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21753208339214325,
      "num_tokens": 70876830.0,
      "step": 40865
    },
    {
      "entropy": 5.8179154872894285,
      "epoch": 3.1798093756078583,
      "grad_norm": 1.25,
      "learning_rate": 0.0003999157189964026,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20854505002498627,
      "num_tokens": 70885952.0,
      "step": 40870
    },
    {
      "entropy": 5.786499166488648,
      "epoch": 3.180198404979576,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003998926661748217,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21354981511831284,
      "num_tokens": 70894237.0,
      "step": 40875
    },
    {
      "entropy": 5.725682020187378,
      "epoch": 3.1805874343512937,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003998696114581933,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2086714819073677,
      "num_tokens": 70903542.0,
      "step": 40880
    },
    {
      "entropy": 5.740878772735596,
      "epoch": 3.180976463723011,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003998465548468674,
      "loss": 4.8532,
      "mean_token_accuracy": 0.21615061312913894,
      "num_tokens": 70913305.0,
      "step": 40885
    },
    {
      "entropy": 5.796900033950806,
      "epoch": 3.1813654930947286,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003998234963411936,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2117125943303108,
      "num_tokens": 70922061.0,
      "step": 40890
    },
    {
      "entropy": 5.756347608566284,
      "epoch": 3.1817545224664463,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003998004359415221,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2064922034740448,
      "num_tokens": 70930653.0,
      "step": 40895
    },
    {
      "entropy": 5.853744602203369,
      "epoch": 3.1821435518381636,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039977737364820247,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20538506507873536,
      "num_tokens": 70939549.0,
      "step": 40900
    },
    {
      "entropy": 5.781472301483154,
      "epoch": 3.1825325812098813,
      "grad_norm": 1.25,
      "learning_rate": 0.00039975430946158493,
      "loss": 4.9053,
      "mean_token_accuracy": 0.20559643357992172,
      "num_tokens": 70948349.0,
      "step": 40905
    },
    {
      "entropy": 5.8041548252105715,
      "epoch": 3.182921610581599,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003997312433820192,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20969062745571138,
      "num_tokens": 70957619.0,
      "step": 40910
    },
    {
      "entropy": 5.850238370895386,
      "epoch": 3.1833106399533166,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039970817540985545,
      "loss": 5.0741,
      "mean_token_accuracy": 0.20132511407136916,
      "num_tokens": 70966408.0,
      "step": 40915
    },
    {
      "entropy": 5.804398012161255,
      "epoch": 3.183699669325034,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003996851055454436,
      "loss": 4.976,
      "mean_token_accuracy": 0.2054872542619705,
      "num_tokens": 70975023.0,
      "step": 40920
    },
    {
      "entropy": 5.769306421279907,
      "epoch": 3.1840886986967516,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003996620337891337,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20351338982582093,
      "num_tokens": 70983495.0,
      "step": 40925
    },
    {
      "entropy": 5.749544477462768,
      "epoch": 3.1844777280684693,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000399638960141276,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21251570284366608,
      "num_tokens": 70992284.0,
      "step": 40930
    },
    {
      "entropy": 5.767158031463623,
      "epoch": 3.1848667574401865,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039961588460222033,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21640600860118867,
      "num_tokens": 71000334.0,
      "step": 40935
    },
    {
      "entropy": 5.742600297927856,
      "epoch": 3.1852557868119042,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000399592807172317,
      "loss": 4.8474,
      "mean_token_accuracy": 0.21630490124225615,
      "num_tokens": 71008694.0,
      "step": 40940
    },
    {
      "entropy": 5.818905067443848,
      "epoch": 3.185644816183622,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039956972785191606,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21220496147871018,
      "num_tokens": 71016702.0,
      "step": 40945
    },
    {
      "entropy": 5.742748260498047,
      "epoch": 3.1860338455553396,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039954664664136787,
      "loss": 4.877,
      "mean_token_accuracy": 0.21565277874469757,
      "num_tokens": 71025088.0,
      "step": 40950
    },
    {
      "entropy": 5.724049806594849,
      "epoch": 3.186422874927057,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003995235635410225,
      "loss": 4.964,
      "mean_token_accuracy": 0.20990327447652818,
      "num_tokens": 71034382.0,
      "step": 40955
    },
    {
      "entropy": 5.736189651489258,
      "epoch": 3.1868119042987746,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003995004785512302,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22395765632390977,
      "num_tokens": 71043464.0,
      "step": 40960
    },
    {
      "entropy": 5.800384044647217,
      "epoch": 3.1872009336704923,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003994773916723414,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20932624191045762,
      "num_tokens": 71052207.0,
      "step": 40965
    },
    {
      "entropy": 5.699514389038086,
      "epoch": 3.1875899630422095,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003994543029047063,
      "loss": 4.9924,
      "mean_token_accuracy": 0.20477917194366455,
      "num_tokens": 71060965.0,
      "step": 40970
    },
    {
      "entropy": 5.780302095413208,
      "epoch": 3.1879789924139272,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003994312122486752,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21124306619167327,
      "num_tokens": 71070171.0,
      "step": 40975
    },
    {
      "entropy": 5.823684978485107,
      "epoch": 3.188368021785645,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003994081197045985,
      "loss": 5.0116,
      "mean_token_accuracy": 0.2028563901782036,
      "num_tokens": 71078440.0,
      "step": 40980
    },
    {
      "entropy": 5.821783590316772,
      "epoch": 3.188757051157362,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003993850252728267,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2088819921016693,
      "num_tokens": 71086903.0,
      "step": 40985
    },
    {
      "entropy": 5.727171421051025,
      "epoch": 3.18914608052908,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039936192895371006,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21178353875875472,
      "num_tokens": 71095950.0,
      "step": 40990
    },
    {
      "entropy": 5.759453010559082,
      "epoch": 3.1895351099007976,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000399338830747599,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21065323501825334,
      "num_tokens": 71103762.0,
      "step": 40995
    },
    {
      "entropy": 5.784603500366211,
      "epoch": 3.1899241392725153,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003993157306548442,
      "loss": 5.0418,
      "mean_token_accuracy": 0.19890278428792954,
      "num_tokens": 71112994.0,
      "step": 41000
    },
    {
      "entropy": 5.721879339218139,
      "epoch": 3.1903131686442325,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039929262867579603,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21019052267074584,
      "num_tokens": 71122140.0,
      "step": 41005
    },
    {
      "entropy": 5.76257209777832,
      "epoch": 3.1907021980159502,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039926952481080495,
      "loss": 4.9083,
      "mean_token_accuracy": 0.20683010220527648,
      "num_tokens": 71130300.0,
      "step": 41010
    },
    {
      "entropy": 5.8049383640289305,
      "epoch": 3.191091227387668,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039924641906022176,
      "loss": 4.9994,
      "mean_token_accuracy": 0.2043612703680992,
      "num_tokens": 71139524.0,
      "step": 41015
    },
    {
      "entropy": 5.840487241744995,
      "epoch": 3.191480256759385,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003992233114243969,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21030060052871705,
      "num_tokens": 71148626.0,
      "step": 41020
    },
    {
      "entropy": 5.787528944015503,
      "epoch": 3.191869286131103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000399200201903681,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20657522231340408,
      "num_tokens": 71157592.0,
      "step": 41025
    },
    {
      "entropy": 5.765617179870605,
      "epoch": 3.1922583155028206,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003991770904984247,
      "loss": 4.8633,
      "mean_token_accuracy": 0.20911939889192582,
      "num_tokens": 71165953.0,
      "step": 41030
    },
    {
      "entropy": 5.68063645362854,
      "epoch": 3.192647344874538,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003991539772089787,
      "loss": 4.829,
      "mean_token_accuracy": 0.22162360399961473,
      "num_tokens": 71174581.0,
      "step": 41035
    },
    {
      "entropy": 5.682666206359864,
      "epoch": 3.1930363742462555,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039913086203569373,
      "loss": 4.882,
      "mean_token_accuracy": 0.21956687718629836,
      "num_tokens": 71182950.0,
      "step": 41040
    },
    {
      "entropy": 5.74949541091919,
      "epoch": 3.1934254036179732,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039910774497892046,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20348596423864365,
      "num_tokens": 71192115.0,
      "step": 41045
    },
    {
      "entropy": 5.870685052871704,
      "epoch": 3.193814432989691,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039908462603900977,
      "loss": 5.0375,
      "mean_token_accuracy": 0.20360516607761384,
      "num_tokens": 71200401.0,
      "step": 41050
    },
    {
      "entropy": 5.782311916351318,
      "epoch": 3.194203462361408,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003990615052163123,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21607262194156646,
      "num_tokens": 71209705.0,
      "step": 41055
    },
    {
      "entropy": 5.6553199768066404,
      "epoch": 3.194592491733126,
      "grad_norm": 1.34375,
      "learning_rate": 0.000399038382511179,
      "loss": 4.8547,
      "mean_token_accuracy": 0.2232244998216629,
      "num_tokens": 71218738.0,
      "step": 41060
    },
    {
      "entropy": 5.777820348739624,
      "epoch": 3.1949815211048436,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003990152579239607,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21410251557826995,
      "num_tokens": 71226454.0,
      "step": 41065
    },
    {
      "entropy": 5.8642816066741945,
      "epoch": 3.195370550476561,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039899213145500815,
      "loss": 5.042,
      "mean_token_accuracy": 0.20737990587949753,
      "num_tokens": 71235764.0,
      "step": 41070
    },
    {
      "entropy": 5.773732042312622,
      "epoch": 3.1957595798482785,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039896900310467244,
      "loss": 4.9988,
      "mean_token_accuracy": 0.20989206433296204,
      "num_tokens": 71244417.0,
      "step": 41075
    },
    {
      "entropy": 5.792234706878662,
      "epoch": 3.196148609219996,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039894587287330434,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21305650919675828,
      "num_tokens": 71252866.0,
      "step": 41080
    },
    {
      "entropy": 5.839225482940674,
      "epoch": 3.1965376385917135,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039892274076125503,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20738569945096968,
      "num_tokens": 71262376.0,
      "step": 41085
    },
    {
      "entropy": 5.786189699172974,
      "epoch": 3.196926667963431,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003988996067688753,
      "loss": 4.8782,
      "mean_token_accuracy": 0.20995887219905854,
      "num_tokens": 71270690.0,
      "step": 41090
    },
    {
      "entropy": 5.7922484397888185,
      "epoch": 3.197315697335149,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003988764708965163,
      "loss": 4.9974,
      "mean_token_accuracy": 0.20728829652070999,
      "num_tokens": 71279474.0,
      "step": 41095
    },
    {
      "entropy": 5.702958679199218,
      "epoch": 3.1977047267068666,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000398853333144529,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21458377242088317,
      "num_tokens": 71288907.0,
      "step": 41100
    },
    {
      "entropy": 5.872563648223877,
      "epoch": 3.198093756078584,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039883019351326447,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21636772900819778,
      "num_tokens": 71296788.0,
      "step": 41105
    },
    {
      "entropy": 5.789545249938965,
      "epoch": 3.1984827854503015,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039880705200307377,
      "loss": 4.895,
      "mean_token_accuracy": 0.21370198726654052,
      "num_tokens": 71305131.0,
      "step": 41110
    },
    {
      "entropy": 5.716982364654541,
      "epoch": 3.198871814822019,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003987839086143083,
      "loss": 4.8757,
      "mean_token_accuracy": 0.21180029064416886,
      "num_tokens": 71313333.0,
      "step": 41115
    },
    {
      "entropy": 5.712088632583618,
      "epoch": 3.1992608441937365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039876076334731885,
      "loss": 4.9295,
      "mean_token_accuracy": 0.20944864749908448,
      "num_tokens": 71321852.0,
      "step": 41120
    },
    {
      "entropy": 5.76860818862915,
      "epoch": 3.199649873565454,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000398737616202457,
      "loss": 4.9441,
      "mean_token_accuracy": 0.20839363038539888,
      "num_tokens": 71330905.0,
      "step": 41125
    },
    {
      "entropy": 5.811459112167358,
      "epoch": 3.200038902937172,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039871446718007364,
      "loss": 5.004,
      "mean_token_accuracy": 0.20433032512664795,
      "num_tokens": 71339479.0,
      "step": 41130
    },
    {
      "entropy": 5.768079233169556,
      "epoch": 3.200427932308889,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003986913162805201,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2148787721991539,
      "num_tokens": 71348921.0,
      "step": 41135
    },
    {
      "entropy": 5.797590494155884,
      "epoch": 3.200816961680607,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039866816350414786,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21401213258504867,
      "num_tokens": 71357695.0,
      "step": 41140
    },
    {
      "entropy": 5.751787900924683,
      "epoch": 3.2012059910523245,
      "grad_norm": 1.3125,
      "learning_rate": 0.000398645008851308,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2225073590874672,
      "num_tokens": 71365780.0,
      "step": 41145
    },
    {
      "entropy": 5.739987754821778,
      "epoch": 3.201595020424042,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039862185232235196,
      "loss": 4.8271,
      "mean_token_accuracy": 0.21892186403274536,
      "num_tokens": 71374219.0,
      "step": 41150
    },
    {
      "entropy": 5.748920440673828,
      "epoch": 3.2019840497957595,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003985986939176311,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2083059936761856,
      "num_tokens": 71382438.0,
      "step": 41155
    },
    {
      "entropy": 5.7911077499389645,
      "epoch": 3.202373079167477,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039857553363749674,
      "loss": 4.9598,
      "mean_token_accuracy": 0.20821693241596223,
      "num_tokens": 71390989.0,
      "step": 41160
    },
    {
      "entropy": 5.7301229476928714,
      "epoch": 3.202762108539195,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039855237148230026,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21288724541664122,
      "num_tokens": 71398914.0,
      "step": 41165
    },
    {
      "entropy": 5.8240100860595705,
      "epoch": 3.203151137910912,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039852920745239343,
      "loss": 4.9757,
      "mean_token_accuracy": 0.2073565348982811,
      "num_tokens": 71407638.0,
      "step": 41170
    },
    {
      "entropy": 5.780116128921509,
      "epoch": 3.20354016728263,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039850604154812727,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2129078611731529,
      "num_tokens": 71417123.0,
      "step": 41175
    },
    {
      "entropy": 5.848075723648071,
      "epoch": 3.2039291966543475,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003984828737698536,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20647359937429427,
      "num_tokens": 71425975.0,
      "step": 41180
    },
    {
      "entropy": 5.844686555862427,
      "epoch": 3.204318226026065,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039845970411792384,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20550137460231782,
      "num_tokens": 71435290.0,
      "step": 41185
    },
    {
      "entropy": 5.754124689102173,
      "epoch": 3.2047072553977825,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003984365325926896,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22372455447912215,
      "num_tokens": 71444305.0,
      "step": 41190
    },
    {
      "entropy": 5.7936595439910885,
      "epoch": 3.2050962847695,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039841335919450245,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21301029473543168,
      "num_tokens": 71453007.0,
      "step": 41195
    },
    {
      "entropy": 5.795546531677246,
      "epoch": 3.205485314141218,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039839018392371397,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2130945309996605,
      "num_tokens": 71461233.0,
      "step": 41200
    },
    {
      "entropy": 5.7807258605957035,
      "epoch": 3.205874343512935,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003983670067806759,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2127334013581276,
      "num_tokens": 71469206.0,
      "step": 41205
    },
    {
      "entropy": 5.762767362594604,
      "epoch": 3.206263372884653,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003983438277657398,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21214614063501358,
      "num_tokens": 71477143.0,
      "step": 41210
    },
    {
      "entropy": 5.75270037651062,
      "epoch": 3.2066524022563705,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003983206468792575,
      "loss": 4.985,
      "mean_token_accuracy": 0.20365626215934754,
      "num_tokens": 71485485.0,
      "step": 41215
    },
    {
      "entropy": 5.747783422470093,
      "epoch": 3.2070414316280877,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003982974641215806,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2188277378678322,
      "num_tokens": 71494713.0,
      "step": 41220
    },
    {
      "entropy": 5.739676809310913,
      "epoch": 3.2074304609998054,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000398274279493061,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21351024359464646,
      "num_tokens": 71503036.0,
      "step": 41225
    },
    {
      "entropy": 5.808224439620972,
      "epoch": 3.207819490371523,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039825109299405036,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2051343321800232,
      "num_tokens": 71511749.0,
      "step": 41230
    },
    {
      "entropy": 5.771705865859985,
      "epoch": 3.2082085197432404,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003982279046249006,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22007878571748735,
      "num_tokens": 71519623.0,
      "step": 41235
    },
    {
      "entropy": 5.81016845703125,
      "epoch": 3.208597549114958,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039820471438596346,
      "loss": 4.9859,
      "mean_token_accuracy": 0.20677320063114166,
      "num_tokens": 71528121.0,
      "step": 41240
    },
    {
      "entropy": 5.78837513923645,
      "epoch": 3.208986578486676,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039818152227759097,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20611997395753862,
      "num_tokens": 71537529.0,
      "step": 41245
    },
    {
      "entropy": 5.851659488677979,
      "epoch": 3.2093756078583935,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003981583283001349,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20417221635580063,
      "num_tokens": 71545624.0,
      "step": 41250
    },
    {
      "entropy": 5.8009765625,
      "epoch": 3.2097646372301107,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003981351324539472,
      "loss": 4.9892,
      "mean_token_accuracy": 0.20762692987918854,
      "num_tokens": 71554578.0,
      "step": 41255
    },
    {
      "entropy": 5.8128331184387205,
      "epoch": 3.2101536666018284,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003981119347393799,
      "loss": 4.9145,
      "mean_token_accuracy": 0.20966263562440873,
      "num_tokens": 71563669.0,
      "step": 41260
    },
    {
      "entropy": 5.835852813720703,
      "epoch": 3.210542695973546,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039808873515678495,
      "loss": 5.0219,
      "mean_token_accuracy": 0.20888498723506926,
      "num_tokens": 71572254.0,
      "step": 41265
    },
    {
      "entropy": 5.727073669433594,
      "epoch": 3.2109317253452634,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003980655337065144,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2171695873141289,
      "num_tokens": 71580329.0,
      "step": 41270
    },
    {
      "entropy": 5.8484110832214355,
      "epoch": 3.211320754716981,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003980423303889201,
      "loss": 4.9839,
      "mean_token_accuracy": 0.20728642791509627,
      "num_tokens": 71588818.0,
      "step": 41275
    },
    {
      "entropy": 5.799424743652343,
      "epoch": 3.211709784088699,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039801912520435437,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2101706326007843,
      "num_tokens": 71597018.0,
      "step": 41280
    },
    {
      "entropy": 5.755481290817261,
      "epoch": 3.2120988134604165,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003979959181531692,
      "loss": 4.8652,
      "mean_token_accuracy": 0.21490960717201232,
      "num_tokens": 71605508.0,
      "step": 41285
    },
    {
      "entropy": 5.819076347351074,
      "epoch": 3.2124878428321337,
      "grad_norm": 1.25,
      "learning_rate": 0.00039797270923571683,
      "loss": 5.0263,
      "mean_token_accuracy": 0.19907890856266022,
      "num_tokens": 71615034.0,
      "step": 41290
    },
    {
      "entropy": 5.727461624145508,
      "epoch": 3.2128768722038514,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039794949845234925,
      "loss": 4.9523,
      "mean_token_accuracy": 0.20956735610961913,
      "num_tokens": 71623745.0,
      "step": 41295
    },
    {
      "entropy": 5.773867511749268,
      "epoch": 3.213265901575569,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039792628580341874,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2133917659521103,
      "num_tokens": 71632009.0,
      "step": 41300
    },
    {
      "entropy": 5.680625057220459,
      "epoch": 3.2136549309472864,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039790307128927745,
      "loss": 4.8187,
      "mean_token_accuracy": 0.2185284599661827,
      "num_tokens": 71640982.0,
      "step": 41305
    },
    {
      "entropy": 5.767665147781372,
      "epoch": 3.214043960319004,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039787985491027773,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2144096463918686,
      "num_tokens": 71649556.0,
      "step": 41310
    },
    {
      "entropy": 5.76659893989563,
      "epoch": 3.2144329896907218,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039785663666677175,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21351449936628342,
      "num_tokens": 71657691.0,
      "step": 41315
    },
    {
      "entropy": 5.865299940109253,
      "epoch": 3.214822019062439,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039783341655911185,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20650093257427216,
      "num_tokens": 71666050.0,
      "step": 41320
    },
    {
      "entropy": 5.800711059570313,
      "epoch": 3.2152110484341567,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003978101945876504,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22236639112234116,
      "num_tokens": 71675225.0,
      "step": 41325
    },
    {
      "entropy": 5.745185422897339,
      "epoch": 3.2156000778058744,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039778697075273977,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21332037895917894,
      "num_tokens": 71683929.0,
      "step": 41330
    },
    {
      "entropy": 5.8070821285247805,
      "epoch": 3.2159891071775917,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039776374505473224,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21182017773389816,
      "num_tokens": 71692192.0,
      "step": 41335
    },
    {
      "entropy": 5.827668762207031,
      "epoch": 3.2163781365493094,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003977405174939803,
      "loss": 5.012,
      "mean_token_accuracy": 0.20461510717868805,
      "num_tokens": 71701091.0,
      "step": 41340
    },
    {
      "entropy": 5.81718807220459,
      "epoch": 3.216767165921027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039771728807083635,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20473780632019042,
      "num_tokens": 71710271.0,
      "step": 41345
    },
    {
      "entropy": 5.7647322654724125,
      "epoch": 3.2171561952927448,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039769405678565287,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2156020849943161,
      "num_tokens": 71719238.0,
      "step": 41350
    },
    {
      "entropy": 5.8681073665618895,
      "epoch": 3.217545224664462,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039767082363878237,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21010494232177734,
      "num_tokens": 71728144.0,
      "step": 41355
    },
    {
      "entropy": 5.833740711212158,
      "epoch": 3.2179342540361797,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003976475886305774,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21396714597940444,
      "num_tokens": 71736724.0,
      "step": 41360
    },
    {
      "entropy": 5.7121439456939695,
      "epoch": 3.2183232834078974,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003976243517613904,
      "loss": 4.9447,
      "mean_token_accuracy": 0.20895344018936157,
      "num_tokens": 71745672.0,
      "step": 41365
    },
    {
      "entropy": 5.7311968326568605,
      "epoch": 3.2187123127796147,
      "grad_norm": 1.375,
      "learning_rate": 0.00039760111303157415,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21089872866868972,
      "num_tokens": 71754401.0,
      "step": 41370
    },
    {
      "entropy": 5.7750989437103275,
      "epoch": 3.2191013421513324,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000397577872441481,
      "loss": 4.9959,
      "mean_token_accuracy": 0.19995273500680924,
      "num_tokens": 71763273.0,
      "step": 41375
    },
    {
      "entropy": 5.847412347793579,
      "epoch": 3.21949037152305,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003975546299914639,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2096270963549614,
      "num_tokens": 71772239.0,
      "step": 41380
    },
    {
      "entropy": 5.795411729812622,
      "epoch": 3.2198794008947678,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003975313856818753,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2073624074459076,
      "num_tokens": 71780900.0,
      "step": 41385
    },
    {
      "entropy": 5.8129322052001955,
      "epoch": 3.220268430266485,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003975081395130679,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2234566479921341,
      "num_tokens": 71788490.0,
      "step": 41390
    },
    {
      "entropy": 5.770750570297241,
      "epoch": 3.2206574596382027,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003974848914853946,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20513194054365158,
      "num_tokens": 71796960.0,
      "step": 41395
    },
    {
      "entropy": 5.767718553543091,
      "epoch": 3.2210464890099204,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039746164159920795,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21339890062808992,
      "num_tokens": 71805694.0,
      "step": 41400
    },
    {
      "entropy": 5.782075643539429,
      "epoch": 3.2214355183816377,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039743838985486075,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21322095394134521,
      "num_tokens": 71814505.0,
      "step": 41405
    },
    {
      "entropy": 5.868829822540283,
      "epoch": 3.2218245477533554,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003974151362527059,
      "loss": 5.0879,
      "mean_token_accuracy": 0.20288662165403365,
      "num_tokens": 71822950.0,
      "step": 41410
    },
    {
      "entropy": 5.732941198348999,
      "epoch": 3.222213577125073,
      "grad_norm": 1.375,
      "learning_rate": 0.00039739188079309626,
      "loss": 4.799,
      "mean_token_accuracy": 0.23948272168636323,
      "num_tokens": 71831681.0,
      "step": 41415
    },
    {
      "entropy": 5.766956758499146,
      "epoch": 3.2226026064967903,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003973686234763846,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20776891857385635,
      "num_tokens": 71839678.0,
      "step": 41420
    },
    {
      "entropy": 5.750882816314697,
      "epoch": 3.222991635868508,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003973453643029239,
      "loss": 4.9466,
      "mean_token_accuracy": 0.20962336510419846,
      "num_tokens": 71848667.0,
      "step": 41425
    },
    {
      "entropy": 5.751950359344482,
      "epoch": 3.2233806652402257,
      "grad_norm": 1.5,
      "learning_rate": 0.0003973221032730669,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21663199812173844,
      "num_tokens": 71858117.0,
      "step": 41430
    },
    {
      "entropy": 5.784395933151245,
      "epoch": 3.2237696946119434,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003972988403871668,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21016646027565003,
      "num_tokens": 71866870.0,
      "step": 41435
    },
    {
      "entropy": 5.750413799285889,
      "epoch": 3.2241587239836607,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003972755756455764,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2143656611442566,
      "num_tokens": 71875228.0,
      "step": 41440
    },
    {
      "entropy": 5.750663232803345,
      "epoch": 3.2245477533553784,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003972523090486487,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21038378328084945,
      "num_tokens": 71883489.0,
      "step": 41445
    },
    {
      "entropy": 5.761793088912964,
      "epoch": 3.224936782727096,
      "grad_norm": 1.375,
      "learning_rate": 0.0003972290405967369,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21483329236507415,
      "num_tokens": 71891968.0,
      "step": 41450
    },
    {
      "entropy": 5.730778932571411,
      "epoch": 3.2253258120988133,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003972057702901939,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2081231579184532,
      "num_tokens": 71900249.0,
      "step": 41455
    },
    {
      "entropy": 5.800330972671508,
      "epoch": 3.225714841470531,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003971824981293729,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21230370700359344,
      "num_tokens": 71908379.0,
      "step": 41460
    },
    {
      "entropy": 5.805521774291992,
      "epoch": 3.2261038708422487,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003971592241146269,
      "loss": 5.033,
      "mean_token_accuracy": 0.2092769369482994,
      "num_tokens": 71917112.0,
      "step": 41465
    },
    {
      "entropy": 5.7395017623901365,
      "epoch": 3.226492900213966,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003971359482463092,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2077142998576164,
      "num_tokens": 71926353.0,
      "step": 41470
    },
    {
      "entropy": 5.804167938232422,
      "epoch": 3.2268819295856837,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003971126705247727,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2055477738380432,
      "num_tokens": 71934948.0,
      "step": 41475
    },
    {
      "entropy": 5.74941349029541,
      "epoch": 3.2272709589574013,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003970893909503709,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21023307293653487,
      "num_tokens": 71943264.0,
      "step": 41480
    },
    {
      "entropy": 5.733584022521972,
      "epoch": 3.227659988329119,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000397066109523457,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2171202689409256,
      "num_tokens": 71951785.0,
      "step": 41485
    },
    {
      "entropy": 5.806576919555664,
      "epoch": 3.2280490177008363,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003970428262443842,
      "loss": 5.0799,
      "mean_token_accuracy": 0.19907734990119935,
      "num_tokens": 71960994.0,
      "step": 41490
    },
    {
      "entropy": 5.807471847534179,
      "epoch": 3.228438047072554,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039701954111350565,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21206496506929398,
      "num_tokens": 71970077.0,
      "step": 41495
    },
    {
      "entropy": 5.839105749130249,
      "epoch": 3.2288270764442717,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003969962541311748,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2114150419831276,
      "num_tokens": 71978605.0,
      "step": 41500
    },
    {
      "entropy": 5.780712509155274,
      "epoch": 3.229216105815989,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039697296529774494,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2058292031288147,
      "num_tokens": 71986707.0,
      "step": 41505
    },
    {
      "entropy": 5.735283231735229,
      "epoch": 3.2296051351877066,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039694967461356944,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22236113250255585,
      "num_tokens": 71994857.0,
      "step": 41510
    },
    {
      "entropy": 5.709937334060669,
      "epoch": 3.2299941645594243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039692638207900177,
      "loss": 4.8675,
      "mean_token_accuracy": 0.21237671822309495,
      "num_tokens": 72004049.0,
      "step": 41515
    },
    {
      "entropy": 5.795295572280883,
      "epoch": 3.2303831939311416,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003969030876943954,
      "loss": 4.9045,
      "mean_token_accuracy": 0.219709812104702,
      "num_tokens": 72012401.0,
      "step": 41520
    },
    {
      "entropy": 5.745581388473511,
      "epoch": 3.2307722233028593,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039687979146010367,
      "loss": 4.924,
      "mean_token_accuracy": 0.2079921320080757,
      "num_tokens": 72020789.0,
      "step": 41525
    },
    {
      "entropy": 5.711466264724732,
      "epoch": 3.231161252674577,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039685649337648,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21568625718355178,
      "num_tokens": 72028446.0,
      "step": 41530
    },
    {
      "entropy": 5.694146299362183,
      "epoch": 3.2315502820462947,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039683319344387813,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2188536375761032,
      "num_tokens": 72036741.0,
      "step": 41535
    },
    {
      "entropy": 5.722387886047363,
      "epoch": 3.231939311418012,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039680989166265135,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21364345401525497,
      "num_tokens": 72045618.0,
      "step": 41540
    },
    {
      "entropy": 5.824947786331177,
      "epoch": 3.2323283407897296,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039678658803315336,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2030959501862526,
      "num_tokens": 72054643.0,
      "step": 41545
    },
    {
      "entropy": 5.760968732833862,
      "epoch": 3.2327173701614473,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039676328255573777,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21390660256147384,
      "num_tokens": 72062721.0,
      "step": 41550
    },
    {
      "entropy": 5.689414119720459,
      "epoch": 3.2331063995331646,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00039673997523075825,
      "loss": 4.9067,
      "mean_token_accuracy": 0.20879996418952942,
      "num_tokens": 72071640.0,
      "step": 41555
    },
    {
      "entropy": 5.754532766342163,
      "epoch": 3.2334954289048823,
      "grad_norm": 1.375,
      "learning_rate": 0.00039671666605856825,
      "loss": 4.851,
      "mean_token_accuracy": 0.22473989129066468,
      "num_tokens": 72080022.0,
      "step": 41560
    },
    {
      "entropy": 5.78931827545166,
      "epoch": 3.2338844582766,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039669335503952165,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20165071338415147,
      "num_tokens": 72088035.0,
      "step": 41565
    },
    {
      "entropy": 5.841640663146973,
      "epoch": 3.2342734876483172,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039667004217397206,
      "loss": 5.0186,
      "mean_token_accuracy": 0.20210168063640593,
      "num_tokens": 72096427.0,
      "step": 41570
    },
    {
      "entropy": 5.790648555755615,
      "epoch": 3.234662517020035,
      "grad_norm": 1.375,
      "learning_rate": 0.00039664672746227326,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2093882143497467,
      "num_tokens": 72105207.0,
      "step": 41575
    },
    {
      "entropy": 5.761158800125122,
      "epoch": 3.2350515463917526,
      "grad_norm": 1.25,
      "learning_rate": 0.0003966234109047789,
      "loss": 4.8404,
      "mean_token_accuracy": 0.22233157455921174,
      "num_tokens": 72113963.0,
      "step": 41580
    },
    {
      "entropy": 5.729147052764892,
      "epoch": 3.2354405757634703,
      "grad_norm": 1.375,
      "learning_rate": 0.000396600092501843,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21021894067525865,
      "num_tokens": 72122585.0,
      "step": 41585
    },
    {
      "entropy": 5.781980037689209,
      "epoch": 3.2358296051351876,
      "grad_norm": 1.390625,
      "learning_rate": 0.00039657677225381915,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21595092862844467,
      "num_tokens": 72131349.0,
      "step": 41590
    },
    {
      "entropy": 5.806960344314575,
      "epoch": 3.2362186345069053,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003965534501610613,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21161748021841048,
      "num_tokens": 72140960.0,
      "step": 41595
    },
    {
      "entropy": 5.832040166854858,
      "epoch": 3.236607663878623,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003965301262239234,
      "loss": 4.923,
      "mean_token_accuracy": 0.21271607428789138,
      "num_tokens": 72150126.0,
      "step": 41600
    },
    {
      "entropy": 5.733131742477417,
      "epoch": 3.2369966932503402,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003965068004427591,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21350779384374619,
      "num_tokens": 72158909.0,
      "step": 41605
    },
    {
      "entropy": 5.89306116104126,
      "epoch": 3.237385722622058,
      "grad_norm": 1.375,
      "learning_rate": 0.0003964834728179226,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20242188274860382,
      "num_tokens": 72168409.0,
      "step": 41610
    },
    {
      "entropy": 5.780468416213989,
      "epoch": 3.2377747519937756,
      "grad_norm": 1.375,
      "learning_rate": 0.00039646014334976783,
      "loss": 4.8875,
      "mean_token_accuracy": 0.20918132215738297,
      "num_tokens": 72176366.0,
      "step": 41615
    },
    {
      "entropy": 5.7993604183197025,
      "epoch": 3.2381637813654933,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039643681203864863,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20882545858621598,
      "num_tokens": 72185905.0,
      "step": 41620
    },
    {
      "entropy": 5.770069408416748,
      "epoch": 3.2385528107372106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039641347888491905,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21048083901405334,
      "num_tokens": 72195108.0,
      "step": 41625
    },
    {
      "entropy": 5.792064476013183,
      "epoch": 3.2389418401089283,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039639014388893337,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20949421226978301,
      "num_tokens": 72204218.0,
      "step": 41630
    },
    {
      "entropy": 5.80674729347229,
      "epoch": 3.239330869480646,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003963668070510453,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21359972208738326,
      "num_tokens": 72212658.0,
      "step": 41635
    },
    {
      "entropy": 5.749911785125732,
      "epoch": 3.2397198988523632,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003963434683716091,
      "loss": 4.936,
      "mean_token_accuracy": 0.2147606998682022,
      "num_tokens": 72221167.0,
      "step": 41640
    },
    {
      "entropy": 5.794390344619751,
      "epoch": 3.240108928224081,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003963201278509789,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20405421704053878,
      "num_tokens": 72230254.0,
      "step": 41645
    },
    {
      "entropy": 5.860540533065796,
      "epoch": 3.2404979575957986,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003962967854895089,
      "loss": 5.0716,
      "mean_token_accuracy": 0.20366933345794677,
      "num_tokens": 72239203.0,
      "step": 41650
    },
    {
      "entropy": 5.775693750381469,
      "epoch": 3.240886986967516,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003962734412875533,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20404946208000183,
      "num_tokens": 72247506.0,
      "step": 41655
    },
    {
      "entropy": 5.8076694965362545,
      "epoch": 3.2412760163392336,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003962500952454662,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21339436024427413,
      "num_tokens": 72256251.0,
      "step": 41660
    },
    {
      "entropy": 5.776844024658203,
      "epoch": 3.2416650457109513,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039622674736360195,
      "loss": 4.9238,
      "mean_token_accuracy": 0.211805360019207,
      "num_tokens": 72264674.0,
      "step": 41665
    },
    {
      "entropy": 5.742752456665039,
      "epoch": 3.2420540750826685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039620339764231467,
      "loss": 4.8552,
      "mean_token_accuracy": 0.21759113818407058,
      "num_tokens": 72273547.0,
      "step": 41670
    },
    {
      "entropy": 5.815679931640625,
      "epoch": 3.242443104454386,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039618004608195877,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21539129316806793,
      "num_tokens": 72281947.0,
      "step": 41675
    },
    {
      "entropy": 5.715179824829102,
      "epoch": 3.242832133826104,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039615669268288863,
      "loss": 4.8832,
      "mean_token_accuracy": 0.21577232033014299,
      "num_tokens": 72290420.0,
      "step": 41680
    },
    {
      "entropy": 5.717205476760864,
      "epoch": 3.2432211631978216,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039613333744545836,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21262929886579512,
      "num_tokens": 72299300.0,
      "step": 41685
    },
    {
      "entropy": 5.77725887298584,
      "epoch": 3.243610192569539,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003961099803700226,
      "loss": 4.9858,
      "mean_token_accuracy": 0.20263432562351227,
      "num_tokens": 72307217.0,
      "step": 41690
    },
    {
      "entropy": 5.790657091140747,
      "epoch": 3.2439992219412566,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003960866214569357,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21433602422475814,
      "num_tokens": 72315304.0,
      "step": 41695
    },
    {
      "entropy": 5.732626628875733,
      "epoch": 3.2443882513129743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039606326070655204,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2085035353899002,
      "num_tokens": 72323661.0,
      "step": 41700
    },
    {
      "entropy": 5.750534582138061,
      "epoch": 3.2447772806846915,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039603989811922596,
      "loss": 4.9252,
      "mean_token_accuracy": 0.20845236629247665,
      "num_tokens": 72332159.0,
      "step": 41705
    },
    {
      "entropy": 5.785015964508057,
      "epoch": 3.245166310056409,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003960165336953122,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21324131190776824,
      "num_tokens": 72340509.0,
      "step": 41710
    },
    {
      "entropy": 5.788167428970337,
      "epoch": 3.245555339428127,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003959931674351651,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21773664206266402,
      "num_tokens": 72349687.0,
      "step": 41715
    },
    {
      "entropy": 5.775322484970093,
      "epoch": 3.2459443687998446,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003959697993391393,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2127616822719574,
      "num_tokens": 72358341.0,
      "step": 41720
    },
    {
      "entropy": 5.8342207908630375,
      "epoch": 3.246333398171562,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039594642940758934,
      "loss": 4.972,
      "mean_token_accuracy": 0.20729655027389526,
      "num_tokens": 72367346.0,
      "step": 41725
    },
    {
      "entropy": 5.757956266403198,
      "epoch": 3.2467224275432796,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039592305764086984,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2115914151072502,
      "num_tokens": 72375603.0,
      "step": 41730
    },
    {
      "entropy": 5.791325902938842,
      "epoch": 3.2471114569149973,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039589968403933533,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21353906989097596,
      "num_tokens": 72383920.0,
      "step": 41735
    },
    {
      "entropy": 5.777887630462646,
      "epoch": 3.2475004862867145,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003958763086033406,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21097012460231782,
      "num_tokens": 72392764.0,
      "step": 41740
    },
    {
      "entropy": 5.777517414093017,
      "epoch": 3.247889515658432,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003958529313332403,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21646185368299484,
      "num_tokens": 72401358.0,
      "step": 41745
    },
    {
      "entropy": 5.871073246002197,
      "epoch": 3.24827854503015,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003958295522293891,
      "loss": 5.0444,
      "mean_token_accuracy": 0.20554718226194382,
      "num_tokens": 72410039.0,
      "step": 41750
    },
    {
      "entropy": 5.797944164276123,
      "epoch": 3.248667574401867,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003958061712921417,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20827229917049409,
      "num_tokens": 72418858.0,
      "step": 41755
    },
    {
      "entropy": 5.7297399044036865,
      "epoch": 3.249056603773585,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000395782788521853,
      "loss": 4.9558,
      "mean_token_accuracy": 0.2091977372765541,
      "num_tokens": 72426897.0,
      "step": 41760
    },
    {
      "entropy": 5.709538459777832,
      "epoch": 3.2494456331453025,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003957594039188778,
      "loss": 4.8581,
      "mean_token_accuracy": 0.21671053916215896,
      "num_tokens": 72436305.0,
      "step": 41765
    },
    {
      "entropy": 5.758968114852905,
      "epoch": 3.24983466251702,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003957360174835708,
      "loss": 4.9327,
      "mean_token_accuracy": 0.21369852870702744,
      "num_tokens": 72444845.0,
      "step": 41770
    },
    {
      "entropy": 5.754583883285522,
      "epoch": 3.2502236918887375,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003957126292162868,
      "loss": 4.8984,
      "mean_token_accuracy": 0.21952549517154693,
      "num_tokens": 72452926.0,
      "step": 41775
    },
    {
      "entropy": 5.758207082748413,
      "epoch": 3.250612721260455,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003956892391173809,
      "loss": 5.032,
      "mean_token_accuracy": 0.20759376883506775,
      "num_tokens": 72460934.0,
      "step": 41780
    },
    {
      "entropy": 5.741777992248535,
      "epoch": 3.251001750632173,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039566584718720793,
      "loss": 4.9121,
      "mean_token_accuracy": 0.21809347867965698,
      "num_tokens": 72470443.0,
      "step": 41785
    },
    {
      "entropy": 5.849734210968018,
      "epoch": 3.25139078000389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003956424534261227,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2082554891705513,
      "num_tokens": 72478948.0,
      "step": 41790
    },
    {
      "entropy": 5.799407339096069,
      "epoch": 3.251779809375608,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039561905783448025,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2126250982284546,
      "num_tokens": 72488298.0,
      "step": 41795
    },
    {
      "entropy": 5.737770032882691,
      "epoch": 3.2521688387473255,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039559566041263565,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20783808678388596,
      "num_tokens": 72497048.0,
      "step": 41800
    },
    {
      "entropy": 5.785961532592774,
      "epoch": 3.252557868119043,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003955722611609438,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2136231318116188,
      "num_tokens": 72505460.0,
      "step": 41805
    },
    {
      "entropy": 5.7440896987915036,
      "epoch": 3.2529468974907605,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039554886007975976,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21886239796876908,
      "num_tokens": 72513838.0,
      "step": 41810
    },
    {
      "entropy": 5.7780615329742435,
      "epoch": 3.253335926862478,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039552545716943873,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20057988315820693,
      "num_tokens": 72522087.0,
      "step": 41815
    },
    {
      "entropy": 5.769866800308227,
      "epoch": 3.253724956234196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003955020524303358,
      "loss": 4.9703,
      "mean_token_accuracy": 0.2101208359003067,
      "num_tokens": 72530950.0,
      "step": 41820
    },
    {
      "entropy": 5.849100875854492,
      "epoch": 3.254113985605913,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039547864586280583,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21138802766799927,
      "num_tokens": 72539916.0,
      "step": 41825
    },
    {
      "entropy": 5.83349642753601,
      "epoch": 3.254503014977631,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003954552374672042,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2174818307161331,
      "num_tokens": 72548193.0,
      "step": 41830
    },
    {
      "entropy": 5.7442240715026855,
      "epoch": 3.2548920443493485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003954318272438862,
      "loss": 4.852,
      "mean_token_accuracy": 0.22028546780347824,
      "num_tokens": 72557512.0,
      "step": 41835
    },
    {
      "entropy": 5.781580018997192,
      "epoch": 3.255281073721066,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003954084151932067,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21178440898656845,
      "num_tokens": 72566127.0,
      "step": 41840
    },
    {
      "entropy": 5.745047235488892,
      "epoch": 3.2556701030927835,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003953850013155214,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21666940599679946,
      "num_tokens": 72574823.0,
      "step": 41845
    },
    {
      "entropy": 5.785233354568481,
      "epoch": 3.256059132464501,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003953615856111852,
      "loss": 4.978,
      "mean_token_accuracy": 0.2044033706188202,
      "num_tokens": 72583664.0,
      "step": 41850
    },
    {
      "entropy": 5.81323652267456,
      "epoch": 3.256448161836219,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039533816808055345,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2029036581516266,
      "num_tokens": 72591844.0,
      "step": 41855
    },
    {
      "entropy": 5.745996809005737,
      "epoch": 3.256837191207936,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039531474872398165,
      "loss": 4.8479,
      "mean_token_accuracy": 0.218050953745842,
      "num_tokens": 72599389.0,
      "step": 41860
    },
    {
      "entropy": 5.730141687393188,
      "epoch": 3.257226220579654,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000395291327541825,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21986958682537078,
      "num_tokens": 72608041.0,
      "step": 41865
    },
    {
      "entropy": 5.773430490493775,
      "epoch": 3.257615249951371,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039526790453443887,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21381855756044388,
      "num_tokens": 72617394.0,
      "step": 41870
    },
    {
      "entropy": 5.75916166305542,
      "epoch": 3.258004279323089,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039524447970217875,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21833214312791824,
      "num_tokens": 72626827.0,
      "step": 41875
    },
    {
      "entropy": 5.775624084472656,
      "epoch": 3.2583933086948065,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039522105304540007,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21388145834207534,
      "num_tokens": 72634777.0,
      "step": 41880
    },
    {
      "entropy": 5.758747529983521,
      "epoch": 3.258782338066524,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003951976245644582,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21434588432312013,
      "num_tokens": 72643440.0,
      "step": 41885
    },
    {
      "entropy": 5.812613010406494,
      "epoch": 3.2591713674382414,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003951741942597087,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21094197928905487,
      "num_tokens": 72651353.0,
      "step": 41890
    },
    {
      "entropy": 5.613944387435913,
      "epoch": 3.259560396809959,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039515076213150714,
      "loss": 4.7704,
      "mean_token_accuracy": 0.21638057976961136,
      "num_tokens": 72659676.0,
      "step": 41895
    },
    {
      "entropy": 5.7297060012817385,
      "epoch": 3.259949426181677,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039512732818020903,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20869381576776505,
      "num_tokens": 72668470.0,
      "step": 41900
    },
    {
      "entropy": 5.667131805419922,
      "epoch": 3.260338455553394,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003951038924061698,
      "loss": 4.8038,
      "mean_token_accuracy": 0.22569464445114135,
      "num_tokens": 72677455.0,
      "step": 41905
    },
    {
      "entropy": 5.758939456939697,
      "epoch": 3.2607274849251118,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039508045480974517,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22165038287639618,
      "num_tokens": 72686405.0,
      "step": 41910
    },
    {
      "entropy": 5.673498630523682,
      "epoch": 3.2611165142968295,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039505701539129087,
      "loss": 4.9044,
      "mean_token_accuracy": 0.21268380135297776,
      "num_tokens": 72695281.0,
      "step": 41915
    },
    {
      "entropy": 5.7695831775665285,
      "epoch": 3.261505543668547,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039503357415116236,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2100575789809227,
      "num_tokens": 72703310.0,
      "step": 41920
    },
    {
      "entropy": 5.757158851623535,
      "epoch": 3.2618945730402644,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039501013108971547,
      "loss": 4.91,
      "mean_token_accuracy": 0.21165044903755187,
      "num_tokens": 72711756.0,
      "step": 41925
    },
    {
      "entropy": 5.791324043273926,
      "epoch": 3.262283602411982,
      "grad_norm": 1.25,
      "learning_rate": 0.0003949866862073059,
      "loss": 4.966,
      "mean_token_accuracy": 0.21065959483385086,
      "num_tokens": 72720569.0,
      "step": 41930
    },
    {
      "entropy": 5.76608419418335,
      "epoch": 3.2626726317837,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039496323950428925,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2075114980340004,
      "num_tokens": 72729245.0,
      "step": 41935
    },
    {
      "entropy": 5.785895967483521,
      "epoch": 3.263061661155417,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003949397909810213,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20666667073965073,
      "num_tokens": 72737727.0,
      "step": 41940
    },
    {
      "entropy": 5.867413711547852,
      "epoch": 3.2634506905271348,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039491634063785813,
      "loss": 5.1081,
      "mean_token_accuracy": 0.200570809841156,
      "num_tokens": 72746090.0,
      "step": 41945
    },
    {
      "entropy": 5.739043235778809,
      "epoch": 3.2638397198988525,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039489288847515525,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21741611063480376,
      "num_tokens": 72755457.0,
      "step": 41950
    },
    {
      "entropy": 5.678177833557129,
      "epoch": 3.26422874927057,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003948694344932687,
      "loss": 4.8905,
      "mean_token_accuracy": 0.21455446481704712,
      "num_tokens": 72764551.0,
      "step": 41955
    },
    {
      "entropy": 5.827851247787476,
      "epoch": 3.2646177786422874,
      "grad_norm": 1.375,
      "learning_rate": 0.0003948459786925542,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22083588540554047,
      "num_tokens": 72773672.0,
      "step": 41960
    },
    {
      "entropy": 5.862974500656128,
      "epoch": 3.265006808014005,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003948225210733677,
      "loss": 5.0227,
      "mean_token_accuracy": 0.20921988040208817,
      "num_tokens": 72782479.0,
      "step": 41965
    },
    {
      "entropy": 5.867896556854248,
      "epoch": 3.265395837385723,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003947990616360652,
      "loss": 5.0371,
      "mean_token_accuracy": 0.20904160141944886,
      "num_tokens": 72790711.0,
      "step": 41970
    },
    {
      "entropy": 5.780977535247803,
      "epoch": 3.26578486675744,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003947756003810026,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20842327326536178,
      "num_tokens": 72798684.0,
      "step": 41975
    },
    {
      "entropy": 5.752422523498535,
      "epoch": 3.2661738961291578,
      "grad_norm": 1.375,
      "learning_rate": 0.0003947521373085359,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2103138878941536,
      "num_tokens": 72806536.0,
      "step": 41980
    },
    {
      "entropy": 5.711889553070068,
      "epoch": 3.2665629255008755,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003947286724190212,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22313413619995118,
      "num_tokens": 72815221.0,
      "step": 41985
    },
    {
      "entropy": 5.75643310546875,
      "epoch": 3.2669519548725927,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039470520571281443,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2182526633143425,
      "num_tokens": 72824072.0,
      "step": 41990
    },
    {
      "entropy": 5.8090746879577635,
      "epoch": 3.2673409842443104,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039468173719027163,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2027595967054367,
      "num_tokens": 72831999.0,
      "step": 41995
    },
    {
      "entropy": 5.704808378219605,
      "epoch": 3.267730013616028,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003946582668517491,
      "loss": 4.8012,
      "mean_token_accuracy": 0.22482135742902756,
      "num_tokens": 72840318.0,
      "step": 42000
    },
    {
      "epoch": 3.267730013616028,
      "eval_entropy": 5.443689430115657,
      "eval_loss": 5.05501651763916,
      "eval_mean_token_accuracy": 0.21525507361643353,
      "eval_num_tokens": 72840318.0,
      "eval_runtime": 21.6833,
      "eval_samples_per_second": 1916.59,
      "eval_steps_per_second": 239.585,
      "step": 42000
    },
    {
      "entropy": 5.7062944889068605,
      "epoch": 3.2681190429877454,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003946347946976026,
      "loss": 4.8533,
      "mean_token_accuracy": 0.21965285241603852,
      "num_tokens": 72848414.0,
      "step": 42005
    },
    {
      "entropy": 5.814571762084961,
      "epoch": 3.268508072359463,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003946113207281887,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2102390006184578,
      "num_tokens": 72857008.0,
      "step": 42010
    },
    {
      "entropy": 5.85265588760376,
      "epoch": 3.2688971017311808,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039458784494386337,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21561134308576585,
      "num_tokens": 72865017.0,
      "step": 42015
    },
    {
      "entropy": 5.744094276428223,
      "epoch": 3.2692861311028985,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039456436734498275,
      "loss": 4.9462,
      "mean_token_accuracy": 0.20419999808073044,
      "num_tokens": 72873812.0,
      "step": 42020
    },
    {
      "entropy": 5.749226427078247,
      "epoch": 3.2696751604746157,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039454088793190316,
      "loss": 4.8538,
      "mean_token_accuracy": 0.21655545979738236,
      "num_tokens": 72882514.0,
      "step": 42025
    },
    {
      "entropy": 5.797271776199341,
      "epoch": 3.2700641898463334,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039451740670498094,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20221734344959258,
      "num_tokens": 72891205.0,
      "step": 42030
    },
    {
      "entropy": 5.77302656173706,
      "epoch": 3.270453219218051,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003944939236645723,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21493741124868393,
      "num_tokens": 72899679.0,
      "step": 42035
    },
    {
      "entropy": 5.811156415939331,
      "epoch": 3.2708422485897684,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003944704388110335,
      "loss": 4.9816,
      "mean_token_accuracy": 0.20956346690654754,
      "num_tokens": 72908518.0,
      "step": 42040
    },
    {
      "entropy": 5.881707239151001,
      "epoch": 3.271231277961486,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000394446952144721,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2081236720085144,
      "num_tokens": 72916967.0,
      "step": 42045
    },
    {
      "entropy": 5.797267723083496,
      "epoch": 3.2716203073332037,
      "grad_norm": 1.25,
      "learning_rate": 0.0003944234636659911,
      "loss": 4.8589,
      "mean_token_accuracy": 0.21798156946897507,
      "num_tokens": 72925539.0,
      "step": 42050
    },
    {
      "entropy": 5.773934936523437,
      "epoch": 3.2720093367049214,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003943999733752003,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21244845688343048,
      "num_tokens": 72933918.0,
      "step": 42055
    },
    {
      "entropy": 5.7244730472564695,
      "epoch": 3.2723983660766387,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003943764812727048,
      "loss": 5.0114,
      "mean_token_accuracy": 0.2057717874646187,
      "num_tokens": 72942472.0,
      "step": 42060
    },
    {
      "entropy": 5.7699785232543945,
      "epoch": 3.2727873954483564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039435298735886133,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2112285777926445,
      "num_tokens": 72951972.0,
      "step": 42065
    },
    {
      "entropy": 5.802233362197876,
      "epoch": 3.273176424820074,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003943294916340261,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21212680339813234,
      "num_tokens": 72960182.0,
      "step": 42070
    },
    {
      "entropy": 5.803643703460693,
      "epoch": 3.2735654541917913,
      "grad_norm": 1.25,
      "learning_rate": 0.0003943059940985559,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20378346592187882,
      "num_tokens": 72968627.0,
      "step": 42075
    },
    {
      "entropy": 5.731094932556152,
      "epoch": 3.273954483563509,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039428249475280706,
      "loss": 4.8814,
      "mean_token_accuracy": 0.21912231147289277,
      "num_tokens": 72977847.0,
      "step": 42080
    },
    {
      "entropy": 5.691620492935181,
      "epoch": 3.2743435129352267,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003942589935971363,
      "loss": 4.8005,
      "mean_token_accuracy": 0.22511896789073943,
      "num_tokens": 72985956.0,
      "step": 42085
    },
    {
      "entropy": 5.7404550075531,
      "epoch": 3.274732542306944,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003942354906319001,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21594283431768418,
      "num_tokens": 72994630.0,
      "step": 42090
    },
    {
      "entropy": 5.749108219146729,
      "epoch": 3.2751215716786617,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003942119858574551,
      "loss": 4.919,
      "mean_token_accuracy": 0.21988788098096848,
      "num_tokens": 73003093.0,
      "step": 42095
    },
    {
      "entropy": 5.716646480560303,
      "epoch": 3.2755106010503794,
      "grad_norm": 1.296875,
      "learning_rate": 0.000394188479274158,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2143036335706711,
      "num_tokens": 73011835.0,
      "step": 42100
    },
    {
      "entropy": 5.764347743988037,
      "epoch": 3.2758996304220966,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039416497088236535,
      "loss": 5.0077,
      "mean_token_accuracy": 0.20877723544836044,
      "num_tokens": 73020226.0,
      "step": 42105
    },
    {
      "entropy": 5.818635654449463,
      "epoch": 3.2762886597938143,
      "grad_norm": 1.296875,
      "learning_rate": 0.000394141460682434,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20762919783592224,
      "num_tokens": 73028508.0,
      "step": 42110
    },
    {
      "entropy": 5.769434976577759,
      "epoch": 3.276677689165532,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003941179486747206,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2116596519947052,
      "num_tokens": 73038138.0,
      "step": 42115
    },
    {
      "entropy": 5.699726295471192,
      "epoch": 3.2770667185372497,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003940944348595819,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21338316053152084,
      "num_tokens": 73046520.0,
      "step": 42120
    },
    {
      "entropy": 5.793140506744384,
      "epoch": 3.277455747908967,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039407091923737475,
      "loss": 5.0161,
      "mean_token_accuracy": 0.20282937586307526,
      "num_tokens": 73055615.0,
      "step": 42125
    },
    {
      "entropy": 5.759832096099854,
      "epoch": 3.2778447772806847,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039404740180845594,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21197818219661713,
      "num_tokens": 73064563.0,
      "step": 42130
    },
    {
      "entropy": 5.721374082565307,
      "epoch": 3.2782338066524024,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039402388257318236,
      "loss": 4.9416,
      "mean_token_accuracy": 0.216379514336586,
      "num_tokens": 73073105.0,
      "step": 42135
    },
    {
      "entropy": 5.811513042449951,
      "epoch": 3.2786228360241196,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003940003615319106,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20713454335927964,
      "num_tokens": 73081976.0,
      "step": 42140
    },
    {
      "entropy": 5.801881170272827,
      "epoch": 3.2790118653958373,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039397683868499784,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21783215552568436,
      "num_tokens": 73090596.0,
      "step": 42145
    },
    {
      "entropy": 5.776477384567261,
      "epoch": 3.279400894767555,
      "grad_norm": 1.328125,
      "learning_rate": 0.000393953314032801,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21627363860607146,
      "num_tokens": 73099348.0,
      "step": 42150
    },
    {
      "entropy": 5.753832530975342,
      "epoch": 3.2797899241392727,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003939297875756769,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21850076764822007,
      "num_tokens": 73107513.0,
      "step": 42155
    },
    {
      "entropy": 5.813546371459961,
      "epoch": 3.28017895351099,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003939062593139825,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20741217881441115,
      "num_tokens": 73115608.0,
      "step": 42160
    },
    {
      "entropy": 5.785907125473022,
      "epoch": 3.2805679828827077,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000393882729248075,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2119610533118248,
      "num_tokens": 73124734.0,
      "step": 42165
    },
    {
      "entropy": 5.781655597686767,
      "epoch": 3.2809570122544254,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039385919737831127,
      "loss": 4.9481,
      "mean_token_accuracy": 0.20915319323539733,
      "num_tokens": 73133826.0,
      "step": 42170
    },
    {
      "entropy": 5.78559250831604,
      "epoch": 3.2813460416261426,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039383566370504837,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21931619495153426,
      "num_tokens": 73142418.0,
      "step": 42175
    },
    {
      "entropy": 5.773282289505005,
      "epoch": 3.2817350709978603,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003938121282286434,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21043350994586946,
      "num_tokens": 73150993.0,
      "step": 42180
    },
    {
      "entropy": 5.797844696044922,
      "epoch": 3.282124100369578,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039378859094945366,
      "loss": 5.0725,
      "mean_token_accuracy": 0.201751047372818,
      "num_tokens": 73160930.0,
      "step": 42185
    },
    {
      "entropy": 5.757339572906494,
      "epoch": 3.2825131297412953,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003937650518678359,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2122027099132538,
      "num_tokens": 73169917.0,
      "step": 42190
    },
    {
      "entropy": 5.797887468338013,
      "epoch": 3.282902159113013,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003937415109841476,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21199240088462828,
      "num_tokens": 73178309.0,
      "step": 42195
    },
    {
      "entropy": 5.739905166625976,
      "epoch": 3.2832911884847307,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003937179682987459,
      "loss": 4.832,
      "mean_token_accuracy": 0.21589563190937042,
      "num_tokens": 73186885.0,
      "step": 42200
    },
    {
      "entropy": 5.688966751098633,
      "epoch": 3.283680217856448,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039369442381198793,
      "loss": 4.9035,
      "mean_token_accuracy": 0.2184841126203537,
      "num_tokens": 73195202.0,
      "step": 42205
    },
    {
      "entropy": 5.735539102554322,
      "epoch": 3.2840692472281656,
      "grad_norm": 1.296875,
      "learning_rate": 0.000393670877524231,
      "loss": 4.9174,
      "mean_token_accuracy": 0.20495409220457078,
      "num_tokens": 73203370.0,
      "step": 42210
    },
    {
      "entropy": 5.801749897003174,
      "epoch": 3.2844582765998833,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003936473294358325,
      "loss": 5.0214,
      "mean_token_accuracy": 0.20448965728282928,
      "num_tokens": 73211878.0,
      "step": 42215
    },
    {
      "entropy": 5.791033315658569,
      "epoch": 3.284847305971601,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003936237795471495,
      "loss": 4.9137,
      "mean_token_accuracy": 0.20994094908237457,
      "num_tokens": 73219756.0,
      "step": 42220
    },
    {
      "entropy": 5.728539085388183,
      "epoch": 3.2852363353433183,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039360022785853957,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2162775620818138,
      "num_tokens": 73227940.0,
      "step": 42225
    },
    {
      "entropy": 5.762232398986816,
      "epoch": 3.285625364715036,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003935766743703599,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21071567982435227,
      "num_tokens": 73236933.0,
      "step": 42230
    },
    {
      "entropy": 5.815197610855103,
      "epoch": 3.2860143940867537,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039355311908296787,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20968196392059327,
      "num_tokens": 73246198.0,
      "step": 42235
    },
    {
      "entropy": 5.799543952941894,
      "epoch": 3.286403423458471,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000393529561996721,
      "loss": 4.9162,
      "mean_token_accuracy": 0.220100037753582,
      "num_tokens": 73255199.0,
      "step": 42240
    },
    {
      "entropy": 5.727614641189575,
      "epoch": 3.2867924528301886,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039350600311197667,
      "loss": 4.88,
      "mean_token_accuracy": 0.21420133858919144,
      "num_tokens": 73263986.0,
      "step": 42245
    },
    {
      "entropy": 5.720384168624878,
      "epoch": 3.2871814822019063,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039348244242909234,
      "loss": 4.9267,
      "mean_token_accuracy": 0.20655566751956939,
      "num_tokens": 73273474.0,
      "step": 42250
    },
    {
      "entropy": 5.708976316452026,
      "epoch": 3.287570511573624,
      "grad_norm": 1.25,
      "learning_rate": 0.00039345887994842566,
      "loss": 4.8714,
      "mean_token_accuracy": 0.21687986850738525,
      "num_tokens": 73281881.0,
      "step": 42255
    },
    {
      "entropy": 5.779132127761841,
      "epoch": 3.2879595409453413,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003934353156703339,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21020979434251785,
      "num_tokens": 73290589.0,
      "step": 42260
    },
    {
      "entropy": 5.773817682266236,
      "epoch": 3.288348570317059,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003934117495951748,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2052561655640602,
      "num_tokens": 73298669.0,
      "step": 42265
    },
    {
      "entropy": 5.664301538467408,
      "epoch": 3.2887375996887767,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003933881817233058,
      "loss": 4.8364,
      "mean_token_accuracy": 0.21965772211551665,
      "num_tokens": 73307622.0,
      "step": 42270
    },
    {
      "entropy": 5.8861607074737545,
      "epoch": 3.289126629060494,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039336461205508456,
      "loss": 4.9967,
      "mean_token_accuracy": 0.20995798259973525,
      "num_tokens": 73315888.0,
      "step": 42275
    },
    {
      "entropy": 5.845155525207519,
      "epoch": 3.2895156584322116,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003933410405908689,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20566961020231248,
      "num_tokens": 73324999.0,
      "step": 42280
    },
    {
      "entropy": 5.779226636886596,
      "epoch": 3.2899046878039293,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039331746733101613,
      "loss": 4.896,
      "mean_token_accuracy": 0.21289982199668883,
      "num_tokens": 73332907.0,
      "step": 42285
    },
    {
      "entropy": 5.758332872390747,
      "epoch": 3.290293717175647,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039329389227588416,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21108102947473525,
      "num_tokens": 73341202.0,
      "step": 42290
    },
    {
      "entropy": 5.754832744598389,
      "epoch": 3.2906827465473643,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039327031542583067,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20636945068836213,
      "num_tokens": 73350162.0,
      "step": 42295
    },
    {
      "entropy": 5.824967861175537,
      "epoch": 3.291071775919082,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039324673678121344,
      "loss": 4.9775,
      "mean_token_accuracy": 0.2101506069302559,
      "num_tokens": 73358701.0,
      "step": 42300
    },
    {
      "entropy": 5.759915542602539,
      "epoch": 3.291460805290799,
      "grad_norm": 1.25,
      "learning_rate": 0.0003932231563423901,
      "loss": 5.0121,
      "mean_token_accuracy": 0.20313595086336136,
      "num_tokens": 73367445.0,
      "step": 42305
    },
    {
      "entropy": 5.809169626235962,
      "epoch": 3.291849834662517,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003931995741097186,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2095741406083107,
      "num_tokens": 73376105.0,
      "step": 42310
    },
    {
      "entropy": 5.754014396667481,
      "epoch": 3.2922388640342346,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039317599008355665,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21127452105283737,
      "num_tokens": 73384382.0,
      "step": 42315
    },
    {
      "entropy": 5.775046491622925,
      "epoch": 3.2926278934059523,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003931524042642622,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21664304435253143,
      "num_tokens": 73393725.0,
      "step": 42320
    },
    {
      "entropy": 5.7508354663848875,
      "epoch": 3.2930169227776696,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003931288166521931,
      "loss": 4.852,
      "mean_token_accuracy": 0.2169520989060402,
      "num_tokens": 73402070.0,
      "step": 42325
    },
    {
      "entropy": 5.717920732498169,
      "epoch": 3.2934059521493873,
      "grad_norm": 1.25,
      "learning_rate": 0.00039310522724770713,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2174407109618187,
      "num_tokens": 73411414.0,
      "step": 42330
    },
    {
      "entropy": 5.809507560729981,
      "epoch": 3.293794981521105,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039308163605116235,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20932976007461548,
      "num_tokens": 73420187.0,
      "step": 42335
    },
    {
      "entropy": 5.783754110336304,
      "epoch": 3.294184010892822,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003930580430629167,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2143746107816696,
      "num_tokens": 73428316.0,
      "step": 42340
    },
    {
      "entropy": 5.7673805236816404,
      "epoch": 3.29457304026454,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003930344482833281,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2082654669880867,
      "num_tokens": 73437069.0,
      "step": 42345
    },
    {
      "entropy": 5.73949236869812,
      "epoch": 3.2949620696362576,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039301085171275475,
      "loss": 4.8973,
      "mean_token_accuracy": 0.21958001852035522,
      "num_tokens": 73445044.0,
      "step": 42350
    },
    {
      "entropy": 5.752908182144165,
      "epoch": 3.2953510990079753,
      "grad_norm": 1.375,
      "learning_rate": 0.0003929872533515545,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20624818801879882,
      "num_tokens": 73453085.0,
      "step": 42355
    },
    {
      "entropy": 5.729835462570191,
      "epoch": 3.2957401283796925,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039296365320008546,
      "loss": 4.9317,
      "mean_token_accuracy": 0.20765060633420945,
      "num_tokens": 73462103.0,
      "step": 42360
    },
    {
      "entropy": 5.723503255844117,
      "epoch": 3.2961291577514102,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039294005125870576,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2199642613530159,
      "num_tokens": 73471347.0,
      "step": 42365
    },
    {
      "entropy": 5.829682016372681,
      "epoch": 3.296518187123128,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039291644752777346,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20506861358880996,
      "num_tokens": 73479673.0,
      "step": 42370
    },
    {
      "entropy": 5.830968189239502,
      "epoch": 3.296907216494845,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039289284200764676,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20436266660690308,
      "num_tokens": 73488760.0,
      "step": 42375
    },
    {
      "entropy": 5.782667589187622,
      "epoch": 3.297296245866563,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003928692346986839,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21267121136188508,
      "num_tokens": 73497700.0,
      "step": 42380
    },
    {
      "entropy": 5.7757350444793705,
      "epoch": 3.2976852752382806,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039284562560124297,
      "loss": 4.8981,
      "mean_token_accuracy": 0.21625377833843232,
      "num_tokens": 73505952.0,
      "step": 42385
    },
    {
      "entropy": 5.768964576721191,
      "epoch": 3.2980743046099983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003928220147156822,
      "loss": 4.9218,
      "mean_token_accuracy": 0.20911584049463272,
      "num_tokens": 73515068.0,
      "step": 42390
    },
    {
      "entropy": 5.852684020996094,
      "epoch": 3.2984633339817155,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039279840204235994,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21259979605674745,
      "num_tokens": 73524520.0,
      "step": 42395
    },
    {
      "entropy": 5.792552947998047,
      "epoch": 3.2988523633534332,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039277478758163443,
      "loss": 4.9001,
      "mean_token_accuracy": 0.21308257728815078,
      "num_tokens": 73533067.0,
      "step": 42400
    },
    {
      "entropy": 5.770245742797852,
      "epoch": 3.299241392725151,
      "grad_norm": 1.34375,
      "learning_rate": 0.000392751171333864,
      "loss": 5.0035,
      "mean_token_accuracy": 0.20173162072896958,
      "num_tokens": 73541550.0,
      "step": 42405
    },
    {
      "entropy": 5.663273334503174,
      "epoch": 3.299630422096868,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039272755329940683,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22782112658023834,
      "num_tokens": 73550645.0,
      "step": 42410
    },
    {
      "entropy": 5.754573154449463,
      "epoch": 3.300019451468586,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039270393347862155,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2059642568230629,
      "num_tokens": 73559005.0,
      "step": 42415
    },
    {
      "entropy": 5.859376001358032,
      "epoch": 3.3004084808403036,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039268031187186635,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20948057621717453,
      "num_tokens": 73567335.0,
      "step": 42420
    },
    {
      "entropy": 5.840239763259888,
      "epoch": 3.300797510212021,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039265668847949977,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21448366343975067,
      "num_tokens": 73575788.0,
      "step": 42425
    },
    {
      "entropy": 5.763343048095703,
      "epoch": 3.3011865395837385,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003926330633018802,
      "loss": 4.9383,
      "mean_token_accuracy": 0.20963292866945266,
      "num_tokens": 73584430.0,
      "step": 42430
    },
    {
      "entropy": 5.871312189102173,
      "epoch": 3.3015755689554562,
      "grad_norm": 1.375,
      "learning_rate": 0.00039260943633936615,
      "loss": 5.076,
      "mean_token_accuracy": 0.20629605352878572,
      "num_tokens": 73593725.0,
      "step": 42435
    },
    {
      "entropy": 5.753225803375244,
      "epoch": 3.3019645983271735,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039258580759231603,
      "loss": 4.93,
      "mean_token_accuracy": 0.2107033759355545,
      "num_tokens": 73602028.0,
      "step": 42440
    },
    {
      "entropy": 5.7495622634887695,
      "epoch": 3.302353627698891,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003925621770610885,
      "loss": 4.8737,
      "mean_token_accuracy": 0.21666236966848373,
      "num_tokens": 73610002.0,
      "step": 42445
    },
    {
      "entropy": 5.759077405929565,
      "epoch": 3.302742657070609,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000392538544746042,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2135808452963829,
      "num_tokens": 73618943.0,
      "step": 42450
    },
    {
      "entropy": 5.838293409347534,
      "epoch": 3.3031316864423266,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039251491064753525,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20381409227848052,
      "num_tokens": 73627381.0,
      "step": 42455
    },
    {
      "entropy": 5.839942836761475,
      "epoch": 3.303520715814044,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003924912747659267,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21090895384550096,
      "num_tokens": 73636434.0,
      "step": 42460
    },
    {
      "entropy": 5.863178396224976,
      "epoch": 3.3039097451857615,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00039246763710157513,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20052160769701005,
      "num_tokens": 73644939.0,
      "step": 42465
    },
    {
      "entropy": 5.780732536315918,
      "epoch": 3.3042987745574792,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003924439976548391,
      "loss": 4.8799,
      "mean_token_accuracy": 0.21577879786491394,
      "num_tokens": 73653284.0,
      "step": 42470
    },
    {
      "entropy": 5.737558174133301,
      "epoch": 3.3046878039291965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039242035642607733,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22135815918445587,
      "num_tokens": 73661883.0,
      "step": 42475
    },
    {
      "entropy": 5.706303930282592,
      "epoch": 3.305076833300914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003923967134156486,
      "loss": 4.8914,
      "mean_token_accuracy": 0.20873292535543442,
      "num_tokens": 73670564.0,
      "step": 42480
    },
    {
      "entropy": 5.713171911239624,
      "epoch": 3.305465862672632,
      "grad_norm": 1.25,
      "learning_rate": 0.0003923730686239115,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21192733347415924,
      "num_tokens": 73679219.0,
      "step": 42485
    },
    {
      "entropy": 5.778691196441651,
      "epoch": 3.3058548920443496,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003923494220512249,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21103958338499068,
      "num_tokens": 73688358.0,
      "step": 42490
    },
    {
      "entropy": 5.888421297073364,
      "epoch": 3.306243921416067,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003923257736979476,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21228283345699311,
      "num_tokens": 73696761.0,
      "step": 42495
    },
    {
      "entropy": 5.857212781906128,
      "epoch": 3.3066329507877845,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039230212356443855,
      "loss": 4.9947,
      "mean_token_accuracy": 0.20153977423906327,
      "num_tokens": 73704549.0,
      "step": 42500
    },
    {
      "entropy": 5.743925857543945,
      "epoch": 3.307021980159502,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039227847165105635,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20347167253494264,
      "num_tokens": 73713533.0,
      "step": 42505
    },
    {
      "entropy": 5.780810260772705,
      "epoch": 3.3074110095312195,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039225481795816016,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22428275346755983,
      "num_tokens": 73722628.0,
      "step": 42510
    },
    {
      "entropy": 5.7934410572052,
      "epoch": 3.307800038902937,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003922311624861086,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22024836391210556,
      "num_tokens": 73730286.0,
      "step": 42515
    },
    {
      "entropy": 5.672967338562012,
      "epoch": 3.308189068274655,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003922075052352608,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21009422838687897,
      "num_tokens": 73738893.0,
      "step": 42520
    },
    {
      "entropy": 5.783132266998291,
      "epoch": 3.308578097646372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003921838462059756,
      "loss": 5.0162,
      "mean_token_accuracy": 0.20755598247051238,
      "num_tokens": 73747249.0,
      "step": 42525
    },
    {
      "entropy": 5.713928604125977,
      "epoch": 3.30896712701809,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039216018539861217,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21549345254898072,
      "num_tokens": 73755828.0,
      "step": 42530
    },
    {
      "entropy": 5.798518371582031,
      "epoch": 3.3093561563898075,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003921365228135293,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21458789259195327,
      "num_tokens": 73763913.0,
      "step": 42535
    },
    {
      "entropy": 5.735788726806641,
      "epoch": 3.3097451857615248,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039211285845108617,
      "loss": 4.88,
      "mean_token_accuracy": 0.20903635770082474,
      "num_tokens": 73772394.0,
      "step": 42540
    },
    {
      "entropy": 5.79608645439148,
      "epoch": 3.3101342151332425,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003920891923116418,
      "loss": 5.0125,
      "mean_token_accuracy": 0.200534188747406,
      "num_tokens": 73782004.0,
      "step": 42545
    },
    {
      "entropy": 5.751929378509521,
      "epoch": 3.31052324450496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039206552439555544,
      "loss": 4.929,
      "mean_token_accuracy": 0.21120332777500153,
      "num_tokens": 73790482.0,
      "step": 42550
    },
    {
      "entropy": 5.814033317565918,
      "epoch": 3.310912273876678,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003920418547031859,
      "loss": 4.961,
      "mean_token_accuracy": 0.21484753787517546,
      "num_tokens": 73799054.0,
      "step": 42555
    },
    {
      "entropy": 5.786821174621582,
      "epoch": 3.311301303248395,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003920181832348926,
      "loss": 4.92,
      "mean_token_accuracy": 0.21317008435726165,
      "num_tokens": 73807873.0,
      "step": 42560
    },
    {
      "entropy": 5.727629280090332,
      "epoch": 3.311690332620113,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003919945099910346,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21818889826536178,
      "num_tokens": 73816625.0,
      "step": 42565
    },
    {
      "entropy": 5.801596927642822,
      "epoch": 3.3120793619918305,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039197083497197107,
      "loss": 4.902,
      "mean_token_accuracy": 0.21404514610767364,
      "num_tokens": 73825411.0,
      "step": 42570
    },
    {
      "entropy": 5.807532691955567,
      "epoch": 3.3124683913635478,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003919471581780613,
      "loss": 5.0297,
      "mean_token_accuracy": 0.20387880951166154,
      "num_tokens": 73834594.0,
      "step": 42575
    },
    {
      "entropy": 5.779193687438965,
      "epoch": 3.3128574207352655,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003919234796096646,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21007105261087416,
      "num_tokens": 73842928.0,
      "step": 42580
    },
    {
      "entropy": 5.751088237762451,
      "epoch": 3.313246450106983,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003918997992671402,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21497204452753066,
      "num_tokens": 73851639.0,
      "step": 42585
    },
    {
      "entropy": 5.754175090789795,
      "epoch": 3.313635479478701,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003918761171508475,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2101210594177246,
      "num_tokens": 73860618.0,
      "step": 42590
    },
    {
      "entropy": 5.748440313339233,
      "epoch": 3.314024508850418,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003918524332611456,
      "loss": 4.8827,
      "mean_token_accuracy": 0.21692682802677155,
      "num_tokens": 73869140.0,
      "step": 42595
    },
    {
      "entropy": 5.769749402999878,
      "epoch": 3.314413538222136,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003918287475983941,
      "loss": 4.9351,
      "mean_token_accuracy": 0.208855564892292,
      "num_tokens": 73878248.0,
      "step": 42600
    },
    {
      "entropy": 5.722035121917725,
      "epoch": 3.3148025675938535,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039180506016295226,
      "loss": 4.9037,
      "mean_token_accuracy": 0.20788008719682693,
      "num_tokens": 73886743.0,
      "step": 42605
    },
    {
      "entropy": 5.833974456787109,
      "epoch": 3.3151915969655708,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039178137095517956,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20891356766223906,
      "num_tokens": 73895941.0,
      "step": 42610
    },
    {
      "entropy": 5.812851810455323,
      "epoch": 3.3155806263372885,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003917576799754354,
      "loss": 5.0416,
      "mean_token_accuracy": 0.19819609224796295,
      "num_tokens": 73904794.0,
      "step": 42615
    },
    {
      "entropy": 5.802658891677856,
      "epoch": 3.315969655709006,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003917339872240793,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20312890261411667,
      "num_tokens": 73913727.0,
      "step": 42620
    },
    {
      "entropy": 5.773225545883179,
      "epoch": 3.3163586850807234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003917102927014708,
      "loss": 5.0405,
      "mean_token_accuracy": 0.20601338744163514,
      "num_tokens": 73923036.0,
      "step": 42625
    },
    {
      "entropy": 5.861292743682862,
      "epoch": 3.316747714452441,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003916865964079694,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21079111695289612,
      "num_tokens": 73932036.0,
      "step": 42630
    },
    {
      "entropy": 5.849609899520874,
      "epoch": 3.317136743824159,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003916628983439345,
      "loss": 5.0467,
      "mean_token_accuracy": 0.19445690065622329,
      "num_tokens": 73941515.0,
      "step": 42635
    },
    {
      "entropy": 5.737329816818237,
      "epoch": 3.317525773195876,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039163919850972585,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21279737502336502,
      "num_tokens": 73949767.0,
      "step": 42640
    },
    {
      "entropy": 5.763401746749878,
      "epoch": 3.3179148025675937,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003916154969057031,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21319681406021118,
      "num_tokens": 73958687.0,
      "step": 42645
    },
    {
      "entropy": 5.751964521408081,
      "epoch": 3.3183038319393114,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039159179353222566,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21392656713724137,
      "num_tokens": 73967935.0,
      "step": 42650
    },
    {
      "entropy": 5.822303056716919,
      "epoch": 3.318692861311029,
      "grad_norm": 1.421875,
      "learning_rate": 0.00039156808838965336,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21493997126817704,
      "num_tokens": 73976230.0,
      "step": 42655
    },
    {
      "entropy": 5.800126695632935,
      "epoch": 3.3190818906827464,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039154438147834595,
      "loss": 4.856,
      "mean_token_accuracy": 0.221036858856678,
      "num_tokens": 73984412.0,
      "step": 42660
    },
    {
      "entropy": 5.754728937149048,
      "epoch": 3.319470920054464,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039152067279866294,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2092828407883644,
      "num_tokens": 73993114.0,
      "step": 42665
    },
    {
      "entropy": 5.718155765533448,
      "epoch": 3.319859949426182,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003914969623509643,
      "loss": 4.9288,
      "mean_token_accuracy": 0.20733820050954818,
      "num_tokens": 74002423.0,
      "step": 42670
    },
    {
      "entropy": 5.807724809646606,
      "epoch": 3.320248978797899,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003914732501356096,
      "loss": 4.914,
      "mean_token_accuracy": 0.2104165256023407,
      "num_tokens": 74011094.0,
      "step": 42675
    },
    {
      "entropy": 5.826362991333008,
      "epoch": 3.3206380081696167,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039144953615295877,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20615897923707963,
      "num_tokens": 74019054.0,
      "step": 42680
    },
    {
      "entropy": 5.782886743545532,
      "epoch": 3.3210270375413344,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039142582040337147,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20404196977615358,
      "num_tokens": 74027970.0,
      "step": 42685
    },
    {
      "entropy": 5.730244541168213,
      "epoch": 3.321416066913052,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003914021028872077,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21534445136785507,
      "num_tokens": 74036919.0,
      "step": 42690
    },
    {
      "entropy": 5.79663987159729,
      "epoch": 3.3218050962847694,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003913783836048274,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21359802782535553,
      "num_tokens": 74045707.0,
      "step": 42695
    },
    {
      "entropy": 5.819616651535034,
      "epoch": 3.322194125656487,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003913546625565902,
      "loss": 5.055,
      "mean_token_accuracy": 0.20681966543197633,
      "num_tokens": 74054858.0,
      "step": 42700
    },
    {
      "entropy": 5.844473028182984,
      "epoch": 3.322583155028205,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039133093974285635,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2052426278591156,
      "num_tokens": 74064253.0,
      "step": 42705
    },
    {
      "entropy": 5.809714412689209,
      "epoch": 3.322972184399922,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039130721516398556,
      "loss": 4.937,
      "mean_token_accuracy": 0.21493516117334366,
      "num_tokens": 74072824.0,
      "step": 42710
    },
    {
      "entropy": 5.821947002410889,
      "epoch": 3.3233612137716397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003912834888203379,
      "loss": 5.0498,
      "mean_token_accuracy": 0.20406683385372162,
      "num_tokens": 74082082.0,
      "step": 42715
    },
    {
      "entropy": 5.824991273880005,
      "epoch": 3.3237502431433574,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003912597607122734,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21919937878847123,
      "num_tokens": 74090748.0,
      "step": 42720
    },
    {
      "entropy": 5.851564073562622,
      "epoch": 3.324139272515075,
      "grad_norm": 1.25,
      "learning_rate": 0.00039123603084015204,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2002374231815338,
      "num_tokens": 74099645.0,
      "step": 42725
    },
    {
      "entropy": 5.75317349433899,
      "epoch": 3.3245283018867924,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039121229920433394,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21946839094161988,
      "num_tokens": 74108572.0,
      "step": 42730
    },
    {
      "entropy": 5.816953945159912,
      "epoch": 3.32491733125851,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003911885658051792,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2089310795068741,
      "num_tokens": 74116567.0,
      "step": 42735
    },
    {
      "entropy": 5.829708480834961,
      "epoch": 3.3253063606302278,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039116483064304794,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21370429247617723,
      "num_tokens": 74126086.0,
      "step": 42740
    },
    {
      "entropy": 5.7831205368042,
      "epoch": 3.325695390001945,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003911410937183002,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20667790472507477,
      "num_tokens": 74135160.0,
      "step": 42745
    },
    {
      "entropy": 5.835002470016479,
      "epoch": 3.3260844193736627,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039111735503129626,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2073177367448807,
      "num_tokens": 74143540.0,
      "step": 42750
    },
    {
      "entropy": 5.7581315517425535,
      "epoch": 3.3264734487453804,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003910936145823962,
      "loss": 4.947,
      "mean_token_accuracy": 0.21322102844715118,
      "num_tokens": 74152542.0,
      "step": 42755
    },
    {
      "entropy": 5.789960527420044,
      "epoch": 3.3268624781170977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003910698723719604,
      "loss": 4.907,
      "mean_token_accuracy": 0.21263383626937865,
      "num_tokens": 74161008.0,
      "step": 42760
    },
    {
      "entropy": 5.730138301849365,
      "epoch": 3.3272515074888154,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039104612840034904,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2155284494161606,
      "num_tokens": 74169637.0,
      "step": 42765
    },
    {
      "entropy": 5.788354206085205,
      "epoch": 3.327640536860533,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039102238266792234,
      "loss": 4.937,
      "mean_token_accuracy": 0.20898367315530778,
      "num_tokens": 74178868.0,
      "step": 42770
    },
    {
      "entropy": 5.8141467571258545,
      "epoch": 3.3280295662322503,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039099863517504064,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21088891327381135,
      "num_tokens": 74187935.0,
      "step": 42775
    },
    {
      "entropy": 5.8787782192230225,
      "epoch": 3.328418595603968,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003909748859220644,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20401460379362107,
      "num_tokens": 74197448.0,
      "step": 42780
    },
    {
      "entropy": 5.824393558502197,
      "epoch": 3.3288076249756857,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039095113490935377,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2054844617843628,
      "num_tokens": 74206188.0,
      "step": 42785
    },
    {
      "entropy": 5.802308082580566,
      "epoch": 3.3291966543474034,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003909273821372692,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21174070835113526,
      "num_tokens": 74214765.0,
      "step": 42790
    },
    {
      "entropy": 5.737492656707763,
      "epoch": 3.3295856837191207,
      "grad_norm": 1.375,
      "learning_rate": 0.00039090362760617114,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21215202659368515,
      "num_tokens": 74223461.0,
      "step": 42795
    },
    {
      "entropy": 5.773314666748047,
      "epoch": 3.3299747130908384,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039087987131642,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21124511808156968,
      "num_tokens": 74232486.0,
      "step": 42800
    },
    {
      "entropy": 5.788905239105224,
      "epoch": 3.330363742462556,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003908561132683762,
      "loss": 4.9972,
      "mean_token_accuracy": 0.20450852513313295,
      "num_tokens": 74241256.0,
      "step": 42805
    },
    {
      "entropy": 5.869234561920166,
      "epoch": 3.3307527718342733,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039083235346240044,
      "loss": 5.057,
      "mean_token_accuracy": 0.20581034421920777,
      "num_tokens": 74249661.0,
      "step": 42810
    },
    {
      "entropy": 5.763320875167847,
      "epoch": 3.331141801205991,
      "grad_norm": 1.171875,
      "learning_rate": 0.000390808591898853,
      "loss": 4.9104,
      "mean_token_accuracy": 0.21130876392126083,
      "num_tokens": 74258719.0,
      "step": 42815
    },
    {
      "entropy": 5.80388650894165,
      "epoch": 3.3315308305777087,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003907848285780945,
      "loss": 4.982,
      "mean_token_accuracy": 0.20749324113130568,
      "num_tokens": 74267181.0,
      "step": 42820
    },
    {
      "entropy": 5.865523338317871,
      "epoch": 3.3319198599494264,
      "grad_norm": 1.375,
      "learning_rate": 0.00039076106350048557,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20920634567737578,
      "num_tokens": 74275498.0,
      "step": 42825
    },
    {
      "entropy": 5.74682207107544,
      "epoch": 3.3323088893211437,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003907372966663867,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21665335744619368,
      "num_tokens": 74284090.0,
      "step": 42830
    },
    {
      "entropy": 5.790553426742553,
      "epoch": 3.3326979186928614,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003907135280761585,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21534194201231002,
      "num_tokens": 74292333.0,
      "step": 42835
    },
    {
      "entropy": 5.731353616714477,
      "epoch": 3.333086948064579,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003906897577301618,
      "loss": 4.8933,
      "mean_token_accuracy": 0.21302371323108674,
      "num_tokens": 74301131.0,
      "step": 42840
    },
    {
      "entropy": 5.726293325424194,
      "epoch": 3.3334759774362963,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003906659856287571,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20427053421735764,
      "num_tokens": 74310340.0,
      "step": 42845
    },
    {
      "entropy": 5.788184356689453,
      "epoch": 3.333865006808014,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039064221177230517,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2026902750134468,
      "num_tokens": 74318919.0,
      "step": 42850
    },
    {
      "entropy": 5.79738039970398,
      "epoch": 3.3342540361797317,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039061843616116683,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21613091081380845,
      "num_tokens": 74327909.0,
      "step": 42855
    },
    {
      "entropy": 5.832695627212525,
      "epoch": 3.334643065551449,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039059465879570264,
      "loss": 5.0437,
      "mean_token_accuracy": 0.209121373295784,
      "num_tokens": 74337017.0,
      "step": 42860
    },
    {
      "entropy": 5.790581130981446,
      "epoch": 3.3350320949231667,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003905708796762736,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20273246020078659,
      "num_tokens": 74345962.0,
      "step": 42865
    },
    {
      "entropy": 5.801639795303345,
      "epoch": 3.3354211242948844,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003905470988032403,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2052523449063301,
      "num_tokens": 74355214.0,
      "step": 42870
    },
    {
      "entropy": 5.82442193031311,
      "epoch": 3.3358101536666016,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003905233161769637,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2013792410492897,
      "num_tokens": 74364892.0,
      "step": 42875
    },
    {
      "entropy": 5.730095148086548,
      "epoch": 3.3361991830383193,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003904995317978048,
      "loss": 4.8157,
      "mean_token_accuracy": 0.21565012335777284,
      "num_tokens": 74373022.0,
      "step": 42880
    },
    {
      "entropy": 5.718376779556275,
      "epoch": 3.336588212410037,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039047574566612416,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21584296822547913,
      "num_tokens": 74381419.0,
      "step": 42885
    },
    {
      "entropy": 5.701881361007691,
      "epoch": 3.3369772417817547,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039045195778228295,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21503207683563233,
      "num_tokens": 74390131.0,
      "step": 42890
    },
    {
      "entropy": 5.840640544891357,
      "epoch": 3.337366271153472,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003904281681466421,
      "loss": 5.066,
      "mean_token_accuracy": 0.20779961943626404,
      "num_tokens": 74399401.0,
      "step": 42895
    },
    {
      "entropy": 5.750869941711426,
      "epoch": 3.3377553005251897,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003904043767595624,
      "loss": 4.8425,
      "mean_token_accuracy": 0.21412189304828644,
      "num_tokens": 74408301.0,
      "step": 42900
    },
    {
      "entropy": 5.767932891845703,
      "epoch": 3.3381443298969073,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039038058362140507,
      "loss": 4.9575,
      "mean_token_accuracy": 0.20647684782743453,
      "num_tokens": 74417718.0,
      "step": 42905
    },
    {
      "entropy": 5.735597467422485,
      "epoch": 3.3385333592686246,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039035678873253096,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2115159809589386,
      "num_tokens": 74427168.0,
      "step": 42910
    },
    {
      "entropy": 5.721801710128784,
      "epoch": 3.3389223886403423,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003903329920933012,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22096198052167892,
      "num_tokens": 74436135.0,
      "step": 42915
    },
    {
      "entropy": 5.7768796443939205,
      "epoch": 3.33931141801206,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003903091937040769,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21123828291893004,
      "num_tokens": 74444532.0,
      "step": 42920
    },
    {
      "entropy": 5.725009727478027,
      "epoch": 3.3397004473837777,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003902853935652191,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22060043066740037,
      "num_tokens": 74452513.0,
      "step": 42925
    },
    {
      "entropy": 5.752758932113648,
      "epoch": 3.340089476755495,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003902615916770889,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20614133179187774,
      "num_tokens": 74461561.0,
      "step": 42930
    },
    {
      "entropy": 5.774969625473022,
      "epoch": 3.3404785061272126,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039023778804004753,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21234968006610871,
      "num_tokens": 74471371.0,
      "step": 42935
    },
    {
      "entropy": 5.809992074966431,
      "epoch": 3.3408675354989303,
      "grad_norm": 1.25,
      "learning_rate": 0.00039021398265445616,
      "loss": 4.9633,
      "mean_token_accuracy": 0.20803303718566896,
      "num_tokens": 74480299.0,
      "step": 42940
    },
    {
      "entropy": 5.823718404769897,
      "epoch": 3.3412565648706476,
      "grad_norm": 1.359375,
      "learning_rate": 0.000390190175520676,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20968054831027985,
      "num_tokens": 74489511.0,
      "step": 42945
    },
    {
      "entropy": 5.811713218688965,
      "epoch": 3.3416455942423653,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003901663666390683,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2028544142842293,
      "num_tokens": 74497542.0,
      "step": 42950
    },
    {
      "entropy": 5.848036861419677,
      "epoch": 3.342034623614083,
      "grad_norm": 1.25,
      "learning_rate": 0.0003901425560099943,
      "loss": 5.0571,
      "mean_token_accuracy": 0.20871859639883042,
      "num_tokens": 74506906.0,
      "step": 42955
    },
    {
      "entropy": 5.75065050125122,
      "epoch": 3.3424236529858002,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039011874363381523,
      "loss": 4.9425,
      "mean_token_accuracy": 0.20543222278356552,
      "num_tokens": 74515688.0,
      "step": 42960
    },
    {
      "entropy": 5.773118877410889,
      "epoch": 3.342812682357518,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039009492951089236,
      "loss": 4.936,
      "mean_token_accuracy": 0.20458152890205383,
      "num_tokens": 74524405.0,
      "step": 42965
    },
    {
      "entropy": 5.803256464004517,
      "epoch": 3.3432017117292356,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003900711136415873,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22878348082304,
      "num_tokens": 74532438.0,
      "step": 42970
    },
    {
      "entropy": 5.784336566925049,
      "epoch": 3.343590741100953,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003900472960262611,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2126283273100853,
      "num_tokens": 74540979.0,
      "step": 42975
    },
    {
      "entropy": 5.747698736190796,
      "epoch": 3.3439797704726706,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039002347666527537,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22063359171152114,
      "num_tokens": 74549752.0,
      "step": 42980
    },
    {
      "entropy": 5.811239099502563,
      "epoch": 3.3443687998443883,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038999965555899145,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21068868041038513,
      "num_tokens": 74557861.0,
      "step": 42985
    },
    {
      "entropy": 5.80413293838501,
      "epoch": 3.344757829216106,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003899758327077709,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21017554849386216,
      "num_tokens": 74566024.0,
      "step": 42990
    },
    {
      "entropy": 5.73810715675354,
      "epoch": 3.3451468585878232,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038995200811197496,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2104707956314087,
      "num_tokens": 74574311.0,
      "step": 42995
    },
    {
      "entropy": 5.798538875579834,
      "epoch": 3.345535887959541,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038992818177196526,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21518526822328568,
      "num_tokens": 74582955.0,
      "step": 43000
    },
    {
      "entropy": 5.781771039962768,
      "epoch": 3.3459249173312586,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003899043536881034,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21596520841121675,
      "num_tokens": 74591132.0,
      "step": 43005
    },
    {
      "entropy": 5.744034194946289,
      "epoch": 3.346313946702976,
      "grad_norm": 1.4375,
      "learning_rate": 0.00038988052386075084,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2144334524869919,
      "num_tokens": 74598904.0,
      "step": 43010
    },
    {
      "entropy": 5.699103784561157,
      "epoch": 3.3467029760746936,
      "grad_norm": 1.375,
      "learning_rate": 0.00038985669229026917,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2130567416548729,
      "num_tokens": 74606774.0,
      "step": 43015
    },
    {
      "entropy": 5.817168283462524,
      "epoch": 3.3470920054464113,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038983285897702005,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21002042144536973,
      "num_tokens": 74614932.0,
      "step": 43020
    },
    {
      "entropy": 5.83107008934021,
      "epoch": 3.347481034818129,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038980902392136496,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2087450295686722,
      "num_tokens": 74623148.0,
      "step": 43025
    },
    {
      "entropy": 5.865071153640747,
      "epoch": 3.3478700641898462,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003897851871236657,
      "loss": 5.0427,
      "mean_token_accuracy": 0.19660421311855317,
      "num_tokens": 74631910.0,
      "step": 43030
    },
    {
      "entropy": 5.778626251220703,
      "epoch": 3.348259093561564,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038976134858428397,
      "loss": 4.89,
      "mean_token_accuracy": 0.22158281803131102,
      "num_tokens": 74640067.0,
      "step": 43035
    },
    {
      "entropy": 5.753057050704956,
      "epoch": 3.3486481229332816,
      "grad_norm": 1.375,
      "learning_rate": 0.00038973750830358144,
      "loss": 4.9578,
      "mean_token_accuracy": 0.20895665436983107,
      "num_tokens": 74648375.0,
      "step": 43040
    },
    {
      "entropy": 5.704550457000733,
      "epoch": 3.349037152304999,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038971366628191986,
      "loss": 4.829,
      "mean_token_accuracy": 0.21641192734241485,
      "num_tokens": 74657078.0,
      "step": 43045
    },
    {
      "entropy": 5.848461055755616,
      "epoch": 3.3494261816767166,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000389689822519661,
      "loss": 5.0355,
      "mean_token_accuracy": 0.20600672960281372,
      "num_tokens": 74666024.0,
      "step": 43050
    },
    {
      "entropy": 5.893258571624756,
      "epoch": 3.3498152110484343,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038966597701716653,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20516044348478318,
      "num_tokens": 74674750.0,
      "step": 43055
    },
    {
      "entropy": 5.725406217575073,
      "epoch": 3.350204240420152,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003896421297747985,
      "loss": 4.8844,
      "mean_token_accuracy": 0.21399151980876924,
      "num_tokens": 74683069.0,
      "step": 43060
    },
    {
      "entropy": 5.747315406799316,
      "epoch": 3.350593269791869,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003896182807929185,
      "loss": 4.9628,
      "mean_token_accuracy": 0.20986505448818207,
      "num_tokens": 74692351.0,
      "step": 43065
    },
    {
      "entropy": 5.772113943099976,
      "epoch": 3.350982299163587,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003895944300718885,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21815911382436753,
      "num_tokens": 74700646.0,
      "step": 43070
    },
    {
      "entropy": 5.768545532226563,
      "epoch": 3.351371328535304,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038957057761207053,
      "loss": 4.9877,
      "mean_token_accuracy": 0.209941466152668,
      "num_tokens": 74709364.0,
      "step": 43075
    },
    {
      "entropy": 5.833791351318359,
      "epoch": 3.351760357907022,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038954672341382635,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20199767649173736,
      "num_tokens": 74717905.0,
      "step": 43080
    },
    {
      "entropy": 5.792632913589477,
      "epoch": 3.3521493872787396,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000389522867477518,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22230317145586015,
      "num_tokens": 74726648.0,
      "step": 43085
    },
    {
      "entropy": 5.856190013885498,
      "epoch": 3.3525384166504573,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003894990098035074,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20897010415792466,
      "num_tokens": 74735362.0,
      "step": 43090
    },
    {
      "entropy": 5.838282346725464,
      "epoch": 3.3529274460221745,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003894751503921566,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21451593637466432,
      "num_tokens": 74744560.0,
      "step": 43095
    },
    {
      "entropy": 5.771103429794311,
      "epoch": 3.353316475393892,
      "grad_norm": 1.375,
      "learning_rate": 0.0003894512892438275,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23024735301733018,
      "num_tokens": 74752951.0,
      "step": 43100
    },
    {
      "entropy": 5.764113473892212,
      "epoch": 3.35370550476561,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038942742635888236,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21502251625061036,
      "num_tokens": 74761664.0,
      "step": 43105
    },
    {
      "entropy": 5.764804983139038,
      "epoch": 3.354094534137327,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038940356173768303,
      "loss": 4.849,
      "mean_token_accuracy": 0.21943265646696092,
      "num_tokens": 74770031.0,
      "step": 43110
    },
    {
      "entropy": 5.675157165527343,
      "epoch": 3.354483563509045,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003893796953805919,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22137391716241836,
      "num_tokens": 74778425.0,
      "step": 43115
    },
    {
      "entropy": 5.738363647460938,
      "epoch": 3.3548725928807626,
      "grad_norm": 1.25,
      "learning_rate": 0.00038935582728797085,
      "loss": 4.9708,
      "mean_token_accuracy": 0.20903128981590272,
      "num_tokens": 74787039.0,
      "step": 43120
    },
    {
      "entropy": 5.8077332973480225,
      "epoch": 3.3552616222524803,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038933195746018213,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2160432368516922,
      "num_tokens": 74795458.0,
      "step": 43125
    },
    {
      "entropy": 5.761938142776489,
      "epoch": 3.3556506516241975,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038930808589758797,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21908112317323686,
      "num_tokens": 74803922.0,
      "step": 43130
    },
    {
      "entropy": 5.819369506835938,
      "epoch": 3.356039680995915,
      "grad_norm": 1.25,
      "learning_rate": 0.0003892842126005506,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20671227276325227,
      "num_tokens": 74813199.0,
      "step": 43135
    },
    {
      "entropy": 5.788135099411011,
      "epoch": 3.356428710367633,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003892603375694321,
      "loss": 4.9376,
      "mean_token_accuracy": 0.21227660328149794,
      "num_tokens": 74822052.0,
      "step": 43140
    },
    {
      "entropy": 5.898581552505493,
      "epoch": 3.35681773973935,
      "grad_norm": 1.453125,
      "learning_rate": 0.00038923646080459484,
      "loss": 5.1211,
      "mean_token_accuracy": 0.1966175228357315,
      "num_tokens": 74829771.0,
      "step": 43145
    },
    {
      "entropy": 5.831902885437012,
      "epoch": 3.357206769111068,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038921258230640123,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20399683862924575,
      "num_tokens": 74839054.0,
      "step": 43150
    },
    {
      "entropy": 5.822342777252198,
      "epoch": 3.3575957984827856,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038918870207521334,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20010855197906494,
      "num_tokens": 74847830.0,
      "step": 43155
    },
    {
      "entropy": 5.83560700416565,
      "epoch": 3.3579848278545033,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038916482011139373,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21636561304330826,
      "num_tokens": 74856621.0,
      "step": 43160
    },
    {
      "entropy": 5.731892108917236,
      "epoch": 3.3583738572262205,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003891409364153047,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21556456387043,
      "num_tokens": 74865094.0,
      "step": 43165
    },
    {
      "entropy": 5.7617998123168945,
      "epoch": 3.358762886597938,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003891170509873086,
      "loss": 4.9803,
      "mean_token_accuracy": 0.20563586801290512,
      "num_tokens": 74874821.0,
      "step": 43170
    },
    {
      "entropy": 5.828053045272827,
      "epoch": 3.359151915969656,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00038909316382776784,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21250070184469222,
      "num_tokens": 74883047.0,
      "step": 43175
    },
    {
      "entropy": 5.790762996673584,
      "epoch": 3.359540945341373,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003890692749370449,
      "loss": 4.9573,
      "mean_token_accuracy": 0.20824091881513596,
      "num_tokens": 74892462.0,
      "step": 43180
    },
    {
      "entropy": 5.683311462402344,
      "epoch": 3.359929974713091,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038904538431550235,
      "loss": 4.8154,
      "mean_token_accuracy": 0.21600456684827804,
      "num_tokens": 74900584.0,
      "step": 43185
    },
    {
      "entropy": 5.7048932075500485,
      "epoch": 3.3603190040848085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003890214919635026,
      "loss": 4.8484,
      "mean_token_accuracy": 0.2169600933790207,
      "num_tokens": 74909300.0,
      "step": 43190
    },
    {
      "entropy": 5.736833190917968,
      "epoch": 3.360708033456526,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003889975978814082,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21156176924705505,
      "num_tokens": 74917781.0,
      "step": 43195
    },
    {
      "entropy": 5.778401470184326,
      "epoch": 3.3610970628282435,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003889737020695817,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21896960586309433,
      "num_tokens": 74926566.0,
      "step": 43200
    },
    {
      "entropy": 5.763939428329468,
      "epoch": 3.361486092199961,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038894980452838573,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22024949938058852,
      "num_tokens": 74934956.0,
      "step": 43205
    },
    {
      "entropy": 5.767733192443847,
      "epoch": 3.3618751215716784,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003889259052581828,
      "loss": 4.941,
      "mean_token_accuracy": 0.21761380434036254,
      "num_tokens": 74943491.0,
      "step": 43210
    },
    {
      "entropy": 5.859687948226929,
      "epoch": 3.362264150943396,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038890200425933545,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21021403074264527,
      "num_tokens": 74952486.0,
      "step": 43215
    },
    {
      "entropy": 5.793098735809326,
      "epoch": 3.362653180315114,
      "grad_norm": 1.375,
      "learning_rate": 0.0003888781015322067,
      "loss": 4.968,
      "mean_token_accuracy": 0.21527902334928511,
      "num_tokens": 74960563.0,
      "step": 43220
    },
    {
      "entropy": 5.731127071380615,
      "epoch": 3.3630422096868315,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038885419707715885,
      "loss": 4.8783,
      "mean_token_accuracy": 0.21830789744853973,
      "num_tokens": 74968974.0,
      "step": 43225
    },
    {
      "entropy": 5.883921480178833,
      "epoch": 3.363431239058549,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038883029089455487,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20537883043289185,
      "num_tokens": 74976905.0,
      "step": 43230
    },
    {
      "entropy": 5.861655044555664,
      "epoch": 3.3638202684302665,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003888063829847574,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20276482701301574,
      "num_tokens": 74986112.0,
      "step": 43235
    },
    {
      "entropy": 5.804697275161743,
      "epoch": 3.364209297801984,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003887824733481292,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21133866906166077,
      "num_tokens": 74994951.0,
      "step": 43240
    },
    {
      "entropy": 5.855419063568116,
      "epoch": 3.3645983271737014,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038875856198503306,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2072945088148117,
      "num_tokens": 75003266.0,
      "step": 43245
    },
    {
      "entropy": 5.797330617904663,
      "epoch": 3.364987356545419,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038873464889583177,
      "loss": 4.9765,
      "mean_token_accuracy": 0.20561964809894562,
      "num_tokens": 75012619.0,
      "step": 43250
    },
    {
      "entropy": 5.7820902347564695,
      "epoch": 3.365376385917137,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038871073408088826,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20851636976003646,
      "num_tokens": 75021269.0,
      "step": 43255
    },
    {
      "entropy": 5.840204572677612,
      "epoch": 3.3657654152888545,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003886868175405653,
      "loss": 5.0197,
      "mean_token_accuracy": 0.20724011063575745,
      "num_tokens": 75029944.0,
      "step": 43260
    },
    {
      "entropy": 5.877769851684571,
      "epoch": 3.366154444660572,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003886628992752259,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21159321814775467,
      "num_tokens": 75038884.0,
      "step": 43265
    },
    {
      "entropy": 5.8844305038452145,
      "epoch": 3.3665434740322895,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003886389792852329,
      "loss": 5.009,
      "mean_token_accuracy": 0.21310233771800996,
      "num_tokens": 75047101.0,
      "step": 43270
    },
    {
      "entropy": 5.737908792495728,
      "epoch": 3.366932503404007,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038861505757094923,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21620312184095383,
      "num_tokens": 75055778.0,
      "step": 43275
    },
    {
      "entropy": 5.771483755111694,
      "epoch": 3.3673215327757244,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003885911341327379,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21473979204893112,
      "num_tokens": 75065329.0,
      "step": 43280
    },
    {
      "entropy": 5.77823805809021,
      "epoch": 3.367710562147442,
      "grad_norm": 1.28125,
      "learning_rate": 0.000388567208970962,
      "loss": 4.9885,
      "mean_token_accuracy": 0.20329664051532745,
      "num_tokens": 75074319.0,
      "step": 43285
    },
    {
      "entropy": 5.690600967407226,
      "epoch": 3.36809959151916,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003885432820859844,
      "loss": 4.8415,
      "mean_token_accuracy": 0.21449536085128784,
      "num_tokens": 75083134.0,
      "step": 43290
    },
    {
      "entropy": 5.814583015441895,
      "epoch": 3.368488620890877,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038851935347816827,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20102193355560302,
      "num_tokens": 75090511.0,
      "step": 43295
    },
    {
      "entropy": 5.801841068267822,
      "epoch": 3.368877650262595,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003884954231478764,
      "loss": 4.9541,
      "mean_token_accuracy": 0.20990679562091827,
      "num_tokens": 75100398.0,
      "step": 43300
    },
    {
      "entropy": 5.787739181518555,
      "epoch": 3.3692666796343125,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003884714910954724,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20340281575918198,
      "num_tokens": 75109565.0,
      "step": 43305
    },
    {
      "entropy": 5.742578744888306,
      "epoch": 3.3696557090060297,
      "grad_norm": 1.328125,
      "learning_rate": 0.000388447557321319,
      "loss": 4.92,
      "mean_token_accuracy": 0.21569794714450835,
      "num_tokens": 75117906.0,
      "step": 43310
    },
    {
      "entropy": 5.854322052001953,
      "epoch": 3.3700447383777474,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003884236218257795,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21277662962675095,
      "num_tokens": 75126880.0,
      "step": 43315
    },
    {
      "entropy": 5.863040161132813,
      "epoch": 3.370433767749465,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003883996846092171,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2069724142551422,
      "num_tokens": 75135235.0,
      "step": 43320
    },
    {
      "entropy": 5.788700342178345,
      "epoch": 3.370822797121183,
      "grad_norm": 1.328125,
      "learning_rate": 0.000388375745671995,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2088899463415146,
      "num_tokens": 75144340.0,
      "step": 43325
    },
    {
      "entropy": 5.778656625747681,
      "epoch": 3.3712118264929,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038835180501447646,
      "loss": 4.9476,
      "mean_token_accuracy": 0.20824449509382248,
      "num_tokens": 75152473.0,
      "step": 43330
    },
    {
      "entropy": 5.912785816192627,
      "epoch": 3.3716008558646178,
      "grad_norm": 1.375,
      "learning_rate": 0.0003883278626370245,
      "loss": 5.0259,
      "mean_token_accuracy": 0.20538419038057326,
      "num_tokens": 75161326.0,
      "step": 43335
    },
    {
      "entropy": 5.818997764587403,
      "epoch": 3.3719898852363355,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038830391854000285,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2076309710741043,
      "num_tokens": 75169798.0,
      "step": 43340
    },
    {
      "entropy": 5.793631076812744,
      "epoch": 3.3723789146080527,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038827997272377447,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21255874037742614,
      "num_tokens": 75178678.0,
      "step": 43345
    },
    {
      "entropy": 5.77736005783081,
      "epoch": 3.3727679439797704,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038825602518870276,
      "loss": 5.0682,
      "mean_token_accuracy": 0.2014079809188843,
      "num_tokens": 75187639.0,
      "step": 43350
    },
    {
      "entropy": 5.82472791671753,
      "epoch": 3.373156973351488,
      "grad_norm": 1.375,
      "learning_rate": 0.0003882320759351512,
      "loss": 4.9585,
      "mean_token_accuracy": 0.20840444266796113,
      "num_tokens": 75195822.0,
      "step": 43355
    },
    {
      "entropy": 5.828814554214477,
      "epoch": 3.373546002723206,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00038820812496348305,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2051539972424507,
      "num_tokens": 75205303.0,
      "step": 43360
    },
    {
      "entropy": 5.79276967048645,
      "epoch": 3.373935032094923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038818417227406185,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2132664516568184,
      "num_tokens": 75214159.0,
      "step": 43365
    },
    {
      "entropy": 5.788076639175415,
      "epoch": 3.3743240614666408,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038816021786725096,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21185071468353273,
      "num_tokens": 75222822.0,
      "step": 43370
    },
    {
      "entropy": 5.836170244216919,
      "epoch": 3.3747130908383585,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038813626174341386,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20198375284671782,
      "num_tokens": 75230832.0,
      "step": 43375
    },
    {
      "entropy": 5.815476608276367,
      "epoch": 3.3751021202100757,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003881123039029141,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20886081457138062,
      "num_tokens": 75239752.0,
      "step": 43380
    },
    {
      "entropy": 5.842246627807617,
      "epoch": 3.3754911495817934,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003880883443461152,
      "loss": 5.0418,
      "mean_token_accuracy": 0.20313954055309297,
      "num_tokens": 75248380.0,
      "step": 43385
    },
    {
      "entropy": 5.823566102981568,
      "epoch": 3.375880178953511,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003880643830733806,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2168781951069832,
      "num_tokens": 75257082.0,
      "step": 43390
    },
    {
      "entropy": 5.789053821563721,
      "epoch": 3.3762692083252284,
      "grad_norm": 1.4375,
      "learning_rate": 0.00038804042008507396,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21087796986103058,
      "num_tokens": 75266097.0,
      "step": 43395
    },
    {
      "entropy": 5.757659959793091,
      "epoch": 3.376658237696946,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038801645538155883,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21146950274705886,
      "num_tokens": 75274658.0,
      "step": 43400
    },
    {
      "entropy": 5.805580949783325,
      "epoch": 3.3770472670686638,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038799248896319896,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2117716059088707,
      "num_tokens": 75283222.0,
      "step": 43405
    },
    {
      "entropy": 5.813009357452392,
      "epoch": 3.377436296440381,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038796852083035794,
      "loss": 4.9395,
      "mean_token_accuracy": 0.21019311398267745,
      "num_tokens": 75291091.0,
      "step": 43410
    },
    {
      "entropy": 5.742338037490844,
      "epoch": 3.3778253258120987,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003879445509833993,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2218306541442871,
      "num_tokens": 75299468.0,
      "step": 43415
    },
    {
      "entropy": 5.854673242568969,
      "epoch": 3.3782143551838164,
      "grad_norm": 1.421875,
      "learning_rate": 0.000387920579422687,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21270017921924592,
      "num_tokens": 75308213.0,
      "step": 43420
    },
    {
      "entropy": 5.820257043838501,
      "epoch": 3.378603384555534,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003878966061485845,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21120356917381286,
      "num_tokens": 75317290.0,
      "step": 43425
    },
    {
      "entropy": 5.8295081615447994,
      "epoch": 3.3789924139272514,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038787263116145577,
      "loss": 4.893,
      "mean_token_accuracy": 0.21402274668216706,
      "num_tokens": 75326100.0,
      "step": 43430
    },
    {
      "entropy": 5.727777767181396,
      "epoch": 3.379381443298969,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038784865446166454,
      "loss": 4.807,
      "mean_token_accuracy": 0.22277100384235382,
      "num_tokens": 75334418.0,
      "step": 43435
    },
    {
      "entropy": 5.706051635742187,
      "epoch": 3.3797704726706868,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038782467604957455,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22450652867555618,
      "num_tokens": 75342760.0,
      "step": 43440
    },
    {
      "entropy": 5.686639785766602,
      "epoch": 3.380159502042404,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038780069592554977,
      "loss": 4.9433,
      "mean_token_accuracy": 0.20826275050640106,
      "num_tokens": 75351440.0,
      "step": 43445
    },
    {
      "entropy": 5.777030277252197,
      "epoch": 3.3805485314141217,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003877767140899539,
      "loss": 4.9399,
      "mean_token_accuracy": 0.20552217811346055,
      "num_tokens": 75359594.0,
      "step": 43450
    },
    {
      "entropy": 5.931403970718383,
      "epoch": 3.3809375607858394,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038775273054315087,
      "loss": 5.1189,
      "mean_token_accuracy": 0.20903530567884446,
      "num_tokens": 75368281.0,
      "step": 43455
    },
    {
      "entropy": 5.703004980087281,
      "epoch": 3.381326590157557,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003877287452855046,
      "loss": 4.8548,
      "mean_token_accuracy": 0.21662104427814483,
      "num_tokens": 75376584.0,
      "step": 43460
    },
    {
      "entropy": 5.776071834564209,
      "epoch": 3.3817156195292744,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038770475831737905,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2007954850792885,
      "num_tokens": 75385041.0,
      "step": 43465
    },
    {
      "entropy": 5.736465740203857,
      "epoch": 3.382104648900992,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038768076963913823,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21868306994438172,
      "num_tokens": 75393885.0,
      "step": 43470
    },
    {
      "entropy": 5.768190526962281,
      "epoch": 3.3824936782727097,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038765677925114614,
      "loss": 4.9897,
      "mean_token_accuracy": 0.20529239773750305,
      "num_tokens": 75402199.0,
      "step": 43475
    },
    {
      "entropy": 5.7986400604248045,
      "epoch": 3.382882707644427,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003876327871537667,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21075936257839203,
      "num_tokens": 75411073.0,
      "step": 43480
    },
    {
      "entropy": 5.83245620727539,
      "epoch": 3.3832717370161447,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038760879334736394,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21377402395009995,
      "num_tokens": 75420245.0,
      "step": 43485
    },
    {
      "entropy": 5.799361371994019,
      "epoch": 3.3836607663878624,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038758479783230197,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21208999156951905,
      "num_tokens": 75428945.0,
      "step": 43490
    },
    {
      "entropy": 5.761487579345703,
      "epoch": 3.38404979575958,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003875608006089449,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21865946650505066,
      "num_tokens": 75437054.0,
      "step": 43495
    },
    {
      "entropy": 5.7555159568786625,
      "epoch": 3.3844388251312973,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003875368016776569,
      "loss": 4.9021,
      "mean_token_accuracy": 0.214740751683712,
      "num_tokens": 75445774.0,
      "step": 43500
    },
    {
      "entropy": 5.851074934005737,
      "epoch": 3.384827854503015,
      "grad_norm": 1.28125,
      "learning_rate": 0.000387512801038802,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20817840248346328,
      "num_tokens": 75454358.0,
      "step": 43505
    },
    {
      "entropy": 5.758426237106323,
      "epoch": 3.3852168838747323,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003874887986927444,
      "loss": 4.9029,
      "mean_token_accuracy": 0.21534413993358612,
      "num_tokens": 75463422.0,
      "step": 43510
    },
    {
      "entropy": 5.714532136917114,
      "epoch": 3.38560591324645,
      "grad_norm": 1.25,
      "learning_rate": 0.0003874647946398484,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21564772129058837,
      "num_tokens": 75472592.0,
      "step": 43515
    },
    {
      "entropy": 5.76764931678772,
      "epoch": 3.3859949426181677,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003874407888804781,
      "loss": 5.0006,
      "mean_token_accuracy": 0.21047221273183822,
      "num_tokens": 75481164.0,
      "step": 43520
    },
    {
      "entropy": 5.920309352874756,
      "epoch": 3.3863839719898854,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038741678141499783,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20157782286405562,
      "num_tokens": 75491061.0,
      "step": 43525
    },
    {
      "entropy": 5.842792272567749,
      "epoch": 3.3867730013616026,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003873927722437718,
      "loss": 4.9787,
      "mean_token_accuracy": 0.21324381828308106,
      "num_tokens": 75499791.0,
      "step": 43530
    },
    {
      "entropy": 5.832587337493896,
      "epoch": 3.3871620307333203,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003873687613671643,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21257290095090867,
      "num_tokens": 75509380.0,
      "step": 43535
    },
    {
      "entropy": 5.769758033752441,
      "epoch": 3.387551060105038,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003873447487855398,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21691685169935226,
      "num_tokens": 75519333.0,
      "step": 43540
    },
    {
      "entropy": 5.731731176376343,
      "epoch": 3.3879400894767553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038732073449926244,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22346869707107545,
      "num_tokens": 75527736.0,
      "step": 43545
    },
    {
      "entropy": 5.742165184020996,
      "epoch": 3.388329118848473,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003872967185086968,
      "loss": 4.9186,
      "mean_token_accuracy": 0.20924646705389022,
      "num_tokens": 75536877.0,
      "step": 43550
    },
    {
      "entropy": 5.817529249191284,
      "epoch": 3.3887181482201907,
      "grad_norm": 1.375,
      "learning_rate": 0.0003872727008142072,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21145665198564528,
      "num_tokens": 75544887.0,
      "step": 43555
    },
    {
      "entropy": 5.852278757095337,
      "epoch": 3.3891071775919084,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038724868141615807,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2157441794872284,
      "num_tokens": 75552951.0,
      "step": 43560
    },
    {
      "entropy": 5.724443864822388,
      "epoch": 3.3894962069636256,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003872246603149138,
      "loss": 4.904,
      "mean_token_accuracy": 0.21671249568462372,
      "num_tokens": 75561828.0,
      "step": 43565
    },
    {
      "entropy": 5.78367247581482,
      "epoch": 3.3898852363353433,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003872006375108389,
      "loss": 4.959,
      "mean_token_accuracy": 0.2085007384419441,
      "num_tokens": 75570552.0,
      "step": 43570
    },
    {
      "entropy": 5.7697676658630375,
      "epoch": 3.390274265707061,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000387176613004298,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21136243939399718,
      "num_tokens": 75579324.0,
      "step": 43575
    },
    {
      "entropy": 5.7815851211547855,
      "epoch": 3.3906632950787783,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003871525867956555,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2158256486058235,
      "num_tokens": 75588027.0,
      "step": 43580
    },
    {
      "entropy": 5.809529876708984,
      "epoch": 3.391052324450496,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038712855888527604,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20674671828746796,
      "num_tokens": 75596374.0,
      "step": 43585
    },
    {
      "entropy": 5.797618103027344,
      "epoch": 3.3914413538222137,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038710452927352414,
      "loss": 4.968,
      "mean_token_accuracy": 0.2059774398803711,
      "num_tokens": 75605820.0,
      "step": 43590
    },
    {
      "entropy": 5.724615859985351,
      "epoch": 3.3918303831939314,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003870804979607645,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21301112025976182,
      "num_tokens": 75615822.0,
      "step": 43595
    },
    {
      "entropy": 5.809083032608032,
      "epoch": 3.3922194125656486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038705646494736163,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21521904021501542,
      "num_tokens": 75624409.0,
      "step": 43600
    },
    {
      "entropy": 5.8270384788513185,
      "epoch": 3.3926084419373663,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003870324302336802,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2111015722155571,
      "num_tokens": 75632190.0,
      "step": 43605
    },
    {
      "entropy": 5.81065731048584,
      "epoch": 3.392997471309084,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003870083938200851,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20971662551164627,
      "num_tokens": 75640033.0,
      "step": 43610
    },
    {
      "entropy": 5.784988355636597,
      "epoch": 3.3933865006808013,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038698435570694084,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2114199310541153,
      "num_tokens": 75648598.0,
      "step": 43615
    },
    {
      "entropy": 5.806748485565185,
      "epoch": 3.393775530052519,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003869603158946122,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2127452477812767,
      "num_tokens": 75656841.0,
      "step": 43620
    },
    {
      "entropy": 5.76475157737732,
      "epoch": 3.3941645594242367,
      "grad_norm": 1.296875,
      "learning_rate": 0.000386936274383464,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22006040066480637,
      "num_tokens": 75664918.0,
      "step": 43625
    },
    {
      "entropy": 5.770832586288452,
      "epoch": 3.394553588795954,
      "grad_norm": 1.265625,
      "learning_rate": 0.000386912231173861,
      "loss": 4.865,
      "mean_token_accuracy": 0.21892206370830536,
      "num_tokens": 75672800.0,
      "step": 43630
    },
    {
      "entropy": 5.753575038909912,
      "epoch": 3.3949426181676716,
      "grad_norm": 1.234375,
      "learning_rate": 0.000386888186266168,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20871349722146987,
      "num_tokens": 75681669.0,
      "step": 43635
    },
    {
      "entropy": 5.861952304840088,
      "epoch": 3.3953316475393893,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003868641396607499,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2131394699215889,
      "num_tokens": 75690109.0,
      "step": 43640
    },
    {
      "entropy": 5.759485483169556,
      "epoch": 3.3957206769111066,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003868400913579715,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21360743939876556,
      "num_tokens": 75698395.0,
      "step": 43645
    },
    {
      "entropy": 5.79687533378601,
      "epoch": 3.3961097062828243,
      "grad_norm": 1.375,
      "learning_rate": 0.00038681604135819774,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21640504747629166,
      "num_tokens": 75707515.0,
      "step": 43650
    },
    {
      "entropy": 5.754794311523438,
      "epoch": 3.396498735654542,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038679198966179353,
      "loss": 4.8289,
      "mean_token_accuracy": 0.22756207585334778,
      "num_tokens": 75715725.0,
      "step": 43655
    },
    {
      "entropy": 5.739500331878662,
      "epoch": 3.3968877650262597,
      "grad_norm": 1.25,
      "learning_rate": 0.00038676793626912375,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2099432736635208,
      "num_tokens": 75723969.0,
      "step": 43660
    },
    {
      "entropy": 5.804638814926148,
      "epoch": 3.397276794397977,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038674388118055354,
      "loss": 5.0012,
      "mean_token_accuracy": 0.20546324849128722,
      "num_tokens": 75732185.0,
      "step": 43665
    },
    {
      "entropy": 5.861702108383179,
      "epoch": 3.3976658237696946,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003867198243964477,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21136565804481505,
      "num_tokens": 75740523.0,
      "step": 43670
    },
    {
      "entropy": 5.770979356765747,
      "epoch": 3.3980548531414123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003866957659171714,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2189035475254059,
      "num_tokens": 75749478.0,
      "step": 43675
    },
    {
      "entropy": 5.730063629150391,
      "epoch": 3.3984438825131296,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003866717057430896,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21349195688962935,
      "num_tokens": 75758202.0,
      "step": 43680
    },
    {
      "entropy": 5.8633922100067135,
      "epoch": 3.3988329118848473,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003866476438745675,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20707566291093826,
      "num_tokens": 75766304.0,
      "step": 43685
    },
    {
      "entropy": 5.818242788314819,
      "epoch": 3.399221941256565,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00038662358031197,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21161001473665236,
      "num_tokens": 75774140.0,
      "step": 43690
    },
    {
      "entropy": 5.830604982376099,
      "epoch": 3.3996109706282827,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003865995150556624,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21068360209465026,
      "num_tokens": 75783055.0,
      "step": 43695
    },
    {
      "entropy": 5.7711621761322025,
      "epoch": 3.4,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038657544810600966,
      "loss": 4.8836,
      "mean_token_accuracy": 0.2177087351679802,
      "num_tokens": 75791158.0,
      "step": 43700
    },
    {
      "entropy": 5.779644727706909,
      "epoch": 3.4003890293717176,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003865513794633773,
      "loss": 4.9761,
      "mean_token_accuracy": 0.20983498692512512,
      "num_tokens": 75800254.0,
      "step": 43705
    },
    {
      "entropy": 5.792922496795654,
      "epoch": 3.4007780587434353,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003865273091281301,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2085151955485344,
      "num_tokens": 75808782.0,
      "step": 43710
    },
    {
      "entropy": 5.797131443023682,
      "epoch": 3.4011670881151526,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003865032371006336,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21661184281110762,
      "num_tokens": 75817767.0,
      "step": 43715
    },
    {
      "entropy": 5.747809410095215,
      "epoch": 3.4015561174868703,
      "grad_norm": 1.375,
      "learning_rate": 0.000386479163381253,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2166934937238693,
      "num_tokens": 75826052.0,
      "step": 43720
    },
    {
      "entropy": 5.7999084949493405,
      "epoch": 3.401945146858588,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038645508797035345,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20328622311353683,
      "num_tokens": 75835189.0,
      "step": 43725
    },
    {
      "entropy": 5.849588632583618,
      "epoch": 3.402334176230305,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038643101086830036,
      "loss": 4.967,
      "mean_token_accuracy": 0.21421560347080232,
      "num_tokens": 75843933.0,
      "step": 43730
    },
    {
      "entropy": 5.737955379486084,
      "epoch": 3.402723205602023,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000386406932075459,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22456381916999818,
      "num_tokens": 75852155.0,
      "step": 43735
    },
    {
      "entropy": 5.7874391078948975,
      "epoch": 3.4031122349737406,
      "grad_norm": 1.375,
      "learning_rate": 0.0003863828515921948,
      "loss": 4.962,
      "mean_token_accuracy": 0.2121907502412796,
      "num_tokens": 75861104.0,
      "step": 43740
    },
    {
      "entropy": 5.825198173522949,
      "epoch": 3.403501264345458,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003863587694188732,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21612623780965806,
      "num_tokens": 75869831.0,
      "step": 43745
    },
    {
      "entropy": 5.820079755783081,
      "epoch": 3.4038902937171756,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038633468555585943,
      "loss": 5.054,
      "mean_token_accuracy": 0.20010418742895125,
      "num_tokens": 75878202.0,
      "step": 43750
    },
    {
      "entropy": 5.778425455093384,
      "epoch": 3.4042793230888932,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003863106000035191,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21483632326126098,
      "num_tokens": 75886730.0,
      "step": 43755
    },
    {
      "entropy": 5.815909767150879,
      "epoch": 3.404668352460611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003862865127622175,
      "loss": 4.9843,
      "mean_token_accuracy": 0.20829276293516158,
      "num_tokens": 75895389.0,
      "step": 43760
    },
    {
      "entropy": 5.876795053482056,
      "epoch": 3.405057381832328,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003862624238323201,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20841635167598724,
      "num_tokens": 75904177.0,
      "step": 43765
    },
    {
      "entropy": 5.826126766204834,
      "epoch": 3.405446411204046,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003862383332141927,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20877449363470077,
      "num_tokens": 75913150.0,
      "step": 43770
    },
    {
      "entropy": 5.831707000732422,
      "epoch": 3.4058354405757636,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003862142409082006,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20937421321868896,
      "num_tokens": 75921801.0,
      "step": 43775
    },
    {
      "entropy": 5.812568235397339,
      "epoch": 3.406224469947481,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003861901469147093,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2069947674870491,
      "num_tokens": 75930261.0,
      "step": 43780
    },
    {
      "entropy": 5.8466959476470945,
      "epoch": 3.4066134993191985,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038616605123408464,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20563859790563582,
      "num_tokens": 75938168.0,
      "step": 43785
    },
    {
      "entropy": 5.800450944900513,
      "epoch": 3.4070025286909162,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038614195386669207,
      "loss": 4.9802,
      "mean_token_accuracy": 0.20323264300823213,
      "num_tokens": 75947095.0,
      "step": 43790
    },
    {
      "entropy": 5.708704042434692,
      "epoch": 3.407391558062634,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003861178548128972,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22108725011348723,
      "num_tokens": 75955546.0,
      "step": 43795
    },
    {
      "entropy": 5.74482741355896,
      "epoch": 3.407780587434351,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038609375407306566,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20377250462770463,
      "num_tokens": 75964434.0,
      "step": 43800
    },
    {
      "entropy": 5.834991407394409,
      "epoch": 3.408169616806069,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038606965164756334,
      "loss": 4.9317,
      "mean_token_accuracy": 0.20649501085281372,
      "num_tokens": 75972657.0,
      "step": 43805
    },
    {
      "entropy": 5.867269802093506,
      "epoch": 3.4085586461777866,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038604554753675585,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2048252433538437,
      "num_tokens": 75981488.0,
      "step": 43810
    },
    {
      "entropy": 5.820777320861817,
      "epoch": 3.408947675549504,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003860214417410089,
      "loss": 4.943,
      "mean_token_accuracy": 0.21125797033309937,
      "num_tokens": 75989901.0,
      "step": 43815
    },
    {
      "entropy": 5.823960685729981,
      "epoch": 3.4093367049212215,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038599733426068826,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2114304259419441,
      "num_tokens": 75999248.0,
      "step": 43820
    },
    {
      "entropy": 5.798661184310913,
      "epoch": 3.4097257342929392,
      "grad_norm": 1.375,
      "learning_rate": 0.00038597322509615977,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2157400667667389,
      "num_tokens": 76007635.0,
      "step": 43825
    },
    {
      "entropy": 5.797388315200806,
      "epoch": 3.4101147636646565,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038594911424778925,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21505948901176453,
      "num_tokens": 76016135.0,
      "step": 43830
    },
    {
      "entropy": 5.738279581069946,
      "epoch": 3.410503793036374,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038592500171594245,
      "loss": 4.8321,
      "mean_token_accuracy": 0.22323838770389556,
      "num_tokens": 76024934.0,
      "step": 43835
    },
    {
      "entropy": 5.7914186954498295,
      "epoch": 3.410892822408092,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038590088750098536,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21057940870523453,
      "num_tokens": 76034242.0,
      "step": 43840
    },
    {
      "entropy": 5.734814834594727,
      "epoch": 3.411281851779809,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038587677160328377,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21776286661624908,
      "num_tokens": 76043352.0,
      "step": 43845
    },
    {
      "entropy": 5.782661199569702,
      "epoch": 3.411670881151527,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003858526540232037,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21871993392705918,
      "num_tokens": 76051890.0,
      "step": 43850
    },
    {
      "entropy": 5.753050851821899,
      "epoch": 3.4120599105232445,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038582853476111094,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2132248669862747,
      "num_tokens": 76061325.0,
      "step": 43855
    },
    {
      "entropy": 5.779659128189087,
      "epoch": 3.4124489398949622,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003858044138173717,
      "loss": 4.9766,
      "mean_token_accuracy": 0.20771912783384322,
      "num_tokens": 76070337.0,
      "step": 43860
    },
    {
      "entropy": 5.847348594665528,
      "epoch": 3.4128379692666795,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038578029119235176,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2039801672101021,
      "num_tokens": 76078745.0,
      "step": 43865
    },
    {
      "entropy": 5.858489656448365,
      "epoch": 3.413226998638397,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038575616688641724,
      "loss": 4.984,
      "mean_token_accuracy": 0.2068471074104309,
      "num_tokens": 76086253.0,
      "step": 43870
    },
    {
      "entropy": 5.742902517318726,
      "epoch": 3.413616028010115,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003857320408999342,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22225331515073776,
      "num_tokens": 76094902.0,
      "step": 43875
    },
    {
      "entropy": 5.8470234870910645,
      "epoch": 3.414005057381832,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003857079132332687,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20211324244737625,
      "num_tokens": 76103341.0,
      "step": 43880
    },
    {
      "entropy": 5.769184160232544,
      "epoch": 3.41439408675355,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038568378388678675,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2213097870349884,
      "num_tokens": 76111286.0,
      "step": 43885
    },
    {
      "entropy": 5.769033145904541,
      "epoch": 3.4147831161252675,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003856596528608546,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20674742460250856,
      "num_tokens": 76120022.0,
      "step": 43890
    },
    {
      "entropy": 5.756644821166992,
      "epoch": 3.4151721454969852,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003856355201558384,
      "loss": 4.8774,
      "mean_token_accuracy": 0.21620626896619796,
      "num_tokens": 76128225.0,
      "step": 43895
    },
    {
      "entropy": 5.8343840599060055,
      "epoch": 3.4155611748687025,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038561138577210426,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21019227504730226,
      "num_tokens": 76136521.0,
      "step": 43900
    },
    {
      "entropy": 5.909796667098999,
      "epoch": 3.41595020424042,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038558724971001835,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21115416884422303,
      "num_tokens": 76145206.0,
      "step": 43905
    },
    {
      "entropy": 5.827983808517456,
      "epoch": 3.416339233612138,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000385563111969947,
      "loss": 4.9978,
      "mean_token_accuracy": 0.20574368834495543,
      "num_tokens": 76154617.0,
      "step": 43910
    },
    {
      "entropy": 5.78784761428833,
      "epoch": 3.416728262983855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038553897255225633,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20978382229804993,
      "num_tokens": 76163403.0,
      "step": 43915
    },
    {
      "entropy": 5.854395627975464,
      "epoch": 3.417117292355573,
      "grad_norm": 1.484375,
      "learning_rate": 0.00038551483145731277,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20514819622039795,
      "num_tokens": 76171919.0,
      "step": 43920
    },
    {
      "entropy": 5.824197626113891,
      "epoch": 3.4175063217272905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003854906886854825,
      "loss": 4.9317,
      "mean_token_accuracy": 0.20727615803480148,
      "num_tokens": 76180295.0,
      "step": 43925
    },
    {
      "entropy": 5.75875449180603,
      "epoch": 3.417895351099008,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038546654423713187,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2146226942539215,
      "num_tokens": 76188853.0,
      "step": 43930
    },
    {
      "entropy": 5.734459161758423,
      "epoch": 3.4182843804707255,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003854423981126273,
      "loss": 4.9595,
      "mean_token_accuracy": 0.20724107325077057,
      "num_tokens": 76196557.0,
      "step": 43935
    },
    {
      "entropy": 5.78241229057312,
      "epoch": 3.418673409842443,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003854182503123351,
      "loss": 4.996,
      "mean_token_accuracy": 0.20995033234357835,
      "num_tokens": 76205653.0,
      "step": 43940
    },
    {
      "entropy": 5.781963777542114,
      "epoch": 3.4190624392141604,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038539410083662166,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21272823065519333,
      "num_tokens": 76213713.0,
      "step": 43945
    },
    {
      "entropy": 5.811822462081909,
      "epoch": 3.419451468585878,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038536994968585355,
      "loss": 5.0034,
      "mean_token_accuracy": 0.20943650007247924,
      "num_tokens": 76222088.0,
      "step": 43950
    },
    {
      "entropy": 5.79152479171753,
      "epoch": 3.419840497957596,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038534579686039705,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21332916468381882,
      "num_tokens": 76231028.0,
      "step": 43955
    },
    {
      "entropy": 5.852616500854492,
      "epoch": 3.4202295273293135,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038532164236061874,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21388441622257232,
      "num_tokens": 76239961.0,
      "step": 43960
    },
    {
      "entropy": 5.838753080368042,
      "epoch": 3.4206185567010308,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000385297486186885,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20876551419496536,
      "num_tokens": 76248495.0,
      "step": 43965
    },
    {
      "entropy": 5.786052417755127,
      "epoch": 3.4210075860727485,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003852733283395626,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21540594100952148,
      "num_tokens": 76257075.0,
      "step": 43970
    },
    {
      "entropy": 5.752816915512085,
      "epoch": 3.421396615444466,
      "grad_norm": 1.328125,
      "learning_rate": 0.000385249168819018,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21494379192590712,
      "num_tokens": 76265429.0,
      "step": 43975
    },
    {
      "entropy": 5.842991924285888,
      "epoch": 3.4217856448161834,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003852250076256177,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20217147171497346,
      "num_tokens": 76273813.0,
      "step": 43980
    },
    {
      "entropy": 5.736404752731323,
      "epoch": 3.422174674187901,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003852008447597283,
      "loss": 4.8792,
      "mean_token_accuracy": 0.21831849813461304,
      "num_tokens": 76282425.0,
      "step": 43985
    },
    {
      "entropy": 5.795290613174439,
      "epoch": 3.422563703559619,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038517668022171656,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20497841984033585,
      "num_tokens": 76290884.0,
      "step": 43990
    },
    {
      "entropy": 5.814934396743775,
      "epoch": 3.4229527329313365,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003851525140119489,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2025263398885727,
      "num_tokens": 76300377.0,
      "step": 43995
    },
    {
      "entropy": 5.787289571762085,
      "epoch": 3.4233417623030538,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038512834613079226,
      "loss": 4.9785,
      "mean_token_accuracy": 0.20442370027303697,
      "num_tokens": 76308728.0,
      "step": 44000
    },
    {
      "entropy": 5.746721458435059,
      "epoch": 3.4237307916747715,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003851041765786133,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21452730745077134,
      "num_tokens": 76317259.0,
      "step": 44005
    },
    {
      "entropy": 5.770110654830932,
      "epoch": 3.424119821046489,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038508000535577864,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21386731863021852,
      "num_tokens": 76325774.0,
      "step": 44010
    },
    {
      "entropy": 5.75904107093811,
      "epoch": 3.4245088504182064,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003850558324626552,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21252703964710234,
      "num_tokens": 76333875.0,
      "step": 44015
    },
    {
      "entropy": 5.792487621307373,
      "epoch": 3.424897879789924,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003850316578996096,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21473785638809204,
      "num_tokens": 76343598.0,
      "step": 44020
    },
    {
      "entropy": 5.764108657836914,
      "epoch": 3.425286909161642,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003850074816670087,
      "loss": 4.8873,
      "mean_token_accuracy": 0.2128181591629982,
      "num_tokens": 76351436.0,
      "step": 44025
    },
    {
      "entropy": 5.778263807296753,
      "epoch": 3.4256759385333595,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038498330376521945,
      "loss": 4.89,
      "mean_token_accuracy": 0.2126704692840576,
      "num_tokens": 76359992.0,
      "step": 44030
    },
    {
      "entropy": 5.805085563659668,
      "epoch": 3.4260649679050768,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003849591241946085,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2104354113340378,
      "num_tokens": 76368066.0,
      "step": 44035
    },
    {
      "entropy": 5.715711641311645,
      "epoch": 3.4264539972767944,
      "grad_norm": 1.40625,
      "learning_rate": 0.000384934942955543,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22145914435386657,
      "num_tokens": 76376462.0,
      "step": 44040
    },
    {
      "entropy": 5.813705253601074,
      "epoch": 3.426843026648512,
      "grad_norm": 1.453125,
      "learning_rate": 0.00038491076004838965,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21307809054851531,
      "num_tokens": 76385420.0,
      "step": 44045
    },
    {
      "entropy": 5.808110761642456,
      "epoch": 3.4272320560202294,
      "grad_norm": 1.375,
      "learning_rate": 0.00038488657547351544,
      "loss": 4.9046,
      "mean_token_accuracy": 0.2159888356924057,
      "num_tokens": 76393636.0,
      "step": 44050
    },
    {
      "entropy": 5.692878484725952,
      "epoch": 3.427621085391947,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003848623892312874,
      "loss": 4.9018,
      "mean_token_accuracy": 0.21461812406778336,
      "num_tokens": 76401879.0,
      "step": 44055
    },
    {
      "entropy": 5.735193729400635,
      "epoch": 3.428010114763665,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003848382013220725,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21041683852672577,
      "num_tokens": 76410316.0,
      "step": 44060
    },
    {
      "entropy": 5.762240028381347,
      "epoch": 3.428399144135382,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003848140117462376,
      "loss": 4.8274,
      "mean_token_accuracy": 0.21778357177972793,
      "num_tokens": 76418623.0,
      "step": 44065
    },
    {
      "entropy": 5.739213180541992,
      "epoch": 3.4287881735070997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003847898205041499,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21151988208293915,
      "num_tokens": 76426890.0,
      "step": 44070
    },
    {
      "entropy": 5.806094551086426,
      "epoch": 3.4291772028788174,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038476562759617644,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20648994892835618,
      "num_tokens": 76435728.0,
      "step": 44075
    },
    {
      "entropy": 5.779249954223633,
      "epoch": 3.4295662322505347,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038474143302268425,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20930425971746444,
      "num_tokens": 76444381.0,
      "step": 44080
    },
    {
      "entropy": 5.7651347637176515,
      "epoch": 3.4299552616222524,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003847172367840406,
      "loss": 4.922,
      "mean_token_accuracy": 0.21533828377723693,
      "num_tokens": 76452812.0,
      "step": 44085
    },
    {
      "entropy": 5.695934104919433,
      "epoch": 3.43034429099397,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003846930388806124,
      "loss": 4.8477,
      "mean_token_accuracy": 0.214813894033432,
      "num_tokens": 76461650.0,
      "step": 44090
    },
    {
      "entropy": 5.795609331130981,
      "epoch": 3.430733320365688,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038466883931276703,
      "loss": 4.9483,
      "mean_token_accuracy": 0.20852570980787277,
      "num_tokens": 76470868.0,
      "step": 44095
    },
    {
      "entropy": 5.888917303085327,
      "epoch": 3.431122349737405,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038464463808087146,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21069361567497252,
      "num_tokens": 76479027.0,
      "step": 44100
    },
    {
      "entropy": 5.781145238876343,
      "epoch": 3.4315113791091227,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003846204351852931,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21569590866565705,
      "num_tokens": 76487745.0,
      "step": 44105
    },
    {
      "entropy": 5.8028566360473635,
      "epoch": 3.4319004084808404,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038459623062639907,
      "loss": 5.044,
      "mean_token_accuracy": 0.20022841095924376,
      "num_tokens": 76496298.0,
      "step": 44110
    },
    {
      "entropy": 5.823739290237427,
      "epoch": 3.4322894378525577,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038457202440455673,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2100173130631447,
      "num_tokens": 76505041.0,
      "step": 44115
    },
    {
      "entropy": 5.862679672241211,
      "epoch": 3.4326784672242754,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003845478165201333,
      "loss": 5.0342,
      "mean_token_accuracy": 0.199405337870121,
      "num_tokens": 76513265.0,
      "step": 44120
    },
    {
      "entropy": 5.815752553939819,
      "epoch": 3.433067496595993,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038452360697349615,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21243662685155867,
      "num_tokens": 76522084.0,
      "step": 44125
    },
    {
      "entropy": 5.79075927734375,
      "epoch": 3.433456525967711,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003844993957650125,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21269137412309647,
      "num_tokens": 76529931.0,
      "step": 44130
    },
    {
      "entropy": 5.856421089172363,
      "epoch": 3.433845555339428,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003844751828950499,
      "loss": 5.0171,
      "mean_token_accuracy": 0.19944206923246383,
      "num_tokens": 76538431.0,
      "step": 44135
    },
    {
      "entropy": 5.804907751083374,
      "epoch": 3.4342345847111457,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038445096836397564,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2018522709608078,
      "num_tokens": 76547124.0,
      "step": 44140
    },
    {
      "entropy": 5.748795032501221,
      "epoch": 3.4346236140828634,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038442675217215713,
      "loss": 4.8833,
      "mean_token_accuracy": 0.21910492330789566,
      "num_tokens": 76555812.0,
      "step": 44145
    },
    {
      "entropy": 5.874452733993531,
      "epoch": 3.4350126434545807,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038440253431996175,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20251107662916185,
      "num_tokens": 76564607.0,
      "step": 44150
    },
    {
      "entropy": 5.70741024017334,
      "epoch": 3.4354016728262984,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038437831480775717,
      "loss": 4.8257,
      "mean_token_accuracy": 0.21325997710227967,
      "num_tokens": 76573482.0,
      "step": 44155
    },
    {
      "entropy": 5.757980680465698,
      "epoch": 3.435790702198016,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038435409363591066,
      "loss": 4.9048,
      "mean_token_accuracy": 0.21582754254341124,
      "num_tokens": 76582816.0,
      "step": 44160
    },
    {
      "entropy": 5.747077512741089,
      "epoch": 3.4361797315697333,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038432987080478986,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22242543399333953,
      "num_tokens": 76591398.0,
      "step": 44165
    },
    {
      "entropy": 5.812362003326416,
      "epoch": 3.436568760941451,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003843056463147623,
      "loss": 4.916,
      "mean_token_accuracy": 0.21590739339590073,
      "num_tokens": 76600308.0,
      "step": 44170
    },
    {
      "entropy": 5.840604782104492,
      "epoch": 3.4369577903131687,
      "grad_norm": 1.5,
      "learning_rate": 0.00038428142016619555,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21056755632162094,
      "num_tokens": 76608381.0,
      "step": 44175
    },
    {
      "entropy": 5.747580575942993,
      "epoch": 3.437346819684886,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003842571923594572,
      "loss": 4.8682,
      "mean_token_accuracy": 0.2201557293534279,
      "num_tokens": 76616909.0,
      "step": 44180
    },
    {
      "entropy": 5.779032230377197,
      "epoch": 3.4377358490566037,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038423296289491477,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21385540813207626,
      "num_tokens": 76625414.0,
      "step": 44185
    },
    {
      "entropy": 5.7763725280761715,
      "epoch": 3.4381248784283214,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038420873177293606,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21221420913934708,
      "num_tokens": 76634081.0,
      "step": 44190
    },
    {
      "entropy": 5.8521315574646,
      "epoch": 3.438513907800039,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003841844989938886,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21607901751995087,
      "num_tokens": 76643054.0,
      "step": 44195
    },
    {
      "entropy": 5.909198760986328,
      "epoch": 3.4389029371717563,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003841602645581402,
      "loss": 5.0752,
      "mean_token_accuracy": 0.20387353152036666,
      "num_tokens": 76651465.0,
      "step": 44200
    },
    {
      "entropy": 5.819275903701782,
      "epoch": 3.439291966543474,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038413602846605854,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2122614338994026,
      "num_tokens": 76660371.0,
      "step": 44205
    },
    {
      "entropy": 5.745811653137207,
      "epoch": 3.4396809959151917,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003841117907180114,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22047587037086486,
      "num_tokens": 76668969.0,
      "step": 44210
    },
    {
      "entropy": 5.8007739067077635,
      "epoch": 3.440070025286909,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038408755131436647,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21058215200901031,
      "num_tokens": 76677096.0,
      "step": 44215
    },
    {
      "entropy": 5.861566209793091,
      "epoch": 3.4404590546586267,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038406331025549157,
      "loss": 5.048,
      "mean_token_accuracy": 0.20430290848016738,
      "num_tokens": 76685761.0,
      "step": 44220
    },
    {
      "entropy": 5.8328831672668455,
      "epoch": 3.4408480840303444,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003840390675417545,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2104451537132263,
      "num_tokens": 76694606.0,
      "step": 44225
    },
    {
      "entropy": 5.767729187011719,
      "epoch": 3.441237113402062,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003840148231735231,
      "loss": 4.8645,
      "mean_token_accuracy": 0.21345752775669097,
      "num_tokens": 76702814.0,
      "step": 44230
    },
    {
      "entropy": 5.733599996566772,
      "epoch": 3.4416261427737793,
      "grad_norm": 1.375,
      "learning_rate": 0.00038399057715116536,
      "loss": 4.8412,
      "mean_token_accuracy": 0.22345227003097534,
      "num_tokens": 76711182.0,
      "step": 44235
    },
    {
      "entropy": 5.821080255508423,
      "epoch": 3.442015172145497,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038396632947504904,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20831616818904877,
      "num_tokens": 76720198.0,
      "step": 44240
    },
    {
      "entropy": 5.835020446777344,
      "epoch": 3.4424042015172147,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003839420801455422,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21948946863412858,
      "num_tokens": 76728195.0,
      "step": 44245
    },
    {
      "entropy": 5.821912384033203,
      "epoch": 3.442793230888932,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003839178291630126,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2109248787164688,
      "num_tokens": 76736921.0,
      "step": 44250
    },
    {
      "entropy": 5.7910377979278564,
      "epoch": 3.4431822602606497,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038389357652782836,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20408324748277665,
      "num_tokens": 76745326.0,
      "step": 44255
    },
    {
      "entropy": 5.70452880859375,
      "epoch": 3.4435712896323674,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038386932224035734,
      "loss": 4.8819,
      "mean_token_accuracy": 0.218246129155159,
      "num_tokens": 76753795.0,
      "step": 44260
    },
    {
      "entropy": 5.756911182403565,
      "epoch": 3.443960319004085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038384506630096765,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2205243691802025,
      "num_tokens": 76762466.0,
      "step": 44265
    },
    {
      "entropy": 5.773748779296875,
      "epoch": 3.4443493483758023,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003838208087100274,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21056657731533052,
      "num_tokens": 76770753.0,
      "step": 44270
    },
    {
      "entropy": 5.814641380310059,
      "epoch": 3.44473837774752,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003837965494679045,
      "loss": 5.042,
      "mean_token_accuracy": 0.2045855239033699,
      "num_tokens": 76779328.0,
      "step": 44275
    },
    {
      "entropy": 5.859599351882935,
      "epoch": 3.4451274071192373,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003837722885749672,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20345864593982696,
      "num_tokens": 76788588.0,
      "step": 44280
    },
    {
      "entropy": 5.815992212295532,
      "epoch": 3.445516436490955,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003837480260315835,
      "loss": 4.8583,
      "mean_token_accuracy": 0.21789378374814988,
      "num_tokens": 76796452.0,
      "step": 44285
    },
    {
      "entropy": 5.738669776916504,
      "epoch": 3.4459054658626727,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038372376183812165,
      "loss": 4.8832,
      "mean_token_accuracy": 0.21622634977102279,
      "num_tokens": 76804773.0,
      "step": 44290
    },
    {
      "entropy": 5.796259117126465,
      "epoch": 3.4462944952343904,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003836994959949497,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20387348532676697,
      "num_tokens": 76813077.0,
      "step": 44295
    },
    {
      "entropy": 5.796672630310058,
      "epoch": 3.4466835246061076,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038367522850243593,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2122795417904854,
      "num_tokens": 76821148.0,
      "step": 44300
    },
    {
      "entropy": 5.791253757476807,
      "epoch": 3.4470725539778253,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038365095936094857,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2096623808145523,
      "num_tokens": 76830494.0,
      "step": 44305
    },
    {
      "entropy": 5.903387928009034,
      "epoch": 3.447461583349543,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003836266885708558,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2095757320523262,
      "num_tokens": 76838597.0,
      "step": 44310
    },
    {
      "entropy": 5.865663385391235,
      "epoch": 3.4478506127212603,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038360241613252593,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20870259404182434,
      "num_tokens": 76846427.0,
      "step": 44315
    },
    {
      "entropy": 5.882029104232788,
      "epoch": 3.448239642092978,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003835781420463273,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20079626888036728,
      "num_tokens": 76855275.0,
      "step": 44320
    },
    {
      "entropy": 5.775842809677124,
      "epoch": 3.4486286714646956,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003835538663126282,
      "loss": 4.916,
      "mean_token_accuracy": 0.21536944508552552,
      "num_tokens": 76863315.0,
      "step": 44325
    },
    {
      "entropy": 5.749256277084351,
      "epoch": 3.4490177008364133,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003835295889317969,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22435604631900788,
      "num_tokens": 76871532.0,
      "step": 44330
    },
    {
      "entropy": 5.791563415527344,
      "epoch": 3.4494067302081306,
      "grad_norm": 1.25,
      "learning_rate": 0.00038350530990420177,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22398627251386644,
      "num_tokens": 76880235.0,
      "step": 44335
    },
    {
      "entropy": 5.845034790039063,
      "epoch": 3.4497957595798483,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003834810292302114,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2070550188422203,
      "num_tokens": 76888518.0,
      "step": 44340
    },
    {
      "entropy": 5.769588804244995,
      "epoch": 3.450184788951566,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000383456746910194,
      "loss": 4.9017,
      "mean_token_accuracy": 0.21496809273958206,
      "num_tokens": 76897294.0,
      "step": 44345
    },
    {
      "entropy": 5.7947005271911625,
      "epoch": 3.4505738183232832,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003834324629445182,
      "loss": 5.057,
      "mean_token_accuracy": 0.20971240550279618,
      "num_tokens": 76906321.0,
      "step": 44350
    },
    {
      "entropy": 5.821798419952392,
      "epoch": 3.450962847695001,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003834081773335522,
      "loss": 4.8579,
      "mean_token_accuracy": 0.21301292777061462,
      "num_tokens": 76914967.0,
      "step": 44355
    },
    {
      "entropy": 5.872678422927857,
      "epoch": 3.4513518770667186,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038338389007766473,
      "loss": 4.9643,
      "mean_token_accuracy": 0.20938342660665513,
      "num_tokens": 76923983.0,
      "step": 44360
    },
    {
      "entropy": 5.754303312301635,
      "epoch": 3.4517409064384363,
      "grad_norm": 1.25,
      "learning_rate": 0.0003833596011772242,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2105577141046524,
      "num_tokens": 76932749.0,
      "step": 44365
    },
    {
      "entropy": 5.767071485519409,
      "epoch": 3.4521299358101536,
      "grad_norm": 1.25,
      "learning_rate": 0.0003833353106325993,
      "loss": 4.923,
      "mean_token_accuracy": 0.21365894079208375,
      "num_tokens": 76943077.0,
      "step": 44370
    },
    {
      "entropy": 5.831573724746704,
      "epoch": 3.4525189651818713,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003833110184441584,
      "loss": 4.895,
      "mean_token_accuracy": 0.21671973317861556,
      "num_tokens": 76951895.0,
      "step": 44375
    },
    {
      "entropy": 5.806667900085449,
      "epoch": 3.452907994553589,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038328672461227024,
      "loss": 4.9764,
      "mean_token_accuracy": 0.2082356631755829,
      "num_tokens": 76961605.0,
      "step": 44380
    },
    {
      "entropy": 5.754713296890259,
      "epoch": 3.4532970239253062,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003832624291373034,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2128053665161133,
      "num_tokens": 76971262.0,
      "step": 44385
    },
    {
      "entropy": 5.750529718399048,
      "epoch": 3.453686053297024,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038323813201962645,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22074589878320694,
      "num_tokens": 76979411.0,
      "step": 44390
    },
    {
      "entropy": 5.792137336730957,
      "epoch": 3.4540750826687416,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038321383325960816,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2003355875611305,
      "num_tokens": 76987957.0,
      "step": 44395
    },
    {
      "entropy": 5.795413684844971,
      "epoch": 3.454464112040459,
      "grad_norm": 1.25,
      "learning_rate": 0.0003831895328576172,
      "loss": 4.899,
      "mean_token_accuracy": 0.20986563414335252,
      "num_tokens": 76996355.0,
      "step": 44400
    },
    {
      "entropy": 5.799166584014893,
      "epoch": 3.4548531414121766,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003831652308140223,
      "loss": 4.9642,
      "mean_token_accuracy": 0.20917538106441497,
      "num_tokens": 77004795.0,
      "step": 44405
    },
    {
      "entropy": 5.846941232681274,
      "epoch": 3.4552421707838943,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003831409271291921,
      "loss": 4.959,
      "mean_token_accuracy": 0.2070248991250992,
      "num_tokens": 77013617.0,
      "step": 44410
    },
    {
      "entropy": 5.738839149475098,
      "epoch": 3.4556312001556115,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038311662180349554,
      "loss": 4.881,
      "mean_token_accuracy": 0.21584440469741822,
      "num_tokens": 77022144.0,
      "step": 44415
    },
    {
      "entropy": 5.763363552093506,
      "epoch": 3.4560202295273292,
      "grad_norm": 1.484375,
      "learning_rate": 0.00038309231483730133,
      "loss": 4.836,
      "mean_token_accuracy": 0.22297521382570268,
      "num_tokens": 77030079.0,
      "step": 44420
    },
    {
      "entropy": 5.804884195327759,
      "epoch": 3.456409258899047,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003830680062309782,
      "loss": 4.994,
      "mean_token_accuracy": 0.20454201251268386,
      "num_tokens": 77038551.0,
      "step": 44425
    },
    {
      "entropy": 5.790785455703736,
      "epoch": 3.4567982882707646,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038304369598489516,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2085963547229767,
      "num_tokens": 77047113.0,
      "step": 44430
    },
    {
      "entropy": 5.8198432445526125,
      "epoch": 3.457187317642482,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000383019384099421,
      "loss": 4.9635,
      "mean_token_accuracy": 0.20834286212921144,
      "num_tokens": 77055789.0,
      "step": 44435
    },
    {
      "entropy": 5.838001680374146,
      "epoch": 3.4575763470141996,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003829950705749246,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22010788321495056,
      "num_tokens": 77064065.0,
      "step": 44440
    },
    {
      "entropy": 5.757308006286621,
      "epoch": 3.4579653763859173,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038297075541177496,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2182082548737526,
      "num_tokens": 77072927.0,
      "step": 44445
    },
    {
      "entropy": 5.8023107051849365,
      "epoch": 3.4583544057576345,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038294643861034087,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21366448402404786,
      "num_tokens": 77081384.0,
      "step": 44450
    },
    {
      "entropy": 5.849637985229492,
      "epoch": 3.4587434351293522,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003829221201709914,
      "loss": 4.9812,
      "mean_token_accuracy": 0.20686790049076081,
      "num_tokens": 77089642.0,
      "step": 44455
    },
    {
      "entropy": 5.798729133605957,
      "epoch": 3.45913246450107,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003828978000940955,
      "loss": 4.9614,
      "mean_token_accuracy": 0.19910217374563216,
      "num_tokens": 77098218.0,
      "step": 44460
    },
    {
      "entropy": 5.778178882598877,
      "epoch": 3.4595214938727876,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003828734783800223,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2029196575284004,
      "num_tokens": 77106651.0,
      "step": 44465
    },
    {
      "entropy": 5.792030906677246,
      "epoch": 3.459910523244505,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003828491550291408,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20978539437055588,
      "num_tokens": 77115378.0,
      "step": 44470
    },
    {
      "entropy": 5.770162868499756,
      "epoch": 3.4602995526162226,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003828248300418199,
      "loss": 4.9248,
      "mean_token_accuracy": 0.20963781774044038,
      "num_tokens": 77124339.0,
      "step": 44475
    },
    {
      "entropy": 5.849606037139893,
      "epoch": 3.4606885819879403,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038280050341842894,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20149501115083696,
      "num_tokens": 77133344.0,
      "step": 44480
    },
    {
      "entropy": 5.834320306777954,
      "epoch": 3.4610776113596575,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003827761751593369,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21177805215120316,
      "num_tokens": 77142321.0,
      "step": 44485
    },
    {
      "entropy": 5.794275760650635,
      "epoch": 3.461466640731375,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003827518452649129,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22131069451570512,
      "num_tokens": 77150675.0,
      "step": 44490
    },
    {
      "entropy": 5.807435035705566,
      "epoch": 3.461855670103093,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003827275137355261,
      "loss": 4.92,
      "mean_token_accuracy": 0.21479563415050507,
      "num_tokens": 77158878.0,
      "step": 44495
    },
    {
      "entropy": 5.8161623001098635,
      "epoch": 3.46224469947481,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003827031805715458,
      "loss": 5.011,
      "mean_token_accuracy": 0.20662875920534135,
      "num_tokens": 77167783.0,
      "step": 44500
    },
    {
      "entropy": 5.829071569442749,
      "epoch": 3.462633728846528,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038267884577334124,
      "loss": 4.955,
      "mean_token_accuracy": 0.20680856555700303,
      "num_tokens": 77176524.0,
      "step": 44505
    },
    {
      "entropy": 5.880359077453614,
      "epoch": 3.4630227582182456,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003826545093412815,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21376874446868896,
      "num_tokens": 77184574.0,
      "step": 44510
    },
    {
      "entropy": 5.8537132263183596,
      "epoch": 3.463411787589963,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038263017127573596,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21018934845924378,
      "num_tokens": 77194389.0,
      "step": 44515
    },
    {
      "entropy": 5.83266921043396,
      "epoch": 3.4638008169616805,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003826058315770739,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2026475504040718,
      "num_tokens": 77202581.0,
      "step": 44520
    },
    {
      "entropy": 5.864630794525146,
      "epoch": 3.464189846333398,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038258149024566465,
      "loss": 5.0013,
      "mean_token_accuracy": 0.19995850920677186,
      "num_tokens": 77211496.0,
      "step": 44525
    },
    {
      "entropy": 5.83262939453125,
      "epoch": 3.464578875705116,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003825571472818774,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21005446016788482,
      "num_tokens": 77220656.0,
      "step": 44530
    },
    {
      "entropy": 5.840118646621704,
      "epoch": 3.464967905076833,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003825328026860817,
      "loss": 5.0251,
      "mean_token_accuracy": 0.20953383892774582,
      "num_tokens": 77230138.0,
      "step": 44535
    },
    {
      "entropy": 5.861305856704712,
      "epoch": 3.465356934448551,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038250845645864686,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2010301470756531,
      "num_tokens": 77238745.0,
      "step": 44540
    },
    {
      "entropy": 5.7308878898620605,
      "epoch": 3.4657459638202686,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003824841085999423,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22596123665571213,
      "num_tokens": 77248246.0,
      "step": 44545
    },
    {
      "entropy": 5.8202677249908445,
      "epoch": 3.466134993191986,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003824597591103375,
      "loss": 4.9318,
      "mean_token_accuracy": 0.20280381590127944,
      "num_tokens": 77256504.0,
      "step": 44550
    },
    {
      "entropy": 5.778558683395386,
      "epoch": 3.4665240225637035,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038243540799020193,
      "loss": 4.9423,
      "mean_token_accuracy": 0.20651733726263047,
      "num_tokens": 77265041.0,
      "step": 44555
    },
    {
      "entropy": 5.841511249542236,
      "epoch": 3.466913051935421,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038241105523990497,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21325500011444093,
      "num_tokens": 77273437.0,
      "step": 44560
    },
    {
      "entropy": 5.7816211700439455,
      "epoch": 3.467302081307139,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003823867008598163,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21341925859451294,
      "num_tokens": 77281228.0,
      "step": 44565
    },
    {
      "entropy": 5.854672145843506,
      "epoch": 3.467691110678856,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038236234485030524,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21599425375461578,
      "num_tokens": 77289855.0,
      "step": 44570
    },
    {
      "entropy": 5.799074125289917,
      "epoch": 3.468080140050574,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038233798721174156,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21770044714212416,
      "num_tokens": 77298437.0,
      "step": 44575
    },
    {
      "entropy": 5.780996465682984,
      "epoch": 3.4684691694222916,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003823136279444948,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2085756853222847,
      "num_tokens": 77307336.0,
      "step": 44580
    },
    {
      "entropy": 5.824058771133423,
      "epoch": 3.468858198794009,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003822892670489345,
      "loss": 4.981,
      "mean_token_accuracy": 0.20721845775842668,
      "num_tokens": 77315661.0,
      "step": 44585
    },
    {
      "entropy": 5.748181676864624,
      "epoch": 3.4692472281657265,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038226490452543036,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2182508274912834,
      "num_tokens": 77324599.0,
      "step": 44590
    },
    {
      "entropy": 5.798062133789062,
      "epoch": 3.469636257537444,
      "grad_norm": 1.25,
      "learning_rate": 0.000382240540374352,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2092011034488678,
      "num_tokens": 77333177.0,
      "step": 44595
    },
    {
      "entropy": 5.797426223754883,
      "epoch": 3.4700252869091615,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003822161745960691,
      "loss": 4.9609,
      "mean_token_accuracy": 0.20942093431949615,
      "num_tokens": 77342356.0,
      "step": 44600
    },
    {
      "entropy": 5.865619754791259,
      "epoch": 3.470414316280879,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003821918071909515,
      "loss": 4.9642,
      "mean_token_accuracy": 0.2107903927564621,
      "num_tokens": 77351476.0,
      "step": 44605
    },
    {
      "entropy": 5.816127157211303,
      "epoch": 3.470803345652597,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003821674381593687,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2110590949654579,
      "num_tokens": 77359887.0,
      "step": 44610
    },
    {
      "entropy": 5.77136378288269,
      "epoch": 3.471192375024314,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003821430675016907,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22130513340234756,
      "num_tokens": 77368550.0,
      "step": 44615
    },
    {
      "entropy": 5.769488000869751,
      "epoch": 3.471581404396032,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038211869521828716,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21810704320669175,
      "num_tokens": 77378034.0,
      "step": 44620
    },
    {
      "entropy": 5.781516075134277,
      "epoch": 3.4719704337677495,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000382094321309528,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21590566635131836,
      "num_tokens": 77387126.0,
      "step": 44625
    },
    {
      "entropy": 5.763752126693726,
      "epoch": 3.472359463139467,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003820699457757829,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2101902797818184,
      "num_tokens": 77395580.0,
      "step": 44630
    },
    {
      "entropy": 5.788371181488037,
      "epoch": 3.4727484925111844,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003820455686174218,
      "loss": 4.988,
      "mean_token_accuracy": 0.2118326798081398,
      "num_tokens": 77404488.0,
      "step": 44635
    },
    {
      "entropy": 5.784642887115479,
      "epoch": 3.473137521882902,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038202118983481456,
      "loss": 4.9891,
      "mean_token_accuracy": 0.20658906251192094,
      "num_tokens": 77413002.0,
      "step": 44640
    },
    {
      "entropy": 5.76349105834961,
      "epoch": 3.47352655125462,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003819968094283311,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21147411018610002,
      "num_tokens": 77421831.0,
      "step": 44645
    },
    {
      "entropy": 5.77415828704834,
      "epoch": 3.473915580626337,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003819724273983414,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2157002329826355,
      "num_tokens": 77430039.0,
      "step": 44650
    },
    {
      "entropy": 5.7870402336120605,
      "epoch": 3.474304609998055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038194804374521536,
      "loss": 4.9518,
      "mean_token_accuracy": 0.207393878698349,
      "num_tokens": 77439469.0,
      "step": 44655
    },
    {
      "entropy": 5.700403547286987,
      "epoch": 3.4746936393697725,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003819236584693229,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21747809797525405,
      "num_tokens": 77448294.0,
      "step": 44660
    },
    {
      "entropy": 5.7471400737762455,
      "epoch": 3.47508266874149,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038189927157103427,
      "loss": 4.845,
      "mean_token_accuracy": 0.2215729162096977,
      "num_tokens": 77456410.0,
      "step": 44665
    },
    {
      "entropy": 5.7760978698730465,
      "epoch": 3.4754716981132074,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003818748830507192,
      "loss": 4.9741,
      "mean_token_accuracy": 0.20822336971759797,
      "num_tokens": 77466081.0,
      "step": 44670
    },
    {
      "entropy": 5.860675859451294,
      "epoch": 3.475860727484925,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000381850492908748,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2087363764643669,
      "num_tokens": 77474020.0,
      "step": 44675
    },
    {
      "entropy": 5.840505170822143,
      "epoch": 3.476249756856643,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038182610114549057,
      "loss": 4.968,
      "mean_token_accuracy": 0.2097627580165863,
      "num_tokens": 77483573.0,
      "step": 44680
    },
    {
      "entropy": 5.766376590728759,
      "epoch": 3.47663878622836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038180170776131714,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20373685508966446,
      "num_tokens": 77492493.0,
      "step": 44685
    },
    {
      "entropy": 5.765184450149536,
      "epoch": 3.477027815600078,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003817773127565977,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21918345838785172,
      "num_tokens": 77501293.0,
      "step": 44690
    },
    {
      "entropy": 5.809802293777466,
      "epoch": 3.4774168449717955,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00038175291613170247,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21846579015254974,
      "num_tokens": 77509795.0,
      "step": 44695
    },
    {
      "entropy": 5.786846017837524,
      "epoch": 3.477805874343513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038172851788700174,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2156926214694977,
      "num_tokens": 77518737.0,
      "step": 44700
    },
    {
      "entropy": 5.768040990829467,
      "epoch": 3.4781949037152304,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003817041180228657,
      "loss": 4.967,
      "mean_token_accuracy": 0.20817156434059142,
      "num_tokens": 77527642.0,
      "step": 44705
    },
    {
      "entropy": 5.740696763992309,
      "epoch": 3.478583933086948,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003816797165396643,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22349030673503875,
      "num_tokens": 77535930.0,
      "step": 44710
    },
    {
      "entropy": 5.758592510223389,
      "epoch": 3.4789729624586654,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038165531343776815,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21824156492948532,
      "num_tokens": 77544627.0,
      "step": 44715
    },
    {
      "entropy": 5.7554105758667,
      "epoch": 3.479361991830383,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003816309087175474,
      "loss": 4.906,
      "mean_token_accuracy": 0.21835257709026337,
      "num_tokens": 77553426.0,
      "step": 44720
    },
    {
      "entropy": 5.787427139282227,
      "epoch": 3.479751021202101,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038160650237937225,
      "loss": 4.983,
      "mean_token_accuracy": 0.21358553767204286,
      "num_tokens": 77561604.0,
      "step": 44725
    },
    {
      "entropy": 5.825679779052734,
      "epoch": 3.4801400505738185,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003815820944236131,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20345656126737593,
      "num_tokens": 77570220.0,
      "step": 44730
    },
    {
      "entropy": 5.780223751068116,
      "epoch": 3.4805290799455357,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038155768485064037,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21499868482351303,
      "num_tokens": 77578913.0,
      "step": 44735
    },
    {
      "entropy": 5.843073225021362,
      "epoch": 3.4809181093172534,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038153327366082434,
      "loss": 4.983,
      "mean_token_accuracy": 0.20757836997509002,
      "num_tokens": 77587649.0,
      "step": 44740
    },
    {
      "entropy": 5.8268921852111815,
      "epoch": 3.481307138688971,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038150886085453546,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21480244696140288,
      "num_tokens": 77596241.0,
      "step": 44745
    },
    {
      "entropy": 5.822463798522949,
      "epoch": 3.4816961680606884,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038148444643214415,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19850883036851882,
      "num_tokens": 77604641.0,
      "step": 44750
    },
    {
      "entropy": 5.806487321853638,
      "epoch": 3.482085197432406,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003814600303940208,
      "loss": 4.918,
      "mean_token_accuracy": 0.20962276607751845,
      "num_tokens": 77613199.0,
      "step": 44755
    },
    {
      "entropy": 5.792320013046265,
      "epoch": 3.4824742268041238,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000381435612740536,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2104360803961754,
      "num_tokens": 77622037.0,
      "step": 44760
    },
    {
      "entropy": 5.78350248336792,
      "epoch": 3.4828632561758415,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038141119347206016,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21449425220489501,
      "num_tokens": 77630541.0,
      "step": 44765
    },
    {
      "entropy": 5.832579469680786,
      "epoch": 3.4832522855475587,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038138677258896383,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2060764640569687,
      "num_tokens": 77638668.0,
      "step": 44770
    },
    {
      "entropy": 5.807975149154663,
      "epoch": 3.4836413149192764,
      "grad_norm": 1.375,
      "learning_rate": 0.0003813623500916176,
      "loss": 4.9932,
      "mean_token_accuracy": 0.214507095515728,
      "num_tokens": 77648123.0,
      "step": 44775
    },
    {
      "entropy": 5.873371601104736,
      "epoch": 3.484030344290994,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038133792598039196,
      "loss": 5.0105,
      "mean_token_accuracy": 0.2119548037648201,
      "num_tokens": 77657368.0,
      "step": 44780
    },
    {
      "entropy": 5.92892107963562,
      "epoch": 3.4844193736627114,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003813135002556575,
      "loss": 5.0552,
      "mean_token_accuracy": 0.20314472913742065,
      "num_tokens": 77666567.0,
      "step": 44785
    },
    {
      "entropy": 5.847765970230102,
      "epoch": 3.484808403034429,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038128907291778497,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20605779588222503,
      "num_tokens": 77674431.0,
      "step": 44790
    },
    {
      "entropy": 5.747258424758911,
      "epoch": 3.4851974324061468,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038126464396714494,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21758346855640412,
      "num_tokens": 77682753.0,
      "step": 44795
    },
    {
      "entropy": 5.771861457824707,
      "epoch": 3.4855864617778645,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038124021340410795,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21649012118577957,
      "num_tokens": 77691072.0,
      "step": 44800
    },
    {
      "entropy": 5.750421571731567,
      "epoch": 3.4859754911495817,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038121578122904496,
      "loss": 4.8491,
      "mean_token_accuracy": 0.21606859415769578,
      "num_tokens": 77699401.0,
      "step": 44805
    },
    {
      "entropy": 5.769607210159302,
      "epoch": 3.4863645205212994,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003811913474423265,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21280055791139602,
      "num_tokens": 77708565.0,
      "step": 44810
    },
    {
      "entropy": 5.812613105773925,
      "epoch": 3.486753549893017,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003811669120443234,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2088782250881195,
      "num_tokens": 77718497.0,
      "step": 44815
    },
    {
      "entropy": 5.785005235671997,
      "epoch": 3.4871425792647344,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003811424750354063,
      "loss": 4.9784,
      "mean_token_accuracy": 0.211370812356472,
      "num_tokens": 77727673.0,
      "step": 44820
    },
    {
      "entropy": 5.748118162155151,
      "epoch": 3.487531608636452,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003811180364159462,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2068103924393654,
      "num_tokens": 77735813.0,
      "step": 44825
    },
    {
      "entropy": 5.755158758163452,
      "epoch": 3.4879206380081698,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038109359618631374,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2209985613822937,
      "num_tokens": 77743529.0,
      "step": 44830
    },
    {
      "entropy": 5.702675676345825,
      "epoch": 3.488309667379887,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00038106915434687983,
      "loss": 4.8698,
      "mean_token_accuracy": 0.212717667222023,
      "num_tokens": 77751251.0,
      "step": 44835
    },
    {
      "entropy": 5.730726289749145,
      "epoch": 3.4886986967516047,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003810447108980153,
      "loss": 4.8961,
      "mean_token_accuracy": 0.21905719935894014,
      "num_tokens": 77759857.0,
      "step": 44840
    },
    {
      "entropy": 5.8087592124938965,
      "epoch": 3.4890877261233224,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003810202658400911,
      "loss": 5.0401,
      "mean_token_accuracy": 0.2116159975528717,
      "num_tokens": 77769109.0,
      "step": 44845
    },
    {
      "entropy": 5.798792839050293,
      "epoch": 3.4894767554950397,
      "grad_norm": 1.25,
      "learning_rate": 0.0003809958191734781,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2070462718605995,
      "num_tokens": 77778795.0,
      "step": 44850
    },
    {
      "entropy": 5.8452733039855955,
      "epoch": 3.4898657848667574,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00038097137089854725,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21140459775924683,
      "num_tokens": 77786901.0,
      "step": 44855
    },
    {
      "entropy": 5.723187685012817,
      "epoch": 3.490254814238475,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038094692101566955,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2157171845436096,
      "num_tokens": 77795998.0,
      "step": 44860
    },
    {
      "entropy": 5.805377674102783,
      "epoch": 3.4906438436101928,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003809224695252159,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20783466547727586,
      "num_tokens": 77805995.0,
      "step": 44865
    },
    {
      "entropy": 5.854003143310547,
      "epoch": 3.49103287298191,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038089801642755746,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2087114691734314,
      "num_tokens": 77814476.0,
      "step": 44870
    },
    {
      "entropy": 5.732790470123291,
      "epoch": 3.4914219023536277,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038087356172306514,
      "loss": 4.8183,
      "mean_token_accuracy": 0.22181419134140015,
      "num_tokens": 77822960.0,
      "step": 44875
    },
    {
      "entropy": 5.752075242996216,
      "epoch": 3.4918109317253454,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038084910541211,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21005307585000993,
      "num_tokens": 77832008.0,
      "step": 44880
    },
    {
      "entropy": 5.75865068435669,
      "epoch": 3.4921999610970627,
      "grad_norm": 1.515625,
      "learning_rate": 0.00038082464749506314,
      "loss": 4.943,
      "mean_token_accuracy": 0.20378140956163407,
      "num_tokens": 77840752.0,
      "step": 44885
    },
    {
      "entropy": 5.779946041107178,
      "epoch": 3.4925889904687804,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038080018797229576,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21439186930656434,
      "num_tokens": 77849122.0,
      "step": 44890
    },
    {
      "entropy": 5.828277349472046,
      "epoch": 3.492978019840498,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003807757268441789,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2160893365740776,
      "num_tokens": 77857114.0,
      "step": 44895
    },
    {
      "entropy": 5.797470378875732,
      "epoch": 3.4933670492122157,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038075126411108367,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21657972335815429,
      "num_tokens": 77866088.0,
      "step": 44900
    },
    {
      "entropy": 5.798373603820801,
      "epoch": 3.493756078583933,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038072679977338143,
      "loss": 4.995,
      "mean_token_accuracy": 0.20964398980140686,
      "num_tokens": 77874161.0,
      "step": 44905
    },
    {
      "entropy": 5.846614837646484,
      "epoch": 3.4941451079556507,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00038070233383144324,
      "loss": 5.0142,
      "mean_token_accuracy": 0.20935287624597548,
      "num_tokens": 77882604.0,
      "step": 44910
    },
    {
      "entropy": 5.890738821029663,
      "epoch": 3.4945341373273684,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003806778662856404,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21055128425359726,
      "num_tokens": 77890944.0,
      "step": 44915
    },
    {
      "entropy": 5.778139114379883,
      "epoch": 3.4949231666990856,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003806533971363441,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21126247197389603,
      "num_tokens": 77899541.0,
      "step": 44920
    },
    {
      "entropy": 5.784995174407959,
      "epoch": 3.4953121960708033,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003806289263839257,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2076089009642601,
      "num_tokens": 77908200.0,
      "step": 44925
    },
    {
      "entropy": 5.812356042861938,
      "epoch": 3.495701225442521,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003806044540287564,
      "loss": 4.914,
      "mean_token_accuracy": 0.21760586351156236,
      "num_tokens": 77916149.0,
      "step": 44930
    },
    {
      "entropy": 5.792283630371093,
      "epoch": 3.4960902548142383,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038057998007120776,
      "loss": 4.8935,
      "mean_token_accuracy": 0.2139286980032921,
      "num_tokens": 77924717.0,
      "step": 44935
    },
    {
      "entropy": 5.736800622940064,
      "epoch": 3.496479284185956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038055550451165086,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21067046076059343,
      "num_tokens": 77933449.0,
      "step": 44940
    },
    {
      "entropy": 5.719511079788208,
      "epoch": 3.4968683135576737,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003805310273504572,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21083937883377074,
      "num_tokens": 77942446.0,
      "step": 44945
    },
    {
      "entropy": 5.831179094314575,
      "epoch": 3.497257342929391,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003805065485879982,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20888094305992128,
      "num_tokens": 77949781.0,
      "step": 44950
    },
    {
      "entropy": 5.803176975250244,
      "epoch": 3.4976463723011086,
      "grad_norm": 1.4375,
      "learning_rate": 0.00038048206822464514,
      "loss": 4.9808,
      "mean_token_accuracy": 0.20000501126050949,
      "num_tokens": 77958638.0,
      "step": 44955
    },
    {
      "entropy": 5.823805046081543,
      "epoch": 3.4980354016728263,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038045758626076965,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21090339869260788,
      "num_tokens": 77967840.0,
      "step": 44960
    },
    {
      "entropy": 5.7874127388000485,
      "epoch": 3.498424431044544,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003804331026967432,
      "loss": 4.9559,
      "mean_token_accuracy": 0.20924318283796312,
      "num_tokens": 77976946.0,
      "step": 44965
    },
    {
      "entropy": 5.797838592529297,
      "epoch": 3.4988134604162613,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003804086175329372,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20023139268159867,
      "num_tokens": 77985644.0,
      "step": 44970
    },
    {
      "entropy": 5.804594087600708,
      "epoch": 3.499202489787979,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003803841307697232,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2115913838148117,
      "num_tokens": 77994118.0,
      "step": 44975
    },
    {
      "entropy": 5.817593336105347,
      "epoch": 3.4995915191596967,
      "grad_norm": 1.25,
      "learning_rate": 0.00038035964240747274,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21989504992961884,
      "num_tokens": 78003281.0,
      "step": 44980
    },
    {
      "entropy": 5.821882581710815,
      "epoch": 3.499980548531414,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003803351524465574,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20938853472471236,
      "num_tokens": 78012075.0,
      "step": 44985
    },
    {
      "entropy": 5.794003582000732,
      "epoch": 3.5003695779031316,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038031066088734883,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22007423788309097,
      "num_tokens": 78020722.0,
      "step": 44990
    },
    {
      "entropy": 5.778456544876098,
      "epoch": 3.5007586072748493,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003802861677302185,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21949015855789183,
      "num_tokens": 78029798.0,
      "step": 44995
    },
    {
      "entropy": 5.799038314819336,
      "epoch": 3.501147636646567,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038026167297553826,
      "loss": 4.892,
      "mean_token_accuracy": 0.21535504311323167,
      "num_tokens": 78038362.0,
      "step": 45000
    },
    {
      "epoch": 3.501147636646567,
      "eval_entropy": 5.546584152280424,
      "eval_loss": 5.047752380371094,
      "eval_mean_token_accuracy": 0.21673591212370397,
      "eval_num_tokens": 78038362.0,
      "eval_runtime": 21.7257,
      "eval_samples_per_second": 1912.853,
      "eval_steps_per_second": 239.118,
      "step": 45000
    },
    {
      "entropy": 5.81340069770813,
      "epoch": 3.5015366660182843,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038023717662367956,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21115302443504333,
      "num_tokens": 78047711.0,
      "step": 45005
    },
    {
      "entropy": 5.77935938835144,
      "epoch": 3.501925695390002,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003802126786750143,
      "loss": 4.9229,
      "mean_token_accuracy": 0.21077784299850463,
      "num_tokens": 78056556.0,
      "step": 45010
    },
    {
      "entropy": 5.793754768371582,
      "epoch": 3.5023147247617192,
      "grad_norm": 1.375,
      "learning_rate": 0.00038018817912991413,
      "loss": 4.8889,
      "mean_token_accuracy": 0.20750444233417512,
      "num_tokens": 78064481.0,
      "step": 45015
    },
    {
      "entropy": 5.72855281829834,
      "epoch": 3.502703754133437,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00038016367798875074,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2189100444316864,
      "num_tokens": 78073310.0,
      "step": 45020
    },
    {
      "entropy": 5.824944257736206,
      "epoch": 3.5030927835051546,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003801391752518959,
      "loss": 5.0036,
      "mean_token_accuracy": 0.20958982706069945,
      "num_tokens": 78082053.0,
      "step": 45025
    },
    {
      "entropy": 5.7840088367462155,
      "epoch": 3.5034818128768723,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038011467091972134,
      "loss": 5.016,
      "mean_token_accuracy": 0.21146827042102814,
      "num_tokens": 78090733.0,
      "step": 45030
    },
    {
      "entropy": 5.829059267044068,
      "epoch": 3.50387084224859,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003800901649925991,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2097789004445076,
      "num_tokens": 78099757.0,
      "step": 45035
    },
    {
      "entropy": 5.841159343719482,
      "epoch": 3.5042598716203073,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038006565747090076,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2133435606956482,
      "num_tokens": 78107876.0,
      "step": 45040
    },
    {
      "entropy": 5.790940427780152,
      "epoch": 3.504648900992025,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038004114835499833,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21235191076993942,
      "num_tokens": 78116109.0,
      "step": 45045
    },
    {
      "entropy": 5.763064384460449,
      "epoch": 3.5050379303637422,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00038001663764526374,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2085259214043617,
      "num_tokens": 78124312.0,
      "step": 45050
    },
    {
      "entropy": 5.811946058273316,
      "epoch": 3.50542695973546,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037999212534206876,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2181302696466446,
      "num_tokens": 78132624.0,
      "step": 45055
    },
    {
      "entropy": 5.784480428695678,
      "epoch": 3.5058159891071776,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003799676114457853,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21810988634824752,
      "num_tokens": 78141020.0,
      "step": 45060
    },
    {
      "entropy": 5.824211931228637,
      "epoch": 3.5062050184788953,
      "grad_norm": 1.375,
      "learning_rate": 0.00037994309595678553,
      "loss": 5.0777,
      "mean_token_accuracy": 0.20995987951755524,
      "num_tokens": 78149937.0,
      "step": 45065
    },
    {
      "entropy": 5.850016927719116,
      "epoch": 3.5065940478506126,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003799185788754413,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21601422727108002,
      "num_tokens": 78158542.0,
      "step": 45070
    },
    {
      "entropy": 5.806027221679687,
      "epoch": 3.5069830772223303,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003798940602021247,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21137935668230057,
      "num_tokens": 78167752.0,
      "step": 45075
    },
    {
      "entropy": 5.759720230102539,
      "epoch": 3.507372106594048,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003798695399372076,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21406469941139222,
      "num_tokens": 78176782.0,
      "step": 45080
    },
    {
      "entropy": 5.874069452285767,
      "epoch": 3.507761135965765,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003798450180810621,
      "loss": 5.0464,
      "mean_token_accuracy": 0.20323964208364487,
      "num_tokens": 78185087.0,
      "step": 45085
    },
    {
      "entropy": 5.756486320495606,
      "epoch": 3.508150165337483,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037982049463406044,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22313541620969773,
      "num_tokens": 78193159.0,
      "step": 45090
    },
    {
      "entropy": 5.806555128097534,
      "epoch": 3.5085391947092006,
      "grad_norm": 1.375,
      "learning_rate": 0.0003797959695965745,
      "loss": 5.0454,
      "mean_token_accuracy": 0.20570686608552932,
      "num_tokens": 78202015.0,
      "step": 45095
    },
    {
      "entropy": 5.75159010887146,
      "epoch": 3.5089282240809183,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037977144296897665,
      "loss": 4.9003,
      "mean_token_accuracy": 0.20878394097089767,
      "num_tokens": 78209644.0,
      "step": 45100
    },
    {
      "entropy": 5.749531650543213,
      "epoch": 3.5093172534526356,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003797469147516388,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2039717838168144,
      "num_tokens": 78218298.0,
      "step": 45105
    },
    {
      "entropy": 5.760928678512573,
      "epoch": 3.5097062828243533,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037972238494493333,
      "loss": 4.896,
      "mean_token_accuracy": 0.21357928216457367,
      "num_tokens": 78227547.0,
      "step": 45110
    },
    {
      "entropy": 5.7835602283477785,
      "epoch": 3.510095312196071,
      "grad_norm": 1.25,
      "learning_rate": 0.0003796978535492323,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21705976724624634,
      "num_tokens": 78236379.0,
      "step": 45115
    },
    {
      "entropy": 5.685134410858154,
      "epoch": 3.510484341567788,
      "grad_norm": 1.34375,
      "learning_rate": 0.000379673320564908,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21963596493005752,
      "num_tokens": 78244672.0,
      "step": 45120
    },
    {
      "entropy": 5.8113706588745115,
      "epoch": 3.510873370939506,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037964878599233264,
      "loss": 5.0888,
      "mean_token_accuracy": 0.20796138644218445,
      "num_tokens": 78253298.0,
      "step": 45125
    },
    {
      "entropy": 5.904376888275147,
      "epoch": 3.5112624003112236,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037962424983187856,
      "loss": 5.0137,
      "mean_token_accuracy": 0.20265062749385834,
      "num_tokens": 78262801.0,
      "step": 45130
    },
    {
      "entropy": 5.903958797454834,
      "epoch": 3.5116514296829413,
      "grad_norm": 1.3125,
      "learning_rate": 0.000379599712083918,
      "loss": 5.02,
      "mean_token_accuracy": 0.21402884721755983,
      "num_tokens": 78271415.0,
      "step": 45135
    },
    {
      "entropy": 5.777164649963379,
      "epoch": 3.5120404590546586,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003795751727488233,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2071382150053978,
      "num_tokens": 78279982.0,
      "step": 45140
    },
    {
      "entropy": 5.740121793746948,
      "epoch": 3.5124294884263763,
      "grad_norm": 1.375,
      "learning_rate": 0.00037955063182696676,
      "loss": 4.913,
      "mean_token_accuracy": 0.2113765373826027,
      "num_tokens": 78288042.0,
      "step": 45145
    },
    {
      "entropy": 5.758187961578369,
      "epoch": 3.5128185177980935,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037952608931872086,
      "loss": 4.8803,
      "mean_token_accuracy": 0.21545160859823226,
      "num_tokens": 78296587.0,
      "step": 45150
    },
    {
      "entropy": 5.863168621063233,
      "epoch": 3.513207547169811,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037950154522445784,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2093970939517021,
      "num_tokens": 78305851.0,
      "step": 45155
    },
    {
      "entropy": 5.791566991806031,
      "epoch": 3.513596576541529,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003794769995445502,
      "loss": 4.967,
      "mean_token_accuracy": 0.21096333861351013,
      "num_tokens": 78314188.0,
      "step": 45160
    },
    {
      "entropy": 5.713402271270752,
      "epoch": 3.5139856059132466,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003794524522793705,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2207271248102188,
      "num_tokens": 78322885.0,
      "step": 45165
    },
    {
      "entropy": 5.806624507904052,
      "epoch": 3.514374635284964,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037942790342929106,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21395778506994248,
      "num_tokens": 78330672.0,
      "step": 45170
    },
    {
      "entropy": 5.81657657623291,
      "epoch": 3.5147636646566816,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037940335299468437,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21521449834108353,
      "num_tokens": 78338853.0,
      "step": 45175
    },
    {
      "entropy": 5.829462194442749,
      "epoch": 3.5151526940283992,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037937880097592294,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20758809745311738,
      "num_tokens": 78346890.0,
      "step": 45180
    },
    {
      "entropy": 5.740259647369385,
      "epoch": 3.5155417234001165,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037935424737337936,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21210171580314635,
      "num_tokens": 78355187.0,
      "step": 45185
    },
    {
      "entropy": 5.793754768371582,
      "epoch": 3.515930752771834,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037932969218742616,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19935035854578018,
      "num_tokens": 78365023.0,
      "step": 45190
    },
    {
      "entropy": 5.814436721801758,
      "epoch": 3.516319782143552,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037930513541843593,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21419403403997422,
      "num_tokens": 78373662.0,
      "step": 45195
    },
    {
      "entropy": 5.847895812988281,
      "epoch": 3.5167088115152696,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037928057706678136,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1986975774168968,
      "num_tokens": 78382161.0,
      "step": 45200
    },
    {
      "entropy": 5.78133454322815,
      "epoch": 3.517097840886987,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037925601713283504,
      "loss": 4.9572,
      "mean_token_accuracy": 0.20672025680541992,
      "num_tokens": 78391165.0,
      "step": 45205
    },
    {
      "entropy": 5.77205810546875,
      "epoch": 3.5174868702587045,
      "grad_norm": 1.375,
      "learning_rate": 0.0003792314556169695,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21259001195430755,
      "num_tokens": 78399432.0,
      "step": 45210
    },
    {
      "entropy": 5.7366846084594725,
      "epoch": 3.5178758996304222,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003792068925195576,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21429156959056855,
      "num_tokens": 78408138.0,
      "step": 45215
    },
    {
      "entropy": 5.727474498748779,
      "epoch": 3.5182649290021395,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003791823278409719,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22176957726478577,
      "num_tokens": 78417684.0,
      "step": 45220
    },
    {
      "entropy": 5.766147708892822,
      "epoch": 3.518653958373857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037915776158158515,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21191220134496688,
      "num_tokens": 78426922.0,
      "step": 45225
    },
    {
      "entropy": 5.722482681274414,
      "epoch": 3.519042987745575,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003791331937417703,
      "loss": 4.8671,
      "mean_token_accuracy": 0.21339316815137863,
      "num_tokens": 78434735.0,
      "step": 45230
    },
    {
      "entropy": 5.799945211410522,
      "epoch": 3.5194320171172926,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037910862432189986,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21255179345607758,
      "num_tokens": 78443263.0,
      "step": 45235
    },
    {
      "entropy": 5.779786920547485,
      "epoch": 3.51982104648901,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003790840533223467,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21179632544517518,
      "num_tokens": 78451498.0,
      "step": 45240
    },
    {
      "entropy": 5.791426801681519,
      "epoch": 3.5202100758607275,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037905948074348387,
      "loss": 4.966,
      "mean_token_accuracy": 0.21130308359861374,
      "num_tokens": 78460097.0,
      "step": 45245
    },
    {
      "entropy": 5.760264778137207,
      "epoch": 3.520599105232445,
      "grad_norm": 1.375,
      "learning_rate": 0.000379034906585684,
      "loss": 4.8884,
      "mean_token_accuracy": 0.2094319701194763,
      "num_tokens": 78468828.0,
      "step": 45250
    },
    {
      "entropy": 5.787218904495239,
      "epoch": 3.5209881346041625,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037901033084932,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21003708839416504,
      "num_tokens": 78477930.0,
      "step": 45255
    },
    {
      "entropy": 5.7490333080291744,
      "epoch": 3.52137716397588,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003789857535347648,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22550841718912124,
      "num_tokens": 78486828.0,
      "step": 45260
    },
    {
      "entropy": 5.76067967414856,
      "epoch": 3.521766193347598,
      "grad_norm": 1.375,
      "learning_rate": 0.0003789611746423912,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2135714128613472,
      "num_tokens": 78495316.0,
      "step": 45265
    },
    {
      "entropy": 5.76439528465271,
      "epoch": 3.5221552227193156,
      "grad_norm": 1.46875,
      "learning_rate": 0.00037893659417257233,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2190175399184227,
      "num_tokens": 78503669.0,
      "step": 45270
    },
    {
      "entropy": 5.837789916992188,
      "epoch": 3.522544252091033,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037891201212568107,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21249043941497803,
      "num_tokens": 78512135.0,
      "step": 45275
    },
    {
      "entropy": 5.744658613204956,
      "epoch": 3.5229332814627505,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037888742850209037,
      "loss": 4.9986,
      "mean_token_accuracy": 0.19927141964435577,
      "num_tokens": 78521537.0,
      "step": 45280
    },
    {
      "entropy": 5.798999977111817,
      "epoch": 3.523322310834468,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003788628433021733,
      "loss": 5.008,
      "mean_token_accuracy": 0.20074418187141418,
      "num_tokens": 78530281.0,
      "step": 45285
    },
    {
      "entropy": 5.857459592819214,
      "epoch": 3.5237113402061855,
      "grad_norm": 1.328125,
      "learning_rate": 0.000378838256526303,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20343895703554155,
      "num_tokens": 78538998.0,
      "step": 45290
    },
    {
      "entropy": 5.809850072860717,
      "epoch": 3.524100369577903,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003788136681748523,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20568218529224397,
      "num_tokens": 78547539.0,
      "step": 45295
    },
    {
      "entropy": 5.785545063018799,
      "epoch": 3.524489398949621,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003787890782481945,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2147335574030876,
      "num_tokens": 78556248.0,
      "step": 45300
    },
    {
      "entropy": 5.762953424453736,
      "epoch": 3.524878428321338,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003787644867467026,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2135144516825676,
      "num_tokens": 78564710.0,
      "step": 45305
    },
    {
      "entropy": 5.79889178276062,
      "epoch": 3.525267457693056,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037873989367074977,
      "loss": 5.02,
      "mean_token_accuracy": 0.2098018780350685,
      "num_tokens": 78573213.0,
      "step": 45310
    },
    {
      "entropy": 5.817741346359253,
      "epoch": 3.5256564870647735,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037871529902070924,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20570724904537202,
      "num_tokens": 78581308.0,
      "step": 45315
    },
    {
      "entropy": 5.679773616790771,
      "epoch": 3.526045516436491,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037869070279695404,
      "loss": 4.811,
      "mean_token_accuracy": 0.2176008015871048,
      "num_tokens": 78590621.0,
      "step": 45320
    },
    {
      "entropy": 5.79523606300354,
      "epoch": 3.5264345458082085,
      "grad_norm": 1.25,
      "learning_rate": 0.00037866610499985755,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21145097017288209,
      "num_tokens": 78599726.0,
      "step": 45325
    },
    {
      "entropy": 5.793423414230347,
      "epoch": 3.526823575179926,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003786415056297929,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21301053762435912,
      "num_tokens": 78609203.0,
      "step": 45330
    },
    {
      "entropy": 5.855488729476929,
      "epoch": 3.527212604551644,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003786169046871333,
      "loss": 4.9597,
      "mean_token_accuracy": 0.20950528532266616,
      "num_tokens": 78617717.0,
      "step": 45335
    },
    {
      "entropy": 5.858923387527466,
      "epoch": 3.527601633923361,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003785923021722521,
      "loss": 5.0613,
      "mean_token_accuracy": 0.20441949218511582,
      "num_tokens": 78626252.0,
      "step": 45340
    },
    {
      "entropy": 5.804244184494019,
      "epoch": 3.527990663295079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037856769808552267,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20624428391456603,
      "num_tokens": 78634641.0,
      "step": 45345
    },
    {
      "entropy": 5.759949588775635,
      "epoch": 3.528379692666796,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003785430924273182,
      "loss": 4.9507,
      "mean_token_accuracy": 0.20903297364711762,
      "num_tokens": 78642765.0,
      "step": 45350
    },
    {
      "entropy": 5.769619417190552,
      "epoch": 3.5287687220385138,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037851848519801214,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2049776777625084,
      "num_tokens": 78652111.0,
      "step": 45355
    },
    {
      "entropy": 5.737068748474121,
      "epoch": 3.5291577514102315,
      "grad_norm": 1.25,
      "learning_rate": 0.0003784938763979778,
      "loss": 4.9043,
      "mean_token_accuracy": 0.21464954763650895,
      "num_tokens": 78661836.0,
      "step": 45360
    },
    {
      "entropy": 5.800971794128418,
      "epoch": 3.529546780781949,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003784692660275886,
      "loss": 4.9678,
      "mean_token_accuracy": 0.20863155126571656,
      "num_tokens": 78670309.0,
      "step": 45365
    },
    {
      "entropy": 5.741023349761963,
      "epoch": 3.529935810153667,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003784446540872181,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2188849076628685,
      "num_tokens": 78678741.0,
      "step": 45370
    },
    {
      "entropy": 5.765968370437622,
      "epoch": 3.530324839525384,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037842004057723957,
      "loss": 4.8843,
      "mean_token_accuracy": 0.21803620010614394,
      "num_tokens": 78686924.0,
      "step": 45375
    },
    {
      "entropy": 5.817259311676025,
      "epoch": 3.530713868897102,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003783954254980265,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21121784597635268,
      "num_tokens": 78695636.0,
      "step": 45380
    },
    {
      "entropy": 5.821703195571899,
      "epoch": 3.531102898268819,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003783708088499524,
      "loss": 4.926,
      "mean_token_accuracy": 0.21488721221685408,
      "num_tokens": 78704375.0,
      "step": 45385
    },
    {
      "entropy": 5.734186315536499,
      "epoch": 3.5314919276405368,
      "grad_norm": 1.484375,
      "learning_rate": 0.00037834619063339085,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21422878354787828,
      "num_tokens": 78712752.0,
      "step": 45390
    },
    {
      "entropy": 5.862602949142456,
      "epoch": 3.5318809570122545,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003783215708487153,
      "loss": 5.1272,
      "mean_token_accuracy": 0.20130405873060225,
      "num_tokens": 78721310.0,
      "step": 45395
    },
    {
      "entropy": 5.746823120117187,
      "epoch": 3.532269986383972,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003782969494962994,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22157032638788224,
      "num_tokens": 78730081.0,
      "step": 45400
    },
    {
      "entropy": 5.8301092147827145,
      "epoch": 3.5326590157556894,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003782723265765167,
      "loss": 5.0403,
      "mean_token_accuracy": 0.20498277842998505,
      "num_tokens": 78738253.0,
      "step": 45405
    },
    {
      "entropy": 5.807328748703003,
      "epoch": 3.533048045127407,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037824770208974085,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22100546807050706,
      "num_tokens": 78746524.0,
      "step": 45410
    },
    {
      "entropy": 5.778104591369629,
      "epoch": 3.533437074499125,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003782230760363454,
      "loss": 4.8433,
      "mean_token_accuracy": 0.2119320124387741,
      "num_tokens": 78755923.0,
      "step": 45415
    },
    {
      "entropy": 5.763756561279297,
      "epoch": 3.533826103870842,
      "grad_norm": 1.375,
      "learning_rate": 0.00037819844841670416,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21955807656049728,
      "num_tokens": 78764862.0,
      "step": 45420
    },
    {
      "entropy": 5.760458374023438,
      "epoch": 3.5342151332425598,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003781738192311906,
      "loss": 4.9498,
      "mean_token_accuracy": 0.20425319373607637,
      "num_tokens": 78774314.0,
      "step": 45425
    },
    {
      "entropy": 5.805817747116089,
      "epoch": 3.5346041626142775,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037814918848017855,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21271311193704606,
      "num_tokens": 78783071.0,
      "step": 45430
    },
    {
      "entropy": 5.790188598632812,
      "epoch": 3.534993191985995,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003781245561640418,
      "loss": 4.983,
      "mean_token_accuracy": 0.2084636002779007,
      "num_tokens": 78791405.0,
      "step": 45435
    },
    {
      "entropy": 5.6832047462463375,
      "epoch": 3.5353822213577124,
      "grad_norm": 1.375,
      "learning_rate": 0.0003780999222831541,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21687202602624894,
      "num_tokens": 78800999.0,
      "step": 45440
    },
    {
      "entropy": 5.845284557342529,
      "epoch": 3.53577125072943,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003780752868378892,
      "loss": 5.0775,
      "mean_token_accuracy": 0.2008301094174385,
      "num_tokens": 78809866.0,
      "step": 45445
    },
    {
      "entropy": 5.799866247177124,
      "epoch": 3.536160280101148,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037805064982862076,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21867633163928984,
      "num_tokens": 78817988.0,
      "step": 45450
    },
    {
      "entropy": 5.866790246963501,
      "epoch": 3.536549309472865,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003780260112557228,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20827557891607285,
      "num_tokens": 78827649.0,
      "step": 45455
    },
    {
      "entropy": 5.800807619094849,
      "epoch": 3.5369383388445828,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037800137111956904,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21047433167696,
      "num_tokens": 78836234.0,
      "step": 45460
    },
    {
      "entropy": 5.800545072555542,
      "epoch": 3.5373273682163004,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037797672942053344,
      "loss": 4.9603,
      "mean_token_accuracy": 0.20911024063825606,
      "num_tokens": 78844348.0,
      "step": 45465
    },
    {
      "entropy": 5.843839883804321,
      "epoch": 3.537716397588018,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037795208615898997,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20615206956863402,
      "num_tokens": 78853111.0,
      "step": 45470
    },
    {
      "entropy": 5.757743406295776,
      "epoch": 3.5381054269597354,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003779274413353124,
      "loss": 4.8934,
      "mean_token_accuracy": 0.21294367164373398,
      "num_tokens": 78861846.0,
      "step": 45475
    },
    {
      "entropy": 5.686294889450073,
      "epoch": 3.538494456331453,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037790279494987475,
      "loss": 4.8115,
      "mean_token_accuracy": 0.2280538097023964,
      "num_tokens": 78870246.0,
      "step": 45480
    },
    {
      "entropy": 5.8200407981872555,
      "epoch": 3.5388834857031704,
      "grad_norm": 1.375,
      "learning_rate": 0.00037787814700305096,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21546923220157624,
      "num_tokens": 78878768.0,
      "step": 45485
    },
    {
      "entropy": 5.77594404220581,
      "epoch": 3.539272515074888,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000377853497495215,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21517941504716873,
      "num_tokens": 78887647.0,
      "step": 45490
    },
    {
      "entropy": 5.8217391014099125,
      "epoch": 3.5396615444466057,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00037782884642674085,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21062049716711045,
      "num_tokens": 78896298.0,
      "step": 45495
    },
    {
      "entropy": 5.816534185409546,
      "epoch": 3.5400505738183234,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003778041937980028,
      "loss": 4.9147,
      "mean_token_accuracy": 0.21762277334928512,
      "num_tokens": 78905075.0,
      "step": 45500
    },
    {
      "entropy": 5.904707956314087,
      "epoch": 3.5404396031900407,
      "grad_norm": 1.375,
      "learning_rate": 0.00037777953960937467,
      "loss": 4.986,
      "mean_token_accuracy": 0.2139839395880699,
      "num_tokens": 78913151.0,
      "step": 45505
    },
    {
      "entropy": 5.825656795501709,
      "epoch": 3.5408286325617584,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037775488386123057,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2101698026061058,
      "num_tokens": 78922400.0,
      "step": 45510
    },
    {
      "entropy": 5.80955753326416,
      "epoch": 3.541217661933476,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037773022655394466,
      "loss": 4.9043,
      "mean_token_accuracy": 0.21352670788764955,
      "num_tokens": 78930481.0,
      "step": 45515
    },
    {
      "entropy": 5.794823932647705,
      "epoch": 3.5416066913051933,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037770556768789115,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2177618607878685,
      "num_tokens": 78939479.0,
      "step": 45520
    },
    {
      "entropy": 5.773702907562256,
      "epoch": 3.541995720676911,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037768090726344414,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20782580673694612,
      "num_tokens": 78948645.0,
      "step": 45525
    },
    {
      "entropy": 5.803742504119873,
      "epoch": 3.5423847500486287,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003776562452809777,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21798108220100404,
      "num_tokens": 78957361.0,
      "step": 45530
    },
    {
      "entropy": 5.894098424911499,
      "epoch": 3.5427737794203464,
      "grad_norm": 1.375,
      "learning_rate": 0.00037763158174086616,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20802320688962936,
      "num_tokens": 78965187.0,
      "step": 45535
    },
    {
      "entropy": 5.784253835678101,
      "epoch": 3.5431628087920637,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037760691664348367,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2132331058382988,
      "num_tokens": 78973821.0,
      "step": 45540
    },
    {
      "entropy": 5.797422933578491,
      "epoch": 3.5435518381637814,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003775822499892046,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2042221650481224,
      "num_tokens": 78982760.0,
      "step": 45545
    },
    {
      "entropy": 5.756853437423706,
      "epoch": 3.543940867535499,
      "grad_norm": 1.25,
      "learning_rate": 0.0003775575817784032,
      "loss": 4.908,
      "mean_token_accuracy": 0.2076610505580902,
      "num_tokens": 78991472.0,
      "step": 45550
    },
    {
      "entropy": 5.846407890319824,
      "epoch": 3.5443298969072163,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037753291201145375,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21355193555355073,
      "num_tokens": 79000160.0,
      "step": 45555
    },
    {
      "entropy": 5.820409059524536,
      "epoch": 3.544718926278934,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003775082406887305,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2084915593266487,
      "num_tokens": 79008774.0,
      "step": 45560
    },
    {
      "entropy": 5.851904296875,
      "epoch": 3.5451079556506517,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037748356781060785,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2060219332575798,
      "num_tokens": 79017463.0,
      "step": 45565
    },
    {
      "entropy": 5.87205319404602,
      "epoch": 3.5454969850223694,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003774588933774602,
      "loss": 4.9893,
      "mean_token_accuracy": 0.21046599000692368,
      "num_tokens": 79026102.0,
      "step": 45570
    },
    {
      "entropy": 5.739949131011963,
      "epoch": 3.5458860143940867,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003774342173896619,
      "loss": 4.8557,
      "mean_token_accuracy": 0.21847089976072312,
      "num_tokens": 79034829.0,
      "step": 45575
    },
    {
      "entropy": 5.780197668075561,
      "epoch": 3.5462750437658044,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037740953984758736,
      "loss": 5.0232,
      "mean_token_accuracy": 0.20312078297138214,
      "num_tokens": 79045120.0,
      "step": 45580
    },
    {
      "entropy": 5.7990889072418215,
      "epoch": 3.5466640731375216,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003773848607516111,
      "loss": 4.902,
      "mean_token_accuracy": 0.21447841376066207,
      "num_tokens": 79054176.0,
      "step": 45585
    },
    {
      "entropy": 5.778203344345092,
      "epoch": 3.5470531025092393,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003773601801021075,
      "loss": 4.926,
      "mean_token_accuracy": 0.21013979166746138,
      "num_tokens": 79063117.0,
      "step": 45590
    },
    {
      "entropy": 5.824786949157715,
      "epoch": 3.547442131880957,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003773354978994512,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20406679511070253,
      "num_tokens": 79071680.0,
      "step": 45595
    },
    {
      "entropy": 5.773075962066651,
      "epoch": 3.5478311612526747,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037731081414401653,
      "loss": 4.8245,
      "mean_token_accuracy": 0.21982985138893127,
      "num_tokens": 79080629.0,
      "step": 45600
    },
    {
      "entropy": 5.787632799148559,
      "epoch": 3.548220190624392,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037728612883617806,
      "loss": 4.9758,
      "mean_token_accuracy": 0.21303411722183227,
      "num_tokens": 79088677.0,
      "step": 45605
    },
    {
      "entropy": 5.734530448913574,
      "epoch": 3.5486092199961097,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003772614419763104,
      "loss": 4.8513,
      "mean_token_accuracy": 0.21452268064022065,
      "num_tokens": 79096835.0,
      "step": 45610
    },
    {
      "entropy": 5.813053321838379,
      "epoch": 3.5489982493678274,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003772367535647882,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2075965777039528,
      "num_tokens": 79105546.0,
      "step": 45615
    },
    {
      "entropy": 5.830370473861694,
      "epoch": 3.5493872787395446,
      "grad_norm": 1.40625,
      "learning_rate": 0.00037721206360198596,
      "loss": 4.9983,
      "mean_token_accuracy": 0.21154278814792632,
      "num_tokens": 79114033.0,
      "step": 45620
    },
    {
      "entropy": 5.792943716049194,
      "epoch": 3.5497763081112623,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003771873720882783,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22482145428657532,
      "num_tokens": 79122536.0,
      "step": 45625
    },
    {
      "entropy": 5.761644983291626,
      "epoch": 3.55016533748298,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037716267902403996,
      "loss": 4.85,
      "mean_token_accuracy": 0.22303423136472703,
      "num_tokens": 79130752.0,
      "step": 45630
    },
    {
      "entropy": 5.722925901412964,
      "epoch": 3.5505543668546977,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003771379844096456,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21498579531908035,
      "num_tokens": 79139552.0,
      "step": 45635
    },
    {
      "entropy": 5.832061767578125,
      "epoch": 3.550943396226415,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003771132882454698,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2147759258747101,
      "num_tokens": 79147739.0,
      "step": 45640
    },
    {
      "entropy": 5.822508811950684,
      "epoch": 3.5513324255981327,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003770885905318874,
      "loss": 5.0367,
      "mean_token_accuracy": 0.20304062217473984,
      "num_tokens": 79157035.0,
      "step": 45645
    },
    {
      "entropy": 5.753520679473877,
      "epoch": 3.5517214549698504,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003770638912692732,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2147069215774536,
      "num_tokens": 79165888.0,
      "step": 45650
    },
    {
      "entropy": 5.757738876342773,
      "epoch": 3.5521104843415676,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003770391904580019,
      "loss": 4.8627,
      "mean_token_accuracy": 0.21802772730588912,
      "num_tokens": 79173908.0,
      "step": 45655
    },
    {
      "entropy": 5.733634424209595,
      "epoch": 3.5524995137132853,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003770144880984482,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21057008057832718,
      "num_tokens": 79182667.0,
      "step": 45660
    },
    {
      "entropy": 5.72923846244812,
      "epoch": 3.552888543085003,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037698978419098704,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2093519002199173,
      "num_tokens": 79191947.0,
      "step": 45665
    },
    {
      "entropy": 5.790074062347412,
      "epoch": 3.5532775724567207,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037696507873599336,
      "loss": 4.9173,
      "mean_token_accuracy": 0.21332127302885057,
      "num_tokens": 79200903.0,
      "step": 45670
    },
    {
      "entropy": 5.894024753570557,
      "epoch": 3.553666601828438,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037694037173384177,
      "loss": 5.0276,
      "mean_token_accuracy": 0.2059596449136734,
      "num_tokens": 79209804.0,
      "step": 45675
    },
    {
      "entropy": 5.819315671920776,
      "epoch": 3.5540556312001557,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003769156631849074,
      "loss": 4.9713,
      "mean_token_accuracy": 0.20824913382530214,
      "num_tokens": 79219023.0,
      "step": 45680
    },
    {
      "entropy": 5.776279592514038,
      "epoch": 3.554444660571873,
      "grad_norm": 1.296875,
      "learning_rate": 0.000376890953089565,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2251773178577423,
      "num_tokens": 79227903.0,
      "step": 45685
    },
    {
      "entropy": 5.73232421875,
      "epoch": 3.5548336899435906,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003768662414481895,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2159269168972969,
      "num_tokens": 79236656.0,
      "step": 45690
    },
    {
      "entropy": 5.778097867965698,
      "epoch": 3.5552227193153083,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037684152826115595,
      "loss": 4.8366,
      "mean_token_accuracy": 0.21674580872058868,
      "num_tokens": 79244506.0,
      "step": 45695
    },
    {
      "entropy": 5.792387914657593,
      "epoch": 3.555611748687026,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003768168135288394,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21014923751354217,
      "num_tokens": 79253434.0,
      "step": 45700
    },
    {
      "entropy": 5.805368041992187,
      "epoch": 3.5560007780587437,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003767920972516147,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21195675432682037,
      "num_tokens": 79261465.0,
      "step": 45705
    },
    {
      "entropy": 5.7188272953033445,
      "epoch": 3.556389807430461,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037676737942985697,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2209422320127487,
      "num_tokens": 79270588.0,
      "step": 45710
    },
    {
      "entropy": 5.867591524124146,
      "epoch": 3.5567788368021787,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037674266006394123,
      "loss": 4.9708,
      "mean_token_accuracy": 0.2134295329451561,
      "num_tokens": 79279531.0,
      "step": 45715
    },
    {
      "entropy": 5.814846706390381,
      "epoch": 3.557167866173896,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037671793915424254,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21679361164569855,
      "num_tokens": 79288625.0,
      "step": 45720
    },
    {
      "entropy": 5.7889608383178714,
      "epoch": 3.5575568955456136,
      "grad_norm": 1.34375,
      "learning_rate": 0.000376693216701136,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2124807357788086,
      "num_tokens": 79296632.0,
      "step": 45725
    },
    {
      "entropy": 5.830271911621094,
      "epoch": 3.5579459249173313,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003766684927049968,
      "loss": 4.9315,
      "mean_token_accuracy": 0.20963082313537598,
      "num_tokens": 79305275.0,
      "step": 45730
    },
    {
      "entropy": 5.865696859359741,
      "epoch": 3.558334954289049,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003766437671662,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20421169102191924,
      "num_tokens": 79314686.0,
      "step": 45735
    },
    {
      "entropy": 5.7749182224273685,
      "epoch": 3.5587239836607663,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003766190400851209,
      "loss": 4.8982,
      "mean_token_accuracy": 0.21576780676841736,
      "num_tokens": 79323594.0,
      "step": 45740
    },
    {
      "entropy": 5.864116764068603,
      "epoch": 3.559113013032484,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003765943114621345,
      "loss": 5.0364,
      "mean_token_accuracy": 0.204942087829113,
      "num_tokens": 79332762.0,
      "step": 45745
    },
    {
      "entropy": 5.870150136947632,
      "epoch": 3.5595020424042016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037656958129761626,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20333393961191176,
      "num_tokens": 79341827.0,
      "step": 45750
    },
    {
      "entropy": 5.818808889389038,
      "epoch": 3.559891071775919,
      "grad_norm": 1.375,
      "learning_rate": 0.00037654484959194126,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21051836609840394,
      "num_tokens": 79350109.0,
      "step": 45755
    },
    {
      "entropy": 5.7596488952636715,
      "epoch": 3.5602801011476366,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037652011634548466,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21303477138280869,
      "num_tokens": 79359115.0,
      "step": 45760
    },
    {
      "entropy": 5.820421552658081,
      "epoch": 3.5606691305193543,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000376495381558622,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20028353333473206,
      "num_tokens": 79368472.0,
      "step": 45765
    },
    {
      "entropy": 5.787241458892822,
      "epoch": 3.561058159891072,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003764706452317284,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21225079596042634,
      "num_tokens": 79376988.0,
      "step": 45770
    },
    {
      "entropy": 5.823366594314575,
      "epoch": 3.5614471892627892,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037644590736517935,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2112113356590271,
      "num_tokens": 79385913.0,
      "step": 45775
    },
    {
      "entropy": 5.790377998352051,
      "epoch": 3.561836218634507,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003764211679593501,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2151399791240692,
      "num_tokens": 79393777.0,
      "step": 45780
    },
    {
      "entropy": 5.743767356872558,
      "epoch": 3.562225248006224,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037639642701461603,
      "loss": 4.7995,
      "mean_token_accuracy": 0.22624871581792833,
      "num_tokens": 79402150.0,
      "step": 45785
    },
    {
      "entropy": 5.779432916641236,
      "epoch": 3.562614277377942,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003763716845313526,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2091594621539116,
      "num_tokens": 79410836.0,
      "step": 45790
    },
    {
      "entropy": 5.760116815567017,
      "epoch": 3.5630033067496596,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003763469405099352,
      "loss": 4.986,
      "mean_token_accuracy": 0.20372284799814225,
      "num_tokens": 79419638.0,
      "step": 45795
    },
    {
      "entropy": 5.757781267166138,
      "epoch": 3.5633923361213773,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003763221949507392,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2117021605372429,
      "num_tokens": 79427749.0,
      "step": 45800
    },
    {
      "entropy": 5.719192981719971,
      "epoch": 3.563781365493095,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00037629744785414027,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21410184800624849,
      "num_tokens": 79436647.0,
      "step": 45805
    },
    {
      "entropy": 5.757330465316772,
      "epoch": 3.5641703948648122,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037627269922051377,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2202155828475952,
      "num_tokens": 79445599.0,
      "step": 45810
    },
    {
      "entropy": 5.873253726959229,
      "epoch": 3.56455942423653,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037624794905023523,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20748904943466187,
      "num_tokens": 79454606.0,
      "step": 45815
    },
    {
      "entropy": 5.791073656082153,
      "epoch": 3.564948453608247,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003762231973436802,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2182355135679245,
      "num_tokens": 79463543.0,
      "step": 45820
    },
    {
      "entropy": 5.767453861236572,
      "epoch": 3.565337482979965,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003761984441012243,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21109696626663207,
      "num_tokens": 79472161.0,
      "step": 45825
    },
    {
      "entropy": 5.81695327758789,
      "epoch": 3.5657265123516826,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003761736893232431,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2139684170484543,
      "num_tokens": 79481300.0,
      "step": 45830
    },
    {
      "entropy": 5.878881740570068,
      "epoch": 3.5661155417234003,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037614893301011223,
      "loss": 5.0965,
      "mean_token_accuracy": 0.2033648356795311,
      "num_tokens": 79490371.0,
      "step": 45835
    },
    {
      "entropy": 5.818865776062012,
      "epoch": 3.5665045710951175,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003761241751622072,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21251972168684005,
      "num_tokens": 79499510.0,
      "step": 45840
    },
    {
      "entropy": 5.811166286468506,
      "epoch": 3.5668936004668352,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003760994157799038,
      "loss": 5.0193,
      "mean_token_accuracy": 0.20051506608724595,
      "num_tokens": 79508705.0,
      "step": 45845
    },
    {
      "entropy": 5.747090435028076,
      "epoch": 3.567282629838553,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003760746548635777,
      "loss": 4.8417,
      "mean_token_accuracy": 0.21802288442850112,
      "num_tokens": 79517351.0,
      "step": 45850
    },
    {
      "entropy": 5.786408805847168,
      "epoch": 3.56767165921027,
      "grad_norm": 1.375,
      "learning_rate": 0.0003760498924136046,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2140167236328125,
      "num_tokens": 79525542.0,
      "step": 45855
    },
    {
      "entropy": 5.806455945968628,
      "epoch": 3.568060688581988,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037602512843036025,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21734960079193116,
      "num_tokens": 79534293.0,
      "step": 45860
    },
    {
      "entropy": 5.904753828048706,
      "epoch": 3.5684497179537056,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003760003629142203,
      "loss": 5.082,
      "mean_token_accuracy": 0.19663709849119188,
      "num_tokens": 79542588.0,
      "step": 45865
    },
    {
      "entropy": 5.792313575744629,
      "epoch": 3.5688387473254233,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037597559586556063,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2166554644703865,
      "num_tokens": 79551651.0,
      "step": 45870
    },
    {
      "entropy": 5.88037805557251,
      "epoch": 3.5692277766971405,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000375950827284757,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20535512268543243,
      "num_tokens": 79560752.0,
      "step": 45875
    },
    {
      "entropy": 5.853493785858154,
      "epoch": 3.5696168060688582,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037592605717218523,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2127401977777481,
      "num_tokens": 79569655.0,
      "step": 45880
    },
    {
      "entropy": 5.831473731994629,
      "epoch": 3.570005835440576,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037590128552822123,
      "loss": 4.8866,
      "mean_token_accuracy": 0.21507970690727235,
      "num_tokens": 79578070.0,
      "step": 45885
    },
    {
      "entropy": 5.787441635131836,
      "epoch": 3.570394864812293,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003758765123532407,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2134702533483505,
      "num_tokens": 79587444.0,
      "step": 45890
    },
    {
      "entropy": 5.787675142288208,
      "epoch": 3.570783894184011,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003758517376476198,
      "loss": 4.963,
      "mean_token_accuracy": 0.21039333194494247,
      "num_tokens": 79596259.0,
      "step": 45895
    },
    {
      "entropy": 5.8233551502227785,
      "epoch": 3.5711729235557286,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037582696141173435,
      "loss": 5.0194,
      "mean_token_accuracy": 0.20775807052850723,
      "num_tokens": 79604984.0,
      "step": 45900
    },
    {
      "entropy": 5.802441740036011,
      "epoch": 3.5715619529274463,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037580218364596015,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21516277492046357,
      "num_tokens": 79613445.0,
      "step": 45905
    },
    {
      "entropy": 5.684372806549073,
      "epoch": 3.5719509822991635,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037577740435067327,
      "loss": 4.7679,
      "mean_token_accuracy": 0.2205743357539177,
      "num_tokens": 79621879.0,
      "step": 45910
    },
    {
      "entropy": 5.727406454086304,
      "epoch": 3.572340011670881,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037575262352624965,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21073272675275803,
      "num_tokens": 79630139.0,
      "step": 45915
    },
    {
      "entropy": 5.7289306163787845,
      "epoch": 3.5727290410425985,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037572784117306535,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21883398294448853,
      "num_tokens": 79638890.0,
      "step": 45920
    },
    {
      "entropy": 5.7716775894165036,
      "epoch": 3.573118070414316,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003757030572914963,
      "loss": 4.9628,
      "mean_token_accuracy": 0.216367968916893,
      "num_tokens": 79647913.0,
      "step": 45925
    },
    {
      "entropy": 5.735942554473877,
      "epoch": 3.573507099786034,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003756782718819188,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23060536831617356,
      "num_tokens": 79656119.0,
      "step": 45930
    },
    {
      "entropy": 5.724011373519898,
      "epoch": 3.5738961291577516,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037565348494470873,
      "loss": 4.8608,
      "mean_token_accuracy": 0.21499150097370148,
      "num_tokens": 79664223.0,
      "step": 45935
    },
    {
      "entropy": 5.7903515815734865,
      "epoch": 3.574285158529469,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003756286964802422,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20209017246961594,
      "num_tokens": 79673227.0,
      "step": 45940
    },
    {
      "entropy": 5.784465646743774,
      "epoch": 3.5746741879011865,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003756039064888954,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21951641738414765,
      "num_tokens": 79682107.0,
      "step": 45945
    },
    {
      "entropy": 5.764537525177002,
      "epoch": 3.575063217272904,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003755791149710444,
      "loss": 4.8682,
      "mean_token_accuracy": 0.21879131644964217,
      "num_tokens": 79690231.0,
      "step": 45950
    },
    {
      "entropy": 5.7277425765991214,
      "epoch": 3.5754522466446215,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00037555432192706547,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2179669439792633,
      "num_tokens": 79698117.0,
      "step": 45955
    },
    {
      "entropy": 5.797821521759033,
      "epoch": 3.575841276016339,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037552952735733466,
      "loss": 4.918,
      "mean_token_accuracy": 0.22032899558544158,
      "num_tokens": 79707301.0,
      "step": 45960
    },
    {
      "entropy": 5.8143998146057125,
      "epoch": 3.576230305388057,
      "grad_norm": 1.375,
      "learning_rate": 0.0003755047312622283,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2020946428179741,
      "num_tokens": 79715547.0,
      "step": 45965
    },
    {
      "entropy": 5.782806491851806,
      "epoch": 3.5766193347597746,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037547993364212267,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2154175654053688,
      "num_tokens": 79724244.0,
      "step": 45970
    },
    {
      "entropy": 5.845580196380615,
      "epoch": 3.577008364131492,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000375455134497394,
      "loss": 4.991,
      "mean_token_accuracy": 0.2138822227716446,
      "num_tokens": 79733001.0,
      "step": 45975
    },
    {
      "entropy": 5.725759267807007,
      "epoch": 3.5773973935032095,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003754303338284186,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21847364604473113,
      "num_tokens": 79741821.0,
      "step": 45980
    },
    {
      "entropy": 5.740760374069214,
      "epoch": 3.577786422874927,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037540553163557263,
      "loss": 4.8798,
      "mean_token_accuracy": 0.21876145750284196,
      "num_tokens": 79750729.0,
      "step": 45985
    },
    {
      "entropy": 5.732310247421265,
      "epoch": 3.5781754522466445,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037538072791923257,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21169619411230087,
      "num_tokens": 79759597.0,
      "step": 45990
    },
    {
      "entropy": 5.813152408599853,
      "epoch": 3.578564481618362,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037535592267977467,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20453510135412217,
      "num_tokens": 79768500.0,
      "step": 45995
    },
    {
      "entropy": 5.885359573364258,
      "epoch": 3.57895351099008,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003753311159175754,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20771556198596955,
      "num_tokens": 79777747.0,
      "step": 46000
    },
    {
      "entropy": 5.835975074768067,
      "epoch": 3.5793425403617976,
      "grad_norm": 1.453125,
      "learning_rate": 0.00037530630763301113,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2184900552034378,
      "num_tokens": 79786033.0,
      "step": 46005
    },
    {
      "entropy": 5.719406032562256,
      "epoch": 3.579731569733515,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003752814978264584,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21667373776435853,
      "num_tokens": 79795143.0,
      "step": 46010
    },
    {
      "entropy": 5.752248096466064,
      "epoch": 3.5801205991052325,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003752566864982934,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21427129209041595,
      "num_tokens": 79804110.0,
      "step": 46015
    },
    {
      "entropy": 5.740458011627197,
      "epoch": 3.5805096284769498,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037523187364889293,
      "loss": 4.7865,
      "mean_token_accuracy": 0.22562515288591384,
      "num_tokens": 79812538.0,
      "step": 46020
    },
    {
      "entropy": 5.798637962341308,
      "epoch": 3.5808986578486675,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003752070592786332,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21472149193286896,
      "num_tokens": 79821730.0,
      "step": 46025
    },
    {
      "entropy": 5.778572511672974,
      "epoch": 3.581287687220385,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037518224338789084,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21569084376096725,
      "num_tokens": 79829950.0,
      "step": 46030
    },
    {
      "entropy": 5.79010124206543,
      "epoch": 3.581676716592103,
      "grad_norm": 1.25,
      "learning_rate": 0.0003751574259770424,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2083486169576645,
      "num_tokens": 79838723.0,
      "step": 46035
    },
    {
      "entropy": 5.841555213928222,
      "epoch": 3.58206574596382,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003751326070464645,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20630229115486146,
      "num_tokens": 79846908.0,
      "step": 46040
    },
    {
      "entropy": 5.781336641311645,
      "epoch": 3.582454775335538,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003751077865965337,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21865664124488832,
      "num_tokens": 79855127.0,
      "step": 46045
    },
    {
      "entropy": 5.890072345733643,
      "epoch": 3.5828438047072555,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037508296462762647,
      "loss": 5.0222,
      "mean_token_accuracy": 0.2119183599948883,
      "num_tokens": 79864144.0,
      "step": 46050
    },
    {
      "entropy": 5.747876214981079,
      "epoch": 3.5832328340789728,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037505814114011956,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22217923402786255,
      "num_tokens": 79872325.0,
      "step": 46055
    },
    {
      "entropy": 5.749766731262207,
      "epoch": 3.5836218634506904,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037503331613438965,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2188735008239746,
      "num_tokens": 79880787.0,
      "step": 46060
    },
    {
      "entropy": 5.77577772140503,
      "epoch": 3.584010892822408,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037500848961081334,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21478435397148132,
      "num_tokens": 79889320.0,
      "step": 46065
    },
    {
      "entropy": 5.784390830993653,
      "epoch": 3.584399922194126,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037498366156976746,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2137455403804779,
      "num_tokens": 79898187.0,
      "step": 46070
    },
    {
      "entropy": 5.769625329971314,
      "epoch": 3.584788951565843,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037495883201162854,
      "loss": 4.9368,
      "mean_token_accuracy": 0.20776156187057496,
      "num_tokens": 79907490.0,
      "step": 46075
    },
    {
      "entropy": 5.721699857711792,
      "epoch": 3.585177980937561,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003749340009367736,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22321154773235322,
      "num_tokens": 79915644.0,
      "step": 46080
    },
    {
      "entropy": 5.814938879013061,
      "epoch": 3.5855670103092785,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037490916834557916,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21523074358701705,
      "num_tokens": 79923991.0,
      "step": 46085
    },
    {
      "entropy": 5.799943971633911,
      "epoch": 3.5859560396809957,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003748843342384221,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20668514221906661,
      "num_tokens": 79932727.0,
      "step": 46090
    },
    {
      "entropy": 5.8021032333374025,
      "epoch": 3.5863450690527134,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037485949861567925,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21377895772457123,
      "num_tokens": 79941978.0,
      "step": 46095
    },
    {
      "entropy": 5.813357210159301,
      "epoch": 3.586734098424431,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003748346614777275,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20584224164485931,
      "num_tokens": 79951050.0,
      "step": 46100
    },
    {
      "entropy": 5.801635789871216,
      "epoch": 3.587123127796149,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003748098228249437,
      "loss": 5.0346,
      "mean_token_accuracy": 0.204413565993309,
      "num_tokens": 79960361.0,
      "step": 46105
    },
    {
      "entropy": 5.771216678619385,
      "epoch": 3.587512157167866,
      "grad_norm": 1.375,
      "learning_rate": 0.00037478498265770463,
      "loss": 4.9994,
      "mean_token_accuracy": 0.21523209512233735,
      "num_tokens": 79969399.0,
      "step": 46110
    },
    {
      "entropy": 5.85726752281189,
      "epoch": 3.587901186539584,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003747601409763872,
      "loss": 4.9715,
      "mean_token_accuracy": 0.20794013738632203,
      "num_tokens": 79977587.0,
      "step": 46115
    },
    {
      "entropy": 5.785594463348389,
      "epoch": 3.588290215911301,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037473529778136853,
      "loss": 4.8531,
      "mean_token_accuracy": 0.21455038338899612,
      "num_tokens": 79985875.0,
      "step": 46120
    },
    {
      "entropy": 5.763126373291016,
      "epoch": 3.5886792452830187,
      "grad_norm": 1.40625,
      "learning_rate": 0.00037471045307302544,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21076056361198425,
      "num_tokens": 79994002.0,
      "step": 46125
    },
    {
      "entropy": 5.746276187896728,
      "epoch": 3.5890682746547364,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037468560685173497,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2198118105530739,
      "num_tokens": 80002008.0,
      "step": 46130
    },
    {
      "entropy": 5.790261840820312,
      "epoch": 3.589457304026454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000374660759117874,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22141229957342148,
      "num_tokens": 80010684.0,
      "step": 46135
    },
    {
      "entropy": 5.870513439178467,
      "epoch": 3.589846333398172,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037463590987181966,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21402999013662338,
      "num_tokens": 80019700.0,
      "step": 46140
    },
    {
      "entropy": 5.7605870246887205,
      "epoch": 3.590235362769889,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037461105911394904,
      "loss": 4.8443,
      "mean_token_accuracy": 0.21756132543087006,
      "num_tokens": 80028175.0,
      "step": 46145
    },
    {
      "entropy": 5.790295743942261,
      "epoch": 3.590624392141607,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037458620684463907,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21332819014787674,
      "num_tokens": 80037149.0,
      "step": 46150
    },
    {
      "entropy": 5.7860071659088135,
      "epoch": 3.591013421513324,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037456135306426693,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21168828159570693,
      "num_tokens": 80045007.0,
      "step": 46155
    },
    {
      "entropy": 5.768174266815185,
      "epoch": 3.5914024508850417,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003745364977732097,
      "loss": 4.9674,
      "mean_token_accuracy": 0.208631631731987,
      "num_tokens": 80053514.0,
      "step": 46160
    },
    {
      "entropy": 5.832877159118652,
      "epoch": 3.5917914802567594,
      "grad_norm": 1.375,
      "learning_rate": 0.0003745116409718446,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2064157873392105,
      "num_tokens": 80062645.0,
      "step": 46165
    },
    {
      "entropy": 5.830390596389771,
      "epoch": 3.592180509628477,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003744867826605488,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2099575400352478,
      "num_tokens": 80070427.0,
      "step": 46170
    },
    {
      "entropy": 5.877482509613037,
      "epoch": 3.5925695390001944,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003744619228396993,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20259436070919037,
      "num_tokens": 80079355.0,
      "step": 46175
    },
    {
      "entropy": 5.768975687026978,
      "epoch": 3.592958568371912,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003744370615096734,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21251513361930846,
      "num_tokens": 80087906.0,
      "step": 46180
    },
    {
      "entropy": 5.8231452941894535,
      "epoch": 3.5933475977436298,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003744121986708485,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20305330455303192,
      "num_tokens": 80096179.0,
      "step": 46185
    },
    {
      "entropy": 5.848571395874023,
      "epoch": 3.593736627115347,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037438733432360163,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20679403096437454,
      "num_tokens": 80104539.0,
      "step": 46190
    },
    {
      "entropy": 5.7761305809021,
      "epoch": 3.5941256564870647,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003743624684683102,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20879078209400176,
      "num_tokens": 80112798.0,
      "step": 46195
    },
    {
      "entropy": 5.845119619369507,
      "epoch": 3.5945146858587824,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003743376011053514,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1998143896460533,
      "num_tokens": 80121480.0,
      "step": 46200
    },
    {
      "entropy": 5.869574928283692,
      "epoch": 3.5949037152305,
      "grad_norm": 1.375,
      "learning_rate": 0.00037431273223510265,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2079964682459831,
      "num_tokens": 80130068.0,
      "step": 46205
    },
    {
      "entropy": 5.776275157928467,
      "epoch": 3.5952927446022174,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003742878618579413,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21802653074264527,
      "num_tokens": 80139434.0,
      "step": 46210
    },
    {
      "entropy": 5.861506366729737,
      "epoch": 3.595681773973935,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003742629899742446,
      "loss": 5.0803,
      "mean_token_accuracy": 0.19995256960392,
      "num_tokens": 80147858.0,
      "step": 46215
    },
    {
      "entropy": 5.795623922348023,
      "epoch": 3.5960708033456523,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037423811658439014,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21821559518575667,
      "num_tokens": 80156803.0,
      "step": 46220
    },
    {
      "entropy": 5.767368984222412,
      "epoch": 3.59645983271737,
      "grad_norm": 1.25,
      "learning_rate": 0.00037421324168875503,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2240346223115921,
      "num_tokens": 80165464.0,
      "step": 46225
    },
    {
      "entropy": 5.825710248947144,
      "epoch": 3.5968488620890877,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000374188365287717,
      "loss": 4.9907,
      "mean_token_accuracy": 0.20965728759765626,
      "num_tokens": 80173875.0,
      "step": 46230
    },
    {
      "entropy": 5.733244323730469,
      "epoch": 3.5972378914608054,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003741634873816534,
      "loss": 4.7986,
      "mean_token_accuracy": 0.22824711352586746,
      "num_tokens": 80183263.0,
      "step": 46235
    },
    {
      "entropy": 5.837101697921753,
      "epoch": 3.597626920832523,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037413860797094175,
      "loss": 5.041,
      "mean_token_accuracy": 0.2051907867193222,
      "num_tokens": 80192295.0,
      "step": 46240
    },
    {
      "entropy": 5.759378290176391,
      "epoch": 3.5980159502042404,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037411372705595947,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21273817718029023,
      "num_tokens": 80200397.0,
      "step": 46245
    },
    {
      "entropy": 5.7736897468566895,
      "epoch": 3.598404979575958,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003740888446370842,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2153225377202034,
      "num_tokens": 80209150.0,
      "step": 46250
    },
    {
      "entropy": 5.872681665420532,
      "epoch": 3.5987940089476753,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003740639607146933,
      "loss": 5.0626,
      "mean_token_accuracy": 0.2034439653158188,
      "num_tokens": 80218218.0,
      "step": 46255
    },
    {
      "entropy": 5.824189853668213,
      "epoch": 3.599183038319393,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037403907528916455,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22059766799211503,
      "num_tokens": 80226811.0,
      "step": 46260
    },
    {
      "entropy": 5.884413957595825,
      "epoch": 3.5995720676911107,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003740141883608755,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20581759810447692,
      "num_tokens": 80234802.0,
      "step": 46265
    },
    {
      "entropy": 5.770146417617798,
      "epoch": 3.5999610970628284,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003739892999302037,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22219782918691636,
      "num_tokens": 80243216.0,
      "step": 46270
    },
    {
      "entropy": 5.80918869972229,
      "epoch": 3.6003501264345457,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003739644099975269,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21303130686283112,
      "num_tokens": 80251293.0,
      "step": 46275
    },
    {
      "entropy": 5.729293251037598,
      "epoch": 3.6007391558062634,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037393951856322263,
      "loss": 4.907,
      "mean_token_accuracy": 0.21701274812221527,
      "num_tokens": 80259455.0,
      "step": 46280
    },
    {
      "entropy": 5.732075881958008,
      "epoch": 3.601128185177981,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003739146256276686,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21778518259525298,
      "num_tokens": 80268207.0,
      "step": 46285
    },
    {
      "entropy": 5.814977931976318,
      "epoch": 3.6015172145496983,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037388973119124267,
      "loss": 4.946,
      "mean_token_accuracy": 0.20817862302064896,
      "num_tokens": 80276700.0,
      "step": 46290
    },
    {
      "entropy": 5.728993129730225,
      "epoch": 3.601906243921416,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003738648352543224,
      "loss": 4.8725,
      "mean_token_accuracy": 0.2171982377767563,
      "num_tokens": 80285322.0,
      "step": 46295
    },
    {
      "entropy": 5.8055178165435795,
      "epoch": 3.6022952732931337,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003738399378172856,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20672914385795593,
      "num_tokens": 80295593.0,
      "step": 46300
    },
    {
      "entropy": 5.906806087493896,
      "epoch": 3.6026843026648514,
      "grad_norm": 1.375,
      "learning_rate": 0.0003738150388805101,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2043824702501297,
      "num_tokens": 80305576.0,
      "step": 46305
    },
    {
      "entropy": 5.850510978698731,
      "epoch": 3.6030733320365687,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037379013844437373,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2131205677986145,
      "num_tokens": 80314181.0,
      "step": 46310
    },
    {
      "entropy": 5.814906215667724,
      "epoch": 3.6034623614082864,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037376523650925415,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2077443853020668,
      "num_tokens": 80322637.0,
      "step": 46315
    },
    {
      "entropy": 5.843758392333984,
      "epoch": 3.603851390780004,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003737403330755293,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21473028212785722,
      "num_tokens": 80331331.0,
      "step": 46320
    },
    {
      "entropy": 5.756258726119995,
      "epoch": 3.6042404201517213,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003737154281435771,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22059810608625413,
      "num_tokens": 80340171.0,
      "step": 46325
    },
    {
      "entropy": 5.827457904815674,
      "epoch": 3.604629449523439,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037369052171377535,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2057574838399887,
      "num_tokens": 80349226.0,
      "step": 46330
    },
    {
      "entropy": 5.794615602493286,
      "epoch": 3.6050184788951567,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003736656137865021,
      "loss": 4.8668,
      "mean_token_accuracy": 0.21670107245445253,
      "num_tokens": 80358383.0,
      "step": 46335
    },
    {
      "entropy": 5.857770729064941,
      "epoch": 3.6054075082668744,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037364070436213514,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2167468249797821,
      "num_tokens": 80366644.0,
      "step": 46340
    },
    {
      "entropy": 5.82479419708252,
      "epoch": 3.6057965376385916,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003736157934410525,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21104977726936341,
      "num_tokens": 80375775.0,
      "step": 46345
    },
    {
      "entropy": 5.8596940517425535,
      "epoch": 3.6061855670103093,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003735908810236322,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2055319294333458,
      "num_tokens": 80384505.0,
      "step": 46350
    },
    {
      "entropy": 5.839061737060547,
      "epoch": 3.6065745963820266,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037356596711025214,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21295370161533356,
      "num_tokens": 80392538.0,
      "step": 46355
    },
    {
      "entropy": 5.797332906723023,
      "epoch": 3.6069636257537443,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003735410517012905,
      "loss": 4.874,
      "mean_token_accuracy": 0.21232632398605347,
      "num_tokens": 80400811.0,
      "step": 46360
    },
    {
      "entropy": 5.8124675273895265,
      "epoch": 3.607352655125462,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003735161347971252,
      "loss": 5.0085,
      "mean_token_accuracy": 0.203411203622818,
      "num_tokens": 80409812.0,
      "step": 46365
    },
    {
      "entropy": 5.812157917022705,
      "epoch": 3.6077416844971797,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003734912163981344,
      "loss": 4.996,
      "mean_token_accuracy": 0.2093289688229561,
      "num_tokens": 80418325.0,
      "step": 46370
    },
    {
      "entropy": 5.795334959030152,
      "epoch": 3.608130713868897,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037346629650469605,
      "loss": 4.9825,
      "mean_token_accuracy": 0.20568234324455262,
      "num_tokens": 80427318.0,
      "step": 46375
    },
    {
      "entropy": 5.833534145355225,
      "epoch": 3.6085197432406146,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037344137511718847,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19779324680566787,
      "num_tokens": 80437238.0,
      "step": 46380
    },
    {
      "entropy": 5.821755218505859,
      "epoch": 3.6089087726123323,
      "grad_norm": 1.25,
      "learning_rate": 0.00037341645223598965,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21192000806331635,
      "num_tokens": 80446497.0,
      "step": 46385
    },
    {
      "entropy": 5.837676858901977,
      "epoch": 3.6092978019840496,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037339152786147777,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21191132962703704,
      "num_tokens": 80455115.0,
      "step": 46390
    },
    {
      "entropy": 5.7890965938568115,
      "epoch": 3.6096868313557673,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037336660199403115,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2142543151974678,
      "num_tokens": 80463596.0,
      "step": 46395
    },
    {
      "entropy": 5.801033592224121,
      "epoch": 3.610075860727485,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003733416746340279,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21460147053003312,
      "num_tokens": 80472160.0,
      "step": 46400
    },
    {
      "entropy": 5.806341648101807,
      "epoch": 3.6104648900992027,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003733167457818463,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2136998549103737,
      "num_tokens": 80481317.0,
      "step": 46405
    },
    {
      "entropy": 5.790742540359497,
      "epoch": 3.61085391947092,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003732918154378645,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2175969362258911,
      "num_tokens": 80489977.0,
      "step": 46410
    },
    {
      "entropy": 5.809862804412842,
      "epoch": 3.6112429488426376,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037326688360246087,
      "loss": 4.9504,
      "mean_token_accuracy": 0.20882742702960969,
      "num_tokens": 80499109.0,
      "step": 46415
    },
    {
      "entropy": 5.770439004898071,
      "epoch": 3.6116319782143553,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003732419502760137,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2172260493040085,
      "num_tokens": 80506981.0,
      "step": 46420
    },
    {
      "entropy": 5.835809659957886,
      "epoch": 3.6120210075860726,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037321701545890126,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21650746315717698,
      "num_tokens": 80515572.0,
      "step": 46425
    },
    {
      "entropy": 5.811970329284668,
      "epoch": 3.6124100369577903,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003731920791515019,
      "loss": 4.9458,
      "mean_token_accuracy": 0.20466053932905198,
      "num_tokens": 80525196.0,
      "step": 46430
    },
    {
      "entropy": 5.825013113021851,
      "epoch": 3.612799066329508,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003731671413541942,
      "loss": 4.9231,
      "mean_token_accuracy": 0.21387933194637299,
      "num_tokens": 80533358.0,
      "step": 46435
    },
    {
      "entropy": 5.812676525115966,
      "epoch": 3.6131880957012257,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00037314220206735626,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2093340665102005,
      "num_tokens": 80541632.0,
      "step": 46440
    },
    {
      "entropy": 5.790643358230591,
      "epoch": 3.613577125072943,
      "grad_norm": 1.40625,
      "learning_rate": 0.00037311726129136665,
      "loss": 5.0993,
      "mean_token_accuracy": 0.19837437719106674,
      "num_tokens": 80550134.0,
      "step": 46445
    },
    {
      "entropy": 5.828024578094483,
      "epoch": 3.6139661544446606,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003730923190266038,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2096709728240967,
      "num_tokens": 80558892.0,
      "step": 46450
    },
    {
      "entropy": 5.934130716323852,
      "epoch": 3.614355183816378,
      "grad_norm": 1.28125,
      "learning_rate": 0.000373067375273446,
      "loss": 5.04,
      "mean_token_accuracy": 0.20332380533218383,
      "num_tokens": 80568613.0,
      "step": 46455
    },
    {
      "entropy": 5.8389081954956055,
      "epoch": 3.6147442131880956,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037304243003227197,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21990666836500167,
      "num_tokens": 80577741.0,
      "step": 46460
    },
    {
      "entropy": 5.8367085456848145,
      "epoch": 3.6151332425598133,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003730174833034601,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2102118343114853,
      "num_tokens": 80586524.0,
      "step": 46465
    },
    {
      "entropy": 5.8092326641082765,
      "epoch": 3.615522271931531,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037299253508738894,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2141026183962822,
      "num_tokens": 80595592.0,
      "step": 46470
    },
    {
      "entropy": 5.820678234100342,
      "epoch": 3.6159113013032487,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037296758538443703,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22017295509576798,
      "num_tokens": 80603632.0,
      "step": 46475
    },
    {
      "entropy": 5.775794172286988,
      "epoch": 3.616300330674966,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037294263419498295,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21951792985200883,
      "num_tokens": 80612255.0,
      "step": 46480
    },
    {
      "entropy": 5.749066114425659,
      "epoch": 3.6166893600466836,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003729176815194053,
      "loss": 4.8748,
      "mean_token_accuracy": 0.2234942778944969,
      "num_tokens": 80621073.0,
      "step": 46485
    },
    {
      "entropy": 5.854146099090576,
      "epoch": 3.617078389418401,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037289272735808267,
      "loss": 5.031,
      "mean_token_accuracy": 0.21090429723262788,
      "num_tokens": 80629376.0,
      "step": 46490
    },
    {
      "entropy": 5.817367744445801,
      "epoch": 3.6174674187901186,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003728677717113936,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20798057317733765,
      "num_tokens": 80637608.0,
      "step": 46495
    },
    {
      "entropy": 5.774727249145508,
      "epoch": 3.6178564481618363,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037284281457971707,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21471129208803177,
      "num_tokens": 80646361.0,
      "step": 46500
    },
    {
      "entropy": 5.810741806030274,
      "epoch": 3.618245477533554,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003728178559634315,
      "loss": 4.9311,
      "mean_token_accuracy": 0.20924106240272522,
      "num_tokens": 80654894.0,
      "step": 46505
    },
    {
      "entropy": 5.846448659896851,
      "epoch": 3.618634506905271,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003727928958629156,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2095610722899437,
      "num_tokens": 80664088.0,
      "step": 46510
    },
    {
      "entropy": 5.810062742233276,
      "epoch": 3.619023536276989,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037276793427854817,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2112276256084442,
      "num_tokens": 80672173.0,
      "step": 46515
    },
    {
      "entropy": 5.858865880966187,
      "epoch": 3.6194125656487066,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037274297121070796,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20318731665611267,
      "num_tokens": 80681744.0,
      "step": 46520
    },
    {
      "entropy": 5.840253829956055,
      "epoch": 3.619801595020424,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003727180066597737,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21166423261165618,
      "num_tokens": 80690443.0,
      "step": 46525
    },
    {
      "entropy": 5.848503541946411,
      "epoch": 3.6201906243921416,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003726930406261243,
      "loss": 5.0658,
      "mean_token_accuracy": 0.20456180721521378,
      "num_tokens": 80698472.0,
      "step": 46530
    },
    {
      "entropy": 5.82731032371521,
      "epoch": 3.6205796537638593,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003726680731101385,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2101417139172554,
      "num_tokens": 80706757.0,
      "step": 46535
    },
    {
      "entropy": 5.811495351791382,
      "epoch": 3.620968683135577,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003726431041121951,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2114171028137207,
      "num_tokens": 80716612.0,
      "step": 46540
    },
    {
      "entropy": 5.778885364532471,
      "epoch": 3.621357712507294,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000372618133632673,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2181735798716545,
      "num_tokens": 80725380.0,
      "step": 46545
    },
    {
      "entropy": 5.865681314468384,
      "epoch": 3.621746741879012,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003725931616719511,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2019834965467453,
      "num_tokens": 80733916.0,
      "step": 46550
    },
    {
      "entropy": 5.837443494796753,
      "epoch": 3.622135771250729,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003725681882304084,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21284728944301606,
      "num_tokens": 80742702.0,
      "step": 46555
    },
    {
      "entropy": 5.77176456451416,
      "epoch": 3.622524800622447,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003725432133084237,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2161381021142006,
      "num_tokens": 80751422.0,
      "step": 46560
    },
    {
      "entropy": 5.832637548446655,
      "epoch": 3.6229138299941646,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003725182369063759,
      "loss": 4.95,
      "mean_token_accuracy": 0.2158133402466774,
      "num_tokens": 80760147.0,
      "step": 46565
    },
    {
      "entropy": 5.769155550003052,
      "epoch": 3.6233028593658823,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003724932590246441,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21189437955617904,
      "num_tokens": 80768509.0,
      "step": 46570
    },
    {
      "entropy": 5.785255479812622,
      "epoch": 3.6236918887376,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003724682796636073,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21475027799606322,
      "num_tokens": 80777226.0,
      "step": 46575
    },
    {
      "entropy": 5.8462152004241945,
      "epoch": 3.624080918109317,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003724432988236445,
      "loss": 4.9848,
      "mean_token_accuracy": 0.209323488175869,
      "num_tokens": 80786029.0,
      "step": 46580
    },
    {
      "entropy": 5.816490936279297,
      "epoch": 3.624469947481035,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003724183165051347,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2037944883108139,
      "num_tokens": 80794524.0,
      "step": 46585
    },
    {
      "entropy": 5.821278858184814,
      "epoch": 3.624858976852752,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037239333270845714,
      "loss": 4.975,
      "mean_token_accuracy": 0.20882696509361268,
      "num_tokens": 80802994.0,
      "step": 46590
    },
    {
      "entropy": 5.765382909774781,
      "epoch": 3.62524800622447,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003723683474339907,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21575449258089066,
      "num_tokens": 80811907.0,
      "step": 46595
    },
    {
      "entropy": 5.844417858123779,
      "epoch": 3.6256370355961876,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003723433606821145,
      "loss": 4.9837,
      "mean_token_accuracy": 0.20974897593259811,
      "num_tokens": 80820709.0,
      "step": 46600
    },
    {
      "entropy": 5.849398946762085,
      "epoch": 3.6260260649679052,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037231837245320777,
      "loss": 4.9651,
      "mean_token_accuracy": 0.20678347647190093,
      "num_tokens": 80829940.0,
      "step": 46605
    },
    {
      "entropy": 5.849050045013428,
      "epoch": 3.6264150943396225,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037229338274764965,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21531798541545868,
      "num_tokens": 80838863.0,
      "step": 46610
    },
    {
      "entropy": 5.8440194606781,
      "epoch": 3.62680412371134,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003722683915658193,
      "loss": 4.9472,
      "mean_token_accuracy": 0.214973084628582,
      "num_tokens": 80847508.0,
      "step": 46615
    },
    {
      "entropy": 5.74535117149353,
      "epoch": 3.627193153083058,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037224339890809587,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20429691225290297,
      "num_tokens": 80856738.0,
      "step": 46620
    },
    {
      "entropy": 5.8455548763275145,
      "epoch": 3.627582182454775,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003722184047748587,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20881297588348388,
      "num_tokens": 80864985.0,
      "step": 46625
    },
    {
      "entropy": 5.833503866195679,
      "epoch": 3.627971211826493,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037219340916648703,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2035933554172516,
      "num_tokens": 80874025.0,
      "step": 46630
    },
    {
      "entropy": 5.837823963165283,
      "epoch": 3.6283602411982105,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037216841208336,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20978901386260987,
      "num_tokens": 80882410.0,
      "step": 46635
    },
    {
      "entropy": 5.865505695343018,
      "epoch": 3.6287492705699282,
      "grad_norm": 1.375,
      "learning_rate": 0.00037214341352585697,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2062186524271965,
      "num_tokens": 80890776.0,
      "step": 46640
    },
    {
      "entropy": 5.856614828109741,
      "epoch": 3.6291382999416455,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003721184134943573,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2127270370721817,
      "num_tokens": 80898985.0,
      "step": 46645
    },
    {
      "entropy": 5.761981201171875,
      "epoch": 3.629527329313363,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037209341198924024,
      "loss": 4.8514,
      "mean_token_accuracy": 0.21824664175510405,
      "num_tokens": 80907046.0,
      "step": 46650
    },
    {
      "entropy": 5.753596830368042,
      "epoch": 3.6299163586850804,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003720684090108852,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21290538012981414,
      "num_tokens": 80915764.0,
      "step": 46655
    },
    {
      "entropy": 5.745474147796631,
      "epoch": 3.630305388056798,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003720434045596716,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22011713087558746,
      "num_tokens": 80925049.0,
      "step": 46660
    },
    {
      "entropy": 5.852899742126465,
      "epoch": 3.630694417428516,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003720183986359788,
      "loss": 5.07,
      "mean_token_accuracy": 0.20584283620119095,
      "num_tokens": 80934928.0,
      "step": 46665
    },
    {
      "entropy": 5.889201545715332,
      "epoch": 3.6310834468002335,
      "grad_norm": 1.25,
      "learning_rate": 0.0003719933912401862,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20568465143442155,
      "num_tokens": 80943388.0,
      "step": 46670
    },
    {
      "entropy": 5.850353002548218,
      "epoch": 3.6314724761719512,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003719683823726732,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21384547054767608,
      "num_tokens": 80952887.0,
      "step": 46675
    },
    {
      "entropy": 5.830696296691895,
      "epoch": 3.6318615055436685,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037194337203381945,
      "loss": 4.939,
      "mean_token_accuracy": 0.21464853733778,
      "num_tokens": 80961284.0,
      "step": 46680
    },
    {
      "entropy": 5.799163579940796,
      "epoch": 3.632250534915386,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037191836022400427,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21478018313646316,
      "num_tokens": 80969963.0,
      "step": 46685
    },
    {
      "entropy": 5.737101221084595,
      "epoch": 3.6326395642871034,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037189334694360724,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21506072133779525,
      "num_tokens": 80978556.0,
      "step": 46690
    },
    {
      "entropy": 5.805430173873901,
      "epoch": 3.633028593658821,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003718683321930079,
      "loss": 4.9439,
      "mean_token_accuracy": 0.207755409181118,
      "num_tokens": 80987615.0,
      "step": 46695
    },
    {
      "entropy": 5.752237701416016,
      "epoch": 3.633417623030539,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003718433159725857,
      "loss": 4.8305,
      "mean_token_accuracy": 0.21764546483755112,
      "num_tokens": 80995729.0,
      "step": 46700
    },
    {
      "entropy": 5.801904582977295,
      "epoch": 3.6338066524022565,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003718182982827205,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20688237100839615,
      "num_tokens": 81004388.0,
      "step": 46705
    },
    {
      "entropy": 5.824491167068482,
      "epoch": 3.634195681773974,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037179327912379166,
      "loss": 4.9925,
      "mean_token_accuracy": 0.20476346611976623,
      "num_tokens": 81013216.0,
      "step": 46710
    },
    {
      "entropy": 5.8132306098937985,
      "epoch": 3.6345847111456915,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037176825849617886,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20231568664312363,
      "num_tokens": 81022829.0,
      "step": 46715
    },
    {
      "entropy": 5.7514279842376705,
      "epoch": 3.634973740517409,
      "grad_norm": 1.34375,
      "learning_rate": 0.00037174323640026173,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22047660648822784,
      "num_tokens": 81031673.0,
      "step": 46720
    },
    {
      "entropy": 5.824792766571045,
      "epoch": 3.6353627698891264,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00037171821283642003,
      "loss": 4.9686,
      "mean_token_accuracy": 0.20837238132953645,
      "num_tokens": 81040368.0,
      "step": 46725
    },
    {
      "entropy": 5.818299913406372,
      "epoch": 3.635751799260844,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003716931878050334,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21674778908491135,
      "num_tokens": 81048436.0,
      "step": 46730
    },
    {
      "entropy": 5.747476387023926,
      "epoch": 3.636140828632562,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037166816130648143,
      "loss": 4.9388,
      "mean_token_accuracy": 0.20761585980653763,
      "num_tokens": 81056487.0,
      "step": 46735
    },
    {
      "entropy": 5.80538854598999,
      "epoch": 3.6365298580042795,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037164313334114407,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20182126611471177,
      "num_tokens": 81064723.0,
      "step": 46740
    },
    {
      "entropy": 5.747455835342407,
      "epoch": 3.636918887375997,
      "grad_norm": 1.375,
      "learning_rate": 0.000371618103909401,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22290216833353044,
      "num_tokens": 81072937.0,
      "step": 46745
    },
    {
      "entropy": 5.766287040710449,
      "epoch": 3.6373079167477145,
      "grad_norm": 1.40625,
      "learning_rate": 0.000371593073011632,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22555731534957885,
      "num_tokens": 81081141.0,
      "step": 46750
    },
    {
      "entropy": 5.765855979919434,
      "epoch": 3.637696946119432,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037156804064821674,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22879180908203126,
      "num_tokens": 81089605.0,
      "step": 46755
    },
    {
      "entropy": 5.799998474121094,
      "epoch": 3.6380859754911494,
      "grad_norm": 1.375,
      "learning_rate": 0.00037154300681953523,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22236431390047073,
      "num_tokens": 81098071.0,
      "step": 46760
    },
    {
      "entropy": 5.89683575630188,
      "epoch": 3.638475004862867,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003715179715259673,
      "loss": 5.0669,
      "mean_token_accuracy": 0.21241187006235124,
      "num_tokens": 81107309.0,
      "step": 46765
    },
    {
      "entropy": 5.848139715194702,
      "epoch": 3.638864034234585,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037149293476789277,
      "loss": 4.9127,
      "mean_token_accuracy": 0.20824284553527833,
      "num_tokens": 81116435.0,
      "step": 46770
    },
    {
      "entropy": 5.858426666259765,
      "epoch": 3.6392530636063025,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003714678965456915,
      "loss": 4.954,
      "mean_token_accuracy": 0.21721839010715485,
      "num_tokens": 81125046.0,
      "step": 46775
    },
    {
      "entropy": 5.802610874176025,
      "epoch": 3.6396420929780198,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037144285685974346,
      "loss": 4.981,
      "mean_token_accuracy": 0.2169325828552246,
      "num_tokens": 81133625.0,
      "step": 46780
    },
    {
      "entropy": 5.76959638595581,
      "epoch": 3.6400311223497375,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037141781571042854,
      "loss": 4.9199,
      "mean_token_accuracy": 0.21787395626306533,
      "num_tokens": 81142461.0,
      "step": 46785
    },
    {
      "entropy": 5.824368953704834,
      "epoch": 3.6404201517214547,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037139277309812676,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21122116595506668,
      "num_tokens": 81151641.0,
      "step": 46790
    },
    {
      "entropy": 5.821697854995728,
      "epoch": 3.6408091810931724,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003713677290232181,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2204246327280998,
      "num_tokens": 81160238.0,
      "step": 46795
    },
    {
      "entropy": 5.803835105895996,
      "epoch": 3.64119821046489,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037134268348608255,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2115943983197212,
      "num_tokens": 81168597.0,
      "step": 46800
    },
    {
      "entropy": 5.851180076599121,
      "epoch": 3.641587239836608,
      "grad_norm": 1.25,
      "learning_rate": 0.00037131763648710003,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20664757788181304,
      "num_tokens": 81177410.0,
      "step": 46805
    },
    {
      "entropy": 5.836679410934448,
      "epoch": 3.641976269208325,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037129258802665076,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21516994535923004,
      "num_tokens": 81186710.0,
      "step": 46810
    },
    {
      "entropy": 5.818358469009399,
      "epoch": 3.6423652985800428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003712675381051147,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21652372181415558,
      "num_tokens": 81195723.0,
      "step": 46815
    },
    {
      "entropy": 5.786525535583496,
      "epoch": 3.6427543279517605,
      "grad_norm": 1.421875,
      "learning_rate": 0.000371242486722872,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21702964305877687,
      "num_tokens": 81203633.0,
      "step": 46820
    },
    {
      "entropy": 5.816823196411133,
      "epoch": 3.6431433573234777,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003712174338803027,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2111393243074417,
      "num_tokens": 81212829.0,
      "step": 46825
    },
    {
      "entropy": 5.826301336288452,
      "epoch": 3.6435323866951954,
      "grad_norm": 1.40625,
      "learning_rate": 0.000371192379577787,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2041080489754677,
      "num_tokens": 81222283.0,
      "step": 46830
    },
    {
      "entropy": 5.820410823822021,
      "epoch": 3.643921416066913,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003711673238157051,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21179766803979874,
      "num_tokens": 81231202.0,
      "step": 46835
    },
    {
      "entropy": 5.870889377593994,
      "epoch": 3.644310445438631,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037114226659443704,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20778151899576186,
      "num_tokens": 81239888.0,
      "step": 46840
    },
    {
      "entropy": 5.8597077369689945,
      "epoch": 3.644699474810348,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003711172079143632,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21262629181146622,
      "num_tokens": 81248529.0,
      "step": 46845
    },
    {
      "entropy": 5.826645469665527,
      "epoch": 3.6450885041820658,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003710921477758637,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21287637650966645,
      "num_tokens": 81257328.0,
      "step": 46850
    },
    {
      "entropy": 5.791141986846924,
      "epoch": 3.6454775335537835,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003710670861793187,
      "loss": 4.909,
      "mean_token_accuracy": 0.21809132993221284,
      "num_tokens": 81265979.0,
      "step": 46855
    },
    {
      "entropy": 5.766604089736939,
      "epoch": 3.6458665629255007,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003710420231251087,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2081373229622841,
      "num_tokens": 81275465.0,
      "step": 46860
    },
    {
      "entropy": 5.819505834579468,
      "epoch": 3.6462555922972184,
      "grad_norm": 1.25,
      "learning_rate": 0.00037101695861361384,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2100534051656723,
      "num_tokens": 81284368.0,
      "step": 46865
    },
    {
      "entropy": 5.82471718788147,
      "epoch": 3.646644621668936,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037099189264521446,
      "loss": 4.9816,
      "mean_token_accuracy": 0.20753067582845688,
      "num_tokens": 81293213.0,
      "step": 46870
    },
    {
      "entropy": 5.760712003707885,
      "epoch": 3.647033651040654,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003709668252202909,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2113695353269577,
      "num_tokens": 81302597.0,
      "step": 46875
    },
    {
      "entropy": 5.763605070114136,
      "epoch": 3.647422680412371,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003709417563392235,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2164827972650528,
      "num_tokens": 81311275.0,
      "step": 46880
    },
    {
      "entropy": 5.824728488922119,
      "epoch": 3.6478117097840888,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003709166860023926,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20988756716251372,
      "num_tokens": 81320596.0,
      "step": 46885
    },
    {
      "entropy": 5.7455894470214846,
      "epoch": 3.648200739155806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037089161421017874,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2112080305814743,
      "num_tokens": 81329579.0,
      "step": 46890
    },
    {
      "entropy": 5.800625371932983,
      "epoch": 3.6485897685275237,
      "grad_norm": 1.25,
      "learning_rate": 0.00037086654096296217,
      "loss": 4.9105,
      "mean_token_accuracy": 0.21128155738115312,
      "num_tokens": 81338386.0,
      "step": 46895
    },
    {
      "entropy": 5.852547454833984,
      "epoch": 3.6489787978992414,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003708414662611235,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2144833594560623,
      "num_tokens": 81347276.0,
      "step": 46900
    },
    {
      "entropy": 5.84384503364563,
      "epoch": 3.649367827270959,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003708163901050432,
      "loss": 4.9212,
      "mean_token_accuracy": 0.20830274671316146,
      "num_tokens": 81356935.0,
      "step": 46905
    },
    {
      "entropy": 5.791039943695068,
      "epoch": 3.649756856642677,
      "grad_norm": 1.375,
      "learning_rate": 0.00037079131249510155,
      "loss": 4.897,
      "mean_token_accuracy": 0.21775818318128587,
      "num_tokens": 81364974.0,
      "step": 46910
    },
    {
      "entropy": 5.727745056152344,
      "epoch": 3.650145886014394,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003707662334316793,
      "loss": 4.8353,
      "mean_token_accuracy": 0.21688156127929686,
      "num_tokens": 81373747.0,
      "step": 46915
    },
    {
      "entropy": 5.886320686340332,
      "epoch": 3.6505349153861117,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003707411529151568,
      "loss": 4.961,
      "mean_token_accuracy": 0.2144709125161171,
      "num_tokens": 81381914.0,
      "step": 46920
    },
    {
      "entropy": 5.816075563430786,
      "epoch": 3.650923944757829,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003707160709459147,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21262144148349763,
      "num_tokens": 81390750.0,
      "step": 46925
    },
    {
      "entropy": 5.837540006637573,
      "epoch": 3.6513129741295467,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003706909875243336,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20285696983337403,
      "num_tokens": 81400046.0,
      "step": 46930
    },
    {
      "entropy": 5.833046245574951,
      "epoch": 3.6517020035012644,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000370665902650794,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21483127772808075,
      "num_tokens": 81408985.0,
      "step": 46935
    },
    {
      "entropy": 5.792379999160767,
      "epoch": 3.652091032872982,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037064081632567675,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21099597960710526,
      "num_tokens": 81417467.0,
      "step": 46940
    },
    {
      "entropy": 5.812753534317016,
      "epoch": 3.6524800622446993,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037061572854936225,
      "loss": 5.0443,
      "mean_token_accuracy": 0.19999421238899232,
      "num_tokens": 81426054.0,
      "step": 46945
    },
    {
      "entropy": 5.774464750289917,
      "epoch": 3.652869091616417,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003705906393222313,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21604742258787155,
      "num_tokens": 81435476.0,
      "step": 46950
    },
    {
      "entropy": 5.858021879196167,
      "epoch": 3.6532581209881347,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003705655486446645,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21023108810186386,
      "num_tokens": 81444613.0,
      "step": 46955
    },
    {
      "entropy": 5.82696270942688,
      "epoch": 3.653647150359852,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003705404565170426,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21793434619903565,
      "num_tokens": 81452995.0,
      "step": 46960
    },
    {
      "entropy": 5.833832693099976,
      "epoch": 3.6540361797315697,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003705153629397463,
      "loss": 4.9206,
      "mean_token_accuracy": 0.21767951399087906,
      "num_tokens": 81461063.0,
      "step": 46965
    },
    {
      "entropy": 5.823679637908936,
      "epoch": 3.6544252091032874,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037049026791315646,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21506234258413315,
      "num_tokens": 81469835.0,
      "step": 46970
    },
    {
      "entropy": 5.73599648475647,
      "epoch": 3.654814238475005,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003704651714376538,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22173629254102706,
      "num_tokens": 81478372.0,
      "step": 46975
    },
    {
      "entropy": 5.781663465499878,
      "epoch": 3.6552032678467223,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003704400735136191,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2134254664182663,
      "num_tokens": 81486870.0,
      "step": 46980
    },
    {
      "entropy": 5.776082515716553,
      "epoch": 3.65559229721844,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003704149741414332,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21472052186727525,
      "num_tokens": 81495566.0,
      "step": 46985
    },
    {
      "entropy": 5.880384731292724,
      "epoch": 3.6559813265901573,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003703898733214769,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21047959923744203,
      "num_tokens": 81503823.0,
      "step": 46990
    },
    {
      "entropy": 5.842616271972656,
      "epoch": 3.656370355961875,
      "grad_norm": 1.375,
      "learning_rate": 0.0003703647710541311,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2037819281220436,
      "num_tokens": 81511870.0,
      "step": 46995
    },
    {
      "entropy": 5.799547004699707,
      "epoch": 3.6567593853335927,
      "grad_norm": 1.25,
      "learning_rate": 0.0003703396673397768,
      "loss": 4.8962,
      "mean_token_accuracy": 0.20831857919692992,
      "num_tokens": 81519959.0,
      "step": 47000
    },
    {
      "entropy": 5.785780715942383,
      "epoch": 3.6571484147053104,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003703145621787947,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21140255481004716,
      "num_tokens": 81529401.0,
      "step": 47005
    },
    {
      "entropy": 5.830147218704224,
      "epoch": 3.657537444077028,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037028945557156587,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21307274997234343,
      "num_tokens": 81538561.0,
      "step": 47010
    },
    {
      "entropy": 5.813931083679199,
      "epoch": 3.6579264734487453,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037026434751847117,
      "loss": 4.972,
      "mean_token_accuracy": 0.21551993787288665,
      "num_tokens": 81547168.0,
      "step": 47015
    },
    {
      "entropy": 5.776926803588867,
      "epoch": 3.658315502820463,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037023923801989163,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21327387392520905,
      "num_tokens": 81556094.0,
      "step": 47020
    },
    {
      "entropy": 5.869414758682251,
      "epoch": 3.6587045321921803,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003702141270762083,
      "loss": 4.942,
      "mean_token_accuracy": 0.2172440081834793,
      "num_tokens": 81564032.0,
      "step": 47025
    },
    {
      "entropy": 5.7849891662597654,
      "epoch": 3.659093561563898,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037018901468780207,
      "loss": 4.988,
      "mean_token_accuracy": 0.21380701363086702,
      "num_tokens": 81572622.0,
      "step": 47030
    },
    {
      "entropy": 5.7666566371917725,
      "epoch": 3.6594825909356157,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003701639008550541,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2166245996952057,
      "num_tokens": 81581800.0,
      "step": 47035
    },
    {
      "entropy": 5.763534212112427,
      "epoch": 3.6598716203073334,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003701387855783454,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22018159180879593,
      "num_tokens": 81590420.0,
      "step": 47040
    },
    {
      "entropy": 5.83518352508545,
      "epoch": 3.6602606496790506,
      "grad_norm": 1.3125,
      "learning_rate": 0.000370113668858057,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2100249260663986,
      "num_tokens": 81598243.0,
      "step": 47045
    },
    {
      "entropy": 5.817593193054199,
      "epoch": 3.6606496790507683,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003700885506945701,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21202801167964935,
      "num_tokens": 81607061.0,
      "step": 47050
    },
    {
      "entropy": 5.76728572845459,
      "epoch": 3.661038708422486,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003700634310882658,
      "loss": 4.8892,
      "mean_token_accuracy": 0.20724625140428543,
      "num_tokens": 81614899.0,
      "step": 47055
    },
    {
      "entropy": 5.815124988555908,
      "epoch": 3.6614277377942033,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037003831003952525,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21611706912517548,
      "num_tokens": 81623819.0,
      "step": 47060
    },
    {
      "entropy": 5.909911203384399,
      "epoch": 3.661816767165921,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037001318754872957,
      "loss": 5.1141,
      "mean_token_accuracy": 0.19993706941604614,
      "num_tokens": 81632852.0,
      "step": 47065
    },
    {
      "entropy": 5.871212339401245,
      "epoch": 3.6622057965376387,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036998806361626,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20822747200727462,
      "num_tokens": 81642098.0,
      "step": 47070
    },
    {
      "entropy": 5.855953168869019,
      "epoch": 3.6625948259093564,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003699629382424979,
      "loss": 5.0486,
      "mean_token_accuracy": 0.20244466811418532,
      "num_tokens": 81651065.0,
      "step": 47075
    },
    {
      "entropy": 5.774432182312012,
      "epoch": 3.6629838552810736,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036993781142782416,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2162814751267433,
      "num_tokens": 81659503.0,
      "step": 47080
    },
    {
      "entropy": 5.7778722763061525,
      "epoch": 3.6633728846527913,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036991268317262033,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2121458038687706,
      "num_tokens": 81667501.0,
      "step": 47085
    },
    {
      "entropy": 5.766786575317383,
      "epoch": 3.663761914024509,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036988755347726766,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22252424955368041,
      "num_tokens": 81676285.0,
      "step": 47090
    },
    {
      "entropy": 5.813645935058593,
      "epoch": 3.6641509433962263,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036986242234214736,
      "loss": 4.9401,
      "mean_token_accuracy": 0.20978277325630187,
      "num_tokens": 81685449.0,
      "step": 47095
    },
    {
      "entropy": 5.781049203872681,
      "epoch": 3.664539972767944,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036983728976764074,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2149082750082016,
      "num_tokens": 81694615.0,
      "step": 47100
    },
    {
      "entropy": 5.8065492630004885,
      "epoch": 3.6649290021396617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036981215575412923,
      "loss": 4.8901,
      "mean_token_accuracy": 0.21141926050186158,
      "num_tokens": 81703105.0,
      "step": 47105
    },
    {
      "entropy": 5.834767293930054,
      "epoch": 3.6653180315113794,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003697870203019942,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2184969365596771,
      "num_tokens": 81711414.0,
      "step": 47110
    },
    {
      "entropy": 5.806837129592895,
      "epoch": 3.6657070608830966,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036976188341161694,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20865327417850493,
      "num_tokens": 81719196.0,
      "step": 47115
    },
    {
      "entropy": 5.821295118331909,
      "epoch": 3.6660960902548143,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003697367450833789,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2094178780913353,
      "num_tokens": 81727604.0,
      "step": 47120
    },
    {
      "entropy": 5.873674011230468,
      "epoch": 3.6664851196265316,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036971160531766155,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21362105160951614,
      "num_tokens": 81735373.0,
      "step": 47125
    },
    {
      "entropy": 5.828206729888916,
      "epoch": 3.6668741489982493,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003696864641148465,
      "loss": 4.929,
      "mean_token_accuracy": 0.2094999670982361,
      "num_tokens": 81743086.0,
      "step": 47130
    },
    {
      "entropy": 5.873082256317138,
      "epoch": 3.667263178369967,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036966132147531487,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21130361407995224,
      "num_tokens": 81751665.0,
      "step": 47135
    },
    {
      "entropy": 5.807980918884278,
      "epoch": 3.6676522077416847,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036963617739944844,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21728401631116867,
      "num_tokens": 81760056.0,
      "step": 47140
    },
    {
      "entropy": 5.705516386032104,
      "epoch": 3.668041237113402,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036961103188762866,
      "loss": 4.812,
      "mean_token_accuracy": 0.22518517971038818,
      "num_tokens": 81768417.0,
      "step": 47145
    },
    {
      "entropy": 5.801606273651123,
      "epoch": 3.6684302664851196,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036958588494023705,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21837680488824845,
      "num_tokens": 81777543.0,
      "step": 47150
    },
    {
      "entropy": 5.83249683380127,
      "epoch": 3.6688192958568373,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003695607365576551,
      "loss": 4.9514,
      "mean_token_accuracy": 0.20461459904909135,
      "num_tokens": 81786456.0,
      "step": 47155
    },
    {
      "entropy": 5.8173939228057865,
      "epoch": 3.6692083252285546,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003695355867402646,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21113619059324265,
      "num_tokens": 81795093.0,
      "step": 47160
    },
    {
      "entropy": 5.740402030944824,
      "epoch": 3.6695973546002723,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000369510435488447,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2270841419696808,
      "num_tokens": 81803488.0,
      "step": 47165
    },
    {
      "entropy": 5.812555599212646,
      "epoch": 3.66998638397199,
      "grad_norm": 1.234375,
      "learning_rate": 0.000369485282802584,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21281418353319168,
      "num_tokens": 81811955.0,
      "step": 47170
    },
    {
      "entropy": 5.857449960708618,
      "epoch": 3.6703754133437076,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036946012868305716,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2120029702782631,
      "num_tokens": 81820336.0,
      "step": 47175
    },
    {
      "entropy": 5.850074911117554,
      "epoch": 3.670764442715425,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003694349731302483,
      "loss": 4.96,
      "mean_token_accuracy": 0.2119382306933403,
      "num_tokens": 81828614.0,
      "step": 47180
    },
    {
      "entropy": 5.759307909011841,
      "epoch": 3.6711534720871426,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003694098161445389,
      "loss": 4.9088,
      "mean_token_accuracy": 0.20689031034708022,
      "num_tokens": 81838092.0,
      "step": 47185
    },
    {
      "entropy": 5.825156879425049,
      "epoch": 3.6715425014588603,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003693846577263109,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21866116523742676,
      "num_tokens": 81846894.0,
      "step": 47190
    },
    {
      "entropy": 5.845710325241089,
      "epoch": 3.6719315308305776,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003693594978759459,
      "loss": 4.8842,
      "mean_token_accuracy": 0.21923942267894744,
      "num_tokens": 81855762.0,
      "step": 47195
    },
    {
      "entropy": 5.898674821853637,
      "epoch": 3.6723205602022952,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036933433659382575,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20316316187381744,
      "num_tokens": 81864977.0,
      "step": 47200
    },
    {
      "entropy": 5.804606676101685,
      "epoch": 3.672709589574013,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003693091738803322,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21092488169670104,
      "num_tokens": 81873892.0,
      "step": 47205
    },
    {
      "entropy": 5.8472809314727785,
      "epoch": 3.6730986189457306,
      "grad_norm": 1.4375,
      "learning_rate": 0.000369284009735847,
      "loss": 4.9964,
      "mean_token_accuracy": 0.20889775604009628,
      "num_tokens": 81882146.0,
      "step": 47210
    },
    {
      "entropy": 5.799028730392456,
      "epoch": 3.673487648317448,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003692588441607521,
      "loss": 4.932,
      "mean_token_accuracy": 0.21479109823703765,
      "num_tokens": 81891342.0,
      "step": 47215
    },
    {
      "entropy": 5.8101013660430905,
      "epoch": 3.6738766776891656,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036923367715542923,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21038652062416077,
      "num_tokens": 81900132.0,
      "step": 47220
    },
    {
      "entropy": 5.762343025207519,
      "epoch": 3.674265707060883,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036920850872026025,
      "loss": 4.907,
      "mean_token_accuracy": 0.2103940427303314,
      "num_tokens": 81908014.0,
      "step": 47225
    },
    {
      "entropy": 5.811326885223389,
      "epoch": 3.6746547364326005,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003691833388556272,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20806120336055756,
      "num_tokens": 81916352.0,
      "step": 47230
    },
    {
      "entropy": 5.786103916168213,
      "epoch": 3.6750437658043182,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003691581675619118,
      "loss": 4.9174,
      "mean_token_accuracy": 0.20671167820692063,
      "num_tokens": 81925071.0,
      "step": 47235
    },
    {
      "entropy": 5.8208764553070065,
      "epoch": 3.675432795176036,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003691329948394961,
      "loss": 5.016,
      "mean_token_accuracy": 0.20420427322387696,
      "num_tokens": 81933734.0,
      "step": 47240
    },
    {
      "entropy": 5.795684909820556,
      "epoch": 3.675821824547753,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00036910782068876204,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2142811670899391,
      "num_tokens": 81942369.0,
      "step": 47245
    },
    {
      "entropy": 5.731361722946167,
      "epoch": 3.676210853919471,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036908264511009155,
      "loss": 4.8222,
      "mean_token_accuracy": 0.21289105713367462,
      "num_tokens": 81951128.0,
      "step": 47250
    },
    {
      "entropy": 5.756579828262329,
      "epoch": 3.6765998832911886,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003690574681038667,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21294140070676804,
      "num_tokens": 81959643.0,
      "step": 47255
    },
    {
      "entropy": 5.845918989181518,
      "epoch": 3.676988912662906,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036903228967046946,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21151593774557115,
      "num_tokens": 81968383.0,
      "step": 47260
    },
    {
      "entropy": 5.917092800140381,
      "epoch": 3.6773779420346235,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003690071098102819,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21381414532661439,
      "num_tokens": 81976917.0,
      "step": 47265
    },
    {
      "entropy": 5.841108942031861,
      "epoch": 3.6777669714063412,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003689819285236861,
      "loss": 4.955,
      "mean_token_accuracy": 0.21099216789007186,
      "num_tokens": 81985437.0,
      "step": 47270
    },
    {
      "entropy": 5.793349123001098,
      "epoch": 3.678156000778059,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036895674581106415,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20784006416797637,
      "num_tokens": 81994024.0,
      "step": 47275
    },
    {
      "entropy": 5.773847198486328,
      "epoch": 3.678545030149776,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036893156167279803,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21197226643562317,
      "num_tokens": 82002394.0,
      "step": 47280
    },
    {
      "entropy": 5.862196779251098,
      "epoch": 3.678934059521494,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003689063761092701,
      "loss": 4.9371,
      "mean_token_accuracy": 0.2083805188536644,
      "num_tokens": 82010986.0,
      "step": 47285
    },
    {
      "entropy": 5.770857858657837,
      "epoch": 3.6793230888932116,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036888118912086225,
      "loss": 4.928,
      "mean_token_accuracy": 0.2211802124977112,
      "num_tokens": 82019272.0,
      "step": 47290
    },
    {
      "entropy": 5.834083986282349,
      "epoch": 3.679712118264929,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003688560007079569,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20346547216176986,
      "num_tokens": 82028293.0,
      "step": 47295
    },
    {
      "entropy": 5.748589372634887,
      "epoch": 3.6801011476366465,
      "grad_norm": 1.21875,
      "learning_rate": 0.000368830810870936,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21349053531885148,
      "num_tokens": 82037787.0,
      "step": 47300
    },
    {
      "entropy": 5.785828018188477,
      "epoch": 3.6804901770083642,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000368805619610182,
      "loss": 4.8319,
      "mean_token_accuracy": 0.2179153397679329,
      "num_tokens": 82046511.0,
      "step": 47305
    },
    {
      "entropy": 5.835928916931152,
      "epoch": 3.680879206380082,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036878042692607693,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21626491844654083,
      "num_tokens": 82055472.0,
      "step": 47310
    },
    {
      "entropy": 5.75044846534729,
      "epoch": 3.681268235751799,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036875523281900324,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21525147557258606,
      "num_tokens": 82063597.0,
      "step": 47315
    },
    {
      "entropy": 5.765591287612915,
      "epoch": 3.681657265123517,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003687300372893431,
      "loss": 4.8104,
      "mean_token_accuracy": 0.21649996042251587,
      "num_tokens": 82071459.0,
      "step": 47320
    },
    {
      "entropy": 5.771722793579102,
      "epoch": 3.682046294495234,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036870484033747883,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2039597749710083,
      "num_tokens": 82080657.0,
      "step": 47325
    },
    {
      "entropy": 5.8430564403533936,
      "epoch": 3.682435323866952,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036867964196379274,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2116810441017151,
      "num_tokens": 82090460.0,
      "step": 47330
    },
    {
      "entropy": 5.888584947586059,
      "epoch": 3.6828243532386695,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003686544421686672,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2205468013882637,
      "num_tokens": 82098463.0,
      "step": 47335
    },
    {
      "entropy": 5.849266242980957,
      "epoch": 3.683213382610387,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036862924095248457,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19971213936805726,
      "num_tokens": 82106857.0,
      "step": 47340
    },
    {
      "entropy": 5.753557538986206,
      "epoch": 3.683602411982105,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036860403831562717,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2093684807419777,
      "num_tokens": 82115849.0,
      "step": 47345
    },
    {
      "entropy": 5.775167560577392,
      "epoch": 3.683991441353822,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003685788342584775,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2207179382443428,
      "num_tokens": 82124280.0,
      "step": 47350
    },
    {
      "entropy": 5.744949293136597,
      "epoch": 3.68438047072554,
      "grad_norm": 1.265625,
      "learning_rate": 0.000368553628781418,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21289092749357225,
      "num_tokens": 82133440.0,
      "step": 47355
    },
    {
      "entropy": 5.850086498260498,
      "epoch": 3.684769500097257,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003685284218848311,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2131542980670929,
      "num_tokens": 82142307.0,
      "step": 47360
    },
    {
      "entropy": 5.833224105834961,
      "epoch": 3.685158529468975,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036850321356909926,
      "loss": 5.049,
      "mean_token_accuracy": 0.19619062095880507,
      "num_tokens": 82151359.0,
      "step": 47365
    },
    {
      "entropy": 5.792849206924439,
      "epoch": 3.6855475588406925,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036847800383460493,
      "loss": 5.0205,
      "mean_token_accuracy": 0.20540768951177596,
      "num_tokens": 82160337.0,
      "step": 47370
    },
    {
      "entropy": 5.782108354568481,
      "epoch": 3.68593658821241,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036845279268173076,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21384851336479188,
      "num_tokens": 82168721.0,
      "step": 47375
    },
    {
      "entropy": 5.828218221664429,
      "epoch": 3.6863256175841275,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036842758011085914,
      "loss": 4.955,
      "mean_token_accuracy": 0.21160003244876863,
      "num_tokens": 82177738.0,
      "step": 47380
    },
    {
      "entropy": 5.886498498916626,
      "epoch": 3.686714646955845,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003684023661223727,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21420596688985824,
      "num_tokens": 82186726.0,
      "step": 47385
    },
    {
      "entropy": 5.803761529922485,
      "epoch": 3.687103676327563,
      "grad_norm": 1.421875,
      "learning_rate": 0.00036837715071665403,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21785073280334472,
      "num_tokens": 82194742.0,
      "step": 47390
    },
    {
      "entropy": 5.793007230758667,
      "epoch": 3.68749270569928,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003683519338940857,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21439505666494368,
      "num_tokens": 82203513.0,
      "step": 47395
    },
    {
      "entropy": 5.7156675338745115,
      "epoch": 3.687881735070998,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036832671565505045,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21286508589982986,
      "num_tokens": 82211963.0,
      "step": 47400
    },
    {
      "entropy": 5.8609624862670895,
      "epoch": 3.6882707644427155,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003683014959999307,
      "loss": 4.9699,
      "mean_token_accuracy": 0.20385728627443314,
      "num_tokens": 82220399.0,
      "step": 47405
    },
    {
      "entropy": 5.782517290115356,
      "epoch": 3.688659793814433,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003682762749291094,
      "loss": 4.8805,
      "mean_token_accuracy": 0.216788911819458,
      "num_tokens": 82228393.0,
      "step": 47410
    },
    {
      "entropy": 5.847663021087646,
      "epoch": 3.6890488231861505,
      "grad_norm": 1.328125,
      "learning_rate": 0.000368251052442969,
      "loss": 4.999,
      "mean_token_accuracy": 0.20513266026973725,
      "num_tokens": 82237289.0,
      "step": 47415
    },
    {
      "entropy": 5.834073400497436,
      "epoch": 3.689437852557868,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036822582854189234,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21279284507036209,
      "num_tokens": 82246744.0,
      "step": 47420
    },
    {
      "entropy": 5.811693382263184,
      "epoch": 3.6898268819295854,
      "grad_norm": 1.359375,
      "learning_rate": 0.000368200603226262,
      "loss": 4.8894,
      "mean_token_accuracy": 0.20954977124929428,
      "num_tokens": 82255467.0,
      "step": 47425
    },
    {
      "entropy": 5.790075016021729,
      "epoch": 3.690215911301303,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000368175376496461,
      "loss": 4.932,
      "mean_token_accuracy": 0.2191600665450096,
      "num_tokens": 82264891.0,
      "step": 47430
    },
    {
      "entropy": 5.76919116973877,
      "epoch": 3.690604940673021,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003681501483528719,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22253467440605162,
      "num_tokens": 82274490.0,
      "step": 47435
    },
    {
      "entropy": 5.80750150680542,
      "epoch": 3.6909939700447385,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003681249187958776,
      "loss": 4.9158,
      "mean_token_accuracy": 0.21334189623594285,
      "num_tokens": 82282934.0,
      "step": 47440
    },
    {
      "entropy": 5.707493591308594,
      "epoch": 3.691382999416456,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036809968782586087,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22531289458274842,
      "num_tokens": 82291443.0,
      "step": 47445
    },
    {
      "entropy": 5.782946062088013,
      "epoch": 3.6917720287881735,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003680744554432045,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21603130251169206,
      "num_tokens": 82299366.0,
      "step": 47450
    },
    {
      "entropy": 5.80762128829956,
      "epoch": 3.692161058159891,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036804922164829153,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21761046946048737,
      "num_tokens": 82308000.0,
      "step": 47455
    },
    {
      "entropy": 5.759347629547119,
      "epoch": 3.6925500875316084,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003680239864415046,
      "loss": 4.9503,
      "mean_token_accuracy": 0.20667369961738585,
      "num_tokens": 82317417.0,
      "step": 47460
    },
    {
      "entropy": 5.742478942871093,
      "epoch": 3.692939116903326,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003679987498232268,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21475998908281327,
      "num_tokens": 82325640.0,
      "step": 47465
    },
    {
      "entropy": 5.841833257675171,
      "epoch": 3.693328146275044,
      "grad_norm": 1.359375,
      "learning_rate": 0.000367973511793841,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2009960815310478,
      "num_tokens": 82334586.0,
      "step": 47470
    },
    {
      "entropy": 5.800191926956177,
      "epoch": 3.6937171756467615,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036794827235373006,
      "loss": 4.863,
      "mean_token_accuracy": 0.21664305478334428,
      "num_tokens": 82343094.0,
      "step": 47475
    },
    {
      "entropy": 5.844036674499511,
      "epoch": 3.6941062050184788,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003679230315032772,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20197262167930602,
      "num_tokens": 82352116.0,
      "step": 47480
    },
    {
      "entropy": 5.831503915786743,
      "epoch": 3.6944952343901964,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003678977892428651,
      "loss": 4.9764,
      "mean_token_accuracy": 0.210683174431324,
      "num_tokens": 82361075.0,
      "step": 47485
    },
    {
      "entropy": 5.733933305740356,
      "epoch": 3.694884263761914,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003678725455728769,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21111406087875367,
      "num_tokens": 82369655.0,
      "step": 47490
    },
    {
      "entropy": 5.8596940517425535,
      "epoch": 3.6952732931336314,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036784730049369577,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20121372491121292,
      "num_tokens": 82378790.0,
      "step": 47495
    },
    {
      "entropy": 5.8146710872650145,
      "epoch": 3.695662322505349,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003678220540057045,
      "loss": 4.9387,
      "mean_token_accuracy": 0.21261062622070312,
      "num_tokens": 82387876.0,
      "step": 47500
    },
    {
      "entropy": 5.816575050354004,
      "epoch": 3.696051351877067,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00036779680610928635,
      "loss": 4.9387,
      "mean_token_accuracy": 0.21852627992630005,
      "num_tokens": 82395701.0,
      "step": 47505
    },
    {
      "entropy": 5.89095139503479,
      "epoch": 3.6964403812487845,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003677715568048244,
      "loss": 5.0663,
      "mean_token_accuracy": 0.20266327857971192,
      "num_tokens": 82404664.0,
      "step": 47510
    },
    {
      "entropy": 5.822957277297974,
      "epoch": 3.6968294106205017,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036774630609270165,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21211827546358109,
      "num_tokens": 82413854.0,
      "step": 47515
    },
    {
      "entropy": 5.882249927520752,
      "epoch": 3.6972184399922194,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00036772105397330147,
      "loss": 5.0331,
      "mean_token_accuracy": 0.21250219643115997,
      "num_tokens": 82422119.0,
      "step": 47520
    },
    {
      "entropy": 5.836722946166992,
      "epoch": 3.697607469363937,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003676958004470067,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2110415741801262,
      "num_tokens": 82429877.0,
      "step": 47525
    },
    {
      "entropy": 5.779804229736328,
      "epoch": 3.6979964987356544,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003676705455142009,
      "loss": 4.9495,
      "mean_token_accuracy": 0.20790804773569108,
      "num_tokens": 82438902.0,
      "step": 47530
    },
    {
      "entropy": 5.830459880828857,
      "epoch": 3.698385528107372,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003676452891752669,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2086479678750038,
      "num_tokens": 82447786.0,
      "step": 47535
    },
    {
      "entropy": 5.844129133224487,
      "epoch": 3.69877455747909,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003676200314305882,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20550404489040375,
      "num_tokens": 82457349.0,
      "step": 47540
    },
    {
      "entropy": 5.881977796554565,
      "epoch": 3.6991635868508075,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003675947722805479,
      "loss": 4.9695,
      "mean_token_accuracy": 0.20824549943208695,
      "num_tokens": 82465823.0,
      "step": 47545
    },
    {
      "entropy": 5.787006473541259,
      "epoch": 3.6995526162225247,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003675695117255293,
      "loss": 4.9121,
      "mean_token_accuracy": 0.20989816784858703,
      "num_tokens": 82475567.0,
      "step": 47550
    },
    {
      "entropy": 5.793598747253418,
      "epoch": 3.6999416455942424,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003675442497659157,
      "loss": 4.8816,
      "mean_token_accuracy": 0.218390853703022,
      "num_tokens": 82483938.0,
      "step": 47555
    },
    {
      "entropy": 5.777826404571533,
      "epoch": 3.7003306749659597,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003675189864020905,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21615647971630098,
      "num_tokens": 82491962.0,
      "step": 47560
    },
    {
      "entropy": 5.808918380737305,
      "epoch": 3.7007197043376774,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003674937216344368,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21517158299684525,
      "num_tokens": 82500856.0,
      "step": 47565
    },
    {
      "entropy": 5.895078659057617,
      "epoch": 3.701108733709395,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003674684554633382,
      "loss": 5.0015,
      "mean_token_accuracy": 0.20611666589975358,
      "num_tokens": 82509305.0,
      "step": 47570
    },
    {
      "entropy": 5.856377649307251,
      "epoch": 3.701497763081113,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003674431878891779,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2118248850107193,
      "num_tokens": 82517872.0,
      "step": 47575
    },
    {
      "entropy": 5.838798856735229,
      "epoch": 3.70188679245283,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003674179189123393,
      "loss": 4.9676,
      "mean_token_accuracy": 0.20874841511249542,
      "num_tokens": 82526172.0,
      "step": 47580
    },
    {
      "entropy": 5.816642808914184,
      "epoch": 3.7022758218245477,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036739264853320606,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2117399051785469,
      "num_tokens": 82535139.0,
      "step": 47585
    },
    {
      "entropy": 5.848377370834351,
      "epoch": 3.7026648511962654,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036736737675216127,
      "loss": 5.0326,
      "mean_token_accuracy": 0.20246654748916626,
      "num_tokens": 82545229.0,
      "step": 47590
    },
    {
      "entropy": 5.9037832736969,
      "epoch": 3.7030538805679827,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003673421035695887,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2042330399155617,
      "num_tokens": 82554516.0,
      "step": 47595
    },
    {
      "entropy": 5.791474485397339,
      "epoch": 3.7034429099397004,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003673168289858716,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22097381055355073,
      "num_tokens": 82562652.0,
      "step": 47600
    },
    {
      "entropy": 5.77741961479187,
      "epoch": 3.703831939311418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003672915530013936,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2128673940896988,
      "num_tokens": 82570803.0,
      "step": 47605
    },
    {
      "entropy": 5.83367657661438,
      "epoch": 3.7042209686831358,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036726627561653815,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2143327221274376,
      "num_tokens": 82578882.0,
      "step": 47610
    },
    {
      "entropy": 5.9022942066192625,
      "epoch": 3.704609998054853,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003672409968316888,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2032284140586853,
      "num_tokens": 82587371.0,
      "step": 47615
    },
    {
      "entropy": 5.92920618057251,
      "epoch": 3.7049990274265707,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036721571664722914,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21020781397819518,
      "num_tokens": 82595889.0,
      "step": 47620
    },
    {
      "entropy": 5.845452213287354,
      "epoch": 3.7053880567982884,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003671904350635428,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21320783346891403,
      "num_tokens": 82604921.0,
      "step": 47625
    },
    {
      "entropy": 5.785266256332397,
      "epoch": 3.7057770861700057,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036716515208101334,
      "loss": 4.9278,
      "mean_token_accuracy": 0.21173641830682755,
      "num_tokens": 82613393.0,
      "step": 47630
    },
    {
      "entropy": 5.801237726211548,
      "epoch": 3.7061661155417234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003671398677000244,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22240647822618484,
      "num_tokens": 82623122.0,
      "step": 47635
    },
    {
      "entropy": 5.798791074752808,
      "epoch": 3.706555144913441,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003671145819209596,
      "loss": 4.8841,
      "mean_token_accuracy": 0.21643069088459016,
      "num_tokens": 82631747.0,
      "step": 47640
    },
    {
      "entropy": 5.867663717269897,
      "epoch": 3.7069441742851588,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003670892947442026,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21093566566705704,
      "num_tokens": 82640690.0,
      "step": 47645
    },
    {
      "entropy": 5.804168462753296,
      "epoch": 3.707333203656876,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003670640061701371,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22058282047510147,
      "num_tokens": 82649032.0,
      "step": 47650
    },
    {
      "entropy": 5.839563941955566,
      "epoch": 3.7077222330285937,
      "grad_norm": 1.375,
      "learning_rate": 0.0003670387161991469,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21046100109815596,
      "num_tokens": 82657970.0,
      "step": 47655
    },
    {
      "entropy": 5.842561721801758,
      "epoch": 3.708111262400311,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003670134248316157,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21391931623220445,
      "num_tokens": 82666534.0,
      "step": 47660
    },
    {
      "entropy": 5.868252182006836,
      "epoch": 3.7085002917720287,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036698813206792724,
      "loss": 5.009,
      "mean_token_accuracy": 0.21215423196554184,
      "num_tokens": 82675015.0,
      "step": 47665
    },
    {
      "entropy": 5.903868913650513,
      "epoch": 3.7088893211437464,
      "grad_norm": 1.421875,
      "learning_rate": 0.00036696283790846524,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20647107809782028,
      "num_tokens": 82683431.0,
      "step": 47670
    },
    {
      "entropy": 5.8910339832305905,
      "epoch": 3.709278350515464,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036693754235361364,
      "loss": 5.0455,
      "mean_token_accuracy": 0.20200192034244538,
      "num_tokens": 82691466.0,
      "step": 47675
    },
    {
      "entropy": 5.858867597579956,
      "epoch": 3.7096673798871818,
      "grad_norm": 1.375,
      "learning_rate": 0.0003669122454037561,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21845878064632415,
      "num_tokens": 82700071.0,
      "step": 47680
    },
    {
      "entropy": 5.823124694824219,
      "epoch": 3.710056409258899,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003668869470592765,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2068781226873398,
      "num_tokens": 82709164.0,
      "step": 47685
    },
    {
      "entropy": 5.826355123519898,
      "epoch": 3.7104454386306167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036686164732055876,
      "loss": 4.9958,
      "mean_token_accuracy": 0.20407401770353317,
      "num_tokens": 82718119.0,
      "step": 47690
    },
    {
      "entropy": 5.833717632293701,
      "epoch": 3.710834468002334,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036683634618798675,
      "loss": 4.9212,
      "mean_token_accuracy": 0.21623031198978424,
      "num_tokens": 82725932.0,
      "step": 47695
    },
    {
      "entropy": 5.784215021133423,
      "epoch": 3.7112234973740517,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003668110436619443,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21834779530763626,
      "num_tokens": 82734183.0,
      "step": 47700
    },
    {
      "entropy": 5.788884782791138,
      "epoch": 3.7116125267457694,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036678573974281546,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21046918481588364,
      "num_tokens": 82742434.0,
      "step": 47705
    },
    {
      "entropy": 5.865456199645996,
      "epoch": 3.712001556117487,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003667604344309841,
      "loss": 5.01,
      "mean_token_accuracy": 0.20261069238185883,
      "num_tokens": 82751859.0,
      "step": 47710
    },
    {
      "entropy": 5.804358005523682,
      "epoch": 3.7123905854892043,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036673512772683424,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2154978707432747,
      "num_tokens": 82760633.0,
      "step": 47715
    },
    {
      "entropy": 5.848613119125366,
      "epoch": 3.712779614860922,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003667098196307498,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21573690921068192,
      "num_tokens": 82769408.0,
      "step": 47720
    },
    {
      "entropy": 5.737537002563476,
      "epoch": 3.7131686442326397,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00036668451014311475,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22020492553710938,
      "num_tokens": 82778565.0,
      "step": 47725
    },
    {
      "entropy": 5.771137428283692,
      "epoch": 3.713557673604357,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036665919926431325,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22024179995059967,
      "num_tokens": 82787322.0,
      "step": 47730
    },
    {
      "entropy": 5.8310027599334715,
      "epoch": 3.7139467029760747,
      "grad_norm": 1.5,
      "learning_rate": 0.0003666338869947293,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21158193051815033,
      "num_tokens": 82794763.0,
      "step": 47735
    },
    {
      "entropy": 5.8416440963745115,
      "epoch": 3.7143357323477924,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003666085733347469,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21041932702064514,
      "num_tokens": 82803651.0,
      "step": 47740
    },
    {
      "entropy": 5.915333652496338,
      "epoch": 3.71472476171951,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036658325828475033,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20295648127794266,
      "num_tokens": 82812064.0,
      "step": 47745
    },
    {
      "entropy": 5.844149208068847,
      "epoch": 3.7151137910912273,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003665579418451235,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21924808174371718,
      "num_tokens": 82821157.0,
      "step": 47750
    },
    {
      "entropy": 5.813853740692139,
      "epoch": 3.715502820462945,
      "grad_norm": 1.375,
      "learning_rate": 0.0003665326240162506,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20995181053876877,
      "num_tokens": 82829875.0,
      "step": 47755
    },
    {
      "entropy": 5.827734899520874,
      "epoch": 3.7158918498346623,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036650730479851593,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22218545377254487,
      "num_tokens": 82838371.0,
      "step": 47760
    },
    {
      "entropy": 5.842835426330566,
      "epoch": 3.71628087920638,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036648198419230344,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20970909148454667,
      "num_tokens": 82846823.0,
      "step": 47765
    },
    {
      "entropy": 5.772879886627197,
      "epoch": 3.7166699085780976,
      "grad_norm": 1.375,
      "learning_rate": 0.00036645666219799753,
      "loss": 4.9312,
      "mean_token_accuracy": 0.21171149164438247,
      "num_tokens": 82855548.0,
      "step": 47770
    },
    {
      "entropy": 5.868474054336548,
      "epoch": 3.7170589379498153,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003664313388159823,
      "loss": 4.928,
      "mean_token_accuracy": 0.21526189148426056,
      "num_tokens": 82864313.0,
      "step": 47775
    },
    {
      "entropy": 5.806881332397461,
      "epoch": 3.717447967321533,
      "grad_norm": 1.3125,
      "learning_rate": 0.000366406014046642,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2130802184343338,
      "num_tokens": 82872603.0,
      "step": 47780
    },
    {
      "entropy": 5.842578983306884,
      "epoch": 3.7178369966932503,
      "grad_norm": 1.609375,
      "learning_rate": 0.000366380687890361,
      "loss": 5.0167,
      "mean_token_accuracy": 0.2084202989935875,
      "num_tokens": 82880962.0,
      "step": 47785
    },
    {
      "entropy": 5.831890678405761,
      "epoch": 3.718226026064968,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003663553603475235,
      "loss": 4.9614,
      "mean_token_accuracy": 0.22131507396697997,
      "num_tokens": 82889876.0,
      "step": 47790
    },
    {
      "entropy": 5.786353397369385,
      "epoch": 3.7186150554366852,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036633003141851373,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21048336923122407,
      "num_tokens": 82899083.0,
      "step": 47795
    },
    {
      "entropy": 5.859846448898315,
      "epoch": 3.719004084808403,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003663047011037162,
      "loss": 4.935,
      "mean_token_accuracy": 0.2125890538096428,
      "num_tokens": 82907207.0,
      "step": 47800
    },
    {
      "entropy": 5.820394468307495,
      "epoch": 3.7193931141801206,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003662793694035151,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21446046531200408,
      "num_tokens": 82915979.0,
      "step": 47805
    },
    {
      "entropy": 5.851028919219971,
      "epoch": 3.7197821435518383,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003662540363182949,
      "loss": 4.9828,
      "mean_token_accuracy": 0.20357959121465682,
      "num_tokens": 82924668.0,
      "step": 47810
    },
    {
      "entropy": 5.770662593841553,
      "epoch": 3.7201711729235556,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003662287018484398,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21677715182304383,
      "num_tokens": 82933065.0,
      "step": 47815
    },
    {
      "entropy": 5.770005512237549,
      "epoch": 3.7205602022952733,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003662033659943345,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2144200697541237,
      "num_tokens": 82942270.0,
      "step": 47820
    },
    {
      "entropy": 5.790526533126831,
      "epoch": 3.720949231666991,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036617802875636333,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21514707505702974,
      "num_tokens": 82950774.0,
      "step": 47825
    },
    {
      "entropy": 5.812875175476075,
      "epoch": 3.7213382610387082,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003661526901349106,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22173648774623872,
      "num_tokens": 82959444.0,
      "step": 47830
    },
    {
      "entropy": 5.82558650970459,
      "epoch": 3.721727290410426,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036612735013036085,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2146499365568161,
      "num_tokens": 82968217.0,
      "step": 47835
    },
    {
      "entropy": 5.801222276687622,
      "epoch": 3.7221163197821436,
      "grad_norm": 1.25,
      "learning_rate": 0.0003661020087430987,
      "loss": 4.9359,
      "mean_token_accuracy": 0.213276706635952,
      "num_tokens": 82977320.0,
      "step": 47840
    },
    {
      "entropy": 5.894186878204346,
      "epoch": 3.7225053491538613,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003660766659735086,
      "loss": 5.0599,
      "mean_token_accuracy": 0.2083798199892044,
      "num_tokens": 82986517.0,
      "step": 47845
    },
    {
      "entropy": 5.939667654037476,
      "epoch": 3.7228943785255786,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003660513218219749,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22739367485046386,
      "num_tokens": 82995419.0,
      "step": 47850
    },
    {
      "entropy": 5.814560270309448,
      "epoch": 3.7232834078972963,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003660259762888824,
      "loss": 4.954,
      "mean_token_accuracy": 0.21166229099035264,
      "num_tokens": 83003459.0,
      "step": 47855
    },
    {
      "entropy": 5.811594676971436,
      "epoch": 3.7236724372690135,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003660006293746156,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21104368716478347,
      "num_tokens": 83011974.0,
      "step": 47860
    },
    {
      "entropy": 5.858108282089233,
      "epoch": 3.7240614666407312,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036597528107955914,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21286963522434235,
      "num_tokens": 83021124.0,
      "step": 47865
    },
    {
      "entropy": 5.829787969589233,
      "epoch": 3.724450496012449,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036594993140409753,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2185853362083435,
      "num_tokens": 83029667.0,
      "step": 47870
    },
    {
      "entropy": 5.831813716888428,
      "epoch": 3.7248395253841666,
      "grad_norm": 1.375,
      "learning_rate": 0.0003659245803486155,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22004348486661912,
      "num_tokens": 83038610.0,
      "step": 47875
    },
    {
      "entropy": 5.863837337493896,
      "epoch": 3.7252285547558843,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003658992279134977,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20435242503881454,
      "num_tokens": 83047280.0,
      "step": 47880
    },
    {
      "entropy": 5.8509824752807615,
      "epoch": 3.7256175841276016,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003658738740991287,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21018191277980805,
      "num_tokens": 83057411.0,
      "step": 47885
    },
    {
      "entropy": 5.854897737503052,
      "epoch": 3.7260066134993193,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036584851890589335,
      "loss": 5.0514,
      "mean_token_accuracy": 0.2017844706773758,
      "num_tokens": 83066224.0,
      "step": 47890
    },
    {
      "entropy": 5.742435312271118,
      "epoch": 3.7263956428710365,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036582316233417636,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21097672134637832,
      "num_tokens": 83075261.0,
      "step": 47895
    },
    {
      "entropy": 5.8045735359191895,
      "epoch": 3.7267846722427542,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036579780438436246,
      "loss": 4.9345,
      "mean_token_accuracy": 0.20901874601840972,
      "num_tokens": 83083796.0,
      "step": 47900
    },
    {
      "entropy": 5.80733003616333,
      "epoch": 3.727173701614472,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003657724450568363,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2082891583442688,
      "num_tokens": 83093349.0,
      "step": 47905
    },
    {
      "entropy": 5.862991809844971,
      "epoch": 3.7275627309861896,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003657470843519829,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2059582069516182,
      "num_tokens": 83102377.0,
      "step": 47910
    },
    {
      "entropy": 5.741253471374511,
      "epoch": 3.727951760357907,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003657217222701868,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22305884659290315,
      "num_tokens": 83110720.0,
      "step": 47915
    },
    {
      "entropy": 5.665694618225098,
      "epoch": 3.7283407897296246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000365696358811833,
      "loss": 4.814,
      "mean_token_accuracy": 0.21523180603981018,
      "num_tokens": 83120765.0,
      "step": 47920
    },
    {
      "entropy": 5.699869918823242,
      "epoch": 3.7287298191013423,
      "grad_norm": 1.25,
      "learning_rate": 0.00036567099397730635,
      "loss": 4.8336,
      "mean_token_accuracy": 0.21737933903932571,
      "num_tokens": 83129100.0,
      "step": 47925
    },
    {
      "entropy": 5.762078189849854,
      "epoch": 3.7291188484730595,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003656456277669917,
      "loss": 4.884,
      "mean_token_accuracy": 0.2194640800356865,
      "num_tokens": 83137918.0,
      "step": 47930
    },
    {
      "entropy": 5.898148870468139,
      "epoch": 3.729507877844777,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036562026018127386,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21034625321626663,
      "num_tokens": 83147227.0,
      "step": 47935
    },
    {
      "entropy": 5.8558440685272215,
      "epoch": 3.729896907216495,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036559489122053786,
      "loss": 4.91,
      "mean_token_accuracy": 0.2174339398741722,
      "num_tokens": 83156019.0,
      "step": 47940
    },
    {
      "entropy": 5.7728844165802,
      "epoch": 3.7302859365882126,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003655695208851686,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22914837449789047,
      "num_tokens": 83164510.0,
      "step": 47945
    },
    {
      "entropy": 5.855320596694947,
      "epoch": 3.73067496595993,
      "grad_norm": 1.484375,
      "learning_rate": 0.00036554414917555096,
      "loss": 4.9943,
      "mean_token_accuracy": 0.20583319962024688,
      "num_tokens": 83173095.0,
      "step": 47950
    },
    {
      "entropy": 5.822858762741089,
      "epoch": 3.7310639953316476,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036551877609207,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21741992980241776,
      "num_tokens": 83181884.0,
      "step": 47955
    },
    {
      "entropy": 5.847262907028198,
      "epoch": 3.7314530247033653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003654934016351107,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2216199517250061,
      "num_tokens": 83190039.0,
      "step": 47960
    },
    {
      "entropy": 5.8938483715057375,
      "epoch": 3.7318420540750825,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036546802580505805,
      "loss": 5.1191,
      "mean_token_accuracy": 0.19443112462759018,
      "num_tokens": 83198757.0,
      "step": 47965
    },
    {
      "entropy": 5.829267740249634,
      "epoch": 3.7322310834468,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003654426486022971,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20871175527572633,
      "num_tokens": 83208304.0,
      "step": 47970
    },
    {
      "entropy": 5.867332983016968,
      "epoch": 3.732620112818518,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003654172700272129,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2139279291033745,
      "num_tokens": 83216702.0,
      "step": 47975
    },
    {
      "entropy": 5.868944406509399,
      "epoch": 3.7330091421902356,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003653918900801905,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2257342129945755,
      "num_tokens": 83225248.0,
      "step": 47980
    },
    {
      "entropy": 5.809030055999756,
      "epoch": 3.733398171561953,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036536650876161514,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22110605090856553,
      "num_tokens": 83233861.0,
      "step": 47985
    },
    {
      "entropy": 5.834697532653808,
      "epoch": 3.7337872009336706,
      "grad_norm": 1.375,
      "learning_rate": 0.0003653411260718718,
      "loss": 4.9898,
      "mean_token_accuracy": 0.20974529832601546,
      "num_tokens": 83242950.0,
      "step": 47990
    },
    {
      "entropy": 5.8395123958587645,
      "epoch": 3.734176230305388,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003653157420113457,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20678835809230806,
      "num_tokens": 83250917.0,
      "step": 47995
    },
    {
      "entropy": 5.871608018875122,
      "epoch": 3.7345652596771055,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003652903565804219,
      "loss": 4.924,
      "mean_token_accuracy": 0.21541347056627275,
      "num_tokens": 83259282.0,
      "step": 48000
    },
    {
      "epoch": 3.7345652596771055,
      "eval_entropy": 5.5707257245112425,
      "eval_loss": 5.04781436920166,
      "eval_mean_token_accuracy": 0.21727151325881425,
      "eval_num_tokens": 83259282.0,
      "eval_runtime": 21.9914,
      "eval_samples_per_second": 1889.736,
      "eval_steps_per_second": 236.228,
      "step": 48000
    },
    {
      "entropy": 5.839330005645752,
      "epoch": 3.734954289048823,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036526496977948574,
      "loss": 5.0314,
      "mean_token_accuracy": 0.210779544711113,
      "num_tokens": 83267674.0,
      "step": 48005
    },
    {
      "entropy": 5.798641347885132,
      "epoch": 3.735343318420541,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036523958160892227,
      "loss": 4.9566,
      "mean_token_accuracy": 0.21812812089920045,
      "num_tokens": 83276084.0,
      "step": 48010
    },
    {
      "entropy": 5.771815443038941,
      "epoch": 3.735732347792258,
      "grad_norm": 1.375,
      "learning_rate": 0.00036521419206911684,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21293993890285492,
      "num_tokens": 83284608.0,
      "step": 48015
    },
    {
      "entropy": 5.852262496948242,
      "epoch": 3.736121377163976,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003651888011604545,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20959798544645308,
      "num_tokens": 83293099.0,
      "step": 48020
    },
    {
      "entropy": 5.832505416870117,
      "epoch": 3.7365104065356936,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036516340888332076,
      "loss": 5.0145,
      "mean_token_accuracy": 0.21204451322555543,
      "num_tokens": 83302225.0,
      "step": 48025
    },
    {
      "entropy": 5.821444845199585,
      "epoch": 3.736899435907411,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036513801523810075,
      "loss": 4.9294,
      "mean_token_accuracy": 0.2141224205493927,
      "num_tokens": 83311070.0,
      "step": 48030
    },
    {
      "entropy": 5.803778886795044,
      "epoch": 3.7372884652791285,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003651126202251799,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2212057366967201,
      "num_tokens": 83319391.0,
      "step": 48035
    },
    {
      "entropy": 5.879584169387817,
      "epoch": 3.737677494650846,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003650872238449434,
      "loss": 5.054,
      "mean_token_accuracy": 0.19815609157085418,
      "num_tokens": 83329011.0,
      "step": 48040
    },
    {
      "entropy": 5.830575084686279,
      "epoch": 3.738066524022564,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003650618260977767,
      "loss": 4.9596,
      "mean_token_accuracy": 0.216921828687191,
      "num_tokens": 83336464.0,
      "step": 48045
    },
    {
      "entropy": 5.802900934219361,
      "epoch": 3.738455553394281,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003650364269840651,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21421520560979843,
      "num_tokens": 83344023.0,
      "step": 48050
    },
    {
      "entropy": 5.7968555927276615,
      "epoch": 3.738844582765999,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003650110265041941,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2138715237379074,
      "num_tokens": 83353397.0,
      "step": 48055
    },
    {
      "entropy": 5.8274883270263675,
      "epoch": 3.7392336121377165,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036498562465854897,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22371768653392793,
      "num_tokens": 83361579.0,
      "step": 48060
    },
    {
      "entropy": 5.810103559494019,
      "epoch": 3.739622641509434,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003649602214475152,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21632397323846816,
      "num_tokens": 83370253.0,
      "step": 48065
    },
    {
      "entropy": 5.817929744720459,
      "epoch": 3.7400116708811515,
      "grad_norm": 1.25,
      "learning_rate": 0.00036493481687147836,
      "loss": 4.9604,
      "mean_token_accuracy": 0.20656620413064958,
      "num_tokens": 83379021.0,
      "step": 48070
    },
    {
      "entropy": 5.9248076438903805,
      "epoch": 3.740400700252869,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003649094109308237,
      "loss": 5.0913,
      "mean_token_accuracy": 0.20106624960899352,
      "num_tokens": 83387285.0,
      "step": 48075
    },
    {
      "entropy": 5.784343814849853,
      "epoch": 3.740789729624587,
      "grad_norm": 1.25,
      "learning_rate": 0.00036488400362593695,
      "loss": 4.867,
      "mean_token_accuracy": 0.21599290668964385,
      "num_tokens": 83396541.0,
      "step": 48080
    },
    {
      "entropy": 5.806223726272583,
      "epoch": 3.741178758996304,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036485859495720337,
      "loss": 4.9747,
      "mean_token_accuracy": 0.20569291859865188,
      "num_tokens": 83404766.0,
      "step": 48085
    },
    {
      "entropy": 5.865656423568725,
      "epoch": 3.741567788368022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036483318492500875,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21017386317253112,
      "num_tokens": 83413162.0,
      "step": 48090
    },
    {
      "entropy": 5.798042058944702,
      "epoch": 3.741956817739739,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036480777352973856,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2187887981534004,
      "num_tokens": 83422386.0,
      "step": 48095
    },
    {
      "entropy": 5.7904381275177,
      "epoch": 3.742345847111457,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036478236077177825,
      "loss": 4.871,
      "mean_token_accuracy": 0.21698054820299148,
      "num_tokens": 83431144.0,
      "step": 48100
    },
    {
      "entropy": 5.79972276687622,
      "epoch": 3.7427348764831745,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036475694665151357,
      "loss": 5.003,
      "mean_token_accuracy": 0.2111394315958023,
      "num_tokens": 83440372.0,
      "step": 48105
    },
    {
      "entropy": 5.830613231658935,
      "epoch": 3.743123905854892,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036473153116933007,
      "loss": 4.938,
      "mean_token_accuracy": 0.21442829221487045,
      "num_tokens": 83448753.0,
      "step": 48110
    },
    {
      "entropy": 5.891182279586792,
      "epoch": 3.74351293522661,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036470611432561334,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21135952025651933,
      "num_tokens": 83457620.0,
      "step": 48115
    },
    {
      "entropy": 5.872762441635132,
      "epoch": 3.743901964598327,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003646806961207492,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20905730575323106,
      "num_tokens": 83467260.0,
      "step": 48120
    },
    {
      "entropy": 5.787105512619019,
      "epoch": 3.744290993970045,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036465527655512323,
      "loss": 4.9201,
      "mean_token_accuracy": 0.21314291208982467,
      "num_tokens": 83476372.0,
      "step": 48125
    },
    {
      "entropy": 5.780849504470825,
      "epoch": 3.744680023341762,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003646298556291211,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2075783520936966,
      "num_tokens": 83484561.0,
      "step": 48130
    },
    {
      "entropy": 5.8040581226348875,
      "epoch": 3.74506905271348,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036460443334312867,
      "loss": 4.9086,
      "mean_token_accuracy": 0.216504967212677,
      "num_tokens": 83493476.0,
      "step": 48135
    },
    {
      "entropy": 5.822963523864746,
      "epoch": 3.7454580820851975,
      "grad_norm": 1.375,
      "learning_rate": 0.0003645790096975315,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21128910928964614,
      "num_tokens": 83501964.0,
      "step": 48140
    },
    {
      "entropy": 5.7914581298828125,
      "epoch": 3.745847111456915,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036455358469271546,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20687039345502853,
      "num_tokens": 83511336.0,
      "step": 48145
    },
    {
      "entropy": 5.8398580074310305,
      "epoch": 3.7462361408286324,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036452815832906625,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21009505838155745,
      "num_tokens": 83519171.0,
      "step": 48150
    },
    {
      "entropy": 5.877081251144409,
      "epoch": 3.74662517020035,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036450273060696984,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21030834764242173,
      "num_tokens": 83527359.0,
      "step": 48155
    },
    {
      "entropy": 5.832540607452392,
      "epoch": 3.747014199572068,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003644773015268118,
      "loss": 4.8763,
      "mean_token_accuracy": 0.21326199620962144,
      "num_tokens": 83535746.0,
      "step": 48160
    },
    {
      "entropy": 5.790484046936035,
      "epoch": 3.747403228943785,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036445187108897825,
      "loss": 4.9043,
      "mean_token_accuracy": 0.21852574348449708,
      "num_tokens": 83544109.0,
      "step": 48165
    },
    {
      "entropy": 5.864397096633911,
      "epoch": 3.747792258315503,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003644264392938549,
      "loss": 5.1022,
      "mean_token_accuracy": 0.20583832114934922,
      "num_tokens": 83553099.0,
      "step": 48170
    },
    {
      "entropy": 5.926620674133301,
      "epoch": 3.7481812876872205,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036440100614182775,
      "loss": 5.0416,
      "mean_token_accuracy": 0.2067590445280075,
      "num_tokens": 83561998.0,
      "step": 48175
    },
    {
      "entropy": 5.865017080307007,
      "epoch": 3.748570317058938,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036437557163328255,
      "loss": 4.974,
      "mean_token_accuracy": 0.20464721620082854,
      "num_tokens": 83570896.0,
      "step": 48180
    },
    {
      "entropy": 5.858434343338013,
      "epoch": 3.7489593464306554,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003643501357686053,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20509726852178572,
      "num_tokens": 83579043.0,
      "step": 48185
    },
    {
      "entropy": 5.772554063796997,
      "epoch": 3.749348375802373,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036432469854818194,
      "loss": 4.8664,
      "mean_token_accuracy": 0.21594008356332778,
      "num_tokens": 83587929.0,
      "step": 48190
    },
    {
      "entropy": 5.80068564414978,
      "epoch": 3.7497374051740904,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003642992599723985,
      "loss": 4.882,
      "mean_token_accuracy": 0.21864021867513656,
      "num_tokens": 83596428.0,
      "step": 48195
    },
    {
      "entropy": 5.839773941040039,
      "epoch": 3.750126434545808,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003642738200416409,
      "loss": 4.961,
      "mean_token_accuracy": 0.2149035930633545,
      "num_tokens": 83606247.0,
      "step": 48200
    },
    {
      "entropy": 5.863554239273071,
      "epoch": 3.7505154639175258,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036424837875629517,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2164909467101097,
      "num_tokens": 83615056.0,
      "step": 48205
    },
    {
      "entropy": 5.836910820007324,
      "epoch": 3.7509044932892435,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036422293611674736,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2165834367275238,
      "num_tokens": 83623452.0,
      "step": 48210
    },
    {
      "entropy": 5.776988887786866,
      "epoch": 3.751293522660961,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036419749212338346,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2041123941540718,
      "num_tokens": 83631727.0,
      "step": 48215
    },
    {
      "entropy": 5.815929985046386,
      "epoch": 3.7516825520326784,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003641720467765897,
      "loss": 4.8657,
      "mean_token_accuracy": 0.21736502200365065,
      "num_tokens": 83639644.0,
      "step": 48220
    },
    {
      "entropy": 5.765953302383423,
      "epoch": 3.752071581404396,
      "grad_norm": 1.46875,
      "learning_rate": 0.000364146600076752,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2102223426103592,
      "num_tokens": 83648291.0,
      "step": 48225
    },
    {
      "entropy": 5.746558046340942,
      "epoch": 3.7524606107761134,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003641211520242565,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22728172242641448,
      "num_tokens": 83657482.0,
      "step": 48230
    },
    {
      "entropy": 5.775856351852417,
      "epoch": 3.752849640147831,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036409570261948943,
      "loss": 4.8659,
      "mean_token_accuracy": 0.21271939277648927,
      "num_tokens": 83666264.0,
      "step": 48235
    },
    {
      "entropy": 5.868864679336548,
      "epoch": 3.7532386695195488,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003640702518628369,
      "loss": 5.0534,
      "mean_token_accuracy": 0.2034298449754715,
      "num_tokens": 83675552.0,
      "step": 48240
    },
    {
      "entropy": 5.863110494613648,
      "epoch": 3.7536276988912665,
      "grad_norm": 1.375,
      "learning_rate": 0.000364044799754685,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20900235027074815,
      "num_tokens": 83684018.0,
      "step": 48245
    },
    {
      "entropy": 5.824972486495971,
      "epoch": 3.7540167282629837,
      "grad_norm": 1.375,
      "learning_rate": 0.00036401934629542,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2145191565155983,
      "num_tokens": 83691769.0,
      "step": 48250
    },
    {
      "entropy": 5.771638345718384,
      "epoch": 3.7544057576347014,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036399389148542813,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2163345918059349,
      "num_tokens": 83700741.0,
      "step": 48255
    },
    {
      "entropy": 5.775435209274292,
      "epoch": 3.754794787006419,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036396843532509563,
      "loss": 4.9363,
      "mean_token_accuracy": 0.20917914062738419,
      "num_tokens": 83710345.0,
      "step": 48260
    },
    {
      "entropy": 5.86752758026123,
      "epoch": 3.7551838163781364,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003639429778148087,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21330667585134505,
      "num_tokens": 83719106.0,
      "step": 48265
    },
    {
      "entropy": 5.874535322189331,
      "epoch": 3.755572845749854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003639175189549536,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2030404344201088,
      "num_tokens": 83727842.0,
      "step": 48270
    },
    {
      "entropy": 5.837495374679565,
      "epoch": 3.7559618751215718,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036389205874591676,
      "loss": 4.98,
      "mean_token_accuracy": 0.20765695720911026,
      "num_tokens": 83736664.0,
      "step": 48275
    },
    {
      "entropy": 5.808910131454468,
      "epoch": 3.7563509044932895,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036386659718808437,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20215073227882385,
      "num_tokens": 83745257.0,
      "step": 48280
    },
    {
      "entropy": 5.863513708114624,
      "epoch": 3.7567399338650067,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036384113428184286,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21522438377141953,
      "num_tokens": 83754073.0,
      "step": 48285
    },
    {
      "entropy": 5.839964866638184,
      "epoch": 3.7571289632367244,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003638156700275784,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20508417189121247,
      "num_tokens": 83763348.0,
      "step": 48290
    },
    {
      "entropy": 5.871837472915649,
      "epoch": 3.757517992608442,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003637902044256777,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2066944047808647,
      "num_tokens": 83772325.0,
      "step": 48295
    },
    {
      "entropy": 5.77579550743103,
      "epoch": 3.7579070219801594,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003637647374765269,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21407010406255722,
      "num_tokens": 83780991.0,
      "step": 48300
    },
    {
      "entropy": 5.792299604415893,
      "epoch": 3.758296051351877,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036373926918051243,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21621909290552138,
      "num_tokens": 83789447.0,
      "step": 48305
    },
    {
      "entropy": 5.808613872528076,
      "epoch": 3.7586850807235948,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036371379953802085,
      "loss": 4.921,
      "mean_token_accuracy": 0.21481148898601532,
      "num_tokens": 83797508.0,
      "step": 48310
    },
    {
      "entropy": 5.857865858078003,
      "epoch": 3.7590741100953124,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00036368832854943856,
      "loss": 4.954,
      "mean_token_accuracy": 0.20573668777942658,
      "num_tokens": 83805861.0,
      "step": 48315
    },
    {
      "entropy": 5.8401069164276125,
      "epoch": 3.7594631394670297,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036366285621515206,
      "loss": 5.0116,
      "mean_token_accuracy": 0.20917888432741166,
      "num_tokens": 83814921.0,
      "step": 48320
    },
    {
      "entropy": 5.761917924880981,
      "epoch": 3.7598521688387474,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003636373825355478,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2338728591799736,
      "num_tokens": 83823143.0,
      "step": 48325
    },
    {
      "entropy": 5.799488592147827,
      "epoch": 3.7602411982104647,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003636119075110124,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21674566417932511,
      "num_tokens": 83831407.0,
      "step": 48330
    },
    {
      "entropy": 5.805618572235107,
      "epoch": 3.7606302275821823,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003635864311419323,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2212614268064499,
      "num_tokens": 83840477.0,
      "step": 48335
    },
    {
      "entropy": 5.723220157623291,
      "epoch": 3.7610192569539,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036356095342869414,
      "loss": 4.7901,
      "mean_token_accuracy": 0.22120074778795243,
      "num_tokens": 83849158.0,
      "step": 48340
    },
    {
      "entropy": 5.7741968631744385,
      "epoch": 3.7614082863256177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003635354743716844,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2144678607583046,
      "num_tokens": 83857400.0,
      "step": 48345
    },
    {
      "entropy": 5.901990365982056,
      "epoch": 3.761797315697335,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003635099939712898,
      "loss": 5.1012,
      "mean_token_accuracy": 0.19687193781137466,
      "num_tokens": 83866694.0,
      "step": 48350
    },
    {
      "entropy": 5.873877239227295,
      "epoch": 3.7621863450690527,
      "grad_norm": 1.25,
      "learning_rate": 0.00036348451222789694,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21617410629987716,
      "num_tokens": 83875592.0,
      "step": 48355
    },
    {
      "entropy": 5.8364586353302,
      "epoch": 3.7625753744407704,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036345902914189246,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2168629601597786,
      "num_tokens": 83883582.0,
      "step": 48360
    },
    {
      "entropy": 5.82803316116333,
      "epoch": 3.7629644038124876,
      "grad_norm": 1.421875,
      "learning_rate": 0.000363433544713663,
      "loss": 5.004,
      "mean_token_accuracy": 0.21136520355939864,
      "num_tokens": 83891607.0,
      "step": 48365
    },
    {
      "entropy": 5.795352220535278,
      "epoch": 3.7633534331842053,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036340805894359524,
      "loss": 4.892,
      "mean_token_accuracy": 0.2167914777994156,
      "num_tokens": 83900186.0,
      "step": 48370
    },
    {
      "entropy": 5.8208897590637205,
      "epoch": 3.763742462555923,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036338257183207587,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21317292451858522,
      "num_tokens": 83909271.0,
      "step": 48375
    },
    {
      "entropy": 5.79356017112732,
      "epoch": 3.7641314919276407,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036335708337949165,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21069495677947997,
      "num_tokens": 83917780.0,
      "step": 48380
    },
    {
      "entropy": 5.825481700897217,
      "epoch": 3.764520521299358,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003633315935862293,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21167555898427964,
      "num_tokens": 83926528.0,
      "step": 48385
    },
    {
      "entropy": 5.8778587818145756,
      "epoch": 3.7649095506710757,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003633061024526757,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21000041514635087,
      "num_tokens": 83935838.0,
      "step": 48390
    },
    {
      "entropy": 5.779851818084717,
      "epoch": 3.7652985800427934,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003632806099792175,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2177865669131279,
      "num_tokens": 83944716.0,
      "step": 48395
    },
    {
      "entropy": 5.772506618499756,
      "epoch": 3.7656876094145106,
      "grad_norm": 1.375,
      "learning_rate": 0.00036325511616624155,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2156564399600029,
      "num_tokens": 83952615.0,
      "step": 48400
    },
    {
      "entropy": 5.7676185131072994,
      "epoch": 3.7660766387862283,
      "grad_norm": 1.5,
      "learning_rate": 0.0003632296210141348,
      "loss": 4.8624,
      "mean_token_accuracy": 0.21846913993358613,
      "num_tokens": 83960834.0,
      "step": 48405
    },
    {
      "entropy": 5.740039110183716,
      "epoch": 3.766465668157946,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036320412452328383,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21922811418771743,
      "num_tokens": 83969085.0,
      "step": 48410
    },
    {
      "entropy": 5.760934162139892,
      "epoch": 3.7668546975296637,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003631786266940757,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22029993385076524,
      "num_tokens": 83978541.0,
      "step": 48415
    },
    {
      "entropy": 5.871448659896851,
      "epoch": 3.767243726901381,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036315312752689726,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20716084837913512,
      "num_tokens": 83987439.0,
      "step": 48420
    },
    {
      "entropy": 5.825130367279053,
      "epoch": 3.7676327562730987,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003631276270221354,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2123594656586647,
      "num_tokens": 83995901.0,
      "step": 48425
    },
    {
      "entropy": 5.925189447402954,
      "epoch": 3.768021785644816,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036310212518017717,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21028683036565782,
      "num_tokens": 84004248.0,
      "step": 48430
    },
    {
      "entropy": 5.846250009536743,
      "epoch": 3.7684108150165336,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036307662200140935,
      "loss": 4.9945,
      "mean_token_accuracy": 0.20550352483987808,
      "num_tokens": 84012365.0,
      "step": 48435
    },
    {
      "entropy": 5.834045314788819,
      "epoch": 3.7687998443882513,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003630511174862189,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20488521903753282,
      "num_tokens": 84020286.0,
      "step": 48440
    },
    {
      "entropy": 5.823097133636475,
      "epoch": 3.769188873759969,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003630256116349931,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21173248440027237,
      "num_tokens": 84029632.0,
      "step": 48445
    },
    {
      "entropy": 5.875268268585205,
      "epoch": 3.7695779031316863,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003630001044481186,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2124560609459877,
      "num_tokens": 84037953.0,
      "step": 48450
    },
    {
      "entropy": 5.8263468742370605,
      "epoch": 3.769966932503404,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036297459592598256,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21769519299268722,
      "num_tokens": 84046860.0,
      "step": 48455
    },
    {
      "entropy": 5.806731557846069,
      "epoch": 3.7703559618751217,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003629490860689721,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2066218838095665,
      "num_tokens": 84055306.0,
      "step": 48460
    },
    {
      "entropy": 5.797434568405151,
      "epoch": 3.770744991246839,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036292357487747426,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21063714772462844,
      "num_tokens": 84064407.0,
      "step": 48465
    },
    {
      "entropy": 5.821697235107422,
      "epoch": 3.7711340206185566,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036289806235187606,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21471987068653106,
      "num_tokens": 84072812.0,
      "step": 48470
    },
    {
      "entropy": 5.7669425964355465,
      "epoch": 3.7715230499902743,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003628725484925647,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21202560216188432,
      "num_tokens": 84081514.0,
      "step": 48475
    },
    {
      "entropy": 5.8390062808990475,
      "epoch": 3.771912079361992,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036284703329992734,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2095906600356102,
      "num_tokens": 84089945.0,
      "step": 48480
    },
    {
      "entropy": 5.883571338653565,
      "epoch": 3.7723011087337093,
      "grad_norm": 1.34375,
      "learning_rate": 0.000362821516774351,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20647194534540175,
      "num_tokens": 84098328.0,
      "step": 48485
    },
    {
      "entropy": 5.8289751529693605,
      "epoch": 3.772690138105427,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036279599891622297,
      "loss": 4.9844,
      "mean_token_accuracy": 0.20670776963233947,
      "num_tokens": 84106688.0,
      "step": 48490
    },
    {
      "entropy": 5.736328792572022,
      "epoch": 3.7730791674771447,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036277047972593036,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21784191578626633,
      "num_tokens": 84115568.0,
      "step": 48495
    },
    {
      "entropy": 5.877501106262207,
      "epoch": 3.773468196848862,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003627449592038604,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2130642667412758,
      "num_tokens": 84123824.0,
      "step": 48500
    },
    {
      "entropy": 5.830891561508179,
      "epoch": 3.7738572262205796,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003627194373504004,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2125416100025177,
      "num_tokens": 84131927.0,
      "step": 48505
    },
    {
      "entropy": 5.845138502120972,
      "epoch": 3.7742462555922973,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036269391416593753,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21750535517930986,
      "num_tokens": 84139855.0,
      "step": 48510
    },
    {
      "entropy": 5.773061227798462,
      "epoch": 3.774635284964015,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003626683896508591,
      "loss": 4.8662,
      "mean_token_accuracy": 0.21444876492023468,
      "num_tokens": 84147705.0,
      "step": 48515
    },
    {
      "entropy": 5.831519031524659,
      "epoch": 3.7750243143357323,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003626428638055524,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20024812519550322,
      "num_tokens": 84156662.0,
      "step": 48520
    },
    {
      "entropy": 5.806011009216308,
      "epoch": 3.77541334370745,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036261733663040473,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20691410154104234,
      "num_tokens": 84165828.0,
      "step": 48525
    },
    {
      "entropy": 5.800295829772949,
      "epoch": 3.775802373079167,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036259180812580347,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21730743646621703,
      "num_tokens": 84174680.0,
      "step": 48530
    },
    {
      "entropy": 5.794473123550415,
      "epoch": 3.776191402450885,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036256627829213585,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22008173912763596,
      "num_tokens": 84183530.0,
      "step": 48535
    },
    {
      "entropy": 5.811361169815063,
      "epoch": 3.7765804318226026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036254074712978947,
      "loss": 4.95,
      "mean_token_accuracy": 0.2191836878657341,
      "num_tokens": 84191856.0,
      "step": 48540
    },
    {
      "entropy": 5.853907203674316,
      "epoch": 3.7769694611943203,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036251521463915144,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20507093220949174,
      "num_tokens": 84199789.0,
      "step": 48545
    },
    {
      "entropy": 5.82916226387024,
      "epoch": 3.777358490566038,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003624896808206093,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21761843115091323,
      "num_tokens": 84207539.0,
      "step": 48550
    },
    {
      "entropy": 5.836031150817871,
      "epoch": 3.7777475199377553,
      "grad_norm": 1.25,
      "learning_rate": 0.0003624641456745507,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21326977759599686,
      "num_tokens": 84216478.0,
      "step": 48555
    },
    {
      "entropy": 5.794806385040284,
      "epoch": 3.778136549309473,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036243860920136283,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21193195581436158,
      "num_tokens": 84225452.0,
      "step": 48560
    },
    {
      "entropy": 5.826672124862671,
      "epoch": 3.77852557868119,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036241307140143323,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2282642036676407,
      "num_tokens": 84235015.0,
      "step": 48565
    },
    {
      "entropy": 5.781571722030639,
      "epoch": 3.778914608052908,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036238753227514934,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2117591917514801,
      "num_tokens": 84243694.0,
      "step": 48570
    },
    {
      "entropy": 5.857712507247925,
      "epoch": 3.7793036374246256,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036236199182289883,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2028741329908371,
      "num_tokens": 84252338.0,
      "step": 48575
    },
    {
      "entropy": 5.793635034561158,
      "epoch": 3.7796926667963433,
      "grad_norm": 1.390625,
      "learning_rate": 0.000362336450045069,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22171851992607117,
      "num_tokens": 84260963.0,
      "step": 48580
    },
    {
      "entropy": 5.890757751464844,
      "epoch": 3.7800816961680606,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036231090694204763,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20527772307395936,
      "num_tokens": 84270012.0,
      "step": 48585
    },
    {
      "entropy": 5.789273738861084,
      "epoch": 3.7804707255397783,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036228536251422226,
      "loss": 4.9521,
      "mean_token_accuracy": 0.20600617974996566,
      "num_tokens": 84278301.0,
      "step": 48590
    },
    {
      "entropy": 5.862462854385376,
      "epoch": 3.780859754911496,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036225981676198045,
      "loss": 5.0169,
      "mean_token_accuracy": 0.20666330307722092,
      "num_tokens": 84287382.0,
      "step": 48595
    },
    {
      "entropy": 5.847435426712036,
      "epoch": 3.781248784283213,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003622342696857098,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21007767021656037,
      "num_tokens": 84296358.0,
      "step": 48600
    },
    {
      "entropy": 5.848305511474609,
      "epoch": 3.781637813654931,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003622087212857979,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22253683358430862,
      "num_tokens": 84304638.0,
      "step": 48605
    },
    {
      "entropy": 5.772209024429321,
      "epoch": 3.7820268430266486,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003621831715626325,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21041425168514252,
      "num_tokens": 84313884.0,
      "step": 48610
    },
    {
      "entropy": 5.899036407470703,
      "epoch": 3.7824158723983663,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036215762051660115,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20718845427036287,
      "num_tokens": 84322937.0,
      "step": 48615
    },
    {
      "entropy": 5.8359044075012205,
      "epoch": 3.7828049017700835,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036213206814809175,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21275405138731002,
      "num_tokens": 84331603.0,
      "step": 48620
    },
    {
      "entropy": 5.8694957256317135,
      "epoch": 3.7831939311418012,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003621065144574919,
      "loss": 5.023,
      "mean_token_accuracy": 0.2151937574148178,
      "num_tokens": 84339948.0,
      "step": 48625
    },
    {
      "entropy": 5.853318548202514,
      "epoch": 3.7835829605135185,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003620809594451893,
      "loss": 4.962,
      "mean_token_accuracy": 0.20590658485889435,
      "num_tokens": 84348489.0,
      "step": 48630
    },
    {
      "entropy": 5.850880098342896,
      "epoch": 3.783971989885236,
      "grad_norm": 1.375,
      "learning_rate": 0.0003620554031115718,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2095250517129898,
      "num_tokens": 84357267.0,
      "step": 48635
    },
    {
      "entropy": 5.835837459564209,
      "epoch": 3.784361019256954,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003620298454570271,
      "loss": 4.9752,
      "mean_token_accuracy": 0.20703328996896744,
      "num_tokens": 84366189.0,
      "step": 48640
    },
    {
      "entropy": 5.874964904785156,
      "epoch": 3.7847500486286716,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003620042864819429,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20731819570064544,
      "num_tokens": 84375327.0,
      "step": 48645
    },
    {
      "entropy": 5.849944305419922,
      "epoch": 3.7851390780003893,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003619787261867072,
      "loss": 4.9295,
      "mean_token_accuracy": 0.20880162566900254,
      "num_tokens": 84384152.0,
      "step": 48650
    },
    {
      "entropy": 5.854048776626587,
      "epoch": 3.7855281073721065,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003619531645717078,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2010303795337677,
      "num_tokens": 84392700.0,
      "step": 48655
    },
    {
      "entropy": 5.8048155307769775,
      "epoch": 3.7859171367438242,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003619276016373325,
      "loss": 5.0098,
      "mean_token_accuracy": 0.20672389268875122,
      "num_tokens": 84401310.0,
      "step": 48660
    },
    {
      "entropy": 5.780894184112549,
      "epoch": 3.7863061661155415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003619020373839693,
      "loss": 4.9679,
      "mean_token_accuracy": 0.20631978213787078,
      "num_tokens": 84410028.0,
      "step": 48665
    },
    {
      "entropy": 5.9214768409729,
      "epoch": 3.786695195487259,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003618764718120059,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21580541133880615,
      "num_tokens": 84419015.0,
      "step": 48670
    },
    {
      "entropy": 5.910969066619873,
      "epoch": 3.787084224858977,
      "grad_norm": 1.25,
      "learning_rate": 0.0003618509049218304,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21167200654745102,
      "num_tokens": 84427148.0,
      "step": 48675
    },
    {
      "entropy": 5.869602155685425,
      "epoch": 3.7874732542306946,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036182533671383056,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2114063784480095,
      "num_tokens": 84436058.0,
      "step": 48680
    },
    {
      "entropy": 5.870849466323852,
      "epoch": 3.787862283602412,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003617997671883944,
      "loss": 4.9654,
      "mean_token_accuracy": 0.209538996219635,
      "num_tokens": 84444494.0,
      "step": 48685
    },
    {
      "entropy": 5.908856010437011,
      "epoch": 3.7882513129741295,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036177419634591,
      "loss": 5.0215,
      "mean_token_accuracy": 0.2145876795053482,
      "num_tokens": 84453558.0,
      "step": 48690
    },
    {
      "entropy": 5.833027124404907,
      "epoch": 3.7886403423458472,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003617486241867654,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2108340248465538,
      "num_tokens": 84462636.0,
      "step": 48695
    },
    {
      "entropy": 5.871826171875,
      "epoch": 3.7890293717175645,
      "grad_norm": 1.375,
      "learning_rate": 0.0003617230507113483,
      "loss": 4.9901,
      "mean_token_accuracy": 0.20720430612564086,
      "num_tokens": 84471559.0,
      "step": 48700
    },
    {
      "entropy": 5.85174674987793,
      "epoch": 3.789418401089282,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036169747592004704,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21520561277866362,
      "num_tokens": 84479976.0,
      "step": 48705
    },
    {
      "entropy": 5.842283535003662,
      "epoch": 3.789807430461,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036167189981324956,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21144671440124513,
      "num_tokens": 84488890.0,
      "step": 48710
    },
    {
      "entropy": 5.799002933502197,
      "epoch": 3.7901964598327176,
      "grad_norm": 1.46875,
      "learning_rate": 0.000361646322391344,
      "loss": 4.9205,
      "mean_token_accuracy": 0.21397174745798112,
      "num_tokens": 84496816.0,
      "step": 48715
    },
    {
      "entropy": 5.779604768753051,
      "epoch": 3.790585489204435,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003616207436547183,
      "loss": 4.9507,
      "mean_token_accuracy": 0.20795523077249528,
      "num_tokens": 84505078.0,
      "step": 48720
    },
    {
      "entropy": 5.7595446586608885,
      "epoch": 3.7909745185761525,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036159516360376084,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2111603155732155,
      "num_tokens": 84513448.0,
      "step": 48725
    },
    {
      "entropy": 5.791185522079468,
      "epoch": 3.7913635479478702,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036156958223885945,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21237398386001588,
      "num_tokens": 84522267.0,
      "step": 48730
    },
    {
      "entropy": 5.83224663734436,
      "epoch": 3.7917525773195875,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003615439995604025,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2129225179553032,
      "num_tokens": 84530757.0,
      "step": 48735
    },
    {
      "entropy": 5.823274564743042,
      "epoch": 3.792141606691305,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003615184155687781,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20139263570308685,
      "num_tokens": 84539332.0,
      "step": 48740
    },
    {
      "entropy": 5.822123432159424,
      "epoch": 3.792530636063023,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036149283026437444,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2194872885942459,
      "num_tokens": 84547983.0,
      "step": 48745
    },
    {
      "entropy": 5.8502130031585695,
      "epoch": 3.7929196654347406,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036146724364757984,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21135388314723969,
      "num_tokens": 84556037.0,
      "step": 48750
    },
    {
      "entropy": 5.792223930358887,
      "epoch": 3.793308694806458,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036144165571878233,
      "loss": 4.9658,
      "mean_token_accuracy": 0.21213280707597731,
      "num_tokens": 84564819.0,
      "step": 48755
    },
    {
      "entropy": 5.829569482803345,
      "epoch": 3.7936977241781755,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003614160664783703,
      "loss": 4.9702,
      "mean_token_accuracy": 0.20574671179056167,
      "num_tokens": 84573768.0,
      "step": 48760
    },
    {
      "entropy": 5.8744720935821535,
      "epoch": 3.7940867535498928,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000361390475926732,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20789277851581572,
      "num_tokens": 84582710.0,
      "step": 48765
    },
    {
      "entropy": 5.911956739425659,
      "epoch": 3.7944757829216105,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003613648840642558,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20781103074550628,
      "num_tokens": 84592285.0,
      "step": 48770
    },
    {
      "entropy": 5.9221021175384525,
      "epoch": 3.794864812293328,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003613392908913299,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20861401110887529,
      "num_tokens": 84600885.0,
      "step": 48775
    },
    {
      "entropy": 5.883755731582641,
      "epoch": 3.795253841665046,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003613136964083427,
      "loss": 4.9751,
      "mean_token_accuracy": 0.21648920327425003,
      "num_tokens": 84609775.0,
      "step": 48780
    },
    {
      "entropy": 5.877562570571899,
      "epoch": 3.795642871036763,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003612881006156826,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20732762068510055,
      "num_tokens": 84618636.0,
      "step": 48785
    },
    {
      "entropy": 5.850293970108032,
      "epoch": 3.796031900408481,
      "grad_norm": 1.5,
      "learning_rate": 0.00036126250351373786,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21516039222478867,
      "num_tokens": 84626950.0,
      "step": 48790
    },
    {
      "entropy": 5.774712181091308,
      "epoch": 3.7964209297801985,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036123690510289684,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22077593058347703,
      "num_tokens": 84635551.0,
      "step": 48795
    },
    {
      "entropy": 5.792155408859253,
      "epoch": 3.7968099591519158,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003612113053835482,
      "loss": 4.8793,
      "mean_token_accuracy": 0.20960229486227036,
      "num_tokens": 84643626.0,
      "step": 48800
    },
    {
      "entropy": 5.874011707305908,
      "epoch": 3.7971989885236335,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003611857043560801,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21501094549894334,
      "num_tokens": 84651891.0,
      "step": 48805
    },
    {
      "entropy": 5.810036563873291,
      "epoch": 3.797588017895351,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003611601020208812,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21245778650045394,
      "num_tokens": 84660501.0,
      "step": 48810
    },
    {
      "entropy": 5.746427822113037,
      "epoch": 3.797977047267069,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003611344983783398,
      "loss": 4.8398,
      "mean_token_accuracy": 0.2225646361708641,
      "num_tokens": 84669134.0,
      "step": 48815
    },
    {
      "entropy": 5.803394508361817,
      "epoch": 3.798366076638786,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003611088934288445,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20939016342163086,
      "num_tokens": 84677241.0,
      "step": 48820
    },
    {
      "entropy": 5.808235502243042,
      "epoch": 3.798755106010504,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036108328717278387,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21756461560726165,
      "num_tokens": 84686244.0,
      "step": 48825
    },
    {
      "entropy": 5.727559757232666,
      "epoch": 3.7991441353822215,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036105767961054636,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21468983888626098,
      "num_tokens": 84694570.0,
      "step": 48830
    },
    {
      "entropy": 5.779499912261963,
      "epoch": 3.7995331647539388,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003610320707425205,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2171340510249138,
      "num_tokens": 84703322.0,
      "step": 48835
    },
    {
      "entropy": 5.896618223190307,
      "epoch": 3.7999221941256565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000361006460569095,
      "loss": 4.969,
      "mean_token_accuracy": 0.21632780283689498,
      "num_tokens": 84711960.0,
      "step": 48840
    },
    {
      "entropy": 5.835455751419067,
      "epoch": 3.800311223497374,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036098084909065814,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2144725054502487,
      "num_tokens": 84720956.0,
      "step": 48845
    },
    {
      "entropy": 5.813449048995972,
      "epoch": 3.800700252869092,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003609552363075989,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21158041805028915,
      "num_tokens": 84729829.0,
      "step": 48850
    },
    {
      "entropy": 5.803310871124268,
      "epoch": 3.801089282240809,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036092962222030573,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21511098891496658,
      "num_tokens": 84738314.0,
      "step": 48855
    },
    {
      "entropy": 5.819625806808472,
      "epoch": 3.801478311612527,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003609040068291673,
      "loss": 4.8519,
      "mean_token_accuracy": 0.21624375134706497,
      "num_tokens": 84746868.0,
      "step": 48860
    },
    {
      "entropy": 5.841644620895385,
      "epoch": 3.801867340984244,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003608783901345724,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21290566623210908,
      "num_tokens": 84755679.0,
      "step": 48865
    },
    {
      "entropy": 5.842495250701904,
      "epoch": 3.8022563703559618,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036085277213690947,
      "loss": 5.015,
      "mean_token_accuracy": 0.21089762300252915,
      "num_tokens": 84764653.0,
      "step": 48870
    },
    {
      "entropy": 5.767722654342651,
      "epoch": 3.8026453997276795,
      "grad_norm": 1.375,
      "learning_rate": 0.0003608271528365675,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2155345693230629,
      "num_tokens": 84773090.0,
      "step": 48875
    },
    {
      "entropy": 5.806365013122559,
      "epoch": 3.803034429099397,
      "grad_norm": 1.40625,
      "learning_rate": 0.000360801532233935,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21938023269176482,
      "num_tokens": 84783203.0,
      "step": 48880
    },
    {
      "entropy": 5.8141663551330565,
      "epoch": 3.8034234584711144,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003607759103294008,
      "loss": 4.939,
      "mean_token_accuracy": 0.2217843621969223,
      "num_tokens": 84791800.0,
      "step": 48885
    },
    {
      "entropy": 5.8646800994873045,
      "epoch": 3.803812487842832,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003607502871233538,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21114450991153716,
      "num_tokens": 84800791.0,
      "step": 48890
    },
    {
      "entropy": 5.841810178756714,
      "epoch": 3.80420151721455,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003607246626161826,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21599150002002715,
      "num_tokens": 84808879.0,
      "step": 48895
    },
    {
      "entropy": 5.8505676746368405,
      "epoch": 3.804590546586267,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000360699036808276,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2027676895260811,
      "num_tokens": 84817388.0,
      "step": 48900
    },
    {
      "entropy": 5.907699871063232,
      "epoch": 3.8049795759579847,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003606734097000231,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21445808857679366,
      "num_tokens": 84826044.0,
      "step": 48905
    },
    {
      "entropy": 5.838197898864746,
      "epoch": 3.8053686053297024,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003606477812918125,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21153876930475235,
      "num_tokens": 84834293.0,
      "step": 48910
    },
    {
      "entropy": 5.784072160720825,
      "epoch": 3.80575763470142,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003606221515840331,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21273598074913025,
      "num_tokens": 84842441.0,
      "step": 48915
    },
    {
      "entropy": 5.870151329040527,
      "epoch": 3.8061466640731374,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036059652057707394,
      "loss": 4.976,
      "mean_token_accuracy": 0.21800754368305206,
      "num_tokens": 84851284.0,
      "step": 48920
    },
    {
      "entropy": 5.876770496368408,
      "epoch": 3.806535693444855,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003605708882713237,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21358221620321274,
      "num_tokens": 84859491.0,
      "step": 48925
    },
    {
      "entropy": 5.754761362075806,
      "epoch": 3.806924722816573,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036054525466717145,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21173783540725707,
      "num_tokens": 84868087.0,
      "step": 48930
    },
    {
      "entropy": 5.712821769714355,
      "epoch": 3.80731375218829,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003605196197650062,
      "loss": 4.8316,
      "mean_token_accuracy": 0.22236887216567994,
      "num_tokens": 84876449.0,
      "step": 48935
    },
    {
      "entropy": 5.900254392623902,
      "epoch": 3.8077027815600077,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003604939835652167,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20556507557630538,
      "num_tokens": 84884643.0,
      "step": 48940
    },
    {
      "entropy": 5.747515487670898,
      "epoch": 3.8080918109317254,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036046834606819224,
      "loss": 4.8442,
      "mean_token_accuracy": 0.2212585911154747,
      "num_tokens": 84892865.0,
      "step": 48945
    },
    {
      "entropy": 5.881759214401245,
      "epoch": 3.808480840303443,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003604427072743216,
      "loss": 5.0627,
      "mean_token_accuracy": 0.209686079621315,
      "num_tokens": 84901810.0,
      "step": 48950
    },
    {
      "entropy": 5.8210920810699465,
      "epoch": 3.8088698696751604,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036041706718399386,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22214832454919814,
      "num_tokens": 84909914.0,
      "step": 48955
    },
    {
      "entropy": 5.835501480102539,
      "epoch": 3.809258899046878,
      "grad_norm": 1.28125,
      "learning_rate": 0.000360391425797598,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22583719938993455,
      "num_tokens": 84919417.0,
      "step": 48960
    },
    {
      "entropy": 5.774426889419556,
      "epoch": 3.8096479284185953,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003603657831155233,
      "loss": 4.82,
      "mean_token_accuracy": 0.22366142719984056,
      "num_tokens": 84927987.0,
      "step": 48965
    },
    {
      "entropy": 5.772368240356445,
      "epoch": 3.810036957790313,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003603401391381587,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2263113334774971,
      "num_tokens": 84937220.0,
      "step": 48970
    },
    {
      "entropy": 5.837231636047363,
      "epoch": 3.8104259871620307,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00036031449386589326,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21509986668825148,
      "num_tokens": 84945886.0,
      "step": 48975
    },
    {
      "entropy": 5.876586389541626,
      "epoch": 3.8108150165337484,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003602888472991162,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2041503444314003,
      "num_tokens": 84954834.0,
      "step": 48980
    },
    {
      "entropy": 5.802380037307739,
      "epoch": 3.811204045905466,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036026319943821664,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2128758653998375,
      "num_tokens": 84963321.0,
      "step": 48985
    },
    {
      "entropy": 5.888273668289185,
      "epoch": 3.8115930752771834,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003602375502835837,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2080211877822876,
      "num_tokens": 84971350.0,
      "step": 48990
    },
    {
      "entropy": 5.786940431594848,
      "epoch": 3.811982104648901,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00036021189983560667,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21136770099401475,
      "num_tokens": 84979753.0,
      "step": 48995
    },
    {
      "entropy": 5.791334009170532,
      "epoch": 3.8123711340206183,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036018624809467466,
      "loss": 4.9026,
      "mean_token_accuracy": 0.20939179956912995,
      "num_tokens": 84988568.0,
      "step": 49000
    },
    {
      "entropy": 5.840465354919433,
      "epoch": 3.812760163392336,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00036016059506117687,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21386846154928207,
      "num_tokens": 84996778.0,
      "step": 49005
    },
    {
      "entropy": 5.862120580673218,
      "epoch": 3.8131491927640537,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003601349407355027,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2073841243982315,
      "num_tokens": 85006043.0,
      "step": 49010
    },
    {
      "entropy": 5.919520711898803,
      "epoch": 3.8135382221357714,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003601092851180412,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2122909277677536,
      "num_tokens": 85013893.0,
      "step": 49015
    },
    {
      "entropy": 5.817442989349365,
      "epoch": 3.8139272515074887,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00036008362820918186,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21655259728431703,
      "num_tokens": 85022091.0,
      "step": 49020
    },
    {
      "entropy": 5.8478844165802,
      "epoch": 3.8143162808792064,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003600579700093138,
      "loss": 4.9183,
      "mean_token_accuracy": 0.20790481865406035,
      "num_tokens": 85030896.0,
      "step": 49025
    },
    {
      "entropy": 5.83958740234375,
      "epoch": 3.814705310250924,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036003231051882646,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2212885871529579,
      "num_tokens": 85039862.0,
      "step": 49030
    },
    {
      "entropy": 5.8776469230651855,
      "epoch": 3.8150943396226413,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036000664973810914,
      "loss": 4.997,
      "mean_token_accuracy": 0.2119227886199951,
      "num_tokens": 85048509.0,
      "step": 49035
    },
    {
      "entropy": 5.891389942169189,
      "epoch": 3.815483368994359,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003599809876675513,
      "loss": 4.9492,
      "mean_token_accuracy": 0.2092523992061615,
      "num_tokens": 85056678.0,
      "step": 49040
    },
    {
      "entropy": 5.853326606750488,
      "epoch": 3.8158723983660767,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035995532430754205,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21691081821918487,
      "num_tokens": 85066123.0,
      "step": 49045
    },
    {
      "entropy": 5.853999423980713,
      "epoch": 3.8162614277377944,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003599296596584711,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20025967061519623,
      "num_tokens": 85075391.0,
      "step": 49050
    },
    {
      "entropy": 5.836101484298706,
      "epoch": 3.8166504571095117,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003599039937207278,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21228709816932678,
      "num_tokens": 85083831.0,
      "step": 49055
    },
    {
      "entropy": 5.837673091888428,
      "epoch": 3.8170394864812294,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003598783264947015,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2147866517305374,
      "num_tokens": 85093086.0,
      "step": 49060
    },
    {
      "entropy": 5.827676248550415,
      "epoch": 3.8174285158529466,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035985265798078165,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21358412653207778,
      "num_tokens": 85101635.0,
      "step": 49065
    },
    {
      "entropy": 5.818188285827636,
      "epoch": 3.8178175452246643,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035982698817935783,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22016266286373137,
      "num_tokens": 85110208.0,
      "step": 49070
    },
    {
      "entropy": 5.860749673843384,
      "epoch": 3.818206574596382,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003598013170908194,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20144019424915313,
      "num_tokens": 85118667.0,
      "step": 49075
    },
    {
      "entropy": 5.800343227386475,
      "epoch": 3.8185956039680997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003597756447155559,
      "loss": 4.8877,
      "mean_token_accuracy": 0.21839531362056733,
      "num_tokens": 85127457.0,
      "step": 49080
    },
    {
      "entropy": 5.817784309387207,
      "epoch": 3.8189846333398174,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035974997105395705,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21559182554483414,
      "num_tokens": 85136216.0,
      "step": 49085
    },
    {
      "entropy": 5.842554521560669,
      "epoch": 3.8193736627115347,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035972429610641224,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21449165493249894,
      "num_tokens": 85145168.0,
      "step": 49090
    },
    {
      "entropy": 5.853761959075928,
      "epoch": 3.8197626920832524,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035969861987331114,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21083917915821077,
      "num_tokens": 85153859.0,
      "step": 49095
    },
    {
      "entropy": 5.753165769577026,
      "epoch": 3.8201517214549696,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003596729423550433,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2162601724267006,
      "num_tokens": 85162460.0,
      "step": 49100
    },
    {
      "entropy": 5.717614603042603,
      "epoch": 3.8205407508266873,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003596472635519983,
      "loss": 4.8014,
      "mean_token_accuracy": 0.2199389949440956,
      "num_tokens": 85170602.0,
      "step": 49105
    },
    {
      "entropy": 5.798894214630127,
      "epoch": 3.820929780198405,
      "grad_norm": 1.25,
      "learning_rate": 0.0003596215834645657,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21284208297729493,
      "num_tokens": 85179127.0,
      "step": 49110
    },
    {
      "entropy": 5.784889221191406,
      "epoch": 3.8213188095701227,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035959590209313537,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22739548981189728,
      "num_tokens": 85187486.0,
      "step": 49115
    },
    {
      "entropy": 5.848506164550781,
      "epoch": 3.82170783894184,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003595702194380968,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20802140980958939,
      "num_tokens": 85196272.0,
      "step": 49120
    },
    {
      "entropy": 5.881946134567261,
      "epoch": 3.8220968683135577,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00035954453549983984,
      "loss": 5.0357,
      "mean_token_accuracy": 0.20911688953638077,
      "num_tokens": 85204459.0,
      "step": 49125
    },
    {
      "entropy": 5.916932773590088,
      "epoch": 3.8224858976852754,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035951885027875406,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22024523913860322,
      "num_tokens": 85212985.0,
      "step": 49130
    },
    {
      "entropy": 5.815216112136841,
      "epoch": 3.8228749270569926,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035949316377522934,
      "loss": 4.9567,
      "mean_token_accuracy": 0.217891925573349,
      "num_tokens": 85222295.0,
      "step": 49135
    },
    {
      "entropy": 5.847072505950928,
      "epoch": 3.8232639564287103,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003594674759896553,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20961818397045134,
      "num_tokens": 85231072.0,
      "step": 49140
    },
    {
      "entropy": 5.853895235061645,
      "epoch": 3.823652985800428,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003594417869224217,
      "loss": 4.9486,
      "mean_token_accuracy": 0.20720356106758117,
      "num_tokens": 85239568.0,
      "step": 49145
    },
    {
      "entropy": 5.936257696151733,
      "epoch": 3.8240420151721457,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035941609657391845,
      "loss": 5.0988,
      "mean_token_accuracy": 0.19757563322782518,
      "num_tokens": 85248029.0,
      "step": 49150
    },
    {
      "entropy": 5.853573322296143,
      "epoch": 3.824431044543863,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003593904049445353,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20667776614427566,
      "num_tokens": 85256385.0,
      "step": 49155
    },
    {
      "entropy": 5.78083381652832,
      "epoch": 3.8248200739155807,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035936471203466206,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2153155103325844,
      "num_tokens": 85264727.0,
      "step": 49160
    },
    {
      "entropy": 5.8222626686096195,
      "epoch": 3.8252091032872984,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003593390178446886,
      "loss": 4.9753,
      "mean_token_accuracy": 0.20932394862174988,
      "num_tokens": 85272906.0,
      "step": 49165
    },
    {
      "entropy": 5.832112121582031,
      "epoch": 3.8255981326590156,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003593133223750047,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21281614899635315,
      "num_tokens": 85281923.0,
      "step": 49170
    },
    {
      "entropy": 5.790166139602661,
      "epoch": 3.8259871620307333,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003592876256260004,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20927174538373947,
      "num_tokens": 85290246.0,
      "step": 49175
    },
    {
      "entropy": 5.850388908386231,
      "epoch": 3.826376191402451,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003592619275980656,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2081973597407341,
      "num_tokens": 85298901.0,
      "step": 49180
    },
    {
      "entropy": 5.809437894821167,
      "epoch": 3.8267652207741687,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035923622829159014,
      "loss": 4.924,
      "mean_token_accuracy": 0.21125389635562897,
      "num_tokens": 85307477.0,
      "step": 49185
    },
    {
      "entropy": 5.78082242012024,
      "epoch": 3.827154250145886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003592105277069639,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22233271598815918,
      "num_tokens": 85316400.0,
      "step": 49190
    },
    {
      "entropy": 5.829493999481201,
      "epoch": 3.8275432795176036,
      "grad_norm": 1.421875,
      "learning_rate": 0.000359184825844577,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2128453552722931,
      "num_tokens": 85324651.0,
      "step": 49195
    },
    {
      "entropy": 5.745706605911255,
      "epoch": 3.827932308889321,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035915912270481935,
      "loss": 4.8608,
      "mean_token_accuracy": 0.21970924735069275,
      "num_tokens": 85333361.0,
      "step": 49200
    },
    {
      "entropy": 5.768571853637695,
      "epoch": 3.8283213382610386,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000359133418288081,
      "loss": 4.8782,
      "mean_token_accuracy": 0.21183487325906752,
      "num_tokens": 85341928.0,
      "step": 49205
    },
    {
      "entropy": 5.8108744621276855,
      "epoch": 3.8287103676327563,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035910771259475187,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21959973722696305,
      "num_tokens": 85350261.0,
      "step": 49210
    },
    {
      "entropy": 5.844123125076294,
      "epoch": 3.829099397004474,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003590820056252222,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21411115378141404,
      "num_tokens": 85358778.0,
      "step": 49215
    },
    {
      "entropy": 5.799669933319092,
      "epoch": 3.8294884263761912,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003590562973798818,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21377661675214768,
      "num_tokens": 85367935.0,
      "step": 49220
    },
    {
      "entropy": 5.8212902545928955,
      "epoch": 3.829877455747909,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003590305878591209,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21365402042865753,
      "num_tokens": 85376961.0,
      "step": 49225
    },
    {
      "entropy": 5.90946946144104,
      "epoch": 3.8302664851196266,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035900487706332963,
      "loss": 4.9926,
      "mean_token_accuracy": 0.20869186520576477,
      "num_tokens": 85386106.0,
      "step": 49230
    },
    {
      "entropy": 5.8505762100219725,
      "epoch": 3.830655514491344,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000358979164992898,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21031726151704788,
      "num_tokens": 85396315.0,
      "step": 49235
    },
    {
      "entropy": 5.78671441078186,
      "epoch": 3.8310445438630616,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035895345164821624,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21571587324142455,
      "num_tokens": 85405210.0,
      "step": 49240
    },
    {
      "entropy": 5.762450265884399,
      "epoch": 3.8314335732347793,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003589277370296745,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2093357890844345,
      "num_tokens": 85413661.0,
      "step": 49245
    },
    {
      "entropy": 5.79524245262146,
      "epoch": 3.831822602606497,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035890202113766283,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21650041788816451,
      "num_tokens": 85423043.0,
      "step": 49250
    },
    {
      "entropy": 5.9207408905029295,
      "epoch": 3.8322116319782142,
      "grad_norm": 1.375,
      "learning_rate": 0.00035887630397257167,
      "loss": 4.985,
      "mean_token_accuracy": 0.20888074040412902,
      "num_tokens": 85431918.0,
      "step": 49255
    },
    {
      "entropy": 5.811710691452026,
      "epoch": 3.832600661349932,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000358850585534791,
      "loss": 4.8662,
      "mean_token_accuracy": 0.21720490008592605,
      "num_tokens": 85441108.0,
      "step": 49260
    },
    {
      "entropy": 5.811617469787597,
      "epoch": 3.8329896907216496,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035882486582471124,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21214938312768936,
      "num_tokens": 85449637.0,
      "step": 49265
    },
    {
      "entropy": 5.807532787322998,
      "epoch": 3.833378720093367,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003587991448427225,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21026830822229386,
      "num_tokens": 85458268.0,
      "step": 49270
    },
    {
      "entropy": 5.824789667129517,
      "epoch": 3.8337677494650846,
      "grad_norm": 1.25,
      "learning_rate": 0.00035877342258921514,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21495382487773895,
      "num_tokens": 85467198.0,
      "step": 49275
    },
    {
      "entropy": 5.836228609085083,
      "epoch": 3.8341567788368023,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035874769906457936,
      "loss": 4.917,
      "mean_token_accuracy": 0.21399617791175843,
      "num_tokens": 85475445.0,
      "step": 49280
    },
    {
      "entropy": 5.879547023773194,
      "epoch": 3.83454580820852,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035872197426920567,
      "loss": 4.9882,
      "mean_token_accuracy": 0.20564845502376555,
      "num_tokens": 85484475.0,
      "step": 49285
    },
    {
      "entropy": 5.82730073928833,
      "epoch": 3.8349348375802372,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003586962482034842,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21386825889348984,
      "num_tokens": 85492220.0,
      "step": 49290
    },
    {
      "entropy": 5.836488580703735,
      "epoch": 3.835323866951955,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035867052086780544,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2090174674987793,
      "num_tokens": 85501302.0,
      "step": 49295
    },
    {
      "entropy": 5.8463907718658445,
      "epoch": 3.835712896323672,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003586447922625596,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21112343817949294,
      "num_tokens": 85509229.0,
      "step": 49300
    },
    {
      "entropy": 5.7910332679748535,
      "epoch": 3.83610192569539,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035861906238813726,
      "loss": 4.9319,
      "mean_token_accuracy": 0.20873789638280868,
      "num_tokens": 85518139.0,
      "step": 49305
    },
    {
      "entropy": 5.824616384506226,
      "epoch": 3.8364909550671076,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035859333124492866,
      "loss": 4.945,
      "mean_token_accuracy": 0.21742900013923644,
      "num_tokens": 85526787.0,
      "step": 49310
    },
    {
      "entropy": 5.846977472305298,
      "epoch": 3.8368799844388253,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003585675988333244,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21198030114173888,
      "num_tokens": 85534537.0,
      "step": 49315
    },
    {
      "entropy": 5.866365766525268,
      "epoch": 3.837269013810543,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003585418651537147,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2069234535098076,
      "num_tokens": 85543568.0,
      "step": 49320
    },
    {
      "entropy": 5.758477115631104,
      "epoch": 3.8376580431822602,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035851613020649036,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23072248995304107,
      "num_tokens": 85552201.0,
      "step": 49325
    },
    {
      "entropy": 5.853129434585571,
      "epoch": 3.838047072553978,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035849039399204157,
      "loss": 4.9358,
      "mean_token_accuracy": 0.2178171992301941,
      "num_tokens": 85560534.0,
      "step": 49330
    },
    {
      "entropy": 5.862930583953857,
      "epoch": 3.838436101925695,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035846465651075897,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2111928343772888,
      "num_tokens": 85569720.0,
      "step": 49335
    },
    {
      "entropy": 5.7834405422210695,
      "epoch": 3.838825131297413,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035843891776303304,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2101634606719017,
      "num_tokens": 85578864.0,
      "step": 49340
    },
    {
      "entropy": 5.853238677978515,
      "epoch": 3.8392141606691306,
      "grad_norm": 1.375,
      "learning_rate": 0.0003584131777492543,
      "loss": 5.0104,
      "mean_token_accuracy": 0.20767427533864974,
      "num_tokens": 85588025.0,
      "step": 49345
    },
    {
      "entropy": 5.909740543365478,
      "epoch": 3.8396031900408483,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003583874364698134,
      "loss": 4.959,
      "mean_token_accuracy": 0.21199358850717545,
      "num_tokens": 85596064.0,
      "step": 49350
    },
    {
      "entropy": 5.829646396636963,
      "epoch": 3.8399922194125655,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003583616939251009,
      "loss": 4.961,
      "mean_token_accuracy": 0.21160644441843032,
      "num_tokens": 85604309.0,
      "step": 49355
    },
    {
      "entropy": 5.782168197631836,
      "epoch": 3.840381248784283,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003583359501155073,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21318109929561616,
      "num_tokens": 85612531.0,
      "step": 49360
    },
    {
      "entropy": 5.7845906734466555,
      "epoch": 3.840770278156001,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003583102050414234,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22119367271661758,
      "num_tokens": 85620736.0,
      "step": 49365
    },
    {
      "entropy": 5.866454076766968,
      "epoch": 3.841159307527718,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035828445870323967,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2062646508216858,
      "num_tokens": 85629062.0,
      "step": 49370
    },
    {
      "entropy": 5.7712372779846195,
      "epoch": 3.841548336899436,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003582587111013469,
      "loss": 4.8397,
      "mean_token_accuracy": 0.21636093258857728,
      "num_tokens": 85637165.0,
      "step": 49375
    },
    {
      "entropy": 5.855878734588623,
      "epoch": 3.8419373662711536,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003582329622361356,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21684902012348176,
      "num_tokens": 85647161.0,
      "step": 49380
    },
    {
      "entropy": 5.857995796203613,
      "epoch": 3.8423263956428713,
      "grad_norm": 1.375,
      "learning_rate": 0.0003582072121079966,
      "loss": 4.9943,
      "mean_token_accuracy": 0.20402364134788514,
      "num_tokens": 85655554.0,
      "step": 49385
    },
    {
      "entropy": 5.830580615997315,
      "epoch": 3.8427154250145885,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035818146071732064,
      "loss": 4.9552,
      "mean_token_accuracy": 0.20660951882600784,
      "num_tokens": 85664087.0,
      "step": 49390
    },
    {
      "entropy": 5.84796724319458,
      "epoch": 3.843104454386306,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035815570806449845,
      "loss": 5.012,
      "mean_token_accuracy": 0.2118282586336136,
      "num_tokens": 85672607.0,
      "step": 49395
    },
    {
      "entropy": 5.94596266746521,
      "epoch": 3.8434934837580235,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035812995414992075,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20967138409614564,
      "num_tokens": 85681431.0,
      "step": 49400
    },
    {
      "entropy": 5.898207759857177,
      "epoch": 3.843882513129741,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035810419897397823,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21092324256896972,
      "num_tokens": 85689751.0,
      "step": 49405
    },
    {
      "entropy": 5.83615550994873,
      "epoch": 3.844271542501459,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003580784425370618,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2151494264602661,
      "num_tokens": 85697995.0,
      "step": 49410
    },
    {
      "entropy": 5.865628242492676,
      "epoch": 3.8446605718731766,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003580526848395622,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21529837399721147,
      "num_tokens": 85706695.0,
      "step": 49415
    },
    {
      "entropy": 5.870561265945435,
      "epoch": 3.8450496012448943,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035802692588187036,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21456843763589858,
      "num_tokens": 85714918.0,
      "step": 49420
    },
    {
      "entropy": 5.876997184753418,
      "epoch": 3.8454386306166115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000358001165664377,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20575084239244462,
      "num_tokens": 85723602.0,
      "step": 49425
    },
    {
      "entropy": 5.904080247879028,
      "epoch": 3.845827659988329,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003579754041874731,
      "loss": 5.1205,
      "mean_token_accuracy": 0.2040577307343483,
      "num_tokens": 85732838.0,
      "step": 49430
    },
    {
      "entropy": 5.823508453369141,
      "epoch": 3.8462166893600465,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003579496414515495,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21526552140712737,
      "num_tokens": 85741814.0,
      "step": 49435
    },
    {
      "entropy": 5.830187892913818,
      "epoch": 3.846605718731764,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035792387745699715,
      "loss": 4.9543,
      "mean_token_accuracy": 0.213832351565361,
      "num_tokens": 85750922.0,
      "step": 49440
    },
    {
      "entropy": 5.842682313919068,
      "epoch": 3.846994748103482,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003578981122042069,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20521206259727479,
      "num_tokens": 85758702.0,
      "step": 49445
    },
    {
      "entropy": 5.81285662651062,
      "epoch": 3.8473837774751996,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003578723456935698,
      "loss": 4.8833,
      "mean_token_accuracy": 0.21340780705213547,
      "num_tokens": 85767027.0,
      "step": 49450
    },
    {
      "entropy": 5.865794658660889,
      "epoch": 3.847772806846917,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003578465779254768,
      "loss": 4.9689,
      "mean_token_accuracy": 0.20797560065984727,
      "num_tokens": 85776043.0,
      "step": 49455
    },
    {
      "entropy": 5.862898492813111,
      "epoch": 3.8481618362186345,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035782080890031876,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20264691412448882,
      "num_tokens": 85785253.0,
      "step": 49460
    },
    {
      "entropy": 5.857070589065552,
      "epoch": 3.848550865590352,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035779503861848677,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21203736811876298,
      "num_tokens": 85793516.0,
      "step": 49465
    },
    {
      "entropy": 5.8103954792022705,
      "epoch": 3.8489398949620695,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003577692670803719,
      "loss": 4.8647,
      "mean_token_accuracy": 0.2175431340932846,
      "num_tokens": 85801852.0,
      "step": 49470
    },
    {
      "entropy": 5.713435697555542,
      "epoch": 3.849328924333787,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003577434942863651,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2161745622754097,
      "num_tokens": 85809857.0,
      "step": 49475
    },
    {
      "entropy": 5.742236804962158,
      "epoch": 3.849717953705505,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003577177202368575,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2152144655585289,
      "num_tokens": 85817999.0,
      "step": 49480
    },
    {
      "entropy": 5.796640682220459,
      "epoch": 3.8501069830772225,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035769194493224014,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20778066515922547,
      "num_tokens": 85827670.0,
      "step": 49485
    },
    {
      "entropy": 5.795357036590576,
      "epoch": 3.85049601244894,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035766616837290413,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2112990066409111,
      "num_tokens": 85836659.0,
      "step": 49490
    },
    {
      "entropy": 5.91450080871582,
      "epoch": 3.8508850418206575,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003576403905592406,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20775260031223297,
      "num_tokens": 85845250.0,
      "step": 49495
    },
    {
      "entropy": 5.89643030166626,
      "epoch": 3.8512740711923747,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035761461149164066,
      "loss": 5.0646,
      "mean_token_accuracy": 0.201773239672184,
      "num_tokens": 85854175.0,
      "step": 49500
    },
    {
      "entropy": 5.944066143035888,
      "epoch": 3.8516631005640924,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035758883117049543,
      "loss": 5.0719,
      "mean_token_accuracy": 0.21372891068458558,
      "num_tokens": 85863493.0,
      "step": 49505
    },
    {
      "entropy": 5.885462856292724,
      "epoch": 3.85205212993581,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003575630495961962,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21525711566209793,
      "num_tokens": 85872798.0,
      "step": 49510
    },
    {
      "entropy": 5.818344926834106,
      "epoch": 3.852441159307528,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035753726676913413,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21522509902715684,
      "num_tokens": 85881355.0,
      "step": 49515
    },
    {
      "entropy": 5.756139183044434,
      "epoch": 3.8528301886792455,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035751148268970046,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22025304138660431,
      "num_tokens": 85890384.0,
      "step": 49520
    },
    {
      "entropy": 5.915133094787597,
      "epoch": 3.853219218050963,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003574856973582863,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21780066043138505,
      "num_tokens": 85898965.0,
      "step": 49525
    },
    {
      "entropy": 5.838275003433227,
      "epoch": 3.8536082474226805,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035745991077528296,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2217951700091362,
      "num_tokens": 85907582.0,
      "step": 49530
    },
    {
      "entropy": 5.766354942321778,
      "epoch": 3.8539972767943977,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035743412294108175,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21845375150442123,
      "num_tokens": 85916343.0,
      "step": 49535
    },
    {
      "entropy": 5.795911359786987,
      "epoch": 3.8543863061661154,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003574083338560739,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21283060908317566,
      "num_tokens": 85925390.0,
      "step": 49540
    },
    {
      "entropy": 5.779561758041382,
      "epoch": 3.854775335537833,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035738254352065074,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2168794736266136,
      "num_tokens": 85934238.0,
      "step": 49545
    },
    {
      "entropy": 5.851829290390015,
      "epoch": 3.855164364909551,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035735675193520364,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21351455003023148,
      "num_tokens": 85942903.0,
      "step": 49550
    },
    {
      "entropy": 5.835073518753052,
      "epoch": 3.855553394281268,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003573309591001239,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21498206555843352,
      "num_tokens": 85951174.0,
      "step": 49555
    },
    {
      "entropy": 5.777992248535156,
      "epoch": 3.855942423652986,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000357305165015803,
      "loss": 4.9608,
      "mean_token_accuracy": 0.20714259147644043,
      "num_tokens": 85960685.0,
      "step": 49560
    },
    {
      "entropy": 5.878385829925537,
      "epoch": 3.8563314530247035,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035727936968263206,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2124515578150749,
      "num_tokens": 85968969.0,
      "step": 49565
    },
    {
      "entropy": 5.88637146949768,
      "epoch": 3.8567204823964207,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003572535731010028,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20684645175933838,
      "num_tokens": 85977576.0,
      "step": 49570
    },
    {
      "entropy": 5.736698961257934,
      "epoch": 3.8571095117681384,
      "grad_norm": 1.375,
      "learning_rate": 0.0003572277752713063,
      "loss": 4.833,
      "mean_token_accuracy": 0.22170543521642685,
      "num_tokens": 85986433.0,
      "step": 49575
    },
    {
      "entropy": 5.847989511489868,
      "epoch": 3.857498541139856,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003572019761939343,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2104301556944847,
      "num_tokens": 85995178.0,
      "step": 49580
    },
    {
      "entropy": 5.800891923904419,
      "epoch": 3.857887570511574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035717617586927814,
      "loss": 4.9465,
      "mean_token_accuracy": 0.21178001016378403,
      "num_tokens": 86004164.0,
      "step": 49585
    },
    {
      "entropy": 5.906930732727051,
      "epoch": 3.858276599883291,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035715037429772935,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2081627756357193,
      "num_tokens": 86012435.0,
      "step": 49590
    },
    {
      "entropy": 5.831538152694702,
      "epoch": 3.8586656292550088,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003571245714796793,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21529198288917542,
      "num_tokens": 86021646.0,
      "step": 49595
    },
    {
      "entropy": 5.77029595375061,
      "epoch": 3.8590546586267265,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003570987674155197,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22439627796411515,
      "num_tokens": 86029746.0,
      "step": 49600
    },
    {
      "entropy": 5.803494644165039,
      "epoch": 3.8594436879984437,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003570729621056419,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2105042576789856,
      "num_tokens": 86038652.0,
      "step": 49605
    },
    {
      "entropy": 5.826426553726196,
      "epoch": 3.8598327173701614,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035704715555043753,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20146183669567108,
      "num_tokens": 86048321.0,
      "step": 49610
    },
    {
      "entropy": 5.879402446746826,
      "epoch": 3.860221746741879,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035702134775029813,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20516550093889235,
      "num_tokens": 86057064.0,
      "step": 49615
    },
    {
      "entropy": 5.917349481582642,
      "epoch": 3.860610776113597,
      "grad_norm": 1.375,
      "learning_rate": 0.0003569955387056154,
      "loss": 5.0464,
      "mean_token_accuracy": 0.21385532170534133,
      "num_tokens": 86065436.0,
      "step": 49620
    },
    {
      "entropy": 5.7958574295043945,
      "epoch": 3.860999805485314,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035696972841678094,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21523911654949188,
      "num_tokens": 86074107.0,
      "step": 49625
    },
    {
      "entropy": 5.712298393249512,
      "epoch": 3.8613888348570318,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003569439168841862,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22960321009159088,
      "num_tokens": 86082822.0,
      "step": 49630
    },
    {
      "entropy": 5.8852170467376705,
      "epoch": 3.861777864228749,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035691810410822303,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2132394328713417,
      "num_tokens": 86092106.0,
      "step": 49635
    },
    {
      "entropy": 5.834230661392212,
      "epoch": 3.8621668936004667,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035689229008928304,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21115613877773284,
      "num_tokens": 86100583.0,
      "step": 49640
    },
    {
      "entropy": 5.84580020904541,
      "epoch": 3.8625559229721844,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035686647482775774,
      "loss": 4.9668,
      "mean_token_accuracy": 0.20675680935382842,
      "num_tokens": 86109257.0,
      "step": 49645
    },
    {
      "entropy": 5.838174200057983,
      "epoch": 3.862944952343902,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035684065832403903,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21273557394742965,
      "num_tokens": 86118731.0,
      "step": 49650
    },
    {
      "entropy": 5.787800264358521,
      "epoch": 3.8633339817156194,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003568148405785187,
      "loss": 4.8195,
      "mean_token_accuracy": 0.21997503489255904,
      "num_tokens": 86127602.0,
      "step": 49655
    },
    {
      "entropy": 5.7841590404510494,
      "epoch": 3.863723011087337,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003567890215915883,
      "loss": 4.901,
      "mean_token_accuracy": 0.22047497630119323,
      "num_tokens": 86135487.0,
      "step": 49660
    },
    {
      "entropy": 5.791139459609985,
      "epoch": 3.8641120404590548,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035676320136363975,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2100493147969246,
      "num_tokens": 86144252.0,
      "step": 49665
    },
    {
      "entropy": 5.866192579269409,
      "epoch": 3.864501069830772,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003567373798950647,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20873358547687532,
      "num_tokens": 86153311.0,
      "step": 49670
    },
    {
      "entropy": 5.913035249710083,
      "epoch": 3.8648900992024897,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035671155718625504,
      "loss": 5.042,
      "mean_token_accuracy": 0.20286408960819244,
      "num_tokens": 86161616.0,
      "step": 49675
    },
    {
      "entropy": 5.806729364395141,
      "epoch": 3.8652791285742074,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003566857332376025,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21394679695367813,
      "num_tokens": 86170213.0,
      "step": 49680
    },
    {
      "entropy": 5.81158561706543,
      "epoch": 3.865668157945925,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035665990804949904,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21919670104980468,
      "num_tokens": 86178459.0,
      "step": 49685
    },
    {
      "entropy": 5.822028255462646,
      "epoch": 3.8660571873176424,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003566340816223364,
      "loss": 4.9871,
      "mean_token_accuracy": 0.209971883893013,
      "num_tokens": 86186540.0,
      "step": 49690
    },
    {
      "entropy": 5.814668464660644,
      "epoch": 3.86644621668936,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003566082539565066,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20956696271896363,
      "num_tokens": 86195831.0,
      "step": 49695
    },
    {
      "entropy": 5.9142174243927,
      "epoch": 3.8668352460610778,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003565824250524013,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2060056820511818,
      "num_tokens": 86204767.0,
      "step": 49700
    },
    {
      "entropy": 5.864923763275146,
      "epoch": 3.867224275432795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035655659491041266,
      "loss": 4.933,
      "mean_token_accuracy": 0.21373645663261415,
      "num_tokens": 86213375.0,
      "step": 49705
    },
    {
      "entropy": 5.843689632415772,
      "epoch": 3.8676133048045127,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003565307635309325,
      "loss": 4.945,
      "mean_token_accuracy": 0.21380660086870193,
      "num_tokens": 86222378.0,
      "step": 49710
    },
    {
      "entropy": 5.807613754272461,
      "epoch": 3.8680023341762304,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003565049309143527,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2259746238589287,
      "num_tokens": 86230519.0,
      "step": 49715
    },
    {
      "entropy": 5.793789386749268,
      "epoch": 3.868391363547948,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035647909706106533,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2155417487025261,
      "num_tokens": 86238536.0,
      "step": 49720
    },
    {
      "entropy": 5.800478172302246,
      "epoch": 3.8687803929196654,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003564532619714624,
      "loss": 5.025,
      "mean_token_accuracy": 0.21009330302476883,
      "num_tokens": 86247309.0,
      "step": 49725
    },
    {
      "entropy": 5.840622186660767,
      "epoch": 3.869169422291383,
      "grad_norm": 1.375,
      "learning_rate": 0.0003564274256459358,
      "loss": 4.9721,
      "mean_token_accuracy": 0.20948131531476974,
      "num_tokens": 86255549.0,
      "step": 49730
    },
    {
      "entropy": 5.976945447921753,
      "epoch": 3.8695584516631003,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003564015880848777,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19420605152845383,
      "num_tokens": 86265868.0,
      "step": 49735
    },
    {
      "entropy": 5.928374910354615,
      "epoch": 3.869947481034818,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035637574928867996,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20456474721431733,
      "num_tokens": 86274504.0,
      "step": 49740
    },
    {
      "entropy": 5.807053136825561,
      "epoch": 3.8703365104065357,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035634990925773475,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2152481719851494,
      "num_tokens": 86283782.0,
      "step": 49745
    },
    {
      "entropy": 5.812450551986695,
      "epoch": 3.8707255397782534,
      "grad_norm": 1.375,
      "learning_rate": 0.0003563240679924342,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21476317048072815,
      "num_tokens": 86292100.0,
      "step": 49750
    },
    {
      "entropy": 5.8804114818572994,
      "epoch": 3.871114569149971,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003562982254931704,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20753551870584488,
      "num_tokens": 86300854.0,
      "step": 49755
    },
    {
      "entropy": 5.855142402648926,
      "epoch": 3.8715035985216883,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003562723817603354,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2071542352437973,
      "num_tokens": 86309297.0,
      "step": 49760
    },
    {
      "entropy": 5.798274564743042,
      "epoch": 3.871892627893406,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035624653679432135,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2090733528137207,
      "num_tokens": 86317820.0,
      "step": 49765
    },
    {
      "entropy": 5.833049011230469,
      "epoch": 3.8722816572651233,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00035622069059552036,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21946779489517212,
      "num_tokens": 86326393.0,
      "step": 49770
    },
    {
      "entropy": 5.9085118770599365,
      "epoch": 3.872670686636841,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003561948431643247,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21040442138910292,
      "num_tokens": 86334947.0,
      "step": 49775
    },
    {
      "entropy": 5.831426286697388,
      "epoch": 3.8730597160085587,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035616899450112655,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21110423952341079,
      "num_tokens": 86343842.0,
      "step": 49780
    },
    {
      "entropy": 5.814009189605713,
      "epoch": 3.8734487453802764,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035614314460631814,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20741496682167054,
      "num_tokens": 86352549.0,
      "step": 49785
    },
    {
      "entropy": 5.8714313983917235,
      "epoch": 3.8738377747519936,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003561172934802917,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21098192483186723,
      "num_tokens": 86361826.0,
      "step": 49790
    },
    {
      "entropy": 5.818069553375244,
      "epoch": 3.8742268041237113,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003560914411234393,
      "loss": 4.8461,
      "mean_token_accuracy": 0.21626058518886565,
      "num_tokens": 86370072.0,
      "step": 49795
    },
    {
      "entropy": 5.81378436088562,
      "epoch": 3.874615833495429,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003560655875361535,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21843957304954528,
      "num_tokens": 86379061.0,
      "step": 49800
    },
    {
      "entropy": 5.923106050491333,
      "epoch": 3.8750048628671463,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003560397327188263,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21071067005395888,
      "num_tokens": 86387704.0,
      "step": 49805
    },
    {
      "entropy": 5.844855308532715,
      "epoch": 3.875393892238864,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003560138766718502,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21367107033729554,
      "num_tokens": 86396740.0,
      "step": 49810
    },
    {
      "entropy": 5.879264211654663,
      "epoch": 3.8757829216105817,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035598801939561755,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21177878379821777,
      "num_tokens": 86405863.0,
      "step": 49815
    },
    {
      "entropy": 5.875940370559692,
      "epoch": 3.8761719509822994,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035596216089052045,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21486801505088807,
      "num_tokens": 86414696.0,
      "step": 49820
    },
    {
      "entropy": 5.76637830734253,
      "epoch": 3.8765609803540166,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035593630115695154,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21385105401277543,
      "num_tokens": 86423566.0,
      "step": 49825
    },
    {
      "entropy": 5.822409439086914,
      "epoch": 3.8769500097257343,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003559104401953031,
      "loss": 4.9797,
      "mean_token_accuracy": 0.20398910492658615,
      "num_tokens": 86431917.0,
      "step": 49830
    },
    {
      "entropy": 5.838342523574829,
      "epoch": 3.8773390390974516,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035588457800596747,
      "loss": 5.032,
      "mean_token_accuracy": 0.2066277638077736,
      "num_tokens": 86440259.0,
      "step": 49835
    },
    {
      "entropy": 5.8199059009552006,
      "epoch": 3.8777280684691693,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003558587145893371,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21519631296396255,
      "num_tokens": 86448873.0,
      "step": 49840
    },
    {
      "entropy": 5.890364503860473,
      "epoch": 3.878117097840887,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003558328499458044,
      "loss": 5.0248,
      "mean_token_accuracy": 0.20627732276916505,
      "num_tokens": 86457551.0,
      "step": 49845
    },
    {
      "entropy": 5.881229066848755,
      "epoch": 3.8785061272126047,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003558069840757619,
      "loss": 5.0217,
      "mean_token_accuracy": 0.20850950628519058,
      "num_tokens": 86467197.0,
      "step": 49850
    },
    {
      "entropy": 5.875233125686646,
      "epoch": 3.8788951565843224,
      "grad_norm": 1.453125,
      "learning_rate": 0.000355781116979602,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20769979059696198,
      "num_tokens": 86476635.0,
      "step": 49855
    },
    {
      "entropy": 5.94115834236145,
      "epoch": 3.8792841859560396,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003557552486577173,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20534246414899826,
      "num_tokens": 86485759.0,
      "step": 49860
    },
    {
      "entropy": 5.848342227935791,
      "epoch": 3.8796732153277573,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003557293791105002,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2171229287981987,
      "num_tokens": 86494165.0,
      "step": 49865
    },
    {
      "entropy": 5.801510810852051,
      "epoch": 3.8800622446994746,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035570350833834334,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22595129311084747,
      "num_tokens": 86502146.0,
      "step": 49870
    },
    {
      "entropy": 5.769648122787475,
      "epoch": 3.8804512740711923,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035567763634163904,
      "loss": 4.8293,
      "mean_token_accuracy": 0.2161725491285324,
      "num_tokens": 86510570.0,
      "step": 49875
    },
    {
      "entropy": 5.769085693359375,
      "epoch": 3.88084030344291,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003556517631207802,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21662116497755052,
      "num_tokens": 86519124.0,
      "step": 49880
    },
    {
      "entropy": 5.870197677612305,
      "epoch": 3.8812293328146277,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035562588867615907,
      "loss": 5.0164,
      "mean_token_accuracy": 0.20443173944950105,
      "num_tokens": 86527754.0,
      "step": 49885
    },
    {
      "entropy": 5.871813535690308,
      "epoch": 3.881618362186345,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035560001300816846,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21121466606855394,
      "num_tokens": 86536265.0,
      "step": 49890
    },
    {
      "entropy": 5.76010046005249,
      "epoch": 3.8820073915580626,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035557413611720095,
      "loss": 4.912,
      "mean_token_accuracy": 0.21914926618337632,
      "num_tokens": 86545845.0,
      "step": 49895
    },
    {
      "entropy": 5.783774185180664,
      "epoch": 3.8823964209297803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035554825800364917,
      "loss": 4.8197,
      "mean_token_accuracy": 0.2215958297252655,
      "num_tokens": 86554291.0,
      "step": 49900
    },
    {
      "entropy": 5.831540632247925,
      "epoch": 3.8827854503014976,
      "grad_norm": 1.375,
      "learning_rate": 0.0003555223786679058,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21128492206335067,
      "num_tokens": 86562913.0,
      "step": 49905
    },
    {
      "entropy": 5.859705209732056,
      "epoch": 3.8831744796732153,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035549649811036346,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20787952840328217,
      "num_tokens": 86571958.0,
      "step": 49910
    },
    {
      "entropy": 5.83790602684021,
      "epoch": 3.883563509044933,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035547061633141487,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22153096944093703,
      "num_tokens": 86580528.0,
      "step": 49915
    },
    {
      "entropy": 5.853646564483642,
      "epoch": 3.8839525384166507,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035544473333145275,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21165041178464888,
      "num_tokens": 86589777.0,
      "step": 49920
    },
    {
      "entropy": 5.9176534652709964,
      "epoch": 3.884341567788368,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003554188491108699,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21207495629787446,
      "num_tokens": 86598802.0,
      "step": 49925
    },
    {
      "entropy": 5.848093891143799,
      "epoch": 3.8847305971600856,
      "grad_norm": 1.328125,
      "learning_rate": 0.000355392963670059,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2093747541308403,
      "num_tokens": 86607353.0,
      "step": 49930
    },
    {
      "entropy": 5.931631755828858,
      "epoch": 3.8851196265318033,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003553670770094129,
      "loss": 5.0663,
      "mean_token_accuracy": 0.2058657705783844,
      "num_tokens": 86616891.0,
      "step": 49935
    },
    {
      "entropy": 5.911854600906372,
      "epoch": 3.8855086559035206,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035534118912932425,
      "loss": 5.022,
      "mean_token_accuracy": 0.2115153044462204,
      "num_tokens": 86626813.0,
      "step": 49940
    },
    {
      "entropy": 5.932025384902954,
      "epoch": 3.8858976852752383,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035531530003018584,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20784884989261626,
      "num_tokens": 86635641.0,
      "step": 49945
    },
    {
      "entropy": 5.8615058898925785,
      "epoch": 3.886286714646956,
      "grad_norm": 1.375,
      "learning_rate": 0.0003552894097123907,
      "loss": 4.9502,
      "mean_token_accuracy": 0.20857900977134705,
      "num_tokens": 86644127.0,
      "step": 49950
    },
    {
      "entropy": 5.8826335906982425,
      "epoch": 3.8866757440186737,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035526351817633157,
      "loss": 4.9358,
      "mean_token_accuracy": 0.21575309187173844,
      "num_tokens": 86652516.0,
      "step": 49955
    },
    {
      "entropy": 5.838472032546997,
      "epoch": 3.887064773390391,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003552376254224013,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21493494063615798,
      "num_tokens": 86661004.0,
      "step": 49960
    },
    {
      "entropy": 5.828317022323608,
      "epoch": 3.8874538027621086,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003552117314509928,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21164486706256866,
      "num_tokens": 86669533.0,
      "step": 49965
    },
    {
      "entropy": 5.882344341278076,
      "epoch": 3.887842832133826,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003551858362624989,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2237582191824913,
      "num_tokens": 86677476.0,
      "step": 49970
    },
    {
      "entropy": 5.883181428909301,
      "epoch": 3.8882318615055436,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003551599398573125,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21064640134572982,
      "num_tokens": 86685271.0,
      "step": 49975
    },
    {
      "entropy": 5.835028791427613,
      "epoch": 3.8886208908772613,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035513404223582666,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20722855180501937,
      "num_tokens": 86694015.0,
      "step": 49980
    },
    {
      "entropy": 5.8857497215271,
      "epoch": 3.889009920248979,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035510814339843427,
      "loss": 5.0084,
      "mean_token_accuracy": 0.20730988532304764,
      "num_tokens": 86702436.0,
      "step": 49985
    },
    {
      "entropy": 5.832481861114502,
      "epoch": 3.889398949620696,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035508224334552835,
      "loss": 4.9665,
      "mean_token_accuracy": 0.20712581276893616,
      "num_tokens": 86711090.0,
      "step": 49990
    },
    {
      "entropy": 5.804567146301269,
      "epoch": 3.889787978992414,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003550563420775018,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2130318194627762,
      "num_tokens": 86720532.0,
      "step": 49995
    },
    {
      "entropy": 5.815111875534058,
      "epoch": 3.8901770083641316,
      "grad_norm": 1.25,
      "learning_rate": 0.00035503043959474775,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22177657186985017,
      "num_tokens": 86728998.0,
      "step": 50000
    },
    {
      "entropy": 5.836471271514893,
      "epoch": 3.890566037735849,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035500453589765906,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21311963945627213,
      "num_tokens": 86737186.0,
      "step": 50005
    },
    {
      "entropy": 5.760082578659057,
      "epoch": 3.8909550671075666,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035497863098662893,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21197385489940643,
      "num_tokens": 86745453.0,
      "step": 50010
    },
    {
      "entropy": 5.791597080230713,
      "epoch": 3.8913440964792843,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035495272486205033,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21475330442190171,
      "num_tokens": 86754463.0,
      "step": 50015
    },
    {
      "entropy": 5.876528978347778,
      "epoch": 3.891733125851002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003549268175243165,
      "loss": 4.987,
      "mean_token_accuracy": 0.21643628776073456,
      "num_tokens": 86763159.0,
      "step": 50020
    },
    {
      "entropy": 5.815583324432373,
      "epoch": 3.892122155222719,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003549009089738204,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2105148121714592,
      "num_tokens": 86772194.0,
      "step": 50025
    },
    {
      "entropy": 5.835250902175903,
      "epoch": 3.892511184594437,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003548749992109551,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2166237637400627,
      "num_tokens": 86780508.0,
      "step": 50030
    },
    {
      "entropy": 5.881607151031494,
      "epoch": 3.8929002139661546,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003548490882361139,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21138374507427216,
      "num_tokens": 86788793.0,
      "step": 50035
    },
    {
      "entropy": 5.81362566947937,
      "epoch": 3.893289243337872,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003548231760496898,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20768265575170516,
      "num_tokens": 86797947.0,
      "step": 50040
    },
    {
      "entropy": 5.862573957443237,
      "epoch": 3.8936782727095895,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035479726265207604,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2016950234770775,
      "num_tokens": 86807325.0,
      "step": 50045
    },
    {
      "entropy": 5.825527191162109,
      "epoch": 3.8940673020813072,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003547713480436659,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21670743077993393,
      "num_tokens": 86816436.0,
      "step": 50050
    },
    {
      "entropy": 5.880147171020508,
      "epoch": 3.894456331453025,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035474543222485254,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2096487522125244,
      "num_tokens": 86824554.0,
      "step": 50055
    },
    {
      "entropy": 5.827675914764404,
      "epoch": 3.894845360824742,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035471951519602926,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2171166568994522,
      "num_tokens": 86833096.0,
      "step": 50060
    },
    {
      "entropy": 5.865033435821533,
      "epoch": 3.89523439019646,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035469359695758905,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2047283336520195,
      "num_tokens": 86841921.0,
      "step": 50065
    },
    {
      "entropy": 5.794017601013183,
      "epoch": 3.895623419568177,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003546676775099254,
      "loss": 4.8617,
      "mean_token_accuracy": 0.21783996522426605,
      "num_tokens": 86850244.0,
      "step": 50070
    },
    {
      "entropy": 5.799462699890137,
      "epoch": 3.896012448939895,
      "grad_norm": 1.375,
      "learning_rate": 0.0003546417568534315,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21636681854724885,
      "num_tokens": 86858445.0,
      "step": 50075
    },
    {
      "entropy": 5.752597284317017,
      "epoch": 3.8964014783116125,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035461583498850083,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22186903208494185,
      "num_tokens": 86867496.0,
      "step": 50080
    },
    {
      "entropy": 5.77060112953186,
      "epoch": 3.8967905076833302,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003545899119155265,
      "loss": 4.8359,
      "mean_token_accuracy": 0.21478314101696014,
      "num_tokens": 86876756.0,
      "step": 50085
    },
    {
      "entropy": 5.845057678222656,
      "epoch": 3.8971795370550475,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035456398763490193,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21137607842683792,
      "num_tokens": 86886125.0,
      "step": 50090
    },
    {
      "entropy": 5.7446348667144775,
      "epoch": 3.897568566426765,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003545380621470205,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2297294795513153,
      "num_tokens": 86895032.0,
      "step": 50095
    },
    {
      "entropy": 5.869444561004639,
      "epoch": 3.897957595798483,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003545121354522756,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1998348355293274,
      "num_tokens": 86904050.0,
      "step": 50100
    },
    {
      "entropy": 5.8282702445983885,
      "epoch": 3.8983466251702,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035448620755106064,
      "loss": 4.974,
      "mean_token_accuracy": 0.2116552010178566,
      "num_tokens": 86912603.0,
      "step": 50105
    },
    {
      "entropy": 5.848015546798706,
      "epoch": 3.898735654541918,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003544602784437689,
      "loss": 4.946,
      "mean_token_accuracy": 0.2169031336903572,
      "num_tokens": 86921693.0,
      "step": 50110
    },
    {
      "entropy": 5.829853057861328,
      "epoch": 3.8991246839136355,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035443434813079393,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2153860241174698,
      "num_tokens": 86929975.0,
      "step": 50115
    },
    {
      "entropy": 5.789067792892456,
      "epoch": 3.8995137132853532,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003544084166125292,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22021604180336,
      "num_tokens": 86938184.0,
      "step": 50120
    },
    {
      "entropy": 5.870650959014893,
      "epoch": 3.8999027426570705,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003543824838893682,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20872282981872559,
      "num_tokens": 86946849.0,
      "step": 50125
    },
    {
      "entropy": 5.852821922302246,
      "epoch": 3.900291772028788,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035435654996170417,
      "loss": 5.0082,
      "mean_token_accuracy": 0.20393765419721605,
      "num_tokens": 86954911.0,
      "step": 50130
    },
    {
      "entropy": 5.817497587203979,
      "epoch": 3.900680801400506,
      "grad_norm": 1.28125,
      "learning_rate": 0.000354330614829931,
      "loss": 4.9606,
      "mean_token_accuracy": 0.20945875644683837,
      "num_tokens": 86963270.0,
      "step": 50135
    },
    {
      "entropy": 5.853619337081909,
      "epoch": 3.901069830772223,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003543046784944419,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22207652330398558,
      "num_tokens": 86971600.0,
      "step": 50140
    },
    {
      "entropy": 5.853485584259033,
      "epoch": 3.901458860143941,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035427874095563054,
      "loss": 4.969,
      "mean_token_accuracy": 0.2131533905863762,
      "num_tokens": 86979743.0,
      "step": 50145
    },
    {
      "entropy": 5.894322061538697,
      "epoch": 3.9018478895156585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003542528022138905,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21634818017482757,
      "num_tokens": 86988652.0,
      "step": 50150
    },
    {
      "entropy": 5.873696374893188,
      "epoch": 3.9022369188873762,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003542268622696153,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20724546015262604,
      "num_tokens": 86997070.0,
      "step": 50155
    },
    {
      "entropy": 5.736970710754394,
      "epoch": 3.9026259482590935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003542009211231986,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21894647777080536,
      "num_tokens": 87005765.0,
      "step": 50160
    },
    {
      "entropy": 5.836826181411743,
      "epoch": 3.903014977630811,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035417497877503396,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2072973892092705,
      "num_tokens": 87014506.0,
      "step": 50165
    },
    {
      "entropy": 5.8156575679779055,
      "epoch": 3.9034040070025284,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035414903522551505,
      "loss": 4.9385,
      "mean_token_accuracy": 0.21120989471673965,
      "num_tokens": 87023666.0,
      "step": 50170
    },
    {
      "entropy": 5.852911376953125,
      "epoch": 3.903793036374246,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035412309047503553,
      "loss": 4.9367,
      "mean_token_accuracy": 0.2155543938279152,
      "num_tokens": 87031942.0,
      "step": 50175
    },
    {
      "entropy": 5.815826177597046,
      "epoch": 3.904182065745964,
      "grad_norm": 1.328125,
      "learning_rate": 0.000354097144523989,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21862224787473677,
      "num_tokens": 87041029.0,
      "step": 50180
    },
    {
      "entropy": 5.835499382019043,
      "epoch": 3.9045710951176815,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003540711973727693,
      "loss": 5.019,
      "mean_token_accuracy": 0.20608026832342147,
      "num_tokens": 87050329.0,
      "step": 50185
    },
    {
      "entropy": 5.859791564941406,
      "epoch": 3.904960124489399,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003540452490217699,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21054288148880004,
      "num_tokens": 87059092.0,
      "step": 50190
    },
    {
      "entropy": 5.857463598251343,
      "epoch": 3.9053491538611165,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003540192994713847,
      "loss": 4.9895,
      "mean_token_accuracy": 0.207786463201046,
      "num_tokens": 87067514.0,
      "step": 50195
    },
    {
      "entropy": 5.852554178237915,
      "epoch": 3.905738183232834,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035399334872200747,
      "loss": 5.0545,
      "mean_token_accuracy": 0.20145612061023713,
      "num_tokens": 87077404.0,
      "step": 50200
    },
    {
      "entropy": 5.887202501296997,
      "epoch": 3.9061272126045514,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003539673967740318,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20996606796979905,
      "num_tokens": 87086406.0,
      "step": 50205
    },
    {
      "entropy": 5.797400045394897,
      "epoch": 3.906516241976269,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003539414436278516,
      "loss": 4.8319,
      "mean_token_accuracy": 0.21830499768257142,
      "num_tokens": 87095029.0,
      "step": 50210
    },
    {
      "entropy": 5.882325077056885,
      "epoch": 3.906905271347987,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035391548928386077,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21221245527267457,
      "num_tokens": 87104412.0,
      "step": 50215
    },
    {
      "entropy": 5.824559354782105,
      "epoch": 3.9072943007197045,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003538895337424529,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21592774242162704,
      "num_tokens": 87113383.0,
      "step": 50220
    },
    {
      "entropy": 5.864041423797607,
      "epoch": 3.9076833300914218,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035386357700402187,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21227518916130067,
      "num_tokens": 87121150.0,
      "step": 50225
    },
    {
      "entropy": 5.762804794311523,
      "epoch": 3.9080723594631395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003538376190689616,
      "loss": 4.913,
      "mean_token_accuracy": 0.21356281489133835,
      "num_tokens": 87129828.0,
      "step": 50230
    },
    {
      "entropy": 5.8977819919586185,
      "epoch": 3.908461388834857,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003538116599376661,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2103077471256256,
      "num_tokens": 87138618.0,
      "step": 50235
    },
    {
      "entropy": 5.85614857673645,
      "epoch": 3.9088504182065744,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035378569961052887,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20946453064680098,
      "num_tokens": 87147576.0,
      "step": 50240
    },
    {
      "entropy": 5.867601442337036,
      "epoch": 3.909239447578292,
      "grad_norm": 1.375,
      "learning_rate": 0.00035375973808794416,
      "loss": 4.8995,
      "mean_token_accuracy": 0.21714259684085846,
      "num_tokens": 87156067.0,
      "step": 50245
    },
    {
      "entropy": 5.840759706497193,
      "epoch": 3.90962847695001,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003537337753703058,
      "loss": 4.9441,
      "mean_token_accuracy": 0.21185347586870193,
      "num_tokens": 87164127.0,
      "step": 50250
    },
    {
      "entropy": 5.812650203704834,
      "epoch": 3.9100175063217275,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003537078114580077,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21386137008666992,
      "num_tokens": 87172143.0,
      "step": 50255
    },
    {
      "entropy": 5.770189428329468,
      "epoch": 3.9104065356934448,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003536818463514438,
      "loss": 4.9703,
      "mean_token_accuracy": 0.20575565695762635,
      "num_tokens": 87180565.0,
      "step": 50260
    },
    {
      "entropy": 5.838504695892334,
      "epoch": 3.9107955650651625,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035365588005100817,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21423009932041168,
      "num_tokens": 87189589.0,
      "step": 50265
    },
    {
      "entropy": 5.818747806549072,
      "epoch": 3.9111845944368797,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003536299125570947,
      "loss": 4.947,
      "mean_token_accuracy": 0.2088731735944748,
      "num_tokens": 87197730.0,
      "step": 50270
    },
    {
      "entropy": 5.784960031509399,
      "epoch": 3.9115736238085974,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003536039438700975,
      "loss": 4.987,
      "mean_token_accuracy": 0.2069752722978592,
      "num_tokens": 87206558.0,
      "step": 50275
    },
    {
      "entropy": 5.819164180755616,
      "epoch": 3.911962653180315,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035357797399041056,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21423240900039672,
      "num_tokens": 87215250.0,
      "step": 50280
    },
    {
      "entropy": 5.852763366699219,
      "epoch": 3.912351682552033,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003535520029184279,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20311967581510543,
      "num_tokens": 87223964.0,
      "step": 50285
    },
    {
      "entropy": 5.8105058670043945,
      "epoch": 3.9127407119237505,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003535260306545437,
      "loss": 4.9399,
      "mean_token_accuracy": 0.20938710123300552,
      "num_tokens": 87232988.0,
      "step": 50290
    },
    {
      "entropy": 5.865932607650757,
      "epoch": 3.9131297412954678,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035350005719915196,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21497756540775298,
      "num_tokens": 87241582.0,
      "step": 50295
    },
    {
      "entropy": 5.822469711303711,
      "epoch": 3.9135187706671855,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035347408255264674,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2128775492310524,
      "num_tokens": 87249963.0,
      "step": 50300
    },
    {
      "entropy": 5.847865104675293,
      "epoch": 3.9139078000389027,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003534481067154223,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2028522565960884,
      "num_tokens": 87258825.0,
      "step": 50305
    },
    {
      "entropy": 5.795261812210083,
      "epoch": 3.9142968294106204,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003534221296878726,
      "loss": 4.878,
      "mean_token_accuracy": 0.21622084677219391,
      "num_tokens": 87268335.0,
      "step": 50310
    },
    {
      "entropy": 5.879649639129639,
      "epoch": 3.914685858782338,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035339615147039196,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20854121595621108,
      "num_tokens": 87276404.0,
      "step": 50315
    },
    {
      "entropy": 5.885884141921997,
      "epoch": 3.915074888154056,
      "grad_norm": 1.25,
      "learning_rate": 0.00035337017206337444,
      "loss": 5.0359,
      "mean_token_accuracy": 0.20820137709379197,
      "num_tokens": 87285735.0,
      "step": 50320
    },
    {
      "entropy": 5.913195037841797,
      "epoch": 3.915463917525773,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003533441914672144,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20526523292064666,
      "num_tokens": 87295030.0,
      "step": 50325
    },
    {
      "entropy": 5.867482900619507,
      "epoch": 3.9158529468974907,
      "grad_norm": 1.375,
      "learning_rate": 0.00035331820968230595,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20309984236955642,
      "num_tokens": 87303530.0,
      "step": 50330
    },
    {
      "entropy": 5.854580640792847,
      "epoch": 3.9162419762692084,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035329222670904325,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21145646423101425,
      "num_tokens": 87312484.0,
      "step": 50335
    },
    {
      "entropy": 5.817336988449097,
      "epoch": 3.9166310056409257,
      "grad_norm": 2.1875,
      "learning_rate": 0.0003532662425478206,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22226296365261078,
      "num_tokens": 87321978.0,
      "step": 50340
    },
    {
      "entropy": 5.814772844314575,
      "epoch": 3.9170200350126434,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035324025719903236,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21494407951831818,
      "num_tokens": 87330819.0,
      "step": 50345
    },
    {
      "entropy": 5.8816934585571286,
      "epoch": 3.917409064384361,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035321427066307276,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20901056081056596,
      "num_tokens": 87338799.0,
      "step": 50350
    },
    {
      "entropy": 5.838096761703492,
      "epoch": 3.917798093756079,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003531882829403361,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21320050954818726,
      "num_tokens": 87347317.0,
      "step": 50355
    },
    {
      "entropy": 5.799908828735352,
      "epoch": 3.918187123127796,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035316229403121666,
      "loss": 4.9025,
      "mean_token_accuracy": 0.21220894753932953,
      "num_tokens": 87355742.0,
      "step": 50360
    },
    {
      "entropy": 5.805998086929321,
      "epoch": 3.9185761524995137,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003531363039361088,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21865929216146468,
      "num_tokens": 87364774.0,
      "step": 50365
    },
    {
      "entropy": 5.8032721996307375,
      "epoch": 3.9189651818712314,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035311031265540695,
      "loss": 4.9218,
      "mean_token_accuracy": 0.21041710972785949,
      "num_tokens": 87373799.0,
      "step": 50370
    },
    {
      "entropy": 5.887723827362061,
      "epoch": 3.9193542112429487,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035308432018950537,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20953630656003952,
      "num_tokens": 87382732.0,
      "step": 50375
    },
    {
      "entropy": 5.859270477294922,
      "epoch": 3.9197432406146664,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035305832653879856,
      "loss": 4.9407,
      "mean_token_accuracy": 0.20847305357456208,
      "num_tokens": 87391452.0,
      "step": 50380
    },
    {
      "entropy": 5.837453317642212,
      "epoch": 3.920132269986384,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035303233170368083,
      "loss": 4.8687,
      "mean_token_accuracy": 0.21647530794143677,
      "num_tokens": 87399998.0,
      "step": 50385
    },
    {
      "entropy": 5.877296352386475,
      "epoch": 3.920521299358102,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035300633568454673,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2190970942378044,
      "num_tokens": 87408558.0,
      "step": 50390
    },
    {
      "entropy": 5.798892831802368,
      "epoch": 3.920910328729819,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035298033848179064,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21400791704654692,
      "num_tokens": 87416715.0,
      "step": 50395
    },
    {
      "entropy": 5.84766263961792,
      "epoch": 3.9212993581015367,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000352954340095807,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2022629678249359,
      "num_tokens": 87425036.0,
      "step": 50400
    },
    {
      "entropy": 5.789225673675537,
      "epoch": 3.921688387473254,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035292834052699036,
      "loss": 4.9156,
      "mean_token_accuracy": 0.21825195401906966,
      "num_tokens": 87433732.0,
      "step": 50405
    },
    {
      "entropy": 5.8573681831359865,
      "epoch": 3.9220774168449717,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003529023397757351,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2049235463142395,
      "num_tokens": 87443088.0,
      "step": 50410
    },
    {
      "entropy": 5.773188209533691,
      "epoch": 3.9224664462166894,
      "grad_norm": 1.375,
      "learning_rate": 0.00035287633784243595,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21383998841047286,
      "num_tokens": 87452568.0,
      "step": 50415
    },
    {
      "entropy": 5.8922370910644535,
      "epoch": 3.922855475588407,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00035285033472748725,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2167195722460747,
      "num_tokens": 87460524.0,
      "step": 50420
    },
    {
      "entropy": 5.813318347930908,
      "epoch": 3.9232445049601243,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003528243304312837,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2119239166378975,
      "num_tokens": 87469863.0,
      "step": 50425
    },
    {
      "entropy": 5.826010227203369,
      "epoch": 3.923633534331842,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003527983249542197,
      "loss": 5.0001,
      "mean_token_accuracy": 0.1991504520177841,
      "num_tokens": 87478026.0,
      "step": 50430
    },
    {
      "entropy": 5.810401248931885,
      "epoch": 3.9240225637035597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035277231829669006,
      "loss": 4.8978,
      "mean_token_accuracy": 0.21176282465457916,
      "num_tokens": 87487267.0,
      "step": 50435
    },
    {
      "entropy": 5.8285423755645756,
      "epoch": 3.924411593075277,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003527463104590892,
      "loss": 4.9647,
      "mean_token_accuracy": 0.20723945796489715,
      "num_tokens": 87496393.0,
      "step": 50440
    },
    {
      "entropy": 5.887448120117187,
      "epoch": 3.9248006224469947,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003527203014418119,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2097357526421547,
      "num_tokens": 87504214.0,
      "step": 50445
    },
    {
      "entropy": 5.826976919174195,
      "epoch": 3.9251896518187124,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003526942912452526,
      "loss": 4.9776,
      "mean_token_accuracy": 0.20684325098991393,
      "num_tokens": 87512837.0,
      "step": 50450
    },
    {
      "entropy": 5.834858798980713,
      "epoch": 3.92557868119043,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003526682798698062,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2075273260474205,
      "num_tokens": 87521107.0,
      "step": 50455
    },
    {
      "entropy": 5.838110637664795,
      "epoch": 3.9259677105621473,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003526422673158673,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21361607015132905,
      "num_tokens": 87530089.0,
      "step": 50460
    },
    {
      "entropy": 5.773072004318237,
      "epoch": 3.926356739933865,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003526162535838305,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21209748834371567,
      "num_tokens": 87539038.0,
      "step": 50465
    },
    {
      "entropy": 5.831792163848877,
      "epoch": 3.9267457693055827,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035259023867409065,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2101265773177147,
      "num_tokens": 87547760.0,
      "step": 50470
    },
    {
      "entropy": 5.754550457000732,
      "epoch": 3.9271347986773,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003525642225870423,
      "loss": 4.8682,
      "mean_token_accuracy": 0.222468826174736,
      "num_tokens": 87556366.0,
      "step": 50475
    },
    {
      "entropy": 5.7899620056152346,
      "epoch": 3.9275238280490177,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035253820532308043,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21100423932075502,
      "num_tokens": 87564411.0,
      "step": 50480
    },
    {
      "entropy": 5.842006349563599,
      "epoch": 3.9279128574207354,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003525121868825997,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20795863419771193,
      "num_tokens": 87572635.0,
      "step": 50485
    },
    {
      "entropy": 5.9257489204406735,
      "epoch": 3.928301886792453,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003524861672659949,
      "loss": 5.0312,
      "mean_token_accuracy": 0.2089160516858101,
      "num_tokens": 87580969.0,
      "step": 50490
    },
    {
      "entropy": 5.932268857955933,
      "epoch": 3.9286909161641703,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003524601464736609,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21564121544361115,
      "num_tokens": 87589477.0,
      "step": 50495
    },
    {
      "entropy": 5.909557342529297,
      "epoch": 3.929079945535888,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003524341245059924,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20403924137353896,
      "num_tokens": 87598047.0,
      "step": 50500
    },
    {
      "entropy": 5.802220392227173,
      "epoch": 3.9294689749076053,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003524081013633843,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21648037880659105,
      "num_tokens": 87606591.0,
      "step": 50505
    },
    {
      "entropy": 5.840164756774902,
      "epoch": 3.929858004279323,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035238207704623147,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21136913895606996,
      "num_tokens": 87615087.0,
      "step": 50510
    },
    {
      "entropy": 5.82448525428772,
      "epoch": 3.9302470336510407,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035235605155492885,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21746322959661485,
      "num_tokens": 87623397.0,
      "step": 50515
    },
    {
      "entropy": 5.880098628997803,
      "epoch": 3.9306360630227584,
      "grad_norm": 1.53125,
      "learning_rate": 0.00035233002488987123,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2122628077864647,
      "num_tokens": 87632064.0,
      "step": 50520
    },
    {
      "entropy": 5.851122140884399,
      "epoch": 3.9310250923944756,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035230399705145365,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21467628628015517,
      "num_tokens": 87640300.0,
      "step": 50525
    },
    {
      "entropy": 5.811364603042603,
      "epoch": 3.9314141217661933,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003522779680400709,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21206727623939514,
      "num_tokens": 87648691.0,
      "step": 50530
    },
    {
      "entropy": 5.875855159759522,
      "epoch": 3.931803151137911,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035225193785611795,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2076011285185814,
      "num_tokens": 87657306.0,
      "step": 50535
    },
    {
      "entropy": 5.79010648727417,
      "epoch": 3.9321921805096283,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003522259064999898,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2164313480257988,
      "num_tokens": 87666347.0,
      "step": 50540
    },
    {
      "entropy": 5.793262100219726,
      "epoch": 3.932581209881346,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00035219987397208137,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22563455104827881,
      "num_tokens": 87674328.0,
      "step": 50545
    },
    {
      "entropy": 5.8699572563171385,
      "epoch": 3.9329702392530637,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003521738402727878,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2102842926979065,
      "num_tokens": 87682769.0,
      "step": 50550
    },
    {
      "entropy": 5.823321056365967,
      "epoch": 3.9333592686247814,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003521478054025041,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21355315595865249,
      "num_tokens": 87691732.0,
      "step": 50555
    },
    {
      "entropy": 5.896574354171753,
      "epoch": 3.9337482979964986,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035212176936162524,
      "loss": 5.0449,
      "mean_token_accuracy": 0.2064281240105629,
      "num_tokens": 87699634.0,
      "step": 50560
    },
    {
      "entropy": 5.846316003799439,
      "epoch": 3.9341373273682163,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003520957321505462,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21642001271247863,
      "num_tokens": 87708807.0,
      "step": 50565
    },
    {
      "entropy": 5.855564022064209,
      "epoch": 3.934526356739934,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035206969376966226,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2092100828886032,
      "num_tokens": 87717338.0,
      "step": 50570
    },
    {
      "entropy": 5.821547269821167,
      "epoch": 3.9349153861116513,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035204365421936826,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2109060063958168,
      "num_tokens": 87726021.0,
      "step": 50575
    },
    {
      "entropy": 5.825818395614624,
      "epoch": 3.935304415483369,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003520176135000594,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22389100044965743,
      "num_tokens": 87735023.0,
      "step": 50580
    },
    {
      "entropy": 5.858525514602661,
      "epoch": 3.9356934448550867,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003519915716121309,
      "loss": 4.9995,
      "mean_token_accuracy": 0.20615150034427643,
      "num_tokens": 87743220.0,
      "step": 50585
    },
    {
      "entropy": 5.860691785812378,
      "epoch": 3.9360824742268044,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003519655285559779,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21434687823057175,
      "num_tokens": 87751772.0,
      "step": 50590
    },
    {
      "entropy": 5.844260120391846,
      "epoch": 3.9364715035985216,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035193948433199547,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21392036229372025,
      "num_tokens": 87760518.0,
      "step": 50595
    },
    {
      "entropy": 5.929533338546753,
      "epoch": 3.9368605329702393,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035191343894057885,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20459677129983903,
      "num_tokens": 87768649.0,
      "step": 50600
    },
    {
      "entropy": 5.845632839202881,
      "epoch": 3.9372495623419566,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035188739238212316,
      "loss": 4.9321,
      "mean_token_accuracy": 0.21569954305887223,
      "num_tokens": 87776908.0,
      "step": 50605
    },
    {
      "entropy": 5.882026147842407,
      "epoch": 3.9376385917136743,
      "grad_norm": 1.375,
      "learning_rate": 0.0003518613446570237,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21475882828235626,
      "num_tokens": 87785271.0,
      "step": 50610
    },
    {
      "entropy": 5.785317039489746,
      "epoch": 3.938027621085392,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035183529576567567,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21779724955558777,
      "num_tokens": 87794290.0,
      "step": 50615
    },
    {
      "entropy": 5.853997850418091,
      "epoch": 3.9384166504571096,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003518092457084743,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21409939974546432,
      "num_tokens": 87802574.0,
      "step": 50620
    },
    {
      "entropy": 5.883027219772339,
      "epoch": 3.9388056798288273,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003517831944858149,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2075902223587036,
      "num_tokens": 87810825.0,
      "step": 50625
    },
    {
      "entropy": 5.745425367355347,
      "epoch": 3.9391947092005446,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035175714209809275,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2138733074069023,
      "num_tokens": 87818624.0,
      "step": 50630
    },
    {
      "entropy": 5.77688889503479,
      "epoch": 3.9395837385722623,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003517310885457032,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2247217983007431,
      "num_tokens": 87826973.0,
      "step": 50635
    },
    {
      "entropy": 5.823399353027344,
      "epoch": 3.9399727679439795,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003517050338290414,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21422916352748872,
      "num_tokens": 87835381.0,
      "step": 50640
    },
    {
      "entropy": 5.8388830661773685,
      "epoch": 3.9403617973156972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035167897794850285,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22009167373180388,
      "num_tokens": 87843898.0,
      "step": 50645
    },
    {
      "entropy": 5.817836713790894,
      "epoch": 3.940750826687415,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035165292090448287,
      "loss": 4.9428,
      "mean_token_accuracy": 0.20568664371967316,
      "num_tokens": 87852600.0,
      "step": 50650
    },
    {
      "entropy": 5.856058406829834,
      "epoch": 3.9411398560591326,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00035162686269737676,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2120719626545906,
      "num_tokens": 87860472.0,
      "step": 50655
    },
    {
      "entropy": 5.886384105682373,
      "epoch": 3.94152888543085,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035160080332758006,
      "loss": 4.9335,
      "mean_token_accuracy": 0.20827173739671706,
      "num_tokens": 87869259.0,
      "step": 50660
    },
    {
      "entropy": 5.846927118301392,
      "epoch": 3.9419179148025676,
      "grad_norm": 1.390625,
      "learning_rate": 0.000351574742795488,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21357918679714202,
      "num_tokens": 87877971.0,
      "step": 50665
    },
    {
      "entropy": 5.815457153320312,
      "epoch": 3.9423069441742853,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003515486811014962,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20273015201091765,
      "num_tokens": 87886837.0,
      "step": 50670
    },
    {
      "entropy": 5.88142580986023,
      "epoch": 3.9426959735460025,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003515226182459999,
      "loss": 5.036,
      "mean_token_accuracy": 0.2064157798886299,
      "num_tokens": 87895971.0,
      "step": 50675
    },
    {
      "entropy": 5.8253381729125975,
      "epoch": 3.9430850029177202,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035149655422939477,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21240597665309907,
      "num_tokens": 87905096.0,
      "step": 50680
    },
    {
      "entropy": 5.88958511352539,
      "epoch": 3.943474032289438,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003514704890520761,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2108414053916931,
      "num_tokens": 87914093.0,
      "step": 50685
    },
    {
      "entropy": 5.900439119338989,
      "epoch": 3.9438630616611556,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003514444227144395,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2029083713889122,
      "num_tokens": 87922724.0,
      "step": 50690
    },
    {
      "entropy": 5.897398471832275,
      "epoch": 3.944252091032873,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003514183552168804,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20852955430746078,
      "num_tokens": 87932478.0,
      "step": 50695
    },
    {
      "entropy": 5.761565446853638,
      "epoch": 3.9446411204045906,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035139228655979457,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22171662598848343,
      "num_tokens": 87941437.0,
      "step": 50700
    },
    {
      "entropy": 5.876450252532959,
      "epoch": 3.945030149776308,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035136621674357717,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2115311086177826,
      "num_tokens": 87950317.0,
      "step": 50705
    },
    {
      "entropy": 5.824373006820679,
      "epoch": 3.9454191791480255,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003513401457686242,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21268172711133956,
      "num_tokens": 87958589.0,
      "step": 50710
    },
    {
      "entropy": 5.90805492401123,
      "epoch": 3.9458082085197432,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00035131407363533083,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20842700749635695,
      "num_tokens": 87967270.0,
      "step": 50715
    },
    {
      "entropy": 5.875395727157593,
      "epoch": 3.946197237891461,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003512880003440929,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21161873489618302,
      "num_tokens": 87975460.0,
      "step": 50720
    },
    {
      "entropy": 5.888176345825196,
      "epoch": 3.9465862672631786,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003512619258953061,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21477845013141633,
      "num_tokens": 87984030.0,
      "step": 50725
    },
    {
      "entropy": 5.881440544128418,
      "epoch": 3.946975296634896,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003512358502893658,
      "loss": 4.99,
      "mean_token_accuracy": 0.21567239612340927,
      "num_tokens": 87992080.0,
      "step": 50730
    },
    {
      "entropy": 5.86673936843872,
      "epoch": 3.9473643260066136,
      "grad_norm": 1.359375,
      "learning_rate": 0.000351209773526668,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21959900557994844,
      "num_tokens": 88000718.0,
      "step": 50735
    },
    {
      "entropy": 5.8251190185546875,
      "epoch": 3.947753355378331,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000351183695607608,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2126056283712387,
      "num_tokens": 88008977.0,
      "step": 50740
    },
    {
      "entropy": 5.798488569259644,
      "epoch": 3.9481423847500485,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035115761653258177,
      "loss": 4.9064,
      "mean_token_accuracy": 0.21691323816776276,
      "num_tokens": 88017071.0,
      "step": 50745
    },
    {
      "entropy": 5.8743188858032225,
      "epoch": 3.9485314141217662,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035113153630198494,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20694128572940826,
      "num_tokens": 88025946.0,
      "step": 50750
    },
    {
      "entropy": 5.82512903213501,
      "epoch": 3.948920443493484,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003511054549162132,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2145151063799858,
      "num_tokens": 88033980.0,
      "step": 50755
    },
    {
      "entropy": 5.841487073898316,
      "epoch": 3.949309472865201,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035107937237566233,
      "loss": 4.9732,
      "mean_token_accuracy": 0.20970526933670045,
      "num_tokens": 88043875.0,
      "step": 50760
    },
    {
      "entropy": 5.828063344955444,
      "epoch": 3.949698502236919,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003510532886807281,
      "loss": 4.8797,
      "mean_token_accuracy": 0.21726407408714293,
      "num_tokens": 88052663.0,
      "step": 50765
    },
    {
      "entropy": 5.871134281158447,
      "epoch": 3.9500875316086366,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003510272038318062,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20958467423915864,
      "num_tokens": 88061570.0,
      "step": 50770
    },
    {
      "entropy": 5.861721229553223,
      "epoch": 3.950476560980354,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003510011178292925,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21702828407287597,
      "num_tokens": 88069376.0,
      "step": 50775
    },
    {
      "entropy": 5.798763942718506,
      "epoch": 3.9508655903520715,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003509750306735828,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22440604716539383,
      "num_tokens": 88077433.0,
      "step": 50780
    },
    {
      "entropy": 5.8551311016082765,
      "epoch": 3.951254619723789,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000350948942365073,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22142754793167113,
      "num_tokens": 88086584.0,
      "step": 50785
    },
    {
      "entropy": 5.820428514480591,
      "epoch": 3.951643649095507,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035092285290415885,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21440511345863342,
      "num_tokens": 88095409.0,
      "step": 50790
    },
    {
      "entropy": 5.791308927536011,
      "epoch": 3.952032678467224,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003508967622912362,
      "loss": 4.888,
      "mean_token_accuracy": 0.21174540966749192,
      "num_tokens": 88103396.0,
      "step": 50795
    },
    {
      "entropy": 5.850898122787475,
      "epoch": 3.952421707838942,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035087067052670115,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21026038080453874,
      "num_tokens": 88112810.0,
      "step": 50800
    },
    {
      "entropy": 5.793637466430664,
      "epoch": 3.9528107372106596,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035084457761094925,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21419971585273742,
      "num_tokens": 88120363.0,
      "step": 50805
    },
    {
      "entropy": 5.80784912109375,
      "epoch": 3.953199766582377,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003508184835443766,
      "loss": 4.8753,
      "mean_token_accuracy": 0.21590973436832428,
      "num_tokens": 88129397.0,
      "step": 50810
    },
    {
      "entropy": 5.898810005187988,
      "epoch": 3.9535887959540945,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035079238832737926,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21472312808036803,
      "num_tokens": 88137712.0,
      "step": 50815
    },
    {
      "entropy": 5.810152053833008,
      "epoch": 3.953977825325812,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000350766291960353,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21627353876829147,
      "num_tokens": 88146521.0,
      "step": 50820
    },
    {
      "entropy": 5.8618003845214846,
      "epoch": 3.95436685469753,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003507401944436939,
      "loss": 5.0563,
      "mean_token_accuracy": 0.20671270787715912,
      "num_tokens": 88155775.0,
      "step": 50825
    },
    {
      "entropy": 5.827570152282715,
      "epoch": 3.954755884069247,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003507140957777979,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21178185790777207,
      "num_tokens": 88164613.0,
      "step": 50830
    },
    {
      "entropy": 5.823754549026489,
      "epoch": 3.955144913440965,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035068799596306096,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2201443701982498,
      "num_tokens": 88174014.0,
      "step": 50835
    },
    {
      "entropy": 5.851794815063476,
      "epoch": 3.955533942812682,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003506618949998791,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20700625479221343,
      "num_tokens": 88182348.0,
      "step": 50840
    },
    {
      "entropy": 5.834593820571899,
      "epoch": 3.9559229721844,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035063579288864853,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21633678823709487,
      "num_tokens": 88190754.0,
      "step": 50845
    },
    {
      "entropy": 5.796058416366577,
      "epoch": 3.9563120015561175,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003506096896297651,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2126050278544426,
      "num_tokens": 88200008.0,
      "step": 50850
    },
    {
      "entropy": 5.721543169021606,
      "epoch": 3.956701030927835,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035058358522362505,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2185722529888153,
      "num_tokens": 88208597.0,
      "step": 50855
    },
    {
      "entropy": 5.759484052658081,
      "epoch": 3.9570900602995525,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035055747967062433,
      "loss": 4.9323,
      "mean_token_accuracy": 0.20800039023160935,
      "num_tokens": 88218243.0,
      "step": 50860
    },
    {
      "entropy": 5.749777841567993,
      "epoch": 3.95747908967127,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003505313729711592,
      "loss": 4.8045,
      "mean_token_accuracy": 0.22602923214435577,
      "num_tokens": 88227028.0,
      "step": 50865
    },
    {
      "entropy": 5.873627519607544,
      "epoch": 3.957868119042988,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035050526512562564,
      "loss": 5.0661,
      "mean_token_accuracy": 0.21386022865772247,
      "num_tokens": 88236294.0,
      "step": 50870
    },
    {
      "entropy": 5.831891632080078,
      "epoch": 3.958257148414705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003504791561344199,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1973951056599617,
      "num_tokens": 88245067.0,
      "step": 50875
    },
    {
      "entropy": 5.776224040985108,
      "epoch": 3.958646177786423,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035045304599793807,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2153935581445694,
      "num_tokens": 88253913.0,
      "step": 50880
    },
    {
      "entropy": 5.8396214008331295,
      "epoch": 3.9590352071581405,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035042693471657637,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20966139286756516,
      "num_tokens": 88262574.0,
      "step": 50885
    },
    {
      "entropy": 5.8632604598999025,
      "epoch": 3.959424236529858,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00035040082229073096,
      "loss": 5.0547,
      "mean_token_accuracy": 0.20820428431034088,
      "num_tokens": 88271128.0,
      "step": 50890
    },
    {
      "entropy": 5.744044399261474,
      "epoch": 3.9598132659015755,
      "grad_norm": 1.25,
      "learning_rate": 0.0003503747087207981,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22177897691726683,
      "num_tokens": 88280087.0,
      "step": 50895
    },
    {
      "entropy": 5.844356918334961,
      "epoch": 3.960202295273293,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035034859400717407,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22156779170036317,
      "num_tokens": 88289816.0,
      "step": 50900
    },
    {
      "entropy": 5.845784282684326,
      "epoch": 3.960591324645011,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035032247815025504,
      "loss": 5.0221,
      "mean_token_accuracy": 0.2092660114169121,
      "num_tokens": 88298985.0,
      "step": 50905
    },
    {
      "entropy": 5.849870586395264,
      "epoch": 3.960980354016728,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035029636115043724,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2189180448651314,
      "num_tokens": 88307978.0,
      "step": 50910
    },
    {
      "entropy": 5.835524415969848,
      "epoch": 3.961369383388446,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035027024300811707,
      "loss": 4.951,
      "mean_token_accuracy": 0.21559546887874603,
      "num_tokens": 88317177.0,
      "step": 50915
    },
    {
      "entropy": 5.939777755737305,
      "epoch": 3.9617584127601635,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003502441237236907,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20750205218791962,
      "num_tokens": 88326490.0,
      "step": 50920
    },
    {
      "entropy": 5.945244836807251,
      "epoch": 3.962147442131881,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003502180032975545,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20755604952573775,
      "num_tokens": 88335282.0,
      "step": 50925
    },
    {
      "entropy": 5.916828536987305,
      "epoch": 3.9625364715035984,
      "grad_norm": 1.375,
      "learning_rate": 0.0003501918817301049,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20322464108467103,
      "num_tokens": 88343982.0,
      "step": 50930
    },
    {
      "entropy": 5.867155504226685,
      "epoch": 3.962925500875316,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035016575902173814,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2092261791229248,
      "num_tokens": 88352961.0,
      "step": 50935
    },
    {
      "entropy": 5.860633707046508,
      "epoch": 3.9633145302470334,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035013963517285064,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21924693137407303,
      "num_tokens": 88361493.0,
      "step": 50940
    },
    {
      "entropy": 5.921890068054199,
      "epoch": 3.963703559618751,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00035011351018383874,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20723384767770767,
      "num_tokens": 88370028.0,
      "step": 50945
    },
    {
      "entropy": 5.808373165130615,
      "epoch": 3.964092588990469,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003500873840550989,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21157557219266893,
      "num_tokens": 88378898.0,
      "step": 50950
    },
    {
      "entropy": 5.7234419822692875,
      "epoch": 3.9644816183621865,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035006125678702755,
      "loss": 4.8368,
      "mean_token_accuracy": 0.2266736462712288,
      "num_tokens": 88388156.0,
      "step": 50955
    },
    {
      "entropy": 5.824250221252441,
      "epoch": 3.964870647733904,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035003512838002103,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2126452937722206,
      "num_tokens": 88396719.0,
      "step": 50960
    },
    {
      "entropy": 5.896205711364746,
      "epoch": 3.9652596771056214,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003500089988344759,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20772356688976287,
      "num_tokens": 88405182.0,
      "step": 50965
    },
    {
      "entropy": 5.828327083587647,
      "epoch": 3.965648706477339,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034998286815078855,
      "loss": 4.8909,
      "mean_token_accuracy": 0.20948072522878647,
      "num_tokens": 88413217.0,
      "step": 50970
    },
    {
      "entropy": 5.8466249942779545,
      "epoch": 3.9660377358490564,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034995673632935557,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21243406534194947,
      "num_tokens": 88421969.0,
      "step": 50975
    },
    {
      "entropy": 5.885087728500366,
      "epoch": 3.966426765220774,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003499306033705735,
      "loss": 5.0333,
      "mean_token_accuracy": 0.20867812633514404,
      "num_tokens": 88430160.0,
      "step": 50980
    },
    {
      "entropy": 5.783443927764893,
      "epoch": 3.966815794592492,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034990446927483877,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21522145718336105,
      "num_tokens": 88437890.0,
      "step": 50985
    },
    {
      "entropy": 5.796192073822022,
      "epoch": 3.9672048239642095,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003498783340425479,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21186937242746354,
      "num_tokens": 88446697.0,
      "step": 50990
    },
    {
      "entropy": 5.7997900485992435,
      "epoch": 3.9675938533359267,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034985219767409747,
      "loss": 4.8731,
      "mean_token_accuracy": 0.21808050572872162,
      "num_tokens": 88455925.0,
      "step": 50995
    },
    {
      "entropy": 5.873678588867188,
      "epoch": 3.9679828827076444,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034982606016988407,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21324298679828643,
      "num_tokens": 88464534.0,
      "step": 51000
    },
    {
      "epoch": 3.9679828827076444,
      "eval_entropy": 5.547831045009404,
      "eval_loss": 5.045764923095703,
      "eval_mean_token_accuracy": 0.2179427005085266,
      "eval_num_tokens": 88464534.0,
      "eval_runtime": 21.6407,
      "eval_samples_per_second": 1920.36,
      "eval_steps_per_second": 240.057,
      "step": 51000
    },
    {
      "entropy": 5.849541330337525,
      "epoch": 3.968371912079362,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034979992153030424,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21181644648313522,
      "num_tokens": 88471931.0,
      "step": 51005
    },
    {
      "entropy": 5.824969100952148,
      "epoch": 3.9687609414510794,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003497737817557548,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21495094746351243,
      "num_tokens": 88480075.0,
      "step": 51010
    },
    {
      "entropy": 5.772866535186767,
      "epoch": 3.969149970822797,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034974764084663213,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2108161985874176,
      "num_tokens": 88488309.0,
      "step": 51015
    },
    {
      "entropy": 5.741270637512207,
      "epoch": 3.9695390001945148,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034972149880333316,
      "loss": 4.8304,
      "mean_token_accuracy": 0.21947142034769057,
      "num_tokens": 88497224.0,
      "step": 51020
    },
    {
      "entropy": 5.799672842025757,
      "epoch": 3.9699280295662325,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003496953556262542,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2100339263677597,
      "num_tokens": 88505980.0,
      "step": 51025
    },
    {
      "entropy": 5.834133958816528,
      "epoch": 3.9703170589379497,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034966921131579216,
      "loss": 5.05,
      "mean_token_accuracy": 0.20662824511528016,
      "num_tokens": 88514904.0,
      "step": 51030
    },
    {
      "entropy": 5.711294794082642,
      "epoch": 3.9707060883096674,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034964306587234367,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22056640535593033,
      "num_tokens": 88523456.0,
      "step": 51035
    },
    {
      "entropy": 5.768588924407959,
      "epoch": 3.9710951176813847,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034961691929630544,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22635228782892228,
      "num_tokens": 88532668.0,
      "step": 51040
    },
    {
      "entropy": 5.914878892898559,
      "epoch": 3.9714841470531024,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003495907715880743,
      "loss": 5.0818,
      "mean_token_accuracy": 0.205003023147583,
      "num_tokens": 88541476.0,
      "step": 51045
    },
    {
      "entropy": 5.837304162979126,
      "epoch": 3.97187317642482,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003495646227480469,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21270519942045213,
      "num_tokens": 88549166.0,
      "step": 51050
    },
    {
      "entropy": 5.848220109939575,
      "epoch": 3.9722622057965378,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034953847277662,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21996228992938996,
      "num_tokens": 88557380.0,
      "step": 51055
    },
    {
      "entropy": 5.85033278465271,
      "epoch": 3.9726512351682555,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034951232167419044,
      "loss": 4.9964,
      "mean_token_accuracy": 0.210194730758667,
      "num_tokens": 88565574.0,
      "step": 51060
    },
    {
      "entropy": 5.776845073699951,
      "epoch": 3.9730402645399727,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034948616944115504,
      "loss": 4.8866,
      "mean_token_accuracy": 0.21639368534088135,
      "num_tokens": 88574590.0,
      "step": 51065
    },
    {
      "entropy": 5.806481075286865,
      "epoch": 3.9734292939116904,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003494600160779105,
      "loss": 4.8774,
      "mean_token_accuracy": 0.22332187443971635,
      "num_tokens": 88583399.0,
      "step": 51070
    },
    {
      "entropy": 5.835177087783814,
      "epoch": 3.9738183232834077,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003494338615848537,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2042484775185585,
      "num_tokens": 88592204.0,
      "step": 51075
    },
    {
      "entropy": 5.836013412475586,
      "epoch": 3.9742073526551254,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003494077059623816,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2080878034234047,
      "num_tokens": 88600813.0,
      "step": 51080
    },
    {
      "entropy": 5.792112112045288,
      "epoch": 3.974596382026843,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034938154921089097,
      "loss": 4.8856,
      "mean_token_accuracy": 0.21374749690294265,
      "num_tokens": 88609053.0,
      "step": 51085
    },
    {
      "entropy": 5.754729080200195,
      "epoch": 3.9749854113985608,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003493553913307788,
      "loss": 4.8783,
      "mean_token_accuracy": 0.21573700606822968,
      "num_tokens": 88618058.0,
      "step": 51090
    },
    {
      "entropy": 5.799743604660034,
      "epoch": 3.975374440770278,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034932923232244183,
      "loss": 4.9082,
      "mean_token_accuracy": 0.21854064017534255,
      "num_tokens": 88626379.0,
      "step": 51095
    },
    {
      "entropy": 5.839680814743042,
      "epoch": 3.9757634701419957,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034930307218627706,
      "loss": 5.0634,
      "mean_token_accuracy": 0.2054445654153824,
      "num_tokens": 88635140.0,
      "step": 51100
    },
    {
      "entropy": 5.861330127716064,
      "epoch": 3.9761524995137134,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003492769109226815,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2038085028529167,
      "num_tokens": 88643409.0,
      "step": 51105
    },
    {
      "entropy": 5.843652009963989,
      "epoch": 3.9765415288854307,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034925074853205194,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21273493468761445,
      "num_tokens": 88651945.0,
      "step": 51110
    },
    {
      "entropy": 5.7568892478942875,
      "epoch": 3.9769305582571484,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034922458501478556,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21843539029359818,
      "num_tokens": 88660251.0,
      "step": 51115
    },
    {
      "entropy": 5.8088569164276125,
      "epoch": 3.977319587628866,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003491984203712792,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21132795065641402,
      "num_tokens": 88669873.0,
      "step": 51120
    },
    {
      "entropy": 5.8370973587036135,
      "epoch": 3.9777086170005838,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003491722546019299,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21427838057279586,
      "num_tokens": 88678242.0,
      "step": 51125
    },
    {
      "entropy": 5.857473087310791,
      "epoch": 3.978097646372301,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003491460877071347,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20006311535835267,
      "num_tokens": 88687719.0,
      "step": 51130
    },
    {
      "entropy": 5.845373916625976,
      "epoch": 3.9784866757440187,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003491199196872906,
      "loss": 4.9986,
      "mean_token_accuracy": 0.20765304565429688,
      "num_tokens": 88696404.0,
      "step": 51135
    },
    {
      "entropy": 5.792873334884644,
      "epoch": 3.978875705115736,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003490937505427947,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21261902898550034,
      "num_tokens": 88705093.0,
      "step": 51140
    },
    {
      "entropy": 5.858883810043335,
      "epoch": 3.9792647344874537,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003490675802740441,
      "loss": 4.9714,
      "mean_token_accuracy": 0.214150370657444,
      "num_tokens": 88713856.0,
      "step": 51145
    },
    {
      "entropy": 5.805819511413574,
      "epoch": 3.9796537638591714,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003490414088814359,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2161180078983307,
      "num_tokens": 88722146.0,
      "step": 51150
    },
    {
      "entropy": 5.8179765224456785,
      "epoch": 3.980042793230889,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003490152363653671,
      "loss": 4.9973,
      "mean_token_accuracy": 0.213023342192173,
      "num_tokens": 88730406.0,
      "step": 51155
    },
    {
      "entropy": 5.7867677211761475,
      "epoch": 3.9804318226026068,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003489890627262349,
      "loss": 4.79,
      "mean_token_accuracy": 0.22346812784671782,
      "num_tokens": 88738859.0,
      "step": 51160
    },
    {
      "entropy": 5.862732648849487,
      "epoch": 3.980820851974324,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034896288796443654,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2084744542837143,
      "num_tokens": 88748111.0,
      "step": 51165
    },
    {
      "entropy": 5.875566244125366,
      "epoch": 3.9812098813460417,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034893671208036895,
      "loss": 5.0262,
      "mean_token_accuracy": 0.20833839178085328,
      "num_tokens": 88757127.0,
      "step": 51170
    },
    {
      "entropy": 5.926524829864502,
      "epoch": 3.981598910717759,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034891053507442944,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2051755055785179,
      "num_tokens": 88765245.0,
      "step": 51175
    },
    {
      "entropy": 5.910472059249878,
      "epoch": 3.9819879400894767,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003488843569470153,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21403139382600783,
      "num_tokens": 88773482.0,
      "step": 51180
    },
    {
      "entropy": 5.906291532516479,
      "epoch": 3.9823769694611943,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034885817769852355,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21241040527820587,
      "num_tokens": 88782460.0,
      "step": 51185
    },
    {
      "entropy": 5.846720504760742,
      "epoch": 3.982765998832912,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034883199732935156,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21643432825803757,
      "num_tokens": 88790713.0,
      "step": 51190
    },
    {
      "entropy": 5.855817794799805,
      "epoch": 3.9831550282046293,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034880581583989655,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21255779266357422,
      "num_tokens": 88800064.0,
      "step": 51195
    },
    {
      "entropy": 5.846401739120483,
      "epoch": 3.983544057576347,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003487796332305557,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22346543222665788,
      "num_tokens": 88808748.0,
      "step": 51200
    },
    {
      "entropy": 5.925779676437378,
      "epoch": 3.9839330869480647,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034875344950172635,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20597105324268342,
      "num_tokens": 88817344.0,
      "step": 51205
    },
    {
      "entropy": 5.814324617385864,
      "epoch": 3.984322116319782,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003487272646538058,
      "loss": 4.959,
      "mean_token_accuracy": 0.2189730554819107,
      "num_tokens": 88825916.0,
      "step": 51210
    },
    {
      "entropy": 5.813282012939453,
      "epoch": 3.9847111456914996,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034870107868719135,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2130565568804741,
      "num_tokens": 88834607.0,
      "step": 51215
    },
    {
      "entropy": 5.831885623931885,
      "epoch": 3.9851001750632173,
      "grad_norm": 1.375,
      "learning_rate": 0.0003486748916022804,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2152194157242775,
      "num_tokens": 88843224.0,
      "step": 51220
    },
    {
      "entropy": 5.807802295684814,
      "epoch": 3.985489204434935,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003486487033994702,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22114470452070237,
      "num_tokens": 88851732.0,
      "step": 51225
    },
    {
      "entropy": 5.792023754119873,
      "epoch": 3.9858782338066523,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034862251407915813,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22244195342063905,
      "num_tokens": 88861017.0,
      "step": 51230
    },
    {
      "entropy": 5.797367858886719,
      "epoch": 3.98626726317837,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034859632364174167,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21483972817659377,
      "num_tokens": 88869731.0,
      "step": 51235
    },
    {
      "entropy": 5.911698627471924,
      "epoch": 3.9866562925500877,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000348570132087618,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21292393803596496,
      "num_tokens": 88878627.0,
      "step": 51240
    },
    {
      "entropy": 5.930258083343506,
      "epoch": 3.987045321921805,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003485439394171847,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21175676286220552,
      "num_tokens": 88887473.0,
      "step": 51245
    },
    {
      "entropy": 5.842200040817261,
      "epoch": 3.9874343512935226,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034851774563083925,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21061163097620011,
      "num_tokens": 88896140.0,
      "step": 51250
    },
    {
      "entropy": 5.814928674697876,
      "epoch": 3.9878233806652403,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003484915507289791,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21731790006160737,
      "num_tokens": 88904630.0,
      "step": 51255
    },
    {
      "entropy": 5.8075474262237545,
      "epoch": 3.988212410036958,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034846535471200153,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21364985406398773,
      "num_tokens": 88913632.0,
      "step": 51260
    },
    {
      "entropy": 5.873779535293579,
      "epoch": 3.9886014394086753,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003484391575803041,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21179973036050798,
      "num_tokens": 88922574.0,
      "step": 51265
    },
    {
      "entropy": 5.79934287071228,
      "epoch": 3.988990468780393,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003484129593342844,
      "loss": 4.881,
      "mean_token_accuracy": 0.21969355195760726,
      "num_tokens": 88930820.0,
      "step": 51270
    },
    {
      "entropy": 5.844461917877197,
      "epoch": 3.9893794981521102,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034838675997433987,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21297162920236587,
      "num_tokens": 88938925.0,
      "step": 51275
    },
    {
      "entropy": 5.843235874176026,
      "epoch": 3.989768527523828,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003483605595008681,
      "loss": 4.9356,
      "mean_token_accuracy": 0.20793224424123763,
      "num_tokens": 88947825.0,
      "step": 51280
    },
    {
      "entropy": 5.923256683349609,
      "epoch": 3.9901575568955456,
      "grad_norm": 1.25,
      "learning_rate": 0.00034833435791426653,
      "loss": 5.1128,
      "mean_token_accuracy": 0.19983021169900894,
      "num_tokens": 88957583.0,
      "step": 51285
    },
    {
      "entropy": 5.970640420913696,
      "epoch": 3.9905465862672633,
      "grad_norm": 1.484375,
      "learning_rate": 0.00034830815521493295,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21258684247732162,
      "num_tokens": 88966514.0,
      "step": 51290
    },
    {
      "entropy": 5.907920408248901,
      "epoch": 3.9909356156389806,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003482819514032647,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21122399419546128,
      "num_tokens": 88974580.0,
      "step": 51295
    },
    {
      "entropy": 5.755581712722778,
      "epoch": 3.9913246450106983,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034825574647965945,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2242751345038414,
      "num_tokens": 88983892.0,
      "step": 51300
    },
    {
      "entropy": 5.872069263458252,
      "epoch": 3.991713674382416,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034822954044451485,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20567660480737687,
      "num_tokens": 88993421.0,
      "step": 51305
    },
    {
      "entropy": 5.896870374679565,
      "epoch": 3.9921027037541332,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003482033332982286,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2138390138745308,
      "num_tokens": 89002246.0,
      "step": 51310
    },
    {
      "entropy": 5.9058653831481935,
      "epoch": 3.992491733125851,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003481771250411982,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2144250527024269,
      "num_tokens": 89012280.0,
      "step": 51315
    },
    {
      "entropy": 5.8697309494018555,
      "epoch": 3.9928807624975686,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003481509156738214,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20556814223527908,
      "num_tokens": 89020776.0,
      "step": 51320
    },
    {
      "entropy": 5.7874444961547855,
      "epoch": 3.9932697918692863,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034812470519649593,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22216229736804963,
      "num_tokens": 89028559.0,
      "step": 51325
    },
    {
      "entropy": 5.847426652908325,
      "epoch": 3.9936588212410036,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003480984936096194,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21203238964080812,
      "num_tokens": 89037093.0,
      "step": 51330
    },
    {
      "entropy": 5.826626396179199,
      "epoch": 3.9940478506127213,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034807228091358964,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21483483165502548,
      "num_tokens": 89046953.0,
      "step": 51335
    },
    {
      "entropy": 5.838588619232178,
      "epoch": 3.994436879984439,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003480460671088043,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21679116636514664,
      "num_tokens": 89055067.0,
      "step": 51340
    },
    {
      "entropy": 5.866257047653198,
      "epoch": 3.9948259093561562,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034801985219566124,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2068975806236267,
      "num_tokens": 89063965.0,
      "step": 51345
    },
    {
      "entropy": 5.869564390182495,
      "epoch": 3.995214938727874,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000347993636174558,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20556047111749648,
      "num_tokens": 89072541.0,
      "step": 51350
    },
    {
      "entropy": 5.820171260833741,
      "epoch": 3.9956039680995916,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034796741904589263,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22106898128986358,
      "num_tokens": 89081063.0,
      "step": 51355
    },
    {
      "entropy": 5.83446307182312,
      "epoch": 3.9959929974713093,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034794120081006275,
      "loss": 4.957,
      "mean_token_accuracy": 0.21530717015266418,
      "num_tokens": 89090028.0,
      "step": 51360
    },
    {
      "entropy": 5.742966747283935,
      "epoch": 3.9963820268430266,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003479149814674663,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2136027619242668,
      "num_tokens": 89098596.0,
      "step": 51365
    },
    {
      "entropy": 5.865492486953736,
      "epoch": 3.9967710562147443,
      "grad_norm": 1.40625,
      "learning_rate": 0.000347888761018501,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21574797332286835,
      "num_tokens": 89106756.0,
      "step": 51370
    },
    {
      "entropy": 5.747814273834228,
      "epoch": 3.9971600855864615,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003478625394635648,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2053950011730194,
      "num_tokens": 89115477.0,
      "step": 51375
    },
    {
      "entropy": 5.817495393753052,
      "epoch": 3.997549114958179,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034783631680305554,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21318839788436889,
      "num_tokens": 89123667.0,
      "step": 51380
    },
    {
      "entropy": 5.811772441864013,
      "epoch": 3.997938144329897,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003478100930373711,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21976623684167862,
      "num_tokens": 89132453.0,
      "step": 51385
    },
    {
      "entropy": 5.906843137741089,
      "epoch": 3.9983271737016146,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003477838681669094,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2093352645635605,
      "num_tokens": 89141214.0,
      "step": 51390
    },
    {
      "entropy": 5.920486831665039,
      "epoch": 3.9987162030733323,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003477576421920682,
      "loss": 5.0498,
      "mean_token_accuracy": 0.20425763875246047,
      "num_tokens": 89149769.0,
      "step": 51395
    },
    {
      "entropy": 5.789428806304931,
      "epoch": 3.9991052324450496,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003477314151132458,
      "loss": 4.9092,
      "mean_token_accuracy": 0.21738403886556626,
      "num_tokens": 89157951.0,
      "step": 51400
    },
    {
      "entropy": 5.904246759414673,
      "epoch": 3.9994942618167673,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034770518693083986,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21435455977916718,
      "num_tokens": 89166697.0,
      "step": 51405
    },
    {
      "entropy": 5.90903787612915,
      "epoch": 3.9998832911884845,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034767895764524845,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22087946236133577,
      "num_tokens": 89175359.0,
      "step": 51410
    },
    {
      "entropy": 5.8504725032382545,
      "epoch": 4.000233417623031,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034765272725686947,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21506953239440918,
      "num_tokens": 89182872.0,
      "step": 51415
    },
    {
      "entropy": 5.843354368209839,
      "epoch": 4.000622446994748,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034762649576610105,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2120300754904747,
      "num_tokens": 89191308.0,
      "step": 51420
    },
    {
      "entropy": 5.83942289352417,
      "epoch": 4.0010114763664655,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034760026317334113,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2227887511253357,
      "num_tokens": 89199177.0,
      "step": 51425
    },
    {
      "entropy": 5.904194116592407,
      "epoch": 4.001400505738183,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003475740294789878,
      "loss": 4.992,
      "mean_token_accuracy": 0.21009704917669297,
      "num_tokens": 89207684.0,
      "step": 51430
    },
    {
      "entropy": 5.784019088745117,
      "epoch": 4.001789535109901,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000347547794683439,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22087557166814803,
      "num_tokens": 89216591.0,
      "step": 51435
    },
    {
      "entropy": 5.923308944702148,
      "epoch": 4.002178564481619,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003475215587870929,
      "loss": 5.0079,
      "mean_token_accuracy": 0.20841668099164962,
      "num_tokens": 89225337.0,
      "step": 51440
    },
    {
      "entropy": 5.829972934722901,
      "epoch": 4.002567593853336,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003474953217903477,
      "loss": 4.8939,
      "mean_token_accuracy": 0.21679709404706954,
      "num_tokens": 89234241.0,
      "step": 51445
    },
    {
      "entropy": 5.810553169250488,
      "epoch": 4.002956623225053,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003474690836936013,
      "loss": 4.8809,
      "mean_token_accuracy": 0.22037444710731507,
      "num_tokens": 89243658.0,
      "step": 51450
    },
    {
      "entropy": 5.863436031341553,
      "epoch": 4.003345652596771,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003474428444972519,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22006140202283858,
      "num_tokens": 89253161.0,
      "step": 51455
    },
    {
      "entropy": 5.808107376098633,
      "epoch": 4.0037346819684885,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003474166042016977,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2114116370677948,
      "num_tokens": 89262376.0,
      "step": 51460
    },
    {
      "entropy": 5.880016088485718,
      "epoch": 4.004123711340206,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00034739036280733674,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21360787600278855,
      "num_tokens": 89270219.0,
      "step": 51465
    },
    {
      "entropy": 5.864965343475342,
      "epoch": 4.004512740711924,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034736412031456714,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2257390648126602,
      "num_tokens": 89278734.0,
      "step": 51470
    },
    {
      "entropy": 5.869487380981445,
      "epoch": 4.004901770083642,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034733787672378734,
      "loss": 4.9168,
      "mean_token_accuracy": 0.21515712141990662,
      "num_tokens": 89288043.0,
      "step": 51475
    },
    {
      "entropy": 5.819180536270141,
      "epoch": 4.005290799455359,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003473116320353953,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22581062614917755,
      "num_tokens": 89296482.0,
      "step": 51480
    },
    {
      "entropy": 5.788719749450683,
      "epoch": 4.005679828827076,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003472853862497895,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2311222493648529,
      "num_tokens": 89304703.0,
      "step": 51485
    },
    {
      "entropy": 5.891383314132691,
      "epoch": 4.006068858198794,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003472591393673679,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20446013361215593,
      "num_tokens": 89312953.0,
      "step": 51490
    },
    {
      "entropy": 5.745553731918335,
      "epoch": 4.0064578875705115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034723289138852885,
      "loss": 4.8317,
      "mean_token_accuracy": 0.222431980073452,
      "num_tokens": 89321727.0,
      "step": 51495
    },
    {
      "entropy": 5.7577221393585205,
      "epoch": 4.006846916942229,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003472066423136707,
      "loss": 4.8628,
      "mean_token_accuracy": 0.22814445197582245,
      "num_tokens": 89330180.0,
      "step": 51500
    },
    {
      "entropy": 5.7535247802734375,
      "epoch": 4.007235946313947,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003471803921431917,
      "loss": 4.9254,
      "mean_token_accuracy": 0.20446530729532242,
      "num_tokens": 89338600.0,
      "step": 51505
    },
    {
      "entropy": 5.868578243255615,
      "epoch": 4.007624975685665,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034715414087749003,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20912158787250518,
      "num_tokens": 89347408.0,
      "step": 51510
    },
    {
      "entropy": 5.792420148849487,
      "epoch": 4.008014005057382,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003471278885169642,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22065535485744475,
      "num_tokens": 89356309.0,
      "step": 51515
    },
    {
      "entropy": 5.825910377502441,
      "epoch": 4.008403034429099,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034710163506201247,
      "loss": 4.9139,
      "mean_token_accuracy": 0.21985791474580765,
      "num_tokens": 89365014.0,
      "step": 51520
    },
    {
      "entropy": 5.921168804168701,
      "epoch": 4.008792063800817,
      "grad_norm": 1.453125,
      "learning_rate": 0.00034707538051303316,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2117210805416107,
      "num_tokens": 89374137.0,
      "step": 51525
    },
    {
      "entropy": 5.875305604934693,
      "epoch": 4.0091810931725345,
      "grad_norm": 1.375,
      "learning_rate": 0.0003470491248704247,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21245753467082978,
      "num_tokens": 89382398.0,
      "step": 51530
    },
    {
      "entropy": 5.854421234130859,
      "epoch": 4.009570122544252,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034702286813458543,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21743429750204085,
      "num_tokens": 89391320.0,
      "step": 51535
    },
    {
      "entropy": 5.757168817520141,
      "epoch": 4.00995915191597,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034699661030591374,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21566785275936126,
      "num_tokens": 89400204.0,
      "step": 51540
    },
    {
      "entropy": 5.808652591705322,
      "epoch": 4.0103481812876876,
      "grad_norm": 1.5,
      "learning_rate": 0.0003469703513848081,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22021826207637787,
      "num_tokens": 89408217.0,
      "step": 51545
    },
    {
      "entropy": 5.83072714805603,
      "epoch": 4.010737210659404,
      "grad_norm": 1.3125,
      "learning_rate": 0.000346944091371667,
      "loss": 4.8628,
      "mean_token_accuracy": 0.21939108073711394,
      "num_tokens": 89417677.0,
      "step": 51550
    },
    {
      "entropy": 5.796425247192383,
      "epoch": 4.011126240031122,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003469178302668888,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2207345873117447,
      "num_tokens": 89426275.0,
      "step": 51555
    },
    {
      "entropy": 5.924442338943481,
      "epoch": 4.01151526940284,
      "grad_norm": 1.375,
      "learning_rate": 0.0003468915680708719,
      "loss": 5.0323,
      "mean_token_accuracy": 0.20831306874752045,
      "num_tokens": 89434855.0,
      "step": 51560
    },
    {
      "entropy": 5.875079393386841,
      "epoch": 4.0119042987745575,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000346865304784015,
      "loss": 4.987,
      "mean_token_accuracy": 0.20980384796857834,
      "num_tokens": 89443317.0,
      "step": 51565
    },
    {
      "entropy": 5.791081666946411,
      "epoch": 4.012293328146275,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003468390404067164,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22150207161903382,
      "num_tokens": 89452181.0,
      "step": 51570
    },
    {
      "entropy": 5.912380981445312,
      "epoch": 4.012682357517993,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003468127749393747,
      "loss": 4.98,
      "mean_token_accuracy": 0.2057015597820282,
      "num_tokens": 89460997.0,
      "step": 51575
    },
    {
      "entropy": 5.891068410873413,
      "epoch": 4.0130713868897105,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003467865083823885,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2083483874797821,
      "num_tokens": 89469431.0,
      "step": 51580
    },
    {
      "entropy": 5.881328153610229,
      "epoch": 4.013460416261427,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003467602407361563,
      "loss": 4.941,
      "mean_token_accuracy": 0.20990949422121047,
      "num_tokens": 89477372.0,
      "step": 51585
    },
    {
      "entropy": 5.867156028747559,
      "epoch": 4.013849445633145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034673397200107663,
      "loss": 4.9791,
      "mean_token_accuracy": 0.210368350148201,
      "num_tokens": 89485542.0,
      "step": 51590
    },
    {
      "entropy": 5.7216509819030765,
      "epoch": 4.014238475004863,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003467077021775481,
      "loss": 4.811,
      "mean_token_accuracy": 0.22827076613903047,
      "num_tokens": 89494041.0,
      "step": 51595
    },
    {
      "entropy": 5.88072566986084,
      "epoch": 4.0146275043765804,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00034668143126596933,
      "loss": 5.037,
      "mean_token_accuracy": 0.21217611134052278,
      "num_tokens": 89502749.0,
      "step": 51600
    },
    {
      "entropy": 5.807913208007813,
      "epoch": 4.015016533748298,
      "grad_norm": 1.46875,
      "learning_rate": 0.00034665515926673903,
      "loss": 4.8417,
      "mean_token_accuracy": 0.22049765288829803,
      "num_tokens": 89510981.0,
      "step": 51605
    },
    {
      "entropy": 5.855193996429444,
      "epoch": 4.015405563120016,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003466288861802556,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21269351243972778,
      "num_tokens": 89519926.0,
      "step": 51610
    },
    {
      "entropy": 5.909092617034912,
      "epoch": 4.0157945924917335,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003466026120069179,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21776532381772995,
      "num_tokens": 89528547.0,
      "step": 51615
    },
    {
      "entropy": 5.82960057258606,
      "epoch": 4.01618362186345,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003465763367471245,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21412865966558456,
      "num_tokens": 89536691.0,
      "step": 51620
    },
    {
      "entropy": 5.791877031326294,
      "epoch": 4.016572651235168,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034655006040127415,
      "loss": 4.853,
      "mean_token_accuracy": 0.21744546443223953,
      "num_tokens": 89545484.0,
      "step": 51625
    },
    {
      "entropy": 5.829100608825684,
      "epoch": 4.016961680606886,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003465237829697655,
      "loss": 4.872,
      "mean_token_accuracy": 0.22078536897897721,
      "num_tokens": 89554295.0,
      "step": 51630
    },
    {
      "entropy": 5.746594667434692,
      "epoch": 4.017350709978603,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034649750445299725,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22994038909673692,
      "num_tokens": 89562633.0,
      "step": 51635
    },
    {
      "entropy": 5.8266387462615965,
      "epoch": 4.017739739350321,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003464712248513682,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20847724974155427,
      "num_tokens": 89571081.0,
      "step": 51640
    },
    {
      "entropy": 5.814377546310425,
      "epoch": 4.018128768722039,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003464449441652771,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22242127358913422,
      "num_tokens": 89579960.0,
      "step": 51645
    },
    {
      "entropy": 5.872315168380737,
      "epoch": 4.018517798093756,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003464186623951227,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21494875103235245,
      "num_tokens": 89589092.0,
      "step": 51650
    },
    {
      "entropy": 5.885835409164429,
      "epoch": 4.018906827465473,
      "grad_norm": 1.375,
      "learning_rate": 0.0003463923795413037,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2062897801399231,
      "num_tokens": 89599430.0,
      "step": 51655
    },
    {
      "entropy": 5.832208776473999,
      "epoch": 4.019295856837191,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034636609560421906,
      "loss": 4.903,
      "mean_token_accuracy": 0.21794089525938035,
      "num_tokens": 89607850.0,
      "step": 51660
    },
    {
      "entropy": 5.812400674819946,
      "epoch": 4.019684886208909,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034633981058426744,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2187441498041153,
      "num_tokens": 89617433.0,
      "step": 51665
    },
    {
      "entropy": 5.819549322128296,
      "epoch": 4.020073915580626,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003463135244818477,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22183749973773956,
      "num_tokens": 89625730.0,
      "step": 51670
    },
    {
      "entropy": 5.830978536605835,
      "epoch": 4.020462944952344,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003462872372973588,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21887386590242386,
      "num_tokens": 89634827.0,
      "step": 51675
    },
    {
      "entropy": 5.808528566360474,
      "epoch": 4.020851974324062,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003462609490311996,
      "loss": 4.9235,
      "mean_token_accuracy": 0.20965302735567093,
      "num_tokens": 89643564.0,
      "step": 51680
    },
    {
      "entropy": 5.839132404327392,
      "epoch": 4.021241003695779,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003462346596837689,
      "loss": 4.8681,
      "mean_token_accuracy": 0.21772970408201217,
      "num_tokens": 89651529.0,
      "step": 51685
    },
    {
      "entropy": 5.8134575366973875,
      "epoch": 4.021630033067496,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003462083692554655,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21801274865865708,
      "num_tokens": 89659139.0,
      "step": 51690
    },
    {
      "entropy": 5.769798469543457,
      "epoch": 4.022019062439214,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034618207774668854,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22245918959379196,
      "num_tokens": 89667060.0,
      "step": 51695
    },
    {
      "entropy": 5.759919548034668,
      "epoch": 4.022408091810932,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034615578515783687,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21999125480651854,
      "num_tokens": 89675179.0,
      "step": 51700
    },
    {
      "entropy": 5.909809303283692,
      "epoch": 4.022797121182649,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034612949148930935,
      "loss": 5.0484,
      "mean_token_accuracy": 0.2021441251039505,
      "num_tokens": 89684661.0,
      "step": 51705
    },
    {
      "entropy": 5.941668605804443,
      "epoch": 4.023186150554367,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000346103196741505,
      "loss": 5.064,
      "mean_token_accuracy": 0.20643474906682968,
      "num_tokens": 89693179.0,
      "step": 51710
    },
    {
      "entropy": 5.8753660202026365,
      "epoch": 4.023575179926085,
      "grad_norm": 1.5,
      "learning_rate": 0.00034607690091482284,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2115110471844673,
      "num_tokens": 89702045.0,
      "step": 51715
    },
    {
      "entropy": 5.825630521774292,
      "epoch": 4.023964209297802,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003460506040096619,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21676555275917053,
      "num_tokens": 89710564.0,
      "step": 51720
    },
    {
      "entropy": 5.794694185256958,
      "epoch": 4.024353238669519,
      "grad_norm": 1.25,
      "learning_rate": 0.0003460243060264211,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2198055401444435,
      "num_tokens": 89719555.0,
      "step": 51725
    },
    {
      "entropy": 5.7684431076049805,
      "epoch": 4.024742268041237,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034599800696549944,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21602105498313903,
      "num_tokens": 89728142.0,
      "step": 51730
    },
    {
      "entropy": 5.892065572738647,
      "epoch": 4.025131297412955,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000345971706827296,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2167395234107971,
      "num_tokens": 89736682.0,
      "step": 51735
    },
    {
      "entropy": 5.79404501914978,
      "epoch": 4.025520326784672,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034594540561221,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21187314242124558,
      "num_tokens": 89745330.0,
      "step": 51740
    },
    {
      "entropy": 5.797719860076905,
      "epoch": 4.02590935615639,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034591910332064027,
      "loss": 4.918,
      "mean_token_accuracy": 0.2186532959342003,
      "num_tokens": 89753891.0,
      "step": 51745
    },
    {
      "entropy": 5.771295022964478,
      "epoch": 4.026298385528108,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003458927999529861,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21554160565137864,
      "num_tokens": 89762562.0,
      "step": 51750
    },
    {
      "entropy": 5.828758001327515,
      "epoch": 4.026687414899825,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034586649550964656,
      "loss": 4.943,
      "mean_token_accuracy": 0.2146461248397827,
      "num_tokens": 89771576.0,
      "step": 51755
    },
    {
      "entropy": 5.848482942581176,
      "epoch": 4.027076444271542,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034584018999102073,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22008802592754365,
      "num_tokens": 89780702.0,
      "step": 51760
    },
    {
      "entropy": 5.84187912940979,
      "epoch": 4.02746547364326,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034581388339750775,
      "loss": 4.8844,
      "mean_token_accuracy": 0.215875343978405,
      "num_tokens": 89789580.0,
      "step": 51765
    },
    {
      "entropy": 5.830805730819702,
      "epoch": 4.027854503014978,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003457875757295067,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21076501458883284,
      "num_tokens": 89797726.0,
      "step": 51770
    },
    {
      "entropy": 5.8494651317596436,
      "epoch": 4.028243532386695,
      "grad_norm": 1.375,
      "learning_rate": 0.00034576126698741697,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21786351054906844,
      "num_tokens": 89805771.0,
      "step": 51775
    },
    {
      "entropy": 5.837840461730957,
      "epoch": 4.028632561758413,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003457349571716376,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2136189341545105,
      "num_tokens": 89814424.0,
      "step": 51780
    },
    {
      "entropy": 5.822020673751831,
      "epoch": 4.02902159113013,
      "grad_norm": 1.453125,
      "learning_rate": 0.00034570864628256795,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20662281066179275,
      "num_tokens": 89823030.0,
      "step": 51785
    },
    {
      "entropy": 5.7909704685211185,
      "epoch": 4.029410620501848,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034568233432060705,
      "loss": 4.8936,
      "mean_token_accuracy": 0.21905582696199416,
      "num_tokens": 89831000.0,
      "step": 51790
    },
    {
      "entropy": 5.902061414718628,
      "epoch": 4.029799649873565,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003456560212861543,
      "loss": 5.005,
      "mean_token_accuracy": 0.20877033919095994,
      "num_tokens": 89839454.0,
      "step": 51795
    },
    {
      "entropy": 5.801869440078735,
      "epoch": 4.030188679245283,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003456297071796087,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2097836822271347,
      "num_tokens": 89848442.0,
      "step": 51800
    },
    {
      "entropy": 5.7670753479003904,
      "epoch": 4.030577708617001,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034560339200136983,
      "loss": 4.8642,
      "mean_token_accuracy": 0.2213040381669998,
      "num_tokens": 89857249.0,
      "step": 51805
    },
    {
      "entropy": 5.835820722579956,
      "epoch": 4.030966737988718,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034557707575183684,
      "loss": 4.8822,
      "mean_token_accuracy": 0.21813945472240448,
      "num_tokens": 89865523.0,
      "step": 51810
    },
    {
      "entropy": 5.827543020248413,
      "epoch": 4.031355767360436,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034555075843140917,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21498481184244156,
      "num_tokens": 89874555.0,
      "step": 51815
    },
    {
      "entropy": 5.876661872863769,
      "epoch": 4.031744796732153,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034552444004048587,
      "loss": 4.944,
      "mean_token_accuracy": 0.21540486365556716,
      "num_tokens": 89882661.0,
      "step": 51820
    },
    {
      "entropy": 5.7983136653900145,
      "epoch": 4.032133826103871,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034549812057946657,
      "loss": 4.902,
      "mean_token_accuracy": 0.22543672770261763,
      "num_tokens": 89892111.0,
      "step": 51825
    },
    {
      "entropy": 5.829182910919189,
      "epoch": 4.032522855475588,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003454718000487505,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21223197281360626,
      "num_tokens": 89900688.0,
      "step": 51830
    },
    {
      "entropy": 5.821658325195313,
      "epoch": 4.032911884847306,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003454454784487369,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2086763098835945,
      "num_tokens": 89909855.0,
      "step": 51835
    },
    {
      "entropy": 5.849876594543457,
      "epoch": 4.033300914219024,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003454191557798254,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21780864596366883,
      "num_tokens": 89918221.0,
      "step": 51840
    },
    {
      "entropy": 5.816346311569214,
      "epoch": 4.033689943590741,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034539283204241525,
      "loss": 4.886,
      "mean_token_accuracy": 0.22041191756725312,
      "num_tokens": 89926898.0,
      "step": 51845
    },
    {
      "entropy": 5.864366483688355,
      "epoch": 4.034078972962459,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034536650723690596,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21079951971769334,
      "num_tokens": 89935555.0,
      "step": 51850
    },
    {
      "entropy": 5.842814254760742,
      "epoch": 4.034468002334176,
      "grad_norm": 1.375,
      "learning_rate": 0.00034534018136369687,
      "loss": 4.9826,
      "mean_token_accuracy": 0.20815846174955369,
      "num_tokens": 89944465.0,
      "step": 51855
    },
    {
      "entropy": 5.832459545135498,
      "epoch": 4.034857031705894,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003453138544231875,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2216557517647743,
      "num_tokens": 89952825.0,
      "step": 51860
    },
    {
      "entropy": 5.903737354278564,
      "epoch": 4.035246061077611,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003452875264157774,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21432709246873854,
      "num_tokens": 89960786.0,
      "step": 51865
    },
    {
      "entropy": 5.821130037307739,
      "epoch": 4.035635090449329,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034526119734186586,
      "loss": 4.9384,
      "mean_token_accuracy": 0.21836579889059066,
      "num_tokens": 89970485.0,
      "step": 51870
    },
    {
      "entropy": 5.814679098129273,
      "epoch": 4.036024119821047,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034523486720185246,
      "loss": 4.8482,
      "mean_token_accuracy": 0.22205869108438492,
      "num_tokens": 89979060.0,
      "step": 51875
    },
    {
      "entropy": 5.8419653415679935,
      "epoch": 4.036413149192764,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034520853599613675,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21774656027555467,
      "num_tokens": 89987105.0,
      "step": 51880
    },
    {
      "entropy": 5.805672454833984,
      "epoch": 4.036802178564481,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034518220372511835,
      "loss": 4.912,
      "mean_token_accuracy": 0.214341701567173,
      "num_tokens": 89995799.0,
      "step": 51885
    },
    {
      "entropy": 5.7886709690094,
      "epoch": 4.037191207936199,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003451558703891967,
      "loss": 4.911,
      "mean_token_accuracy": 0.21123124063014984,
      "num_tokens": 90003906.0,
      "step": 51890
    },
    {
      "entropy": 5.758745527267456,
      "epoch": 4.037580237307917,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003451295359887713,
      "loss": 4.884,
      "mean_token_accuracy": 0.21914792358875274,
      "num_tokens": 90012884.0,
      "step": 51895
    },
    {
      "entropy": 5.791924142837525,
      "epoch": 4.037969266679634,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003451032005242418,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21860792934894563,
      "num_tokens": 90021875.0,
      "step": 51900
    },
    {
      "entropy": 5.82041335105896,
      "epoch": 4.038358296051352,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034507686399600786,
      "loss": 4.871,
      "mean_token_accuracy": 0.21730118691921235,
      "num_tokens": 90030187.0,
      "step": 51905
    },
    {
      "entropy": 5.876274061203003,
      "epoch": 4.03874732542307,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034505052640446907,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2063641294836998,
      "num_tokens": 90038904.0,
      "step": 51910
    },
    {
      "entropy": 5.782050466537475,
      "epoch": 4.039136354794787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000345024187750025,
      "loss": 4.869,
      "mean_token_accuracy": 0.2273107334971428,
      "num_tokens": 90047933.0,
      "step": 51915
    },
    {
      "entropy": 5.768761110305786,
      "epoch": 4.039525384166504,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034499784803307534,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22211536020040512,
      "num_tokens": 90056901.0,
      "step": 51920
    },
    {
      "entropy": 5.778373289108276,
      "epoch": 4.039914413538222,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003449715072540198,
      "loss": 4.7921,
      "mean_token_accuracy": 0.2246643289923668,
      "num_tokens": 90065755.0,
      "step": 51925
    },
    {
      "entropy": 5.8750319480896,
      "epoch": 4.04030344290994,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034494516541325795,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20744441896677018,
      "num_tokens": 90075039.0,
      "step": 51930
    },
    {
      "entropy": 5.8306890487670895,
      "epoch": 4.040692472281657,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003449188225111895,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2039992779493332,
      "num_tokens": 90083435.0,
      "step": 51935
    },
    {
      "entropy": 5.867086601257324,
      "epoch": 4.041081501653375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034489247854821426,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21878828704357148,
      "num_tokens": 90092559.0,
      "step": 51940
    },
    {
      "entropy": 5.823929500579834,
      "epoch": 4.041470531025093,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034486613352473195,
      "loss": 4.92,
      "mean_token_accuracy": 0.2125306621193886,
      "num_tokens": 90101623.0,
      "step": 51945
    },
    {
      "entropy": 5.87227692604065,
      "epoch": 4.04185956039681,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034483978744114224,
      "loss": 4.9194,
      "mean_token_accuracy": 0.21183342337608338,
      "num_tokens": 90109497.0,
      "step": 51950
    },
    {
      "entropy": 5.887918949127197,
      "epoch": 4.042248589768527,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003448134402978449,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2148037686944008,
      "num_tokens": 90118113.0,
      "step": 51955
    },
    {
      "entropy": 5.887177276611328,
      "epoch": 4.042637619140245,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034478709209523974,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2037634715437889,
      "num_tokens": 90126996.0,
      "step": 51960
    },
    {
      "entropy": 5.870947980880738,
      "epoch": 4.043026648511963,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003447607428337265,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21998502910137177,
      "num_tokens": 90136265.0,
      "step": 51965
    },
    {
      "entropy": 5.854250383377075,
      "epoch": 4.04341567788368,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034473439251370505,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2173912599682808,
      "num_tokens": 90145452.0,
      "step": 51970
    },
    {
      "entropy": 5.81551570892334,
      "epoch": 4.043804707255398,
      "grad_norm": 1.375,
      "learning_rate": 0.0003447080411355752,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2147714078426361,
      "num_tokens": 90154305.0,
      "step": 51975
    },
    {
      "entropy": 5.8714135646820065,
      "epoch": 4.044193736627116,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034468168869973673,
      "loss": 4.9475,
      "mean_token_accuracy": 0.20947336703538894,
      "num_tokens": 90162839.0,
      "step": 51980
    },
    {
      "entropy": 5.816893243789673,
      "epoch": 4.0445827659988325,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034465533520658967,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2175579071044922,
      "num_tokens": 90171935.0,
      "step": 51985
    },
    {
      "entropy": 5.856818675994873,
      "epoch": 4.04497179537055,
      "grad_norm": 1.375,
      "learning_rate": 0.00034462898065653366,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21408382505178453,
      "num_tokens": 90180963.0,
      "step": 51990
    },
    {
      "entropy": 5.907380199432373,
      "epoch": 4.045360824742268,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003446026250499687,
      "loss": 4.9744,
      "mean_token_accuracy": 0.20971859842538834,
      "num_tokens": 90189146.0,
      "step": 51995
    },
    {
      "entropy": 5.843783235549926,
      "epoch": 4.045749854113986,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003445762683872946,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2196521818637848,
      "num_tokens": 90197107.0,
      "step": 52000
    },
    {
      "entropy": 5.8998040676116945,
      "epoch": 4.046138883485703,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003445499106689115,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20930359065532683,
      "num_tokens": 90205389.0,
      "step": 52005
    },
    {
      "entropy": 5.87999792098999,
      "epoch": 4.046527912857421,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034452355189521915,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21960538774728774,
      "num_tokens": 90214119.0,
      "step": 52010
    },
    {
      "entropy": 5.85208306312561,
      "epoch": 4.046916942229139,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003444971920666176,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2189662501215935,
      "num_tokens": 90222019.0,
      "step": 52015
    },
    {
      "entropy": 5.852917146682739,
      "epoch": 4.0473059716008555,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003444708311835068,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21542755663394927,
      "num_tokens": 90230150.0,
      "step": 52020
    },
    {
      "entropy": 5.814383840560913,
      "epoch": 4.047695000972573,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034444446924628664,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2231388047337532,
      "num_tokens": 90238320.0,
      "step": 52025
    },
    {
      "entropy": 5.858529663085937,
      "epoch": 4.048084030344291,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034441810625535725,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2090259611606598,
      "num_tokens": 90247695.0,
      "step": 52030
    },
    {
      "entropy": 5.8809877872467045,
      "epoch": 4.048473059716009,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003443917422111185,
      "loss": 5.0071,
      "mean_token_accuracy": 0.2172122120857239,
      "num_tokens": 90257006.0,
      "step": 52035
    },
    {
      "entropy": 5.88595609664917,
      "epoch": 4.048862089087726,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034436537711397054,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2106809437274933,
      "num_tokens": 90265869.0,
      "step": 52040
    },
    {
      "entropy": 5.912539863586426,
      "epoch": 4.049251118459444,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003443390109643134,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20577946305274963,
      "num_tokens": 90274370.0,
      "step": 52045
    },
    {
      "entropy": 5.832812643051147,
      "epoch": 4.049640147831162,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003443126437625472,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2231080561876297,
      "num_tokens": 90282626.0,
      "step": 52050
    },
    {
      "entropy": 5.892820358276367,
      "epoch": 4.0500291772028785,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003442862755090719,
      "loss": 4.904,
      "mean_token_accuracy": 0.22122337967157363,
      "num_tokens": 90290874.0,
      "step": 52055
    },
    {
      "entropy": 5.781940221786499,
      "epoch": 4.050418206574596,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034425990620428763,
      "loss": 4.8049,
      "mean_token_accuracy": 0.22734990566968918,
      "num_tokens": 90299486.0,
      "step": 52060
    },
    {
      "entropy": 5.878848695755005,
      "epoch": 4.050807235946314,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003442335358485946,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2167822912335396,
      "num_tokens": 90307971.0,
      "step": 52065
    },
    {
      "entropy": 5.816729354858398,
      "epoch": 4.051196265318032,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003442071644423928,
      "loss": 4.9019,
      "mean_token_accuracy": 0.21419641524553298,
      "num_tokens": 90316622.0,
      "step": 52070
    },
    {
      "entropy": 5.863480186462402,
      "epoch": 4.051585294689749,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003441807919860824,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21321547776460648,
      "num_tokens": 90325434.0,
      "step": 52075
    },
    {
      "entropy": 5.8772166728973385,
      "epoch": 4.051974324061467,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00034415441848006364,
      "loss": 4.9929,
      "mean_token_accuracy": 0.20999720692634583,
      "num_tokens": 90333927.0,
      "step": 52080
    },
    {
      "entropy": 5.833747434616089,
      "epoch": 4.052363353433184,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034412804392473665,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2228098377585411,
      "num_tokens": 90342203.0,
      "step": 52085
    },
    {
      "entropy": 5.857844018936158,
      "epoch": 4.0527523828049015,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003441016683205016,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22196027338504792,
      "num_tokens": 90350176.0,
      "step": 52090
    },
    {
      "entropy": 5.881503534317017,
      "epoch": 4.053141412176619,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034407529166775874,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2073973923921585,
      "num_tokens": 90358919.0,
      "step": 52095
    },
    {
      "entropy": 5.856105184555053,
      "epoch": 4.053530441548337,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003440489139669083,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21496287882328033,
      "num_tokens": 90367544.0,
      "step": 52100
    },
    {
      "entropy": 5.833160543441773,
      "epoch": 4.053919470920055,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034402253521835036,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21351415365934373,
      "num_tokens": 90377428.0,
      "step": 52105
    },
    {
      "entropy": 5.915317058563232,
      "epoch": 4.054308500291772,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003439961554224855,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21403395831584932,
      "num_tokens": 90385816.0,
      "step": 52110
    },
    {
      "entropy": 5.827901792526245,
      "epoch": 4.05469752966349,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003439697745797137,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22619945406913758,
      "num_tokens": 90394273.0,
      "step": 52115
    },
    {
      "entropy": 5.7623570442199705,
      "epoch": 4.055086559035207,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034394339269043533,
      "loss": 4.8073,
      "mean_token_accuracy": 0.22323550581932067,
      "num_tokens": 90402561.0,
      "step": 52120
    },
    {
      "entropy": 5.791211462020874,
      "epoch": 4.0554755884069245,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003439170097550507,
      "loss": 4.976,
      "mean_token_accuracy": 0.2046961486339569,
      "num_tokens": 90410834.0,
      "step": 52125
    },
    {
      "entropy": 5.875929498672486,
      "epoch": 4.055864617778642,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003438906257739602,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22148752063512803,
      "num_tokens": 90419938.0,
      "step": 52130
    },
    {
      "entropy": 5.905559349060058,
      "epoch": 4.05625364715036,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003438642407475641,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21299238204956056,
      "num_tokens": 90427980.0,
      "step": 52135
    },
    {
      "entropy": 5.866780853271484,
      "epoch": 4.0566426765220776,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003438378546762627,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21856786012649537,
      "num_tokens": 90435816.0,
      "step": 52140
    },
    {
      "entropy": 5.864970064163208,
      "epoch": 4.057031705893795,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003438114675604565,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21690584272146224,
      "num_tokens": 90444456.0,
      "step": 52145
    },
    {
      "entropy": 5.792308568954468,
      "epoch": 4.057420735265513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003437850794005457,
      "loss": 4.8823,
      "mean_token_accuracy": 0.21493453681468963,
      "num_tokens": 90453773.0,
      "step": 52150
    },
    {
      "entropy": 5.811292600631714,
      "epoch": 4.05780976463723,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003437586901969309,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2103012889623642,
      "num_tokens": 90462970.0,
      "step": 52155
    },
    {
      "entropy": 5.84728102684021,
      "epoch": 4.0581987940089475,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003437322999500124,
      "loss": 4.935,
      "mean_token_accuracy": 0.21935613602399825,
      "num_tokens": 90471799.0,
      "step": 52160
    },
    {
      "entropy": 5.833136463165284,
      "epoch": 4.058587823380665,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034370590866019063,
      "loss": 4.88,
      "mean_token_accuracy": 0.2238041177392006,
      "num_tokens": 90480032.0,
      "step": 52165
    },
    {
      "entropy": 5.842194938659668,
      "epoch": 4.058976852752383,
      "grad_norm": 1.4375,
      "learning_rate": 0.000343679516327866,
      "loss": 4.8896,
      "mean_token_accuracy": 0.21448867172002792,
      "num_tokens": 90488216.0,
      "step": 52170
    },
    {
      "entropy": 5.831223821640014,
      "epoch": 4.0593658821241005,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003436531229534391,
      "loss": 4.984,
      "mean_token_accuracy": 0.21151898205280303,
      "num_tokens": 90497209.0,
      "step": 52175
    },
    {
      "entropy": 5.886144065856934,
      "epoch": 4.059754911495818,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003436267285373103,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2138896703720093,
      "num_tokens": 90505255.0,
      "step": 52180
    },
    {
      "entropy": 5.850253963470459,
      "epoch": 4.060143940867536,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034360033307988014,
      "loss": 4.8415,
      "mean_token_accuracy": 0.22359334230422973,
      "num_tokens": 90513508.0,
      "step": 52185
    },
    {
      "entropy": 5.768360900878906,
      "epoch": 4.060532970239253,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034357393658154914,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2197498008608818,
      "num_tokens": 90522482.0,
      "step": 52190
    },
    {
      "entropy": 5.777441596984863,
      "epoch": 4.0609219996109704,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003435475390427178,
      "loss": 4.974,
      "mean_token_accuracy": 0.21254347115755082,
      "num_tokens": 90532024.0,
      "step": 52195
    },
    {
      "entropy": 5.776740837097168,
      "epoch": 4.061311028982688,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034352114046378655,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22101172804832458,
      "num_tokens": 90540403.0,
      "step": 52200
    },
    {
      "entropy": 5.869071531295776,
      "epoch": 4.061700058354406,
      "grad_norm": 1.453125,
      "learning_rate": 0.00034349474084515607,
      "loss": 4.8846,
      "mean_token_accuracy": 0.21936811208724977,
      "num_tokens": 90548840.0,
      "step": 52205
    },
    {
      "entropy": 5.831562757492065,
      "epoch": 4.0620890877261235,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034346834018722706,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2122933954000473,
      "num_tokens": 90557145.0,
      "step": 52210
    },
    {
      "entropy": 5.919728803634643,
      "epoch": 4.062478117097841,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034344193849039997,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20833453983068467,
      "num_tokens": 90566109.0,
      "step": 52215
    },
    {
      "entropy": 5.8889837741851805,
      "epoch": 4.062867146469558,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003434155357550753,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21623438894748687,
      "num_tokens": 90574652.0,
      "step": 52220
    },
    {
      "entropy": 5.855609655380249,
      "epoch": 4.063256175841276,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034338913198165373,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21853173971176149,
      "num_tokens": 90582723.0,
      "step": 52225
    },
    {
      "entropy": 5.920698833465576,
      "epoch": 4.063645205212993,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000343362727170536,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21547368317842483,
      "num_tokens": 90591844.0,
      "step": 52230
    },
    {
      "entropy": 5.81311001777649,
      "epoch": 4.064034234584711,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003433363213221227,
      "loss": 4.8241,
      "mean_token_accuracy": 0.22275959551334382,
      "num_tokens": 90600102.0,
      "step": 52235
    },
    {
      "entropy": 5.826549291610718,
      "epoch": 4.064423263956429,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034330991443681444,
      "loss": 4.8712,
      "mean_token_accuracy": 0.21934038251638413,
      "num_tokens": 90607909.0,
      "step": 52240
    },
    {
      "entropy": 5.7224109172821045,
      "epoch": 4.0648122933281465,
      "grad_norm": 1.421875,
      "learning_rate": 0.000343283506515012,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22111040204763413,
      "num_tokens": 90616110.0,
      "step": 52245
    },
    {
      "entropy": 5.8070876121521,
      "epoch": 4.065201322699864,
      "grad_norm": 1.53125,
      "learning_rate": 0.00034325709755711606,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22416778951883315,
      "num_tokens": 90625096.0,
      "step": 52250
    },
    {
      "entropy": 5.842015647888184,
      "epoch": 4.065590352071581,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034323068756352725,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21566929668188095,
      "num_tokens": 90633369.0,
      "step": 52255
    },
    {
      "entropy": 5.822929334640503,
      "epoch": 4.065979381443299,
      "grad_norm": 1.375,
      "learning_rate": 0.0003432042765346464,
      "loss": 4.8732,
      "mean_token_accuracy": 0.21982914358377456,
      "num_tokens": 90641414.0,
      "step": 52260
    },
    {
      "entropy": 5.868851995468139,
      "epoch": 4.066368410815016,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003431778644708742,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21670421212911606,
      "num_tokens": 90649403.0,
      "step": 52265
    },
    {
      "entropy": 5.830531406402588,
      "epoch": 4.066757440186734,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003431514513726114,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2148990511894226,
      "num_tokens": 90657590.0,
      "step": 52270
    },
    {
      "entropy": 5.8385009765625,
      "epoch": 4.067146469558452,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003431250372402588,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2237924724817276,
      "num_tokens": 90666448.0,
      "step": 52275
    },
    {
      "entropy": 5.7893718719482425,
      "epoch": 4.0675354989301695,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034309862207421724,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2164943262934685,
      "num_tokens": 90675452.0,
      "step": 52280
    },
    {
      "entropy": 5.836388397216797,
      "epoch": 4.067924528301887,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034307220587488743,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21053467243909835,
      "num_tokens": 90684446.0,
      "step": 52285
    },
    {
      "entropy": 5.850708770751953,
      "epoch": 4.068313557673604,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034304578864267026,
      "loss": 4.867,
      "mean_token_accuracy": 0.2224138155579567,
      "num_tokens": 90693348.0,
      "step": 52290
    },
    {
      "entropy": 5.810446786880493,
      "epoch": 4.068702587045322,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00034301937037796654,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21514556407928467,
      "num_tokens": 90701611.0,
      "step": 52295
    },
    {
      "entropy": 5.857712507247925,
      "epoch": 4.069091616417039,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034299295108117716,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20824792832136155,
      "num_tokens": 90710992.0,
      "step": 52300
    },
    {
      "entropy": 5.85392279624939,
      "epoch": 4.069480645788757,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034296653075270296,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21850242465734482,
      "num_tokens": 90719684.0,
      "step": 52305
    },
    {
      "entropy": 5.811822557449341,
      "epoch": 4.069869675160475,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034294010939294486,
      "loss": 4.87,
      "mean_token_accuracy": 0.2256393924355507,
      "num_tokens": 90728579.0,
      "step": 52310
    },
    {
      "entropy": 5.803230047225952,
      "epoch": 4.0702587045321925,
      "grad_norm": 1.5,
      "learning_rate": 0.00034291368700230376,
      "loss": 4.847,
      "mean_token_accuracy": 0.2259753942489624,
      "num_tokens": 90736044.0,
      "step": 52315
    },
    {
      "entropy": 5.7889467716217045,
      "epoch": 4.070647733903909,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003428872635811804,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21657328605651854,
      "num_tokens": 90745388.0,
      "step": 52320
    },
    {
      "entropy": 5.802664136886596,
      "epoch": 4.071036763275627,
      "grad_norm": 1.359375,
      "learning_rate": 0.000342860839129976,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2154616117477417,
      "num_tokens": 90753321.0,
      "step": 52325
    },
    {
      "entropy": 5.79578127861023,
      "epoch": 4.071425792647345,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003428344136490914,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22226570546627045,
      "num_tokens": 90762196.0,
      "step": 52330
    },
    {
      "entropy": 5.868144464492798,
      "epoch": 4.071814822019062,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034280798713892744,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21248525381088257,
      "num_tokens": 90770892.0,
      "step": 52335
    },
    {
      "entropy": 5.8818295955657955,
      "epoch": 4.07220385139078,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003427815595998853,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21553434580564498,
      "num_tokens": 90779457.0,
      "step": 52340
    },
    {
      "entropy": 5.846967172622681,
      "epoch": 4.072592880762498,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003427551310323658,
      "loss": 4.9023,
      "mean_token_accuracy": 0.21872878670692444,
      "num_tokens": 90787945.0,
      "step": 52345
    },
    {
      "entropy": 5.8519257545471195,
      "epoch": 4.0729819101342155,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034272870143677015,
      "loss": 4.9417,
      "mean_token_accuracy": 0.20986691415309905,
      "num_tokens": 90796613.0,
      "step": 52350
    },
    {
      "entropy": 5.8603569030761715,
      "epoch": 4.073370939505932,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034270227081349913,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21690395921468736,
      "num_tokens": 90805175.0,
      "step": 52355
    },
    {
      "entropy": 5.795033168792725,
      "epoch": 4.07375996887765,
      "grad_norm": 1.4375,
      "learning_rate": 0.000342675839162954,
      "loss": 4.8329,
      "mean_token_accuracy": 0.21831076592206955,
      "num_tokens": 90813686.0,
      "step": 52360
    },
    {
      "entropy": 5.787351274490357,
      "epoch": 4.074148998249368,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034264940648553565,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21448871195316316,
      "num_tokens": 90822692.0,
      "step": 52365
    },
    {
      "entropy": 5.84473819732666,
      "epoch": 4.074538027621085,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003426229727816453,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21837498098611832,
      "num_tokens": 90831191.0,
      "step": 52370
    },
    {
      "entropy": 5.788085842132569,
      "epoch": 4.074927056992803,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003425965380516839,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22172005772590636,
      "num_tokens": 90839597.0,
      "step": 52375
    },
    {
      "entropy": 5.8115699768066404,
      "epoch": 4.075316086364521,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003425701022960527,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2167083814740181,
      "num_tokens": 90848208.0,
      "step": 52380
    },
    {
      "entropy": 5.785486364364624,
      "epoch": 4.0757051157362385,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034254366551515276,
      "loss": 4.8159,
      "mean_token_accuracy": 0.22201936244964598,
      "num_tokens": 90856747.0,
      "step": 52385
    },
    {
      "entropy": 5.829539489746094,
      "epoch": 4.076094145107955,
      "grad_norm": 1.375,
      "learning_rate": 0.0003425172277093852,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2162596568465233,
      "num_tokens": 90864908.0,
      "step": 52390
    },
    {
      "entropy": 5.720896911621094,
      "epoch": 4.076483174479673,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000342490788879151,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22386638671159745,
      "num_tokens": 90873950.0,
      "step": 52395
    },
    {
      "entropy": 5.786791610717773,
      "epoch": 4.076872203851391,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034246434902485156,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21612803041934966,
      "num_tokens": 90882418.0,
      "step": 52400
    },
    {
      "entropy": 5.888134098052978,
      "epoch": 4.077261233223108,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034243790814688816,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20419965088367462,
      "num_tokens": 90892109.0,
      "step": 52405
    },
    {
      "entropy": 5.920904970169067,
      "epoch": 4.077650262594826,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034241146624566175,
      "loss": 4.927,
      "mean_token_accuracy": 0.21371452808380126,
      "num_tokens": 90899949.0,
      "step": 52410
    },
    {
      "entropy": 5.929858922958374,
      "epoch": 4.078039291966544,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003423850233215737,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21624796837568283,
      "num_tokens": 90909302.0,
      "step": 52415
    },
    {
      "entropy": 5.855721855163575,
      "epoch": 4.078428321338261,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003423585793750251,
      "loss": 4.965,
      "mean_token_accuracy": 0.21351767033338548,
      "num_tokens": 90918394.0,
      "step": 52420
    },
    {
      "entropy": 5.828109312057495,
      "epoch": 4.078817350709978,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034233213440641724,
      "loss": 4.877,
      "mean_token_accuracy": 0.2179236575961113,
      "num_tokens": 90926495.0,
      "step": 52425
    },
    {
      "entropy": 5.786381816864013,
      "epoch": 4.079206380081696,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034230568841615145,
      "loss": 4.914,
      "mean_token_accuracy": 0.22172889709472657,
      "num_tokens": 90935036.0,
      "step": 52430
    },
    {
      "entropy": 5.8707057476043705,
      "epoch": 4.079595409453414,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000342279241404629,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2085822969675064,
      "num_tokens": 90943304.0,
      "step": 52435
    },
    {
      "entropy": 5.8498293399810795,
      "epoch": 4.079984438825131,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034225279337225106,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21728845238685607,
      "num_tokens": 90953310.0,
      "step": 52440
    },
    {
      "entropy": 5.793787288665771,
      "epoch": 4.080373468196849,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003422263443194191,
      "loss": 4.8503,
      "mean_token_accuracy": 0.2220021277666092,
      "num_tokens": 90961621.0,
      "step": 52445
    },
    {
      "entropy": 5.833256816864013,
      "epoch": 4.080762497568567,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003421998942465344,
      "loss": 4.986,
      "mean_token_accuracy": 0.20948358327150346,
      "num_tokens": 90970196.0,
      "step": 52450
    },
    {
      "entropy": 5.797112321853637,
      "epoch": 4.081151526940284,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034217344315399823,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21189952939748763,
      "num_tokens": 90978855.0,
      "step": 52455
    },
    {
      "entropy": 5.8323376178741455,
      "epoch": 4.081540556312001,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000342146991042212,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21682592779397963,
      "num_tokens": 90986890.0,
      "step": 52460
    },
    {
      "entropy": 5.841513729095459,
      "epoch": 4.081929585683719,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034212053791157695,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2155874952673912,
      "num_tokens": 90995791.0,
      "step": 52465
    },
    {
      "entropy": 5.802629375457764,
      "epoch": 4.082318615055437,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034209408376249464,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22121082693338395,
      "num_tokens": 91003719.0,
      "step": 52470
    },
    {
      "entropy": 5.809089231491089,
      "epoch": 4.082707644427154,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003420676285953664,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2166763037443161,
      "num_tokens": 91012330.0,
      "step": 52475
    },
    {
      "entropy": 5.815928411483765,
      "epoch": 4.083096673798872,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003420411724105937,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2213539496064186,
      "num_tokens": 91020945.0,
      "step": 52480
    },
    {
      "entropy": 5.89399905204773,
      "epoch": 4.08348570317059,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034201471520857793,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2155792385339737,
      "num_tokens": 91030602.0,
      "step": 52485
    },
    {
      "entropy": 5.796461153030395,
      "epoch": 4.083874732542307,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003419882569897204,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22210535705089568,
      "num_tokens": 91039429.0,
      "step": 52490
    },
    {
      "entropy": 5.78185625076294,
      "epoch": 4.084263761914024,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00034196179775442274,
      "loss": 4.8713,
      "mean_token_accuracy": 0.21495497673749925,
      "num_tokens": 91047332.0,
      "step": 52495
    },
    {
      "entropy": 5.819876766204834,
      "epoch": 4.084652791285742,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003419353375030864,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2224995404481888,
      "num_tokens": 91055883.0,
      "step": 52500
    },
    {
      "entropy": 5.852490949630737,
      "epoch": 4.08504182065746,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003419088762361128,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21160476803779601,
      "num_tokens": 91065520.0,
      "step": 52505
    },
    {
      "entropy": 5.8193052291870115,
      "epoch": 4.085430850029177,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003418824139539035,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2181520029902458,
      "num_tokens": 91074182.0,
      "step": 52510
    },
    {
      "entropy": 5.860599184036255,
      "epoch": 4.085819879400895,
      "grad_norm": 1.25,
      "learning_rate": 0.0003418559506568601,
      "loss": 4.9796,
      "mean_token_accuracy": 0.2113502189517021,
      "num_tokens": 91083721.0,
      "step": 52515
    },
    {
      "entropy": 5.891854381561279,
      "epoch": 4.086208908772612,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034182948634538403,
      "loss": 4.95,
      "mean_token_accuracy": 0.2084916740655899,
      "num_tokens": 91091986.0,
      "step": 52520
    },
    {
      "entropy": 5.857538032531738,
      "epoch": 4.08659793814433,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003418030210198769,
      "loss": 4.919,
      "mean_token_accuracy": 0.21881368160247802,
      "num_tokens": 91101183.0,
      "step": 52525
    },
    {
      "entropy": 5.908865976333618,
      "epoch": 4.086986967516047,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034177655468074027,
      "loss": 5.0355,
      "mean_token_accuracy": 0.20624226182699204,
      "num_tokens": 91109574.0,
      "step": 52530
    },
    {
      "entropy": 5.851313781738281,
      "epoch": 4.087375996887765,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003417500873283756,
      "loss": 4.9145,
      "mean_token_accuracy": 0.20936193615198134,
      "num_tokens": 91117820.0,
      "step": 52535
    },
    {
      "entropy": 5.890816116333008,
      "epoch": 4.087765026259483,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003417236189631846,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2096324861049652,
      "num_tokens": 91126958.0,
      "step": 52540
    },
    {
      "entropy": 5.79441466331482,
      "epoch": 4.0881540556312,
      "grad_norm": 1.390625,
      "learning_rate": 0.000341697149585569,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2166468933224678,
      "num_tokens": 91135727.0,
      "step": 52545
    },
    {
      "entropy": 5.777029800415039,
      "epoch": 4.088543085002918,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003416706791959302,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2108830139040947,
      "num_tokens": 91145284.0,
      "step": 52550
    },
    {
      "entropy": 5.861325740814209,
      "epoch": 4.088932114374635,
      "grad_norm": 1.375,
      "learning_rate": 0.00034164420779467003,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21862999200820923,
      "num_tokens": 91154106.0,
      "step": 52555
    },
    {
      "entropy": 5.905989265441894,
      "epoch": 4.089321143746353,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003416177353821901,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21375168412923812,
      "num_tokens": 91162936.0,
      "step": 52560
    },
    {
      "entropy": 5.840847778320312,
      "epoch": 4.08971017311807,
      "grad_norm": 1.421875,
      "learning_rate": 0.000341591261958892,
      "loss": 4.9161,
      "mean_token_accuracy": 0.213680000603199,
      "num_tokens": 91171902.0,
      "step": 52565
    },
    {
      "entropy": 5.8928282260894775,
      "epoch": 4.090099202489788,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00034156478752517754,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21248565018177032,
      "num_tokens": 91180398.0,
      "step": 52570
    },
    {
      "entropy": 5.865040969848633,
      "epoch": 4.090488231861506,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034153831208144837,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22061435729265214,
      "num_tokens": 91188433.0,
      "step": 52575
    },
    {
      "entropy": 5.881752681732178,
      "epoch": 4.090877261233223,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003415118356281062,
      "loss": 5.0708,
      "mean_token_accuracy": 0.19840618073940278,
      "num_tokens": 91196578.0,
      "step": 52580
    },
    {
      "entropy": 5.881213617324829,
      "epoch": 4.091266290604941,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003414853581655528,
      "loss": 5.0081,
      "mean_token_accuracy": 0.21125536262989045,
      "num_tokens": 91206085.0,
      "step": 52585
    },
    {
      "entropy": 5.86044979095459,
      "epoch": 4.091655319976658,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034145887969419,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21598075330257416,
      "num_tokens": 91215222.0,
      "step": 52590
    },
    {
      "entropy": 5.841968297958374,
      "epoch": 4.092044349348376,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034143240021441935,
      "loss": 4.9457,
      "mean_token_accuracy": 0.20829401463270186,
      "num_tokens": 91224249.0,
      "step": 52595
    },
    {
      "entropy": 5.835330533981323,
      "epoch": 4.092433378720093,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003414059197266428,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22637263387441636,
      "num_tokens": 91233157.0,
      "step": 52600
    },
    {
      "entropy": 5.890692520141601,
      "epoch": 4.092822408091811,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034137943823126215,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21913779079914092,
      "num_tokens": 91242639.0,
      "step": 52605
    },
    {
      "entropy": 5.812398147583008,
      "epoch": 4.093211437463529,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003413529557286792,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21963949948549272,
      "num_tokens": 91251642.0,
      "step": 52610
    },
    {
      "entropy": 5.894945478439331,
      "epoch": 4.093600466835246,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003413264722192957,
      "loss": 4.9971,
      "mean_token_accuracy": 0.2151159882545471,
      "num_tokens": 91260706.0,
      "step": 52615
    },
    {
      "entropy": 5.966597080230713,
      "epoch": 4.093989496206964,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003412999877035136,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20358191132545472,
      "num_tokens": 91269840.0,
      "step": 52620
    },
    {
      "entropy": 5.872618293762207,
      "epoch": 4.094378525578681,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00034127350218173463,
      "loss": 4.9888,
      "mean_token_accuracy": 0.20761465430259704,
      "num_tokens": 91278379.0,
      "step": 52625
    },
    {
      "entropy": 5.867011070251465,
      "epoch": 4.094767554950399,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034124701565436076,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20673823356628418,
      "num_tokens": 91286620.0,
      "step": 52630
    },
    {
      "entropy": 5.81561770439148,
      "epoch": 4.095156584322116,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000341220528121794,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2230576291680336,
      "num_tokens": 91294559.0,
      "step": 52635
    },
    {
      "entropy": 5.863329744338989,
      "epoch": 4.095545613693834,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000341194039584436,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21729220002889632,
      "num_tokens": 91302807.0,
      "step": 52640
    },
    {
      "entropy": 5.84819917678833,
      "epoch": 4.095934643065552,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003411675500426888,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2161501467227936,
      "num_tokens": 91310903.0,
      "step": 52645
    },
    {
      "entropy": 5.84290041923523,
      "epoch": 4.096323672437269,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034114105949695445,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21008631438016892,
      "num_tokens": 91320037.0,
      "step": 52650
    },
    {
      "entropy": 5.822517442703247,
      "epoch": 4.096712701808986,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003411145679476347,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21835718899965287,
      "num_tokens": 91329257.0,
      "step": 52655
    },
    {
      "entropy": 5.842272329330444,
      "epoch": 4.097101731180704,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003410880753951317,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22420183420181275,
      "num_tokens": 91337225.0,
      "step": 52660
    },
    {
      "entropy": 5.834420108795166,
      "epoch": 4.097490760552422,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003410615818398473,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21519297808408738,
      "num_tokens": 91346009.0,
      "step": 52665
    },
    {
      "entropy": 5.885254049301148,
      "epoch": 4.097879789924139,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003410350872821835,
      "loss": 5.0042,
      "mean_token_accuracy": 0.2128468245267868,
      "num_tokens": 91356050.0,
      "step": 52670
    },
    {
      "entropy": 5.888118314743042,
      "epoch": 4.098268819295857,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00034100859172254245,
      "loss": 4.9868,
      "mean_token_accuracy": 0.20906075984239578,
      "num_tokens": 91364926.0,
      "step": 52675
    },
    {
      "entropy": 5.81418080329895,
      "epoch": 4.098657848667575,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00034098209516132607,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2179558113217354,
      "num_tokens": 91374129.0,
      "step": 52680
    },
    {
      "entropy": 5.833598470687866,
      "epoch": 4.099046878039292,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003409555975989363,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2185434028506279,
      "num_tokens": 91383201.0,
      "step": 52685
    },
    {
      "entropy": 5.757265233993531,
      "epoch": 4.099435907411009,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034092909903577547,
      "loss": 4.8752,
      "mean_token_accuracy": 0.21811351031064988,
      "num_tokens": 91391979.0,
      "step": 52690
    },
    {
      "entropy": 5.852102565765381,
      "epoch": 4.099824936782727,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003409025994722454,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2155961886048317,
      "num_tokens": 91400127.0,
      "step": 52695
    },
    {
      "entropy": 5.852536296844482,
      "epoch": 4.100213966154445,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003408760989087482,
      "loss": 4.848,
      "mean_token_accuracy": 0.22196063846349717,
      "num_tokens": 91408329.0,
      "step": 52700
    },
    {
      "entropy": 5.772577476501465,
      "epoch": 4.100602995526162,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00034084959734568616,
      "loss": 4.8612,
      "mean_token_accuracy": 0.22442913800477982,
      "num_tokens": 91417156.0,
      "step": 52705
    },
    {
      "entropy": 5.815363454818725,
      "epoch": 4.10099202489788,
      "grad_norm": 1.34375,
      "learning_rate": 0.00034082309478346127,
      "loss": 4.8499,
      "mean_token_accuracy": 0.21924446672201156,
      "num_tokens": 91425776.0,
      "step": 52710
    },
    {
      "entropy": 5.819711780548095,
      "epoch": 4.101381054269598,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00034079659122247574,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2179477781057358,
      "num_tokens": 91435178.0,
      "step": 52715
    },
    {
      "entropy": 5.822615003585815,
      "epoch": 4.101770083641315,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003407700866631317,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2139988884329796,
      "num_tokens": 91445104.0,
      "step": 52720
    },
    {
      "entropy": 5.878048610687256,
      "epoch": 4.102159113013032,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003407435811058312,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2178898572921753,
      "num_tokens": 91453221.0,
      "step": 52725
    },
    {
      "entropy": 5.807293891906738,
      "epoch": 4.10254814238475,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003407170745509764,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22158105075359344,
      "num_tokens": 91462157.0,
      "step": 52730
    },
    {
      "entropy": 5.839710664749146,
      "epoch": 4.1029371717564675,
      "grad_norm": 1.390625,
      "learning_rate": 0.00034069056699896974,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2061545416712761,
      "num_tokens": 91471070.0,
      "step": 52735
    },
    {
      "entropy": 5.849022912979126,
      "epoch": 4.103326201128185,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003406640584502132,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21637336760759354,
      "num_tokens": 91480388.0,
      "step": 52740
    },
    {
      "entropy": 5.8086577415466305,
      "epoch": 4.103715230499903,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003406375489051091,
      "loss": 4.8236,
      "mean_token_accuracy": 0.22090755105018617,
      "num_tokens": 91488498.0,
      "step": 52745
    },
    {
      "entropy": 5.81243052482605,
      "epoch": 4.104104259871621,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003406110383640597,
      "loss": 4.9598,
      "mean_token_accuracy": 0.218596975505352,
      "num_tokens": 91497273.0,
      "step": 52750
    },
    {
      "entropy": 5.856308794021606,
      "epoch": 4.1044932892433375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034058452682746734,
      "loss": 4.955,
      "mean_token_accuracy": 0.20881213247776031,
      "num_tokens": 91505727.0,
      "step": 52755
    },
    {
      "entropy": 5.784570407867432,
      "epoch": 4.104882318615055,
      "grad_norm": 1.296875,
      "learning_rate": 0.000340558014295734,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22463632076978685,
      "num_tokens": 91514551.0,
      "step": 52760
    },
    {
      "entropy": 5.824643325805664,
      "epoch": 4.105271347986773,
      "grad_norm": 1.546875,
      "learning_rate": 0.00034053150076926214,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22397587299346924,
      "num_tokens": 91522709.0,
      "step": 52765
    },
    {
      "entropy": 5.775627279281617,
      "epoch": 4.1056603773584905,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003405049862484541,
      "loss": 4.8748,
      "mean_token_accuracy": 0.21368742138147354,
      "num_tokens": 91531187.0,
      "step": 52770
    },
    {
      "entropy": 5.852588653564453,
      "epoch": 4.106049406730208,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003404784707337122,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21157725900411606,
      "num_tokens": 91539698.0,
      "step": 52775
    },
    {
      "entropy": 5.884338665008545,
      "epoch": 4.106438436101926,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034045195422543865,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21355078518390655,
      "num_tokens": 91548391.0,
      "step": 52780
    },
    {
      "entropy": 5.812026166915894,
      "epoch": 4.106827465473644,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00034042543672403594,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22040676772594453,
      "num_tokens": 91556680.0,
      "step": 52785
    },
    {
      "entropy": 5.799113702774048,
      "epoch": 4.10721649484536,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034039891822990634,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2183515578508377,
      "num_tokens": 91566899.0,
      "step": 52790
    },
    {
      "entropy": 5.853789854049682,
      "epoch": 4.107605524217078,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003403723987434523,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20488266795873641,
      "num_tokens": 91575271.0,
      "step": 52795
    },
    {
      "entropy": 5.898279905319214,
      "epoch": 4.107994553588796,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034034587826507604,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21194139420986174,
      "num_tokens": 91583580.0,
      "step": 52800
    },
    {
      "entropy": 5.819307041168213,
      "epoch": 4.1083835829605135,
      "grad_norm": 1.3125,
      "learning_rate": 0.00034031935679518016,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2176297202706337,
      "num_tokens": 91592565.0,
      "step": 52805
    },
    {
      "entropy": 5.794135808944702,
      "epoch": 4.108772612332231,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000340292834334167,
      "loss": 4.9758,
      "mean_token_accuracy": 0.20166356563568116,
      "num_tokens": 91600873.0,
      "step": 52810
    },
    {
      "entropy": 5.812370872497558,
      "epoch": 4.109161641703949,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000340266310882439,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2180996611714363,
      "num_tokens": 91609761.0,
      "step": 52815
    },
    {
      "entropy": 5.839925146102905,
      "epoch": 4.109550671075667,
      "grad_norm": 1.4375,
      "learning_rate": 0.00034023978644039864,
      "loss": 4.888,
      "mean_token_accuracy": 0.22318151891231536,
      "num_tokens": 91619498.0,
      "step": 52820
    },
    {
      "entropy": 5.888211393356324,
      "epoch": 4.109939700447383,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003402132610084483,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21416703313589097,
      "num_tokens": 91628665.0,
      "step": 52825
    },
    {
      "entropy": 5.901396226882935,
      "epoch": 4.110328729819101,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034018673458699055,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21156488060951234,
      "num_tokens": 91636850.0,
      "step": 52830
    },
    {
      "entropy": 5.801259517669678,
      "epoch": 4.110717759190819,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003401602071764279,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22276766151189803,
      "num_tokens": 91646142.0,
      "step": 52835
    },
    {
      "entropy": 5.783494472503662,
      "epoch": 4.1111067885625365,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034013367877716284,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2112602949142456,
      "num_tokens": 91654637.0,
      "step": 52840
    },
    {
      "entropy": 5.84419994354248,
      "epoch": 4.111495817934254,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003401071493895977,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21010540425777435,
      "num_tokens": 91663208.0,
      "step": 52845
    },
    {
      "entropy": 5.796760320663452,
      "epoch": 4.111884847305972,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003400806190141354,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2216264098882675,
      "num_tokens": 91671530.0,
      "step": 52850
    },
    {
      "entropy": 5.82997145652771,
      "epoch": 4.112273876677689,
      "grad_norm": 1.421875,
      "learning_rate": 0.00034005408765117815,
      "loss": 4.8779,
      "mean_token_accuracy": 0.21789925247430803,
      "num_tokens": 91680056.0,
      "step": 52855
    },
    {
      "entropy": 5.852447462081909,
      "epoch": 4.112662906049406,
      "grad_norm": 1.40625,
      "learning_rate": 0.00034002755530112877,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21510377377271653,
      "num_tokens": 91688501.0,
      "step": 52860
    },
    {
      "entropy": 5.803278923034668,
      "epoch": 4.113051935421124,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003400010219643897,
      "loss": 4.8682,
      "mean_token_accuracy": 0.21593920290470123,
      "num_tokens": 91696869.0,
      "step": 52865
    },
    {
      "entropy": 5.804745626449585,
      "epoch": 4.113440964792842,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003399744876413636,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22157642394304275,
      "num_tokens": 91705300.0,
      "step": 52870
    },
    {
      "entropy": 5.850197887420654,
      "epoch": 4.1138299941645595,
      "grad_norm": 1.40625,
      "learning_rate": 0.000339947952332453,
      "loss": 4.9912,
      "mean_token_accuracy": 0.21330345273017884,
      "num_tokens": 91714218.0,
      "step": 52875
    },
    {
      "entropy": 5.88608341217041,
      "epoch": 4.114219023536277,
      "grad_norm": 1.5,
      "learning_rate": 0.00033992141603806064,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2196979805827141,
      "num_tokens": 91723185.0,
      "step": 52880
    },
    {
      "entropy": 5.9192382335662845,
      "epoch": 4.114608052907995,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033989487875858915,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20224591046571733,
      "num_tokens": 91732035.0,
      "step": 52885
    },
    {
      "entropy": 5.860176038742066,
      "epoch": 4.114997082279712,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033986834049444113,
      "loss": 4.8915,
      "mean_token_accuracy": 0.2140567734837532,
      "num_tokens": 91740657.0,
      "step": 52890
    },
    {
      "entropy": 5.8893311500549315,
      "epoch": 4.115386111651429,
      "grad_norm": 1.375,
      "learning_rate": 0.00033984180124601936,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20481197088956832,
      "num_tokens": 91749894.0,
      "step": 52895
    },
    {
      "entropy": 5.7940216064453125,
      "epoch": 4.115775141023147,
      "grad_norm": 1.578125,
      "learning_rate": 0.00033981526101372635,
      "loss": 4.8671,
      "mean_token_accuracy": 0.21960383653640747,
      "num_tokens": 91759132.0,
      "step": 52900
    },
    {
      "entropy": 5.851868295669556,
      "epoch": 4.116164170394865,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000339788719797965,
      "loss": 4.9795,
      "mean_token_accuracy": 0.20549817979335785,
      "num_tokens": 91767589.0,
      "step": 52905
    },
    {
      "entropy": 5.840805578231811,
      "epoch": 4.1165531997665825,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003397621775991379,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21690140217542647,
      "num_tokens": 91776444.0,
      "step": 52910
    },
    {
      "entropy": 5.921496200561523,
      "epoch": 4.1169422291383,
      "grad_norm": 1.5,
      "learning_rate": 0.0003397356344176479,
      "loss": 5.0591,
      "mean_token_accuracy": 0.20775261521339417,
      "num_tokens": 91785332.0,
      "step": 52915
    },
    {
      "entropy": 5.834307241439819,
      "epoch": 4.117331258510018,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003397090902538976,
      "loss": 4.876,
      "mean_token_accuracy": 0.22727614492177964,
      "num_tokens": 91793798.0,
      "step": 52920
    },
    {
      "entropy": 5.8834545612335205,
      "epoch": 4.117720287881735,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033968254510828995,
      "loss": 5.0275,
      "mean_token_accuracy": 0.21723342537879944,
      "num_tokens": 91802517.0,
      "step": 52925
    },
    {
      "entropy": 5.841442966461182,
      "epoch": 4.118109317253452,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003396559989812275,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2159962922334671,
      "num_tokens": 91811038.0,
      "step": 52930
    },
    {
      "entropy": 5.872584390640259,
      "epoch": 4.11849834662517,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003396294518731133,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2143898382782936,
      "num_tokens": 91819987.0,
      "step": 52935
    },
    {
      "entropy": 5.782623529434204,
      "epoch": 4.118887375996888,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033960290378435,
      "loss": 4.9766,
      "mean_token_accuracy": 0.21360776871442794,
      "num_tokens": 91827919.0,
      "step": 52940
    },
    {
      "entropy": 5.840401887893677,
      "epoch": 4.1192764053686055,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033957635471534045,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2157055541872978,
      "num_tokens": 91836162.0,
      "step": 52945
    },
    {
      "entropy": 5.884769773483276,
      "epoch": 4.119665434740323,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003395498046664875,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21341006755828856,
      "num_tokens": 91844968.0,
      "step": 52950
    },
    {
      "entropy": 5.871742391586304,
      "epoch": 4.120054464112041,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000339523253638194,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2136014521121979,
      "num_tokens": 91852993.0,
      "step": 52955
    },
    {
      "entropy": 5.879800844192505,
      "epoch": 4.120443493483758,
      "grad_norm": 1.375,
      "learning_rate": 0.0003394967016308629,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21868855357170106,
      "num_tokens": 91861435.0,
      "step": 52960
    },
    {
      "entropy": 5.829244995117188,
      "epoch": 4.120832522855475,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003394701486448968,
      "loss": 4.9001,
      "mean_token_accuracy": 0.21847460120916368,
      "num_tokens": 91870149.0,
      "step": 52965
    },
    {
      "entropy": 5.793951511383057,
      "epoch": 4.121221552227193,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033944359468069903,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21318036913871766,
      "num_tokens": 91879110.0,
      "step": 52970
    },
    {
      "entropy": 5.837325763702393,
      "epoch": 4.121610581598911,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033941703973867216,
      "loss": 4.8379,
      "mean_token_accuracy": 0.21839770227670668,
      "num_tokens": 91887433.0,
      "step": 52975
    },
    {
      "entropy": 5.832529211044312,
      "epoch": 4.1219996109706285,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033939048381921935,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21573232263326644,
      "num_tokens": 91896251.0,
      "step": 52980
    },
    {
      "entropy": 5.783792018890381,
      "epoch": 4.122388640342346,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003393639269227434,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22016567289829253,
      "num_tokens": 91905153.0,
      "step": 52985
    },
    {
      "entropy": 5.831999397277832,
      "epoch": 4.122777669714063,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003393373690496473,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2128165066242218,
      "num_tokens": 91913695.0,
      "step": 52990
    },
    {
      "entropy": 5.802449178695679,
      "epoch": 4.123166699085781,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000339310810200334,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21917189955711364,
      "num_tokens": 91922308.0,
      "step": 52995
    },
    {
      "entropy": 5.891654968261719,
      "epoch": 4.123555728457498,
      "grad_norm": 1.375,
      "learning_rate": 0.00033928425037520643,
      "loss": 5.0158,
      "mean_token_accuracy": 0.2106761395931244,
      "num_tokens": 91931267.0,
      "step": 53000
    },
    {
      "entropy": 5.847154569625855,
      "epoch": 4.123944757829216,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003392576895746678,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21056586056947707,
      "num_tokens": 91939943.0,
      "step": 53005
    },
    {
      "entropy": 5.763431549072266,
      "epoch": 4.124333787200934,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003392311277991209,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2137506499886513,
      "num_tokens": 91947886.0,
      "step": 53010
    },
    {
      "entropy": 5.874006557464599,
      "epoch": 4.1247228165726515,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033920456504896895,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20958210229873658,
      "num_tokens": 91956917.0,
      "step": 53015
    },
    {
      "entropy": 5.866233777999878,
      "epoch": 4.125111845944369,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033917800132461493,
      "loss": 4.8788,
      "mean_token_accuracy": 0.21473198384046555,
      "num_tokens": 91965798.0,
      "step": 53020
    },
    {
      "entropy": 5.835514640808105,
      "epoch": 4.125500875316086,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003391514366264619,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22021978050470353,
      "num_tokens": 91974625.0,
      "step": 53025
    },
    {
      "entropy": 5.740242862701416,
      "epoch": 4.125889904687804,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003391248709549129,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22300089448690413,
      "num_tokens": 91983090.0,
      "step": 53030
    },
    {
      "entropy": 5.753971099853516,
      "epoch": 4.126278934059521,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003390983043103711,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21649895012378692,
      "num_tokens": 91991558.0,
      "step": 53035
    },
    {
      "entropy": 5.847105312347412,
      "epoch": 4.126667963431239,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003390717366932395,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21226462721824646,
      "num_tokens": 91999892.0,
      "step": 53040
    },
    {
      "entropy": 5.902355432510376,
      "epoch": 4.127056992802957,
      "grad_norm": 1.375,
      "learning_rate": 0.0003390451681039212,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21563661396503447,
      "num_tokens": 92008562.0,
      "step": 53045
    },
    {
      "entropy": 5.829295825958252,
      "epoch": 4.1274460221746745,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033901859854281954,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21464522927999496,
      "num_tokens": 92017217.0,
      "step": 53050
    },
    {
      "entropy": 5.85615234375,
      "epoch": 4.127835051546391,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003389920280103375,
      "loss": 4.972,
      "mean_token_accuracy": 0.2147740587592125,
      "num_tokens": 92025940.0,
      "step": 53055
    },
    {
      "entropy": 5.780269813537598,
      "epoch": 4.128224080918109,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033896545650687824,
      "loss": 4.896,
      "mean_token_accuracy": 0.2167141169309616,
      "num_tokens": 92034884.0,
      "step": 53060
    },
    {
      "entropy": 5.881720399856567,
      "epoch": 4.128613110289827,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003389388840328451,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21618855744600296,
      "num_tokens": 92043934.0,
      "step": 53065
    },
    {
      "entropy": 5.862233018875122,
      "epoch": 4.129002139661544,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003389123105886411,
      "loss": 4.951,
      "mean_token_accuracy": 0.2094405084848404,
      "num_tokens": 92052078.0,
      "step": 53070
    },
    {
      "entropy": 5.812042331695556,
      "epoch": 4.129391169033262,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003388857361746695,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21577164232730867,
      "num_tokens": 92061054.0,
      "step": 53075
    },
    {
      "entropy": 5.903557538986206,
      "epoch": 4.12978019840498,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033885916079133354,
      "loss": 5.056,
      "mean_token_accuracy": 0.20536187291145325,
      "num_tokens": 92069911.0,
      "step": 53080
    },
    {
      "entropy": 5.873223447799683,
      "epoch": 4.1301692277766975,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033883258443903646,
      "loss": 4.8505,
      "mean_token_accuracy": 0.21839212775230407,
      "num_tokens": 92078450.0,
      "step": 53085
    },
    {
      "entropy": 5.727860069274902,
      "epoch": 4.130558257148414,
      "grad_norm": 1.5,
      "learning_rate": 0.00033880600711818157,
      "loss": 4.901,
      "mean_token_accuracy": 0.21819747984409332,
      "num_tokens": 92087438.0,
      "step": 53090
    },
    {
      "entropy": 5.841617298126221,
      "epoch": 4.130947286520132,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033877942882917203,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22281511425971984,
      "num_tokens": 92095921.0,
      "step": 53095
    },
    {
      "entropy": 5.857618188858032,
      "epoch": 4.13133631589185,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033875284957241117,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2091695934534073,
      "num_tokens": 92104579.0,
      "step": 53100
    },
    {
      "entropy": 5.778086519241333,
      "epoch": 4.131725345263567,
      "grad_norm": 1.375,
      "learning_rate": 0.0003387262693483023,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22095065712928771,
      "num_tokens": 92112749.0,
      "step": 53105
    },
    {
      "entropy": 5.85557336807251,
      "epoch": 4.132114374635285,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033869968815724866,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2153204619884491,
      "num_tokens": 92120977.0,
      "step": 53110
    },
    {
      "entropy": 5.7384131908416744,
      "epoch": 4.132503404007003,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003386731059996537,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2297930747270584,
      "num_tokens": 92129086.0,
      "step": 53115
    },
    {
      "entropy": 5.82343111038208,
      "epoch": 4.1328924333787205,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033864652287592074,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2170750230550766,
      "num_tokens": 92137831.0,
      "step": 53120
    },
    {
      "entropy": 5.889012670516967,
      "epoch": 4.133281462750437,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033861993878645305,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21085717529058456,
      "num_tokens": 92146441.0,
      "step": 53125
    },
    {
      "entropy": 5.912505626678467,
      "epoch": 4.133670492122155,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000338593353731654,
      "loss": 4.9352,
      "mean_token_accuracy": 0.218081296980381,
      "num_tokens": 92154775.0,
      "step": 53130
    },
    {
      "entropy": 5.955947637557983,
      "epoch": 4.134059521493873,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003385667677119271,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2031143769621849,
      "num_tokens": 92163870.0,
      "step": 53135
    },
    {
      "entropy": 5.8521833419799805,
      "epoch": 4.13444855086559,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033854018072767574,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22268462181091309,
      "num_tokens": 92173265.0,
      "step": 53140
    },
    {
      "entropy": 5.943776416778564,
      "epoch": 4.134837580237308,
      "grad_norm": 1.546875,
      "learning_rate": 0.00033851359277930316,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20918085277080536,
      "num_tokens": 92182459.0,
      "step": 53145
    },
    {
      "entropy": 5.883854532241822,
      "epoch": 4.135226609609026,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003384870038672129,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21723269820213317,
      "num_tokens": 92191414.0,
      "step": 53150
    },
    {
      "entropy": 5.88086199760437,
      "epoch": 4.1356156389807435,
      "grad_norm": 1.375,
      "learning_rate": 0.0003384604139918085,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21233344972133636,
      "num_tokens": 92199926.0,
      "step": 53155
    },
    {
      "entropy": 5.876166868209839,
      "epoch": 4.13600466835246,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00033843382315349325,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21622217148542405,
      "num_tokens": 92208399.0,
      "step": 53160
    },
    {
      "entropy": 5.8664548873901365,
      "epoch": 4.136393697724178,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033840723135267073,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20267845541238785,
      "num_tokens": 92218110.0,
      "step": 53165
    },
    {
      "entropy": 5.852568054199219,
      "epoch": 4.136782727095896,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033838063858974434,
      "loss": 4.9694,
      "mean_token_accuracy": 0.20983310788869858,
      "num_tokens": 92227322.0,
      "step": 53170
    },
    {
      "entropy": 5.847979640960693,
      "epoch": 4.137171756467613,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003383540448651177,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22001961320638658,
      "num_tokens": 92235459.0,
      "step": 53175
    },
    {
      "entropy": 5.867280101776123,
      "epoch": 4.137560785839331,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033832745017919424,
      "loss": 5.0072,
      "mean_token_accuracy": 0.20514819175004959,
      "num_tokens": 92244930.0,
      "step": 53180
    },
    {
      "entropy": 5.833106660842896,
      "epoch": 4.137949815211049,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033830085453237757,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2099720448255539,
      "num_tokens": 92254199.0,
      "step": 53185
    },
    {
      "entropy": 5.843534708023071,
      "epoch": 4.138338844582766,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003382742579250712,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2092881605029106,
      "num_tokens": 92262762.0,
      "step": 53190
    },
    {
      "entropy": 5.8687989711761475,
      "epoch": 4.138727873954483,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003382476603576785,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21518648266792298,
      "num_tokens": 92270705.0,
      "step": 53195
    },
    {
      "entropy": 5.862856388092041,
      "epoch": 4.139116903326201,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003382210618306034,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2125919371843338,
      "num_tokens": 92279778.0,
      "step": 53200
    },
    {
      "entropy": 5.827776861190796,
      "epoch": 4.139505932697919,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003381944623442492,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21817974001169205,
      "num_tokens": 92288317.0,
      "step": 53205
    },
    {
      "entropy": 5.887194871902466,
      "epoch": 4.139894962069636,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003381678618990197,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21382184624671935,
      "num_tokens": 92297206.0,
      "step": 53210
    },
    {
      "entropy": 5.90402307510376,
      "epoch": 4.140283991441354,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003381412604953185,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2074090749025345,
      "num_tokens": 92306240.0,
      "step": 53215
    },
    {
      "entropy": 5.826938009262085,
      "epoch": 4.140673020813072,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003381146581335491,
      "loss": 4.7797,
      "mean_token_accuracy": 0.22848791629076004,
      "num_tokens": 92314241.0,
      "step": 53220
    },
    {
      "entropy": 5.827952575683594,
      "epoch": 4.141062050184789,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003380880548141152,
      "loss": 4.9549,
      "mean_token_accuracy": 0.20842324793338776,
      "num_tokens": 92323071.0,
      "step": 53225
    },
    {
      "entropy": 5.799980163574219,
      "epoch": 4.141451079556506,
      "grad_norm": 1.375,
      "learning_rate": 0.00033806145053742043,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21716368943452835,
      "num_tokens": 92331993.0,
      "step": 53230
    },
    {
      "entropy": 5.869219350814819,
      "epoch": 4.141840108928224,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003380348453038686,
      "loss": 4.9276,
      "mean_token_accuracy": 0.21501462012529374,
      "num_tokens": 92340335.0,
      "step": 53235
    },
    {
      "entropy": 5.835952425003052,
      "epoch": 4.142229138299942,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003380082391138633,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2227776825428009,
      "num_tokens": 92348668.0,
      "step": 53240
    },
    {
      "entropy": 5.820970630645752,
      "epoch": 4.142618167671659,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003379816319678083,
      "loss": 4.911,
      "mean_token_accuracy": 0.2181873142719269,
      "num_tokens": 92358153.0,
      "step": 53245
    },
    {
      "entropy": 5.857251453399658,
      "epoch": 4.143007197043377,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003379550238661072,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21793729960918426,
      "num_tokens": 92367441.0,
      "step": 53250
    },
    {
      "entropy": 5.860657596588135,
      "epoch": 4.143396226415095,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033792841480916394,
      "loss": 4.953,
      "mean_token_accuracy": 0.21567545533180238,
      "num_tokens": 92375784.0,
      "step": 53255
    },
    {
      "entropy": 5.806259632110596,
      "epoch": 4.143785255786812,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033790180479738205,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22313224375247956,
      "num_tokens": 92383776.0,
      "step": 53260
    },
    {
      "entropy": 5.77839617729187,
      "epoch": 4.144174285158529,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003378751938311654,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2058481901884079,
      "num_tokens": 92393482.0,
      "step": 53265
    },
    {
      "entropy": 5.9274969577789305,
      "epoch": 4.144563314530247,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003378485819109177,
      "loss": 5.0833,
      "mean_token_accuracy": 0.2126658946275711,
      "num_tokens": 92401991.0,
      "step": 53270
    },
    {
      "entropy": 5.938638067245483,
      "epoch": 4.144952343901965,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033782196903704286,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2131844937801361,
      "num_tokens": 92410624.0,
      "step": 53275
    },
    {
      "entropy": 5.848576784133911,
      "epoch": 4.145341373273682,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003377953552099447,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21264789253473282,
      "num_tokens": 92419846.0,
      "step": 53280
    },
    {
      "entropy": 5.705717468261719,
      "epoch": 4.1457304026454,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003377687404300269,
      "loss": 4.815,
      "mean_token_accuracy": 0.2155806988477707,
      "num_tokens": 92428361.0,
      "step": 53285
    },
    {
      "entropy": 5.77975549697876,
      "epoch": 4.146119432017118,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003377421246976934,
      "loss": 4.8401,
      "mean_token_accuracy": 0.21758217960596085,
      "num_tokens": 92437587.0,
      "step": 53290
    },
    {
      "entropy": 5.871543312072754,
      "epoch": 4.1465084613888346,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033771550801334806,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20645171403884888,
      "num_tokens": 92447374.0,
      "step": 53295
    },
    {
      "entropy": 5.9529203414917,
      "epoch": 4.146897490760552,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033768889037739464,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2081824541091919,
      "num_tokens": 92456308.0,
      "step": 53300
    },
    {
      "entropy": 5.830790996551514,
      "epoch": 4.14728652013227,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033766227179023704,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22505487203598024,
      "num_tokens": 92465416.0,
      "step": 53305
    },
    {
      "entropy": 5.833309507369995,
      "epoch": 4.147675549503988,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033763565225227934,
      "loss": 4.8229,
      "mean_token_accuracy": 0.22772733867168427,
      "num_tokens": 92473622.0,
      "step": 53310
    },
    {
      "entropy": 5.884306383132935,
      "epoch": 4.148064578875705,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003376090317639252,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2079775735735893,
      "num_tokens": 92482122.0,
      "step": 53315
    },
    {
      "entropy": 5.866298246383667,
      "epoch": 4.148453608247423,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033758241032557875,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2140104964375496,
      "num_tokens": 92490241.0,
      "step": 53320
    },
    {
      "entropy": 5.837467527389526,
      "epoch": 4.14884263761914,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033755578793764375,
      "loss": 4.8492,
      "mean_token_accuracy": 0.2124623104929924,
      "num_tokens": 92498176.0,
      "step": 53325
    },
    {
      "entropy": 5.752054166793823,
      "epoch": 4.1492316669908575,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003375291646005243,
      "loss": 4.8523,
      "mean_token_accuracy": 0.21876205652952194,
      "num_tokens": 92506767.0,
      "step": 53330
    },
    {
      "entropy": 5.856043767929077,
      "epoch": 4.149620696362575,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033750254031462425,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22181223779916764,
      "num_tokens": 92515942.0,
      "step": 53335
    },
    {
      "entropy": 5.8891524314880375,
      "epoch": 4.150009725734293,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033747591508034765,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21169962584972382,
      "num_tokens": 92524135.0,
      "step": 53340
    },
    {
      "entropy": 5.751809406280517,
      "epoch": 4.150398755106011,
      "grad_norm": 1.375,
      "learning_rate": 0.0003374492888980985,
      "loss": 4.7557,
      "mean_token_accuracy": 0.22829776853322983,
      "num_tokens": 92532445.0,
      "step": 53345
    },
    {
      "entropy": 5.756174373626709,
      "epoch": 4.150787784477728,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033742266176828073,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22552729547023773,
      "num_tokens": 92540901.0,
      "step": 53350
    },
    {
      "entropy": 5.767401456832886,
      "epoch": 4.151176813849446,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033739603369129846,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22306810617446898,
      "num_tokens": 92548858.0,
      "step": 53355
    },
    {
      "entropy": 5.886229419708252,
      "epoch": 4.151565843221163,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033736940466755566,
      "loss": 4.9914,
      "mean_token_accuracy": 0.20706396996974946,
      "num_tokens": 92558068.0,
      "step": 53360
    },
    {
      "entropy": 5.9186217308044435,
      "epoch": 4.1519548725928805,
      "grad_norm": 1.375,
      "learning_rate": 0.00033734277469745644,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21656395345926285,
      "num_tokens": 92567736.0,
      "step": 53365
    },
    {
      "entropy": 5.851899719238281,
      "epoch": 4.152343901964598,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003373161437814049,
      "loss": 4.8962,
      "mean_token_accuracy": 0.21408206075429917,
      "num_tokens": 92576994.0,
      "step": 53370
    },
    {
      "entropy": 5.824721336364746,
      "epoch": 4.152732931336316,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00033728951191980505,
      "loss": 4.9184,
      "mean_token_accuracy": 0.20799511224031447,
      "num_tokens": 92585228.0,
      "step": 53375
    },
    {
      "entropy": 5.797149753570556,
      "epoch": 4.153121960708034,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033726287911306095,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2187606394290924,
      "num_tokens": 92593754.0,
      "step": 53380
    },
    {
      "entropy": 5.902160882949829,
      "epoch": 4.153510990079751,
      "grad_norm": 1.5,
      "learning_rate": 0.0003372362453615768,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21380991190671922,
      "num_tokens": 92603502.0,
      "step": 53385
    },
    {
      "entropy": 5.911434316635132,
      "epoch": 4.153900019451468,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003372096106657566,
      "loss": 5.0838,
      "mean_token_accuracy": 0.2103756234049797,
      "num_tokens": 92613639.0,
      "step": 53390
    },
    {
      "entropy": 5.918667030334473,
      "epoch": 4.154289048823186,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033718297502600465,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19921103864908218,
      "num_tokens": 92622846.0,
      "step": 53395
    },
    {
      "entropy": 5.882130098342896,
      "epoch": 4.1546780781949035,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033715633844272504,
      "loss": 4.9,
      "mean_token_accuracy": 0.220849846303463,
      "num_tokens": 92631592.0,
      "step": 53400
    },
    {
      "entropy": 5.858701992034912,
      "epoch": 4.155067107566621,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033712970091632187,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21262859106063842,
      "num_tokens": 92640152.0,
      "step": 53405
    },
    {
      "entropy": 5.824657392501831,
      "epoch": 4.155456136938339,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033710306244719954,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21695710122585296,
      "num_tokens": 92648018.0,
      "step": 53410
    },
    {
      "entropy": 5.841556406021118,
      "epoch": 4.155845166310057,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003370764230357619,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21142674535512923,
      "num_tokens": 92656426.0,
      "step": 53415
    },
    {
      "entropy": 5.778595590591431,
      "epoch": 4.156234195681774,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003370497826824134,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2120707780122757,
      "num_tokens": 92665694.0,
      "step": 53420
    },
    {
      "entropy": 5.841378927230835,
      "epoch": 4.156623225053491,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003370231413875582,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21781778037548066,
      "num_tokens": 92674970.0,
      "step": 53425
    },
    {
      "entropy": 5.833185625076294,
      "epoch": 4.157012254425209,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033699649915160053,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22101002782583237,
      "num_tokens": 92682490.0,
      "step": 53430
    },
    {
      "entropy": 5.814183712005615,
      "epoch": 4.1574012837969265,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033696985597494474,
      "loss": 4.917,
      "mean_token_accuracy": 0.21510798037052153,
      "num_tokens": 92690499.0,
      "step": 53435
    },
    {
      "entropy": 5.803383541107178,
      "epoch": 4.157790313168644,
      "grad_norm": 1.53125,
      "learning_rate": 0.000336943211857995,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21198660284280776,
      "num_tokens": 92698457.0,
      "step": 53440
    },
    {
      "entropy": 5.864917325973511,
      "epoch": 4.158179342540362,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003369165668011556,
      "loss": 5.0628,
      "mean_token_accuracy": 0.20996655970811845,
      "num_tokens": 92707621.0,
      "step": 53445
    },
    {
      "entropy": 5.914461469650268,
      "epoch": 4.15856837191208,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003368899208048308,
      "loss": 5.0407,
      "mean_token_accuracy": 0.2060912072658539,
      "num_tokens": 92716667.0,
      "step": 53450
    },
    {
      "entropy": 5.856558322906494,
      "epoch": 4.158957401283797,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033686327386942504,
      "loss": 4.92,
      "mean_token_accuracy": 0.21002083122730256,
      "num_tokens": 92725243.0,
      "step": 53455
    },
    {
      "entropy": 5.888833332061767,
      "epoch": 4.159346430655514,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003368366259953425,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20744064450263977,
      "num_tokens": 92733849.0,
      "step": 53460
    },
    {
      "entropy": 5.922597503662109,
      "epoch": 4.159735460027232,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033680997718298756,
      "loss": 5.037,
      "mean_token_accuracy": 0.20525020658969878,
      "num_tokens": 92742348.0,
      "step": 53465
    },
    {
      "entropy": 5.891862154006958,
      "epoch": 4.1601244893989495,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033678332743276466,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21823608726263047,
      "num_tokens": 92751362.0,
      "step": 53470
    },
    {
      "entropy": 5.818622303009033,
      "epoch": 4.160513518770667,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003367566767450781,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2180062174797058,
      "num_tokens": 92760352.0,
      "step": 53475
    },
    {
      "entropy": 5.76175537109375,
      "epoch": 4.160902548142385,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033673002512033223,
      "loss": 4.8802,
      "mean_token_accuracy": 0.21987295001745225,
      "num_tokens": 92768638.0,
      "step": 53480
    },
    {
      "entropy": 5.748768043518067,
      "epoch": 4.161291577514103,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003367033725589315,
      "loss": 4.8453,
      "mean_token_accuracy": 0.22255823761224747,
      "num_tokens": 92777457.0,
      "step": 53485
    },
    {
      "entropy": 5.822329187393189,
      "epoch": 4.16168060688582,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003366767190612804,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2154759496450424,
      "num_tokens": 92786385.0,
      "step": 53490
    },
    {
      "entropy": 5.909303045272827,
      "epoch": 4.162069636257537,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003366500646277831,
      "loss": 4.976,
      "mean_token_accuracy": 0.21173890084028243,
      "num_tokens": 92794724.0,
      "step": 53495
    },
    {
      "entropy": 5.850612163543701,
      "epoch": 4.162458665629255,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003366234092588443,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2123957246541977,
      "num_tokens": 92803229.0,
      "step": 53500
    },
    {
      "entropy": 5.851525974273682,
      "epoch": 4.1628476950009725,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033659675295486826,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2129154920578003,
      "num_tokens": 92812285.0,
      "step": 53505
    },
    {
      "entropy": 5.821687316894531,
      "epoch": 4.16323672437269,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033657009571625964,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21083459705114366,
      "num_tokens": 92820751.0,
      "step": 53510
    },
    {
      "entropy": 5.8375657081604,
      "epoch": 4.163625753744408,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003365434375434227,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2214834585785866,
      "num_tokens": 92829160.0,
      "step": 53515
    },
    {
      "entropy": 5.817561769485474,
      "epoch": 4.164014783116126,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033651677843676213,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2109537184238434,
      "num_tokens": 92838260.0,
      "step": 53520
    },
    {
      "entropy": 5.748615646362305,
      "epoch": 4.164403812487842,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003364901183966824,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22379323840141296,
      "num_tokens": 92847272.0,
      "step": 53525
    },
    {
      "entropy": 5.83947434425354,
      "epoch": 4.16479284185956,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003364634574235879,
      "loss": 4.882,
      "mean_token_accuracy": 0.21802281737327575,
      "num_tokens": 92856038.0,
      "step": 53530
    },
    {
      "entropy": 5.888909864425659,
      "epoch": 4.165181871231278,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033643679551788325,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2189861997961998,
      "num_tokens": 92864038.0,
      "step": 53535
    },
    {
      "entropy": 5.846358776092529,
      "epoch": 4.1655709006029955,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033641013267997317,
      "loss": 5.052,
      "mean_token_accuracy": 0.21188123673200607,
      "num_tokens": 92872312.0,
      "step": 53540
    },
    {
      "entropy": 5.820207262039185,
      "epoch": 4.165959929974713,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033638346891026196,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20977292209863663,
      "num_tokens": 92881193.0,
      "step": 53545
    },
    {
      "entropy": 5.9242878437042235,
      "epoch": 4.166348959346431,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033635680420915424,
      "loss": 4.9907,
      "mean_token_accuracy": 0.216770438849926,
      "num_tokens": 92889482.0,
      "step": 53550
    },
    {
      "entropy": 5.883403730392456,
      "epoch": 4.166737988718149,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003363301385770548,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21447993367910384,
      "num_tokens": 92899155.0,
      "step": 53555
    },
    {
      "entropy": 5.877398204803467,
      "epoch": 4.167127018089865,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033630347201436806,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21235758513212205,
      "num_tokens": 92907379.0,
      "step": 53560
    },
    {
      "entropy": 5.901299476623535,
      "epoch": 4.167516047461583,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033627680452149873,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2131791114807129,
      "num_tokens": 92916456.0,
      "step": 53565
    },
    {
      "entropy": 5.876926708221435,
      "epoch": 4.167905076833301,
      "grad_norm": 1.375,
      "learning_rate": 0.0003362501360988514,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21066430807113648,
      "num_tokens": 92925005.0,
      "step": 53570
    },
    {
      "entropy": 5.880606460571289,
      "epoch": 4.1682941062050185,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003362234667468308,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20902488082647325,
      "num_tokens": 92934126.0,
      "step": 53575
    },
    {
      "entropy": 5.873444843292236,
      "epoch": 4.168683135576736,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003361967964658415,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21513715237379075,
      "num_tokens": 92943366.0,
      "step": 53580
    },
    {
      "entropy": 5.86112756729126,
      "epoch": 4.169072164948454,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003361701252562882,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2103131115436554,
      "num_tokens": 92952471.0,
      "step": 53585
    },
    {
      "entropy": 5.863969707489014,
      "epoch": 4.169461194320172,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033614345311857566,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21950817853212357,
      "num_tokens": 92960830.0,
      "step": 53590
    },
    {
      "entropy": 5.849753427505493,
      "epoch": 4.169850223691888,
      "grad_norm": 1.5,
      "learning_rate": 0.0003361167800531085,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2255066379904747,
      "num_tokens": 92969387.0,
      "step": 53595
    },
    {
      "entropy": 5.79392442703247,
      "epoch": 4.170239253063606,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003360901060602915,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2176664113998413,
      "num_tokens": 92977928.0,
      "step": 53600
    },
    {
      "entropy": 5.836771059036255,
      "epoch": 4.170628282435324,
      "grad_norm": 1.5625,
      "learning_rate": 0.00033606343114052936,
      "loss": 4.9312,
      "mean_token_accuracy": 0.20728082954883575,
      "num_tokens": 92986233.0,
      "step": 53605
    },
    {
      "entropy": 5.819192743301391,
      "epoch": 4.1710173118070415,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033603675529422685,
      "loss": 4.8731,
      "mean_token_accuracy": 0.21623292565345764,
      "num_tokens": 92995324.0,
      "step": 53610
    },
    {
      "entropy": 5.824749898910523,
      "epoch": 4.171406341178759,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003360100785217887,
      "loss": 4.8937,
      "mean_token_accuracy": 0.21251485049724578,
      "num_tokens": 93003475.0,
      "step": 53615
    },
    {
      "entropy": 5.81056022644043,
      "epoch": 4.171795370550477,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003359834008236198,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2147168517112732,
      "num_tokens": 93011999.0,
      "step": 53620
    },
    {
      "entropy": 5.798842906951904,
      "epoch": 4.172184399922194,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003359567222001248,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21532589197158813,
      "num_tokens": 93021188.0,
      "step": 53625
    },
    {
      "entropy": 5.849372529983521,
      "epoch": 4.172573429293911,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003359300426517085,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2131170317530632,
      "num_tokens": 93029861.0,
      "step": 53630
    },
    {
      "entropy": 5.838745927810669,
      "epoch": 4.172962458665629,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003359033621787759,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2125028043985367,
      "num_tokens": 93037657.0,
      "step": 53635
    },
    {
      "entropy": 5.845746946334839,
      "epoch": 4.173351488037347,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003358766807817317,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2090184882283211,
      "num_tokens": 93046295.0,
      "step": 53640
    },
    {
      "entropy": 5.898080682754516,
      "epoch": 4.1737405174090645,
      "grad_norm": 1.5,
      "learning_rate": 0.0003358499984609807,
      "loss": 4.956,
      "mean_token_accuracy": 0.21833345144987107,
      "num_tokens": 93053831.0,
      "step": 53645
    },
    {
      "entropy": 5.8439373016357425,
      "epoch": 4.174129546780782,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003358233152169278,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21727929413318633,
      "num_tokens": 93062580.0,
      "step": 53650
    },
    {
      "entropy": 5.838990926742554,
      "epoch": 4.1745185761525,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033579663104997803,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21563037484884262,
      "num_tokens": 93072090.0,
      "step": 53655
    },
    {
      "entropy": 5.8875500679016115,
      "epoch": 4.174907605524217,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033576994596053606,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2098407581448555,
      "num_tokens": 93080363.0,
      "step": 53660
    },
    {
      "entropy": 5.866995143890381,
      "epoch": 4.175296634895934,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003357432599490069,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21835999488830565,
      "num_tokens": 93088648.0,
      "step": 53665
    },
    {
      "entropy": 5.8161375522613525,
      "epoch": 4.175685664267652,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033571657301579547,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22200871706008912,
      "num_tokens": 93097534.0,
      "step": 53670
    },
    {
      "entropy": 5.8168755054473875,
      "epoch": 4.17607469363937,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003356898851613067,
      "loss": 4.9147,
      "mean_token_accuracy": 0.20999386608600618,
      "num_tokens": 93106902.0,
      "step": 53675
    },
    {
      "entropy": 5.8229578018188475,
      "epoch": 4.1764637230110875,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003356631963859455,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22281166166067123,
      "num_tokens": 93115208.0,
      "step": 53680
    },
    {
      "entropy": 5.844051837921143,
      "epoch": 4.176852752382805,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003356365066901168,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2190563216805458,
      "num_tokens": 93124493.0,
      "step": 53685
    },
    {
      "entropy": 5.785353279113769,
      "epoch": 4.177241781754523,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033560981607422567,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21268073171377183,
      "num_tokens": 93133984.0,
      "step": 53690
    },
    {
      "entropy": 5.8772913932800295,
      "epoch": 4.17763081112624,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033558312453867703,
      "loss": 5.03,
      "mean_token_accuracy": 0.20563017427921296,
      "num_tokens": 93143134.0,
      "step": 53695
    },
    {
      "entropy": 5.890900993347168,
      "epoch": 4.178019840497957,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033555643208387593,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21998907774686813,
      "num_tokens": 93151192.0,
      "step": 53700
    },
    {
      "entropy": 5.853398942947388,
      "epoch": 4.178408869869675,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033552973871022737,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21735426783561707,
      "num_tokens": 93160300.0,
      "step": 53705
    },
    {
      "entropy": 5.776984214782715,
      "epoch": 4.178797899241393,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033550304441813634,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21758211851119996,
      "num_tokens": 93169441.0,
      "step": 53710
    },
    {
      "entropy": 5.827760553359985,
      "epoch": 4.1791869286131105,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003354763492080079,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2240286648273468,
      "num_tokens": 93177763.0,
      "step": 53715
    },
    {
      "entropy": 5.885996770858765,
      "epoch": 4.179575957984828,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003354496530802472,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21505932062864302,
      "num_tokens": 93186665.0,
      "step": 53720
    },
    {
      "entropy": 5.764900779724121,
      "epoch": 4.179964987356545,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003354229560352592,
      "loss": 4.8688,
      "mean_token_accuracy": 0.21828958094120027,
      "num_tokens": 93195574.0,
      "step": 53725
    },
    {
      "entropy": 5.7927014350891115,
      "epoch": 4.180354016728263,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003353962580734489,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21195422261953353,
      "num_tokens": 93204933.0,
      "step": 53730
    },
    {
      "entropy": 5.844230031967163,
      "epoch": 4.18074304609998,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033536955919522167,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22718557417392732,
      "num_tokens": 93212644.0,
      "step": 53735
    },
    {
      "entropy": 5.81821608543396,
      "epoch": 4.181132075471698,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003353428594009824,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21441629230976106,
      "num_tokens": 93221094.0,
      "step": 53740
    },
    {
      "entropy": 5.878925609588623,
      "epoch": 4.181521104843416,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033531615869113624,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20636970698833465,
      "num_tokens": 93229681.0,
      "step": 53745
    },
    {
      "entropy": 5.784231042861938,
      "epoch": 4.1819101342151335,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033528945706608844,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2103512853384018,
      "num_tokens": 93237976.0,
      "step": 53750
    },
    {
      "entropy": 5.887962770462036,
      "epoch": 4.182299163586851,
      "grad_norm": 1.484375,
      "learning_rate": 0.00033526275452624407,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21543764770030976,
      "num_tokens": 93245614.0,
      "step": 53755
    },
    {
      "entropy": 5.811726379394531,
      "epoch": 4.182688192958568,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033523605107200826,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2218205988407135,
      "num_tokens": 93253715.0,
      "step": 53760
    },
    {
      "entropy": 5.850217676162719,
      "epoch": 4.183077222330286,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00033520934670378634,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2168344721198082,
      "num_tokens": 93262166.0,
      "step": 53765
    },
    {
      "entropy": 5.802445268630981,
      "epoch": 4.183466251702003,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033518264142198335,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22215252965688706,
      "num_tokens": 93271440.0,
      "step": 53770
    },
    {
      "entropy": 5.739235734939575,
      "epoch": 4.183855281073721,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003351559352270047,
      "loss": 4.8661,
      "mean_token_accuracy": 0.21989529579877853,
      "num_tokens": 93279910.0,
      "step": 53775
    },
    {
      "entropy": 5.7950512886047365,
      "epoch": 4.184244310445439,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033512922811925535,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21518078297376633,
      "num_tokens": 93288099.0,
      "step": 53780
    },
    {
      "entropy": 5.851618766784668,
      "epoch": 4.1846333398171565,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003351025200991406,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2104900985956192,
      "num_tokens": 93297029.0,
      "step": 53785
    },
    {
      "entropy": 5.880706071853638,
      "epoch": 4.185022369188874,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003350758111670659,
      "loss": 4.9717,
      "mean_token_accuracy": 0.20797078162431717,
      "num_tokens": 93305763.0,
      "step": 53790
    },
    {
      "entropy": 5.823510599136353,
      "epoch": 4.185411398560591,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003350491013234364,
      "loss": 4.8776,
      "mean_token_accuracy": 0.21601666510105133,
      "num_tokens": 93314445.0,
      "step": 53795
    },
    {
      "entropy": 5.7323085308074955,
      "epoch": 4.185800427932309,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003350223905686573,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2247539058327675,
      "num_tokens": 93323260.0,
      "step": 53800
    },
    {
      "entropy": 5.8788743019104,
      "epoch": 4.186189457304026,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000334995678903134,
      "loss": 5.0045,
      "mean_token_accuracy": 0.20845166742801666,
      "num_tokens": 93331084.0,
      "step": 53805
    },
    {
      "entropy": 5.795471429824829,
      "epoch": 4.186578486675744,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003349689663272718,
      "loss": 4.8596,
      "mean_token_accuracy": 0.22532219886779786,
      "num_tokens": 93339653.0,
      "step": 53810
    },
    {
      "entropy": 5.774336051940918,
      "epoch": 4.186967516047462,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003349422528414759,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21584501266479492,
      "num_tokens": 93349169.0,
      "step": 53815
    },
    {
      "entropy": 5.8463050365448,
      "epoch": 4.1873565454191795,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003349155384461517,
      "loss": 4.9738,
      "mean_token_accuracy": 0.213828244805336,
      "num_tokens": 93358152.0,
      "step": 53820
    },
    {
      "entropy": 5.807898378372192,
      "epoch": 4.187745574790897,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033488882314170463,
      "loss": 4.906,
      "mean_token_accuracy": 0.2142057090997696,
      "num_tokens": 93366388.0,
      "step": 53825
    },
    {
      "entropy": 5.788700866699219,
      "epoch": 4.188134604162614,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00033486210692854,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22182716727256774,
      "num_tokens": 93374984.0,
      "step": 53830
    },
    {
      "entropy": 5.8824609279632565,
      "epoch": 4.188523633534332,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003348353898070631,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21138239353895188,
      "num_tokens": 93383588.0,
      "step": 53835
    },
    {
      "entropy": 5.854922437667847,
      "epoch": 4.188912662906049,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003348086717776795,
      "loss": 4.8491,
      "mean_token_accuracy": 0.21564164608716965,
      "num_tokens": 93392161.0,
      "step": 53840
    },
    {
      "entropy": 5.853561544418335,
      "epoch": 4.189301692277767,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033478195284079447,
      "loss": 4.9682,
      "mean_token_accuracy": 0.211666975915432,
      "num_tokens": 93400934.0,
      "step": 53845
    },
    {
      "entropy": 5.79205732345581,
      "epoch": 4.189690721649485,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003347552329968134,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21470278054475783,
      "num_tokens": 93409780.0,
      "step": 53850
    },
    {
      "entropy": 5.804937744140625,
      "epoch": 4.1900797510212024,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003347285122461419,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2191269114613533,
      "num_tokens": 93418407.0,
      "step": 53855
    },
    {
      "entropy": 5.899067497253418,
      "epoch": 4.190468780392919,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003347017905891852,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20703840404748916,
      "num_tokens": 93426526.0,
      "step": 53860
    },
    {
      "entropy": 5.889360523223877,
      "epoch": 4.190857809764637,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003346750680263488,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21740897595882416,
      "num_tokens": 93434826.0,
      "step": 53865
    },
    {
      "entropy": 5.747199249267578,
      "epoch": 4.191246839136355,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003346483445580384,
      "loss": 4.7996,
      "mean_token_accuracy": 0.22174112796783446,
      "num_tokens": 93443702.0,
      "step": 53870
    },
    {
      "entropy": 5.894687080383301,
      "epoch": 4.191635868508072,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033462162018465917,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2039017528295517,
      "num_tokens": 93452477.0,
      "step": 53875
    },
    {
      "entropy": 5.895621204376221,
      "epoch": 4.19202489787979,
      "grad_norm": 1.375,
      "learning_rate": 0.00033459489490661676,
      "loss": 4.956,
      "mean_token_accuracy": 0.210696217417717,
      "num_tokens": 93460752.0,
      "step": 53880
    },
    {
      "entropy": 5.920444965362549,
      "epoch": 4.192413927251508,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033456816872431673,
      "loss": 5.054,
      "mean_token_accuracy": 0.20734283328056335,
      "num_tokens": 93469917.0,
      "step": 53885
    },
    {
      "entropy": 5.853944730758667,
      "epoch": 4.192802956623225,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003345414416381645,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21851826310157776,
      "num_tokens": 93478530.0,
      "step": 53890
    },
    {
      "entropy": 5.820482778549194,
      "epoch": 4.193191985994942,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033451471364856565,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2198634549975395,
      "num_tokens": 93487137.0,
      "step": 53895
    },
    {
      "entropy": 5.793654251098633,
      "epoch": 4.19358101536666,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033448798475592585,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2114912375807762,
      "num_tokens": 93495163.0,
      "step": 53900
    },
    {
      "entropy": 5.816314220428467,
      "epoch": 4.193970044738378,
      "grad_norm": 1.375,
      "learning_rate": 0.0003344612549606505,
      "loss": 4.882,
      "mean_token_accuracy": 0.22631251364946364,
      "num_tokens": 93503652.0,
      "step": 53905
    },
    {
      "entropy": 5.8094501972198485,
      "epoch": 4.194359074110095,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003344345242631453,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22077007889747619,
      "num_tokens": 93512869.0,
      "step": 53910
    },
    {
      "entropy": 5.807117223739624,
      "epoch": 4.194748103481813,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003344077926638157,
      "loss": 4.8917,
      "mean_token_accuracy": 0.21143667846918107,
      "num_tokens": 93521658.0,
      "step": 53915
    },
    {
      "entropy": 5.848456764221192,
      "epoch": 4.195137132853531,
      "grad_norm": 1.5,
      "learning_rate": 0.00033438106016306753,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21177508383989335,
      "num_tokens": 93530113.0,
      "step": 53920
    },
    {
      "entropy": 5.87568941116333,
      "epoch": 4.195526162225248,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003343543267613062,
      "loss": 4.965,
      "mean_token_accuracy": 0.21584281772375108,
      "num_tokens": 93539052.0,
      "step": 53925
    },
    {
      "entropy": 5.9074928760528564,
      "epoch": 4.195915191596965,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003343275924589374,
      "loss": 4.9007,
      "mean_token_accuracy": 0.21203228682279587,
      "num_tokens": 93547113.0,
      "step": 53930
    },
    {
      "entropy": 5.797955131530761,
      "epoch": 4.196304220968683,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033430085725636685,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2186915934085846,
      "num_tokens": 93556126.0,
      "step": 53935
    },
    {
      "entropy": 5.866074848175049,
      "epoch": 4.196693250340401,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003342741211540002,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2220117300748825,
      "num_tokens": 93564407.0,
      "step": 53940
    },
    {
      "entropy": 5.875904321670532,
      "epoch": 4.197082279712118,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003342473841522431,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21630416363477706,
      "num_tokens": 93572980.0,
      "step": 53945
    },
    {
      "entropy": 5.834396934509277,
      "epoch": 4.197471309083836,
      "grad_norm": 1.390625,
      "learning_rate": 0.00033422064625150124,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21560702621936798,
      "num_tokens": 93581996.0,
      "step": 53950
    },
    {
      "entropy": 5.848588275909424,
      "epoch": 4.197860338455554,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003341939074521804,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21536921113729476,
      "num_tokens": 93590812.0,
      "step": 53955
    },
    {
      "entropy": 5.886680364608765,
      "epoch": 4.1982493678272705,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003341671677546861,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22141322791576384,
      "num_tokens": 93599306.0,
      "step": 53960
    },
    {
      "entropy": 5.8576418399810795,
      "epoch": 4.198638397198988,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033414042715942423,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21486985832452773,
      "num_tokens": 93607384.0,
      "step": 53965
    },
    {
      "entropy": 5.804290103912353,
      "epoch": 4.199027426570706,
      "grad_norm": 1.375,
      "learning_rate": 0.0003341136856668005,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2213495120406151,
      "num_tokens": 93615502.0,
      "step": 53970
    },
    {
      "entropy": 5.855285263061523,
      "epoch": 4.199416455942424,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003340869432772207,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22057077735662461,
      "num_tokens": 93623672.0,
      "step": 53975
    },
    {
      "entropy": 5.828197956085205,
      "epoch": 4.199805485314141,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033406019999109065,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21406349241733552,
      "num_tokens": 93632061.0,
      "step": 53980
    },
    {
      "entropy": 5.918289756774902,
      "epoch": 4.200194514685859,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033403345580881606,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21302729547023774,
      "num_tokens": 93641307.0,
      "step": 53985
    },
    {
      "entropy": 5.768211746215821,
      "epoch": 4.200583544057577,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033400671073080274,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2144381284713745,
      "num_tokens": 93650152.0,
      "step": 53990
    },
    {
      "entropy": 5.823934841156006,
      "epoch": 4.2009725734292935,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003339799647574564,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21355070918798447,
      "num_tokens": 93659102.0,
      "step": 53995
    },
    {
      "entropy": 5.858927488327026,
      "epoch": 4.201361602801011,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003339532178891831,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21594687551259995,
      "num_tokens": 93667620.0,
      "step": 54000
    },
    {
      "epoch": 4.201361602801011,
      "eval_entropy": 5.596676927810684,
      "eval_loss": 5.047396659851074,
      "eval_mean_token_accuracy": 0.21937320121959267,
      "eval_num_tokens": 93667620.0,
      "eval_runtime": 21.6731,
      "eval_samples_per_second": 1917.494,
      "eval_steps_per_second": 239.698,
      "step": 54000
    },
    {
      "entropy": 5.971887111663818,
      "epoch": 4.201750632172729,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003339264701263885,
      "loss": 5.0756,
      "mean_token_accuracy": 0.2112485021352768,
      "num_tokens": 93676315.0,
      "step": 54005
    },
    {
      "entropy": 5.739187479019165,
      "epoch": 4.202139661544447,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003338997214694784,
      "loss": 4.8139,
      "mean_token_accuracy": 0.22412844598293305,
      "num_tokens": 93685095.0,
      "step": 54010
    },
    {
      "entropy": 5.790797233581543,
      "epoch": 4.202528690916164,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033387297191885896,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21456769853830338,
      "num_tokens": 93693831.0,
      "step": 54015
    },
    {
      "entropy": 5.89804310798645,
      "epoch": 4.202917720287882,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003338462214749358,
      "loss": 4.9612,
      "mean_token_accuracy": 0.20885361284017562,
      "num_tokens": 93703015.0,
      "step": 54020
    },
    {
      "entropy": 5.83752818107605,
      "epoch": 4.2033067496596,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003338194701381148,
      "loss": 4.8376,
      "mean_token_accuracy": 0.21939356476068497,
      "num_tokens": 93711004.0,
      "step": 54025
    },
    {
      "entropy": 5.787057828903198,
      "epoch": 4.2036957790313165,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033379271790880203,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22041780650615692,
      "num_tokens": 93718887.0,
      "step": 54030
    },
    {
      "entropy": 5.727867174148559,
      "epoch": 4.204084808403034,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003337659647874034,
      "loss": 4.794,
      "mean_token_accuracy": 0.22765171378850937,
      "num_tokens": 93727216.0,
      "step": 54035
    },
    {
      "entropy": 5.814126300811767,
      "epoch": 4.204473837774752,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003337392107743248,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22085388749837875,
      "num_tokens": 93736193.0,
      "step": 54040
    },
    {
      "entropy": 5.786495018005371,
      "epoch": 4.20486286714647,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003337124558699721,
      "loss": 4.7835,
      "mean_token_accuracy": 0.2273172467947006,
      "num_tokens": 93743946.0,
      "step": 54045
    },
    {
      "entropy": 5.882463884353638,
      "epoch": 4.205251896518187,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033368570007475133,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21057265549898146,
      "num_tokens": 93752802.0,
      "step": 54050
    },
    {
      "entropy": 5.762263774871826,
      "epoch": 4.205640925889905,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003336589433890685,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21930093467235565,
      "num_tokens": 93761933.0,
      "step": 54055
    },
    {
      "entropy": 5.8612583637237545,
      "epoch": 4.206029955261622,
      "grad_norm": 1.484375,
      "learning_rate": 0.00033363218581332956,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21067806631326674,
      "num_tokens": 93770665.0,
      "step": 54060
    },
    {
      "entropy": 5.825173997879029,
      "epoch": 4.2064189846333395,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003336054273479405,
      "loss": 4.912,
      "mean_token_accuracy": 0.21339046508073806,
      "num_tokens": 93780162.0,
      "step": 54065
    },
    {
      "entropy": 5.822549486160279,
      "epoch": 4.206808014005057,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033357866799330743,
      "loss": 4.8851,
      "mean_token_accuracy": 0.21630495488643647,
      "num_tokens": 93788281.0,
      "step": 54070
    },
    {
      "entropy": 5.83717794418335,
      "epoch": 4.207197043376775,
      "grad_norm": 1.375,
      "learning_rate": 0.00033355190774983626,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21329033374786377,
      "num_tokens": 93796980.0,
      "step": 54075
    },
    {
      "entropy": 5.84740948677063,
      "epoch": 4.207586072748493,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033352514661793306,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21021244972944259,
      "num_tokens": 93806048.0,
      "step": 54080
    },
    {
      "entropy": 5.873120880126953,
      "epoch": 4.20797510212021,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033349838459800394,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21483903974294663,
      "num_tokens": 93814203.0,
      "step": 54085
    },
    {
      "entropy": 5.8491216659545895,
      "epoch": 4.208364131491928,
      "grad_norm": 1.40625,
      "learning_rate": 0.000333471621690455,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21548207402229308,
      "num_tokens": 93822652.0,
      "step": 54090
    },
    {
      "entropy": 5.810828733444214,
      "epoch": 4.208753160863645,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003334448578956921,
      "loss": 4.8523,
      "mean_token_accuracy": 0.222181935608387,
      "num_tokens": 93831173.0,
      "step": 54095
    },
    {
      "entropy": 5.826335382461548,
      "epoch": 4.2091421902353625,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003334180932141217,
      "loss": 4.861,
      "mean_token_accuracy": 0.22131213843822478,
      "num_tokens": 93840248.0,
      "step": 54100
    },
    {
      "entropy": 5.791978406906128,
      "epoch": 4.20953121960708,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003333913276461497,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21856824457645416,
      "num_tokens": 93848714.0,
      "step": 54105
    },
    {
      "entropy": 5.809845972061157,
      "epoch": 4.209920248978798,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033336456119218214,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2125195473432541,
      "num_tokens": 93857433.0,
      "step": 54110
    },
    {
      "entropy": 5.887536907196045,
      "epoch": 4.210309278350516,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033333779385262525,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21444258689880372,
      "num_tokens": 93865586.0,
      "step": 54115
    },
    {
      "entropy": 5.776325321197509,
      "epoch": 4.210698307722233,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033331102562788527,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21979486793279648,
      "num_tokens": 93874566.0,
      "step": 54120
    },
    {
      "entropy": 5.839349937438965,
      "epoch": 4.211087337093951,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003332842565183682,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21956595331430434,
      "num_tokens": 93883187.0,
      "step": 54125
    },
    {
      "entropy": 5.892572021484375,
      "epoch": 4.211476366465668,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033325748652448037,
      "loss": 4.9407,
      "mean_token_accuracy": 0.21716114580631257,
      "num_tokens": 93891281.0,
      "step": 54130
    },
    {
      "entropy": 5.89418306350708,
      "epoch": 4.2118653958373855,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033323071564662796,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21288414299488068,
      "num_tokens": 93900281.0,
      "step": 54135
    },
    {
      "entropy": 5.865328788757324,
      "epoch": 4.212254425209103,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033320394388521707,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21161502301692964,
      "num_tokens": 93908937.0,
      "step": 54140
    },
    {
      "entropy": 5.8832580089569095,
      "epoch": 4.212643454580821,
      "grad_norm": 1.484375,
      "learning_rate": 0.000333177171240654,
      "loss": 4.9105,
      "mean_token_accuracy": 0.21832273602485658,
      "num_tokens": 93916471.0,
      "step": 54145
    },
    {
      "entropy": 5.833485841751099,
      "epoch": 4.213032483952539,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033315039771334494,
      "loss": 5.0155,
      "mean_token_accuracy": 0.20712808817625045,
      "num_tokens": 93926163.0,
      "step": 54150
    },
    {
      "entropy": 5.883976888656616,
      "epoch": 4.213421513324256,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003331236233036961,
      "loss": 5.007,
      "mean_token_accuracy": 0.21158347129821778,
      "num_tokens": 93934405.0,
      "step": 54155
    },
    {
      "entropy": 5.876963949203491,
      "epoch": 4.213810542695974,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003330968480121138,
      "loss": 4.9326,
      "mean_token_accuracy": 0.217067751288414,
      "num_tokens": 93943108.0,
      "step": 54160
    },
    {
      "entropy": 5.871871280670166,
      "epoch": 4.214199572067691,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003330700718390043,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21251060515642167,
      "num_tokens": 93951624.0,
      "step": 54165
    },
    {
      "entropy": 5.850645875930786,
      "epoch": 4.2145886014394085,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033304329478477395,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21017044484615327,
      "num_tokens": 93960232.0,
      "step": 54170
    },
    {
      "entropy": 5.924334859848022,
      "epoch": 4.214977630811126,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033301651684982894,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2110401511192322,
      "num_tokens": 93968362.0,
      "step": 54175
    },
    {
      "entropy": 5.832405662536621,
      "epoch": 4.215366660182844,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003329897380345757,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2140204504132271,
      "num_tokens": 93977767.0,
      "step": 54180
    },
    {
      "entropy": 5.817509889602661,
      "epoch": 4.215755689554562,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003329629583394204,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2231801137328148,
      "num_tokens": 93986371.0,
      "step": 54185
    },
    {
      "entropy": 5.889754056930542,
      "epoch": 4.216144718926279,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003329361777647695,
      "loss": 5.0133,
      "mean_token_accuracy": 0.20866560488939284,
      "num_tokens": 93995883.0,
      "step": 54190
    },
    {
      "entropy": 5.942185258865356,
      "epoch": 4.216533748297996,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003329093963110293,
      "loss": 4.9308,
      "mean_token_accuracy": 0.21899665296077728,
      "num_tokens": 94003846.0,
      "step": 54195
    },
    {
      "entropy": 5.902861738204956,
      "epoch": 4.216922777669714,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033288261397860613,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21024268865585327,
      "num_tokens": 94013427.0,
      "step": 54200
    },
    {
      "entropy": 5.818203353881836,
      "epoch": 4.2173118070414315,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033285583076790646,
      "loss": 4.8575,
      "mean_token_accuracy": 0.22237660735845566,
      "num_tokens": 94021856.0,
      "step": 54205
    },
    {
      "entropy": 5.849095249176026,
      "epoch": 4.217700836413149,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033282904667933675,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2067559391260147,
      "num_tokens": 94029865.0,
      "step": 54210
    },
    {
      "entropy": 5.825262498855591,
      "epoch": 4.218089865784867,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003328022617133032,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2172183394432068,
      "num_tokens": 94038961.0,
      "step": 54215
    },
    {
      "entropy": 5.857781028747558,
      "epoch": 4.218478895156585,
      "grad_norm": 1.375,
      "learning_rate": 0.0003327754758702124,
      "loss": 4.9469,
      "mean_token_accuracy": 0.21627720296382905,
      "num_tokens": 94048184.0,
      "step": 54220
    },
    {
      "entropy": 5.8971795558929445,
      "epoch": 4.218867924528302,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033274868915047064,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2094110742211342,
      "num_tokens": 94056098.0,
      "step": 54225
    },
    {
      "entropy": 5.843860530853272,
      "epoch": 4.219256953900019,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033272190155448446,
      "loss": 4.9089,
      "mean_token_accuracy": 0.21378776729106902,
      "num_tokens": 94064744.0,
      "step": 54230
    },
    {
      "entropy": 5.81847653388977,
      "epoch": 4.219645983271737,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003326951130826603,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21451397538185119,
      "num_tokens": 94074039.0,
      "step": 54235
    },
    {
      "entropy": 5.832022619247437,
      "epoch": 4.2200350126434545,
      "grad_norm": 1.34375,
      "learning_rate": 0.00033266832373540463,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2155937671661377,
      "num_tokens": 94082455.0,
      "step": 54240
    },
    {
      "entropy": 5.9056727409362795,
      "epoch": 4.220424042015172,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003326415335131239,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20685748755931854,
      "num_tokens": 94091456.0,
      "step": 54245
    },
    {
      "entropy": 5.800530099868775,
      "epoch": 4.22081307138689,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00033261474241622475,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22232575416564943,
      "num_tokens": 94100054.0,
      "step": 54250
    },
    {
      "entropy": 5.909888124465942,
      "epoch": 4.221202100758608,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003325879504451135,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21943352818489076,
      "num_tokens": 94107582.0,
      "step": 54255
    },
    {
      "entropy": 5.875159072875976,
      "epoch": 4.221591130130324,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033256115760019677,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21664048731327057,
      "num_tokens": 94116146.0,
      "step": 54260
    },
    {
      "entropy": 5.796509170532227,
      "epoch": 4.221980159502042,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033253436388188106,
      "loss": 4.8458,
      "mean_token_accuracy": 0.21939308047294617,
      "num_tokens": 94124159.0,
      "step": 54265
    },
    {
      "entropy": 5.8165966987609865,
      "epoch": 4.22236918887376,
      "grad_norm": 1.3125,
      "learning_rate": 0.000332507569290573,
      "loss": 4.9352,
      "mean_token_accuracy": 0.20942832827568053,
      "num_tokens": 94133942.0,
      "step": 54270
    },
    {
      "entropy": 5.81131010055542,
      "epoch": 4.2227582182454775,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033248077382667913,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2225460335612297,
      "num_tokens": 94142617.0,
      "step": 54275
    },
    {
      "entropy": 5.9180214405059814,
      "epoch": 4.223147247617195,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000332453977490606,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21270446181297303,
      "num_tokens": 94151159.0,
      "step": 54280
    },
    {
      "entropy": 5.931451845169067,
      "epoch": 4.223536276988913,
      "grad_norm": 1.375,
      "learning_rate": 0.0003324271802827602,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2201889857649803,
      "num_tokens": 94160174.0,
      "step": 54285
    },
    {
      "entropy": 5.8430641174316404,
      "epoch": 4.223925306360631,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003324003822035483,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21524726897478103,
      "num_tokens": 94169155.0,
      "step": 54290
    },
    {
      "entropy": 5.827041721343994,
      "epoch": 4.224314335732347,
      "grad_norm": 1.5,
      "learning_rate": 0.0003323735832533771,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21363459825515746,
      "num_tokens": 94177706.0,
      "step": 54295
    },
    {
      "entropy": 5.862002277374268,
      "epoch": 4.224703365104065,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033234678343265297,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2212420240044594,
      "num_tokens": 94186175.0,
      "step": 54300
    },
    {
      "entropy": 5.832858610153198,
      "epoch": 4.225092394475783,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003323199827417827,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21001080870628358,
      "num_tokens": 94194583.0,
      "step": 54305
    },
    {
      "entropy": 5.852217102050782,
      "epoch": 4.2254814238475005,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033229318118117293,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2111903816461563,
      "num_tokens": 94203932.0,
      "step": 54310
    },
    {
      "entropy": 5.9370091438293455,
      "epoch": 4.225870453219218,
      "grad_norm": 1.53125,
      "learning_rate": 0.00033226637875123026,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2143762394785881,
      "num_tokens": 94212230.0,
      "step": 54315
    },
    {
      "entropy": 5.851756381988525,
      "epoch": 4.226259482590936,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033223957545236156,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2104286327958107,
      "num_tokens": 94221062.0,
      "step": 54320
    },
    {
      "entropy": 5.857554817199707,
      "epoch": 4.226648511962654,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003322127712849733,
      "loss": 4.8689,
      "mean_token_accuracy": 0.21470119059085846,
      "num_tokens": 94229527.0,
      "step": 54325
    },
    {
      "entropy": 5.888439178466797,
      "epoch": 4.22703754133437,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003321859662494724,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2110230416059494,
      "num_tokens": 94238179.0,
      "step": 54330
    },
    {
      "entropy": 5.734419822692871,
      "epoch": 4.227426570706088,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003321591603462654,
      "loss": 4.8107,
      "mean_token_accuracy": 0.22955417186021804,
      "num_tokens": 94246200.0,
      "step": 54335
    },
    {
      "entropy": 5.805563831329346,
      "epoch": 4.227815600077806,
      "grad_norm": 1.5,
      "learning_rate": 0.0003321323535757592,
      "loss": 4.95,
      "mean_token_accuracy": 0.20972746908664702,
      "num_tokens": 94254446.0,
      "step": 54340
    },
    {
      "entropy": 5.875521230697632,
      "epoch": 4.2282046294495235,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003321055459383604,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21168266087770463,
      "num_tokens": 94263122.0,
      "step": 54345
    },
    {
      "entropy": 5.9250740051269535,
      "epoch": 4.228593658821241,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033207873743447587,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2156611755490303,
      "num_tokens": 94271386.0,
      "step": 54350
    },
    {
      "entropy": 5.8797022819519045,
      "epoch": 4.228982688192959,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00033205192806451227,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2231452763080597,
      "num_tokens": 94280336.0,
      "step": 54355
    },
    {
      "entropy": 5.824539470672607,
      "epoch": 4.229371717564677,
      "grad_norm": 1.359375,
      "learning_rate": 0.00033202511782887655,
      "loss": 4.907,
      "mean_token_accuracy": 0.21747318059206008,
      "num_tokens": 94289049.0,
      "step": 54360
    },
    {
      "entropy": 5.847026348114014,
      "epoch": 4.229760746936393,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00033199830672797545,
      "loss": 4.9768,
      "mean_token_accuracy": 0.20996737778186797,
      "num_tokens": 94296947.0,
      "step": 54365
    },
    {
      "entropy": 5.823387336730957,
      "epoch": 4.230149776308111,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003319714947622157,
      "loss": 4.8585,
      "mean_token_accuracy": 0.2210083469748497,
      "num_tokens": 94307121.0,
      "step": 54370
    },
    {
      "entropy": 5.862875127792359,
      "epoch": 4.230538805679829,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003319446819320043,
      "loss": 4.9257,
      "mean_token_accuracy": 0.21987499594688414,
      "num_tokens": 94316770.0,
      "step": 54375
    },
    {
      "entropy": 5.868021774291992,
      "epoch": 4.2309278350515465,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003319178682377479,
      "loss": 4.8867,
      "mean_token_accuracy": 0.21835479587316514,
      "num_tokens": 94325324.0,
      "step": 54380
    },
    {
      "entropy": 5.827633810043335,
      "epoch": 4.231316864423264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033189105367985343,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21604911535978316,
      "num_tokens": 94334273.0,
      "step": 54385
    },
    {
      "entropy": 5.789850807189941,
      "epoch": 4.231705893794982,
      "grad_norm": 1.5625,
      "learning_rate": 0.00033186423825872787,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22379010170698166,
      "num_tokens": 94342380.0,
      "step": 54390
    },
    {
      "entropy": 5.874048900604248,
      "epoch": 4.232094923166699,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003318374219747779,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22113455086946487,
      "num_tokens": 94350728.0,
      "step": 54395
    },
    {
      "entropy": 5.873372411727905,
      "epoch": 4.232483952538416,
      "grad_norm": 1.40625,
      "learning_rate": 0.00033181060482841054,
      "loss": 4.947,
      "mean_token_accuracy": 0.2182196781039238,
      "num_tokens": 94359016.0,
      "step": 54400
    },
    {
      "entropy": 5.87741985321045,
      "epoch": 4.232872981910134,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003317837868200327,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22041266411542892,
      "num_tokens": 94367652.0,
      "step": 54405
    },
    {
      "entropy": 5.858276844024658,
      "epoch": 4.233262011281852,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033175696795005133,
      "loss": 4.9045,
      "mean_token_accuracy": 0.21588704586029053,
      "num_tokens": 94376318.0,
      "step": 54410
    },
    {
      "entropy": 5.8965339183807375,
      "epoch": 4.2336510406535695,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003317301482188732,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2077641174197197,
      "num_tokens": 94384967.0,
      "step": 54415
    },
    {
      "entropy": 5.900251245498657,
      "epoch": 4.234040070025287,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003317033276269054,
      "loss": 4.8967,
      "mean_token_accuracy": 0.21918740421533583,
      "num_tokens": 94393258.0,
      "step": 54420
    },
    {
      "entropy": 5.835453128814697,
      "epoch": 4.234429099397005,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033167650617455484,
      "loss": 4.818,
      "mean_token_accuracy": 0.22985647469758988,
      "num_tokens": 94402232.0,
      "step": 54425
    },
    {
      "entropy": 5.88439793586731,
      "epoch": 4.234818128768722,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033164968386222856,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20924014896154403,
      "num_tokens": 94411234.0,
      "step": 54430
    },
    {
      "entropy": 5.837527704238892,
      "epoch": 4.235207158140439,
      "grad_norm": 1.375,
      "learning_rate": 0.0003316228606903334,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2272052139043808,
      "num_tokens": 94420048.0,
      "step": 54435
    },
    {
      "entropy": 5.802836990356445,
      "epoch": 4.235596187512157,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033159603665927655,
      "loss": 4.8794,
      "mean_token_accuracy": 0.21734994500875474,
      "num_tokens": 94428999.0,
      "step": 54440
    },
    {
      "entropy": 5.834684228897094,
      "epoch": 4.235985216883875,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00033156921176946496,
      "loss": 4.8897,
      "mean_token_accuracy": 0.21609367728233336,
      "num_tokens": 94436796.0,
      "step": 54445
    },
    {
      "entropy": 5.811357593536377,
      "epoch": 4.2363742462555924,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003315423860213057,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21386505514383317,
      "num_tokens": 94445477.0,
      "step": 54450
    },
    {
      "entropy": 5.854596090316773,
      "epoch": 4.23676327562731,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003315155594152055,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2135547861456871,
      "num_tokens": 94454828.0,
      "step": 54455
    },
    {
      "entropy": 5.876994609832764,
      "epoch": 4.237152304999028,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003314887319515717,
      "loss": 4.9654,
      "mean_token_accuracy": 0.21273151487112046,
      "num_tokens": 94463549.0,
      "step": 54460
    },
    {
      "entropy": 5.872981882095337,
      "epoch": 4.237541334370745,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003314619036308113,
      "loss": 4.8749,
      "mean_token_accuracy": 0.21386970281600953,
      "num_tokens": 94471581.0,
      "step": 54465
    },
    {
      "entropy": 5.89142861366272,
      "epoch": 4.237930363742462,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033143507445333145,
      "loss": 5.0125,
      "mean_token_accuracy": 0.20487603545188904,
      "num_tokens": 94480230.0,
      "step": 54470
    },
    {
      "entropy": 5.853812265396118,
      "epoch": 4.23831939311418,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00033140824441953913,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21911082416772842,
      "num_tokens": 94488726.0,
      "step": 54475
    },
    {
      "entropy": 5.829182386398315,
      "epoch": 4.238708422485898,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033138141352984145,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2204552710056305,
      "num_tokens": 94497381.0,
      "step": 54480
    },
    {
      "entropy": 5.759686803817749,
      "epoch": 4.239097451857615,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033135458178464565,
      "loss": 4.8755,
      "mean_token_accuracy": 0.21797899305820465,
      "num_tokens": 94505780.0,
      "step": 54485
    },
    {
      "entropy": 5.861319160461425,
      "epoch": 4.239486481229333,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003313277491843587,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21102966815233232,
      "num_tokens": 94514539.0,
      "step": 54490
    },
    {
      "entropy": 5.878709363937378,
      "epoch": 4.239875510601051,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033130091572938786,
      "loss": 4.8879,
      "mean_token_accuracy": 0.21624512672424318,
      "num_tokens": 94523067.0,
      "step": 54495
    },
    {
      "entropy": 5.8554362773895265,
      "epoch": 4.240264539972768,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003312740814201402,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21569064557552337,
      "num_tokens": 94532030.0,
      "step": 54500
    },
    {
      "entropy": 5.911686134338379,
      "epoch": 4.240653569344485,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003312472462570229,
      "loss": 5.0431,
      "mean_token_accuracy": 0.21060900688171386,
      "num_tokens": 94540838.0,
      "step": 54505
    },
    {
      "entropy": 5.839062595367432,
      "epoch": 4.241042598716203,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033122041024044325,
      "loss": 4.9246,
      "mean_token_accuracy": 0.21736403107643126,
      "num_tokens": 94549603.0,
      "step": 54510
    },
    {
      "entropy": 5.779233646392822,
      "epoch": 4.241431628087921,
      "grad_norm": 1.375,
      "learning_rate": 0.00033119357337080827,
      "loss": 4.824,
      "mean_token_accuracy": 0.23067578822374343,
      "num_tokens": 94558246.0,
      "step": 54515
    },
    {
      "entropy": 5.796691370010376,
      "epoch": 4.241820657459638,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003311667356485253,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2093515843153,
      "num_tokens": 94567293.0,
      "step": 54520
    },
    {
      "entropy": 5.911427974700928,
      "epoch": 4.242209686831356,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003311398970740015,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21720293015241623,
      "num_tokens": 94576249.0,
      "step": 54525
    },
    {
      "entropy": 5.851956176757812,
      "epoch": 4.242598716203073,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003311130576476441,
      "loss": 4.919,
      "mean_token_accuracy": 0.22292238771915435,
      "num_tokens": 94584206.0,
      "step": 54530
    },
    {
      "entropy": 5.853858757019043,
      "epoch": 4.242987745574791,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003310862173698604,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22328370809555054,
      "num_tokens": 94592962.0,
      "step": 54535
    },
    {
      "entropy": 5.785601949691772,
      "epoch": 4.243376774946508,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003310593762410576,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2257447674870491,
      "num_tokens": 94601305.0,
      "step": 54540
    },
    {
      "entropy": 5.824986934661865,
      "epoch": 4.243765804318226,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000331032534261643,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2049812361598015,
      "num_tokens": 94609844.0,
      "step": 54545
    },
    {
      "entropy": 5.8097621440887455,
      "epoch": 4.244154833689944,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003310056914320239,
      "loss": 4.947,
      "mean_token_accuracy": 0.21114800721406937,
      "num_tokens": 94618068.0,
      "step": 54550
    },
    {
      "entropy": 5.797760915756226,
      "epoch": 4.244543863061661,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033097884775260754,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22113310396671296,
      "num_tokens": 94627140.0,
      "step": 54555
    },
    {
      "entropy": 5.863585996627807,
      "epoch": 4.244932892433379,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033095200322380124,
      "loss": 4.8959,
      "mean_token_accuracy": 0.21619660407304764,
      "num_tokens": 94636416.0,
      "step": 54560
    },
    {
      "entropy": 5.8374744892120365,
      "epoch": 4.245321921805096,
      "grad_norm": 1.375,
      "learning_rate": 0.00033092515784601245,
      "loss": 4.9701,
      "mean_token_accuracy": 0.2134205147624016,
      "num_tokens": 94645268.0,
      "step": 54565
    },
    {
      "entropy": 5.916219091415405,
      "epoch": 4.245710951176814,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003308983116196483,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21586241871118544,
      "num_tokens": 94653868.0,
      "step": 54570
    },
    {
      "entropy": 5.899639844894409,
      "epoch": 4.246099980548531,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003308714645451163,
      "loss": 4.9072,
      "mean_token_accuracy": 0.21756876558065413,
      "num_tokens": 94662274.0,
      "step": 54575
    },
    {
      "entropy": 5.896301174163819,
      "epoch": 4.246489009920249,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003308446166228237,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21535287350416182,
      "num_tokens": 94670313.0,
      "step": 54580
    },
    {
      "entropy": 5.81079421043396,
      "epoch": 4.246878039291967,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033081776785317803,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21901050359010696,
      "num_tokens": 94679733.0,
      "step": 54585
    },
    {
      "entropy": 5.780589532852173,
      "epoch": 4.247267068663684,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033079091823658645,
      "loss": 4.8294,
      "mean_token_accuracy": 0.224543260037899,
      "num_tokens": 94688081.0,
      "step": 54590
    },
    {
      "entropy": 5.92640700340271,
      "epoch": 4.247656098035401,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003307640677734566,
      "loss": 5.0488,
      "mean_token_accuracy": 0.2048206701874733,
      "num_tokens": 94696852.0,
      "step": 54595
    },
    {
      "entropy": 5.8876622200012205,
      "epoch": 4.248045127407119,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003307372164641957,
      "loss": 4.9151,
      "mean_token_accuracy": 0.21751775443553925,
      "num_tokens": 94706195.0,
      "step": 54600
    },
    {
      "entropy": 5.82441930770874,
      "epoch": 4.248434156778837,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033071036430921133,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21767239272594452,
      "num_tokens": 94714514.0,
      "step": 54605
    },
    {
      "entropy": 5.869266605377197,
      "epoch": 4.248823186150554,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00033068351130891076,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22421156615018845,
      "num_tokens": 94722898.0,
      "step": 54610
    },
    {
      "entropy": 5.94199504852295,
      "epoch": 4.249212215522272,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003306566574637016,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20383311063051224,
      "num_tokens": 94731266.0,
      "step": 54615
    },
    {
      "entropy": 5.834261131286621,
      "epoch": 4.24960124489399,
      "grad_norm": 1.421875,
      "learning_rate": 0.00033062980277399125,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21068195104599,
      "num_tokens": 94740317.0,
      "step": 54620
    },
    {
      "entropy": 5.828776121139526,
      "epoch": 4.249990274265707,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003306029472401871,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21891319751739502,
      "num_tokens": 94749388.0,
      "step": 54625
    },
    {
      "entropy": 5.784021377563477,
      "epoch": 4.250379303637424,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003305760908626968,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22689557522535325,
      "num_tokens": 94758357.0,
      "step": 54630
    },
    {
      "entropy": 5.996087217330933,
      "epoch": 4.250768333009142,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003305492336419277,
      "loss": 5.0637,
      "mean_token_accuracy": 0.20431413650512695,
      "num_tokens": 94767045.0,
      "step": 54635
    },
    {
      "entropy": 5.8504119396209715,
      "epoch": 4.25115736238086,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033052237557828763,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21591024547815324,
      "num_tokens": 94776322.0,
      "step": 54640
    },
    {
      "entropy": 5.76273021697998,
      "epoch": 4.251546391752577,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003304955166721837,
      "loss": 4.7768,
      "mean_token_accuracy": 0.2248910188674927,
      "num_tokens": 94785512.0,
      "step": 54645
    },
    {
      "entropy": 5.84410548210144,
      "epoch": 4.251935421124295,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003304686569240237,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2198322057723999,
      "num_tokens": 94794792.0,
      "step": 54650
    },
    {
      "entropy": 5.836422872543335,
      "epoch": 4.252324450496013,
      "grad_norm": 1.3125,
      "learning_rate": 0.000330441796334215,
      "loss": 4.9731,
      "mean_token_accuracy": 0.215606252849102,
      "num_tokens": 94803810.0,
      "step": 54655
    },
    {
      "entropy": 5.878294897079468,
      "epoch": 4.25271347986773,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003304149349031653,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22142995297908782,
      "num_tokens": 94812516.0,
      "step": 54660
    },
    {
      "entropy": 5.866740751266479,
      "epoch": 4.253102509239447,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003303880726312822,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21123270690441132,
      "num_tokens": 94821296.0,
      "step": 54665
    },
    {
      "entropy": 5.905311250686646,
      "epoch": 4.253491538611165,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00033036120951897334,
      "loss": 4.985,
      "mean_token_accuracy": 0.2080875962972641,
      "num_tokens": 94830201.0,
      "step": 54670
    },
    {
      "entropy": 5.845305442810059,
      "epoch": 4.253880567982883,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003303343455666462,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21716515570878983,
      "num_tokens": 94838797.0,
      "step": 54675
    },
    {
      "entropy": 5.7114630222320555,
      "epoch": 4.2542695973546,
      "grad_norm": 1.453125,
      "learning_rate": 0.00033030748077470846,
      "loss": 4.7326,
      "mean_token_accuracy": 0.23398550152778624,
      "num_tokens": 94848090.0,
      "step": 54680
    },
    {
      "entropy": 5.813529300689697,
      "epoch": 4.254658626726318,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00033028061514356766,
      "loss": 4.952,
      "mean_token_accuracy": 0.2135101556777954,
      "num_tokens": 94856238.0,
      "step": 54685
    },
    {
      "entropy": 5.866329479217529,
      "epoch": 4.255047656098036,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003302537486736315,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2133047863841057,
      "num_tokens": 94864950.0,
      "step": 54690
    },
    {
      "entropy": 5.741536426544189,
      "epoch": 4.255436685469753,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003302268813653077,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22160002887248992,
      "num_tokens": 94873616.0,
      "step": 54695
    },
    {
      "entropy": 5.8455893993377686,
      "epoch": 4.25582571484147,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003302000132190039,
      "loss": 4.9893,
      "mean_token_accuracy": 0.21261759400367736,
      "num_tokens": 94882568.0,
      "step": 54700
    },
    {
      "entropy": 5.877289628982544,
      "epoch": 4.256214744213188,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003301731442351277,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21729569137096405,
      "num_tokens": 94891079.0,
      "step": 54705
    },
    {
      "entropy": 5.892396116256714,
      "epoch": 4.256603773584906,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003301462744140869,
      "loss": 5.0,
      "mean_token_accuracy": 0.2129331946372986,
      "num_tokens": 94900342.0,
      "step": 54710
    },
    {
      "entropy": 5.777550172805786,
      "epoch": 4.256992802956623,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003301194037562892,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22190909832715988,
      "num_tokens": 94908580.0,
      "step": 54715
    },
    {
      "entropy": 5.835763549804687,
      "epoch": 4.257381832328341,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003300925322621422,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22201373875141145,
      "num_tokens": 94917166.0,
      "step": 54720
    },
    {
      "entropy": 5.843785047531128,
      "epoch": 4.257770861700059,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003300656599320537,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21539368480443954,
      "num_tokens": 94925761.0,
      "step": 54725
    },
    {
      "entropy": 5.81359224319458,
      "epoch": 4.2581598910717755,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003300387867664315,
      "loss": 4.8611,
      "mean_token_accuracy": 0.21661551296710968,
      "num_tokens": 94934037.0,
      "step": 54730
    },
    {
      "entropy": 5.816027736663818,
      "epoch": 4.258548920443493,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003300119127656833,
      "loss": 4.923,
      "mean_token_accuracy": 0.21549272388219834,
      "num_tokens": 94942414.0,
      "step": 54735
    },
    {
      "entropy": 5.809495496749878,
      "epoch": 4.258937949815211,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032998503793021694,
      "loss": 4.8496,
      "mean_token_accuracy": 0.22374883145093918,
      "num_tokens": 94950666.0,
      "step": 54740
    },
    {
      "entropy": 5.798728942871094,
      "epoch": 4.259326979186929,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003299581622604401,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2106999233365059,
      "num_tokens": 94959266.0,
      "step": 54745
    },
    {
      "entropy": 5.836488485336304,
      "epoch": 4.259716008558646,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032993128575676064,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20847178846597672,
      "num_tokens": 94967991.0,
      "step": 54750
    },
    {
      "entropy": 5.846051216125488,
      "epoch": 4.260105037930364,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032990440841958635,
      "loss": 4.935,
      "mean_token_accuracy": 0.21453418880701064,
      "num_tokens": 94976881.0,
      "step": 54755
    },
    {
      "entropy": 5.79466757774353,
      "epoch": 4.260494067302082,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000329877530249325,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22567687481641768,
      "num_tokens": 94985651.0,
      "step": 54760
    },
    {
      "entropy": 5.784200859069824,
      "epoch": 4.2608830966737985,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003298506512463845,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21447030305862427,
      "num_tokens": 94994355.0,
      "step": 54765
    },
    {
      "entropy": 5.791024589538575,
      "epoch": 4.261272126045516,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003298237714111727,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21618910878896713,
      "num_tokens": 95003747.0,
      "step": 54770
    },
    {
      "entropy": 5.85491533279419,
      "epoch": 4.261661155417234,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003297968907440974,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21203519105911256,
      "num_tokens": 95012285.0,
      "step": 54775
    },
    {
      "entropy": 5.813201189041138,
      "epoch": 4.262050184788952,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003297700092455665,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21497584581375123,
      "num_tokens": 95020372.0,
      "step": 54780
    },
    {
      "entropy": 5.874746894836425,
      "epoch": 4.262439214160669,
      "grad_norm": 1.375,
      "learning_rate": 0.00032974312691598794,
      "loss": 4.9177,
      "mean_token_accuracy": 0.20970916301012038,
      "num_tokens": 95028489.0,
      "step": 54785
    },
    {
      "entropy": 5.851466083526612,
      "epoch": 4.262828243532387,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032971624375576956,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22772799879312516,
      "num_tokens": 95036014.0,
      "step": 54790
    },
    {
      "entropy": 5.7969505310058596,
      "epoch": 4.263217272904104,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003296893597653192,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22391374558210372,
      "num_tokens": 95044544.0,
      "step": 54795
    },
    {
      "entropy": 5.805670213699341,
      "epoch": 4.2636063022758215,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003296624749450448,
      "loss": 4.9401,
      "mean_token_accuracy": 0.2121354177594185,
      "num_tokens": 95053307.0,
      "step": 54800
    },
    {
      "entropy": 5.813626003265381,
      "epoch": 4.263995331647539,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003296355892953545,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2173500806093216,
      "num_tokens": 95061900.0,
      "step": 54805
    },
    {
      "entropy": 5.8369749069213865,
      "epoch": 4.264384361019257,
      "grad_norm": 1.53125,
      "learning_rate": 0.000329608702816656,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21374083906412125,
      "num_tokens": 95070077.0,
      "step": 54810
    },
    {
      "entropy": 5.78107213973999,
      "epoch": 4.264773390390975,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003295818155093574,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2171681985259056,
      "num_tokens": 95078921.0,
      "step": 54815
    },
    {
      "entropy": 5.762372064590454,
      "epoch": 4.265162419762692,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032955492737386663,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23465891927480698,
      "num_tokens": 95086666.0,
      "step": 54820
    },
    {
      "entropy": 5.8002910137176515,
      "epoch": 4.26555144913441,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003295280384105916,
      "loss": 4.98,
      "mean_token_accuracy": 0.21793352961540222,
      "num_tokens": 95095383.0,
      "step": 54825
    },
    {
      "entropy": 5.8189630031585695,
      "epoch": 4.265940478506128,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00032950114861994047,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22275084555149077,
      "num_tokens": 95103934.0,
      "step": 54830
    },
    {
      "entropy": 5.881512451171875,
      "epoch": 4.2663295078778445,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032947425800232107,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2168121337890625,
      "num_tokens": 95113373.0,
      "step": 54835
    },
    {
      "entropy": 5.881409502029419,
      "epoch": 4.266718537249562,
      "grad_norm": 1.375,
      "learning_rate": 0.0003294473665581415,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21540509909391403,
      "num_tokens": 95121780.0,
      "step": 54840
    },
    {
      "entropy": 5.849076986312866,
      "epoch": 4.26710756662128,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00032942047428780985,
      "loss": 4.9954,
      "mean_token_accuracy": 0.20862145274877547,
      "num_tokens": 95130257.0,
      "step": 54845
    },
    {
      "entropy": 5.862047529220581,
      "epoch": 4.267496595992998,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003293935811917342,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21740685850381852,
      "num_tokens": 95138505.0,
      "step": 54850
    },
    {
      "entropy": 5.738326501846314,
      "epoch": 4.267885625364715,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003293666872703224,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22058558315038682,
      "num_tokens": 95147514.0,
      "step": 54855
    },
    {
      "entropy": 5.788561248779297,
      "epoch": 4.268274654736433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003293397925239827,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21905076503753662,
      "num_tokens": 95156613.0,
      "step": 54860
    },
    {
      "entropy": 5.883743333816528,
      "epoch": 4.26866368410815,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003293128969531231,
      "loss": 5.011,
      "mean_token_accuracy": 0.21260046809911728,
      "num_tokens": 95165522.0,
      "step": 54865
    },
    {
      "entropy": 5.934041738510132,
      "epoch": 4.2690527134798675,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003292860005581518,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21669624596834183,
      "num_tokens": 95174159.0,
      "step": 54870
    },
    {
      "entropy": 5.838985824584961,
      "epoch": 4.269441742851585,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00032925910333947694,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21921789199113845,
      "num_tokens": 95182500.0,
      "step": 54875
    },
    {
      "entropy": 5.7371663570404055,
      "epoch": 4.269830772223303,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003292322052975064,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2208938866853714,
      "num_tokens": 95191297.0,
      "step": 54880
    },
    {
      "entropy": 5.8271307945251465,
      "epoch": 4.270219801595021,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003292053064326485,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21705079972743987,
      "num_tokens": 95199716.0,
      "step": 54885
    },
    {
      "entropy": 5.924205541610718,
      "epoch": 4.270608830966738,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003291784067453114,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20631819367408752,
      "num_tokens": 95208700.0,
      "step": 54890
    },
    {
      "entropy": 5.8480369567871096,
      "epoch": 4.270997860338456,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032915150623590304,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21619944274425507,
      "num_tokens": 95217242.0,
      "step": 54895
    },
    {
      "entropy": 5.893847894668579,
      "epoch": 4.271386889710173,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032912460490483193,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21353087127208709,
      "num_tokens": 95225969.0,
      "step": 54900
    },
    {
      "entropy": 5.855908823013306,
      "epoch": 4.2717759190818905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003290977027525061,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21352205127477647,
      "num_tokens": 95234804.0,
      "step": 54905
    },
    {
      "entropy": 5.883644962310791,
      "epoch": 4.272164948453608,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032907079977933373,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20870378166437148,
      "num_tokens": 95243012.0,
      "step": 54910
    },
    {
      "entropy": 5.883186101913452,
      "epoch": 4.272553977825326,
      "grad_norm": 1.625,
      "learning_rate": 0.00032904389598572295,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22056670486927032,
      "num_tokens": 95251099.0,
      "step": 54915
    },
    {
      "entropy": 5.911118698120117,
      "epoch": 4.272943007197044,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032901699137208215,
      "loss": 5.0038,
      "mean_token_accuracy": 0.20982222259044647,
      "num_tokens": 95259357.0,
      "step": 54920
    },
    {
      "entropy": 5.891458415985108,
      "epoch": 4.273332036568761,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003289900859388194,
      "loss": 5.0124,
      "mean_token_accuracy": 0.20951216965913771,
      "num_tokens": 95268407.0,
      "step": 54925
    },
    {
      "entropy": 5.868859481811524,
      "epoch": 4.273721065940478,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032896317968634305,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21499414145946502,
      "num_tokens": 95277452.0,
      "step": 54930
    },
    {
      "entropy": 5.810532426834106,
      "epoch": 4.274110095312196,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003289362726150614,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22041234523057937,
      "num_tokens": 95287306.0,
      "step": 54935
    },
    {
      "entropy": 5.81248288154602,
      "epoch": 4.2744991246839135,
      "grad_norm": 1.59375,
      "learning_rate": 0.00032890936472538255,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22181656807661057,
      "num_tokens": 95296384.0,
      "step": 54940
    },
    {
      "entropy": 5.877361869812011,
      "epoch": 4.274888154055631,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000328882456017715,
      "loss": 4.9552,
      "mean_token_accuracy": 0.22213151305913925,
      "num_tokens": 95304818.0,
      "step": 54945
    },
    {
      "entropy": 5.833614683151245,
      "epoch": 4.275277183427349,
      "grad_norm": 1.484375,
      "learning_rate": 0.00032885554649246697,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21347809582948685,
      "num_tokens": 95313333.0,
      "step": 54950
    },
    {
      "entropy": 5.894190835952759,
      "epoch": 4.275666212799067,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003288286361500467,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21490027606487275,
      "num_tokens": 95321504.0,
      "step": 54955
    },
    {
      "entropy": 5.892676830291748,
      "epoch": 4.276055242170784,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032880172499086247,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21883108615875244,
      "num_tokens": 95329874.0,
      "step": 54960
    },
    {
      "entropy": 5.88817458152771,
      "epoch": 4.276444271542501,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003287748130153228,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20791219621896745,
      "num_tokens": 95338273.0,
      "step": 54965
    },
    {
      "entropy": 5.822776746749878,
      "epoch": 4.276833300914219,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003287479002238359,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22205694764852524,
      "num_tokens": 95346945.0,
      "step": 54970
    },
    {
      "entropy": 5.81406421661377,
      "epoch": 4.2772223302859365,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003287209866168102,
      "loss": 4.8527,
      "mean_token_accuracy": 0.22102379202842712,
      "num_tokens": 95355828.0,
      "step": 54975
    },
    {
      "entropy": 5.9019030094146725,
      "epoch": 4.277611359657654,
      "grad_norm": 1.34375,
      "learning_rate": 0.000328694072194654,
      "loss": 4.9645,
      "mean_token_accuracy": 0.20816127359867095,
      "num_tokens": 95364379.0,
      "step": 54980
    },
    {
      "entropy": 5.859149408340454,
      "epoch": 4.278000389029372,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032866715695777565,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21812351793050766,
      "num_tokens": 95373290.0,
      "step": 54985
    },
    {
      "entropy": 5.827863502502441,
      "epoch": 4.2783894184010896,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003286402409065837,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21077560931444167,
      "num_tokens": 95381641.0,
      "step": 54990
    },
    {
      "entropy": 5.798708152770996,
      "epoch": 4.278778447772807,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032861332404148647,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2138545662164688,
      "num_tokens": 95390428.0,
      "step": 54995
    },
    {
      "entropy": 5.855465984344482,
      "epoch": 4.279167477144524,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032858640636289235,
      "loss": 4.8809,
      "mean_token_accuracy": 0.22131982296705247,
      "num_tokens": 95398677.0,
      "step": 55000
    },
    {
      "entropy": 5.918880271911621,
      "epoch": 4.279556506516242,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003285594878712098,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2090146228671074,
      "num_tokens": 95406890.0,
      "step": 55005
    },
    {
      "entropy": 5.8172605514526365,
      "epoch": 4.2799455358879595,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003285325685668473,
      "loss": 4.8707,
      "mean_token_accuracy": 0.21443295776844024,
      "num_tokens": 95415937.0,
      "step": 55010
    },
    {
      "entropy": 5.761719846725464,
      "epoch": 4.280334565259677,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032850564845021327,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22111608386039733,
      "num_tokens": 95424509.0,
      "step": 55015
    },
    {
      "entropy": 5.829315090179444,
      "epoch": 4.280723594631395,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032847872752171615,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21944970339536668,
      "num_tokens": 95432776.0,
      "step": 55020
    },
    {
      "entropy": 5.950258588790893,
      "epoch": 4.2811126240031125,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032845180578176456,
      "loss": 5.0044,
      "mean_token_accuracy": 0.20795313864946366,
      "num_tokens": 95442632.0,
      "step": 55025
    },
    {
      "entropy": 5.8630283832550045,
      "epoch": 4.28150165337483,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032842488323076675,
      "loss": 4.8926,
      "mean_token_accuracy": 0.21726803928613664,
      "num_tokens": 95450883.0,
      "step": 55030
    },
    {
      "entropy": 5.755566215515136,
      "epoch": 4.281890682746547,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032839795986913146,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2193460702896118,
      "num_tokens": 95459596.0,
      "step": 55035
    },
    {
      "entropy": 5.899817848205567,
      "epoch": 4.282279712118265,
      "grad_norm": 1.375,
      "learning_rate": 0.00032837103569726713,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20928156822919847,
      "num_tokens": 95468601.0,
      "step": 55040
    },
    {
      "entropy": 5.846908044815064,
      "epoch": 4.2826687414899824,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003283441107155822,
      "loss": 4.9153,
      "mean_token_accuracy": 0.2145295336842537,
      "num_tokens": 95477596.0,
      "step": 55045
    },
    {
      "entropy": 5.800001430511474,
      "epoch": 4.2830577708617,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003283171849244854,
      "loss": 4.8221,
      "mean_token_accuracy": 0.22546463012695311,
      "num_tokens": 95486895.0,
      "step": 55050
    },
    {
      "entropy": 5.744867897033691,
      "epoch": 4.283446800233418,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032829025832438507,
      "loss": 4.7902,
      "mean_token_accuracy": 0.23085368424654007,
      "num_tokens": 95495302.0,
      "step": 55055
    },
    {
      "entropy": 5.832707357406616,
      "epoch": 4.2838358296051355,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032826333091569,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21548306196928024,
      "num_tokens": 95504284.0,
      "step": 55060
    },
    {
      "entropy": 5.801221084594727,
      "epoch": 4.284224858976852,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003282364026988087,
      "loss": 4.8651,
      "mean_token_accuracy": 0.22157531827688218,
      "num_tokens": 95512990.0,
      "step": 55065
    },
    {
      "entropy": 5.841635894775391,
      "epoch": 4.28461388834857,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032820947367414953,
      "loss": 4.9941,
      "mean_token_accuracy": 0.20947542786598206,
      "num_tokens": 95521742.0,
      "step": 55070
    },
    {
      "entropy": 5.80087628364563,
      "epoch": 4.285002917720288,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003281825438421215,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2205156445503235,
      "num_tokens": 95529680.0,
      "step": 55075
    },
    {
      "entropy": 5.791516065597534,
      "epoch": 4.285391947092005,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032815561320313286,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22506801933050155,
      "num_tokens": 95538102.0,
      "step": 55080
    },
    {
      "entropy": 5.843248748779297,
      "epoch": 4.285780976463723,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032812868175759245,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2143519103527069,
      "num_tokens": 95547149.0,
      "step": 55085
    },
    {
      "entropy": 5.883685350418091,
      "epoch": 4.286170005835441,
      "grad_norm": 1.25,
      "learning_rate": 0.0003281017495059088,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2152955412864685,
      "num_tokens": 95555670.0,
      "step": 55090
    },
    {
      "entropy": 5.923160791397095,
      "epoch": 4.2865590352071585,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003280748164484907,
      "loss": 5.0063,
      "mean_token_accuracy": 0.20940128415822984,
      "num_tokens": 95565080.0,
      "step": 55095
    },
    {
      "entropy": 5.7549244403839115,
      "epoch": 4.286948064578875,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003280478825857468,
      "loss": 4.9028,
      "mean_token_accuracy": 0.220510533452034,
      "num_tokens": 95574563.0,
      "step": 55100
    },
    {
      "entropy": 5.853540134429932,
      "epoch": 4.287337093950593,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00032802094791808565,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21017315834760666,
      "num_tokens": 95583149.0,
      "step": 55105
    },
    {
      "entropy": 5.8665729522705075,
      "epoch": 4.287726123322311,
      "grad_norm": 1.328125,
      "learning_rate": 0.000327994012445916,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21795314252376558,
      "num_tokens": 95591921.0,
      "step": 55110
    },
    {
      "entropy": 5.920933103561401,
      "epoch": 4.288115152694028,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003279670761696465,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21067111641168595,
      "num_tokens": 95601421.0,
      "step": 55115
    },
    {
      "entropy": 5.879787349700928,
      "epoch": 4.288504182065746,
      "grad_norm": 1.375,
      "learning_rate": 0.00032794013908968593,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2180783972144127,
      "num_tokens": 95610362.0,
      "step": 55120
    },
    {
      "entropy": 5.790458726882934,
      "epoch": 4.288893211437464,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003279132012064431,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2208317667245865,
      "num_tokens": 95618432.0,
      "step": 55125
    },
    {
      "entropy": 5.862422513961792,
      "epoch": 4.289282240809181,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003278862625203266,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21313380897045137,
      "num_tokens": 95626563.0,
      "step": 55130
    },
    {
      "entropy": 5.866970586776733,
      "epoch": 4.289671270180898,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032785932303174524,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2146655187010765,
      "num_tokens": 95635589.0,
      "step": 55135
    },
    {
      "entropy": 5.830287027359009,
      "epoch": 4.290060299552616,
      "grad_norm": 1.375,
      "learning_rate": 0.00032783238274110787,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22286026328802108,
      "num_tokens": 95643841.0,
      "step": 55140
    },
    {
      "entropy": 5.8444130420684814,
      "epoch": 4.290449328924334,
      "grad_norm": 1.484375,
      "learning_rate": 0.00032780544164882317,
      "loss": 4.996,
      "mean_token_accuracy": 0.2208150401711464,
      "num_tokens": 95652059.0,
      "step": 55145
    },
    {
      "entropy": 5.858502531051636,
      "epoch": 4.290838358296051,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003277784997552999,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21006468534469605,
      "num_tokens": 95661424.0,
      "step": 55150
    },
    {
      "entropy": 5.877637815475464,
      "epoch": 4.291227387667769,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032775155706094684,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21560871750116348,
      "num_tokens": 95669617.0,
      "step": 55155
    },
    {
      "entropy": 5.873055934906006,
      "epoch": 4.291616417039487,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003277246135661729,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22072556614875793,
      "num_tokens": 95677831.0,
      "step": 55160
    },
    {
      "entropy": 5.880334711074829,
      "epoch": 4.2920054464112045,
      "grad_norm": 1.375,
      "learning_rate": 0.0003276976692713869,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2101071372628212,
      "num_tokens": 95686320.0,
      "step": 55165
    },
    {
      "entropy": 5.796175861358643,
      "epoch": 4.292394475782921,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003276707241769976,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22798438519239425,
      "num_tokens": 95694504.0,
      "step": 55170
    },
    {
      "entropy": 5.911861562728882,
      "epoch": 4.292783505154639,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000327643778283414,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2051316499710083,
      "num_tokens": 95702981.0,
      "step": 55175
    },
    {
      "entropy": 5.85449481010437,
      "epoch": 4.293172534526357,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032761683159104475,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22112558782100677,
      "num_tokens": 95711447.0,
      "step": 55180
    },
    {
      "entropy": 5.861120986938476,
      "epoch": 4.293561563898074,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003275898841002989,
      "loss": 4.8729,
      "mean_token_accuracy": 0.21780747622251512,
      "num_tokens": 95719539.0,
      "step": 55185
    },
    {
      "entropy": 5.7952686786651615,
      "epoch": 4.293950593269792,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032756293581158517,
      "loss": 4.9111,
      "mean_token_accuracy": 0.220981465280056,
      "num_tokens": 95728512.0,
      "step": 55190
    },
    {
      "entropy": 5.832409524917603,
      "epoch": 4.29433962264151,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003275359867253127,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21669438779354094,
      "num_tokens": 95737153.0,
      "step": 55195
    },
    {
      "entropy": 5.942195701599121,
      "epoch": 4.294728652013227,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003275090368418901,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21149560660123826,
      "num_tokens": 95745942.0,
      "step": 55200
    },
    {
      "entropy": 5.83469409942627,
      "epoch": 4.295117681384944,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003274820861617265,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21120262295007705,
      "num_tokens": 95755074.0,
      "step": 55205
    },
    {
      "entropy": 5.833517169952392,
      "epoch": 4.295506710756662,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003274551346852308,
      "loss": 4.951,
      "mean_token_accuracy": 0.21271960139274598,
      "num_tokens": 95763590.0,
      "step": 55210
    },
    {
      "entropy": 5.851959848403931,
      "epoch": 4.29589574012838,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003274281824128119,
      "loss": 4.9432,
      "mean_token_accuracy": 0.20896997600793837,
      "num_tokens": 95772036.0,
      "step": 55215
    },
    {
      "entropy": 5.856668901443482,
      "epoch": 4.296284769500097,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032740122934487876,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22545751333236694,
      "num_tokens": 95780559.0,
      "step": 55220
    },
    {
      "entropy": 5.852445363998413,
      "epoch": 4.296673798871815,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003273742754818405,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21392373144626617,
      "num_tokens": 95789022.0,
      "step": 55225
    },
    {
      "entropy": 5.821959590911865,
      "epoch": 4.297062828243533,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003273473208241058,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2134683445096016,
      "num_tokens": 95797341.0,
      "step": 55230
    },
    {
      "entropy": 5.795345497131348,
      "epoch": 4.29745185761525,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032732036537208386,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2209555670619011,
      "num_tokens": 95805961.0,
      "step": 55235
    },
    {
      "entropy": 5.831258630752563,
      "epoch": 4.297840886986967,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003272934091261837,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21515632718801497,
      "num_tokens": 95814429.0,
      "step": 55240
    },
    {
      "entropy": 5.843974351882935,
      "epoch": 4.298229916358685,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032726645208681427,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2207690268754959,
      "num_tokens": 95822612.0,
      "step": 55245
    },
    {
      "entropy": 5.80584020614624,
      "epoch": 4.298618945730403,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032723949425438465,
      "loss": 4.828,
      "mean_token_accuracy": 0.2241604059934616,
      "num_tokens": 95830957.0,
      "step": 55250
    },
    {
      "entropy": 5.865558004379272,
      "epoch": 4.29900797510212,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032721253562930383,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21349277794361116,
      "num_tokens": 95839676.0,
      "step": 55255
    },
    {
      "entropy": 5.878401517868042,
      "epoch": 4.299397004473838,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032718557621198087,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21890877932310104,
      "num_tokens": 95848093.0,
      "step": 55260
    },
    {
      "entropy": 5.878016233444214,
      "epoch": 4.299786033845555,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003271586160028249,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2076333001255989,
      "num_tokens": 95856971.0,
      "step": 55265
    },
    {
      "entropy": 5.793928527832032,
      "epoch": 4.300175063217273,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003271316550022449,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21493091583251953,
      "num_tokens": 95865978.0,
      "step": 55270
    },
    {
      "entropy": 5.864184379577637,
      "epoch": 4.30056409258899,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032710469321065,
      "loss": 4.9742,
      "mean_token_accuracy": 0.21658277213573457,
      "num_tokens": 95874605.0,
      "step": 55275
    },
    {
      "entropy": 5.891385459899903,
      "epoch": 4.300953121960708,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003270777306284493,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2117101162672043,
      "num_tokens": 95883460.0,
      "step": 55280
    },
    {
      "entropy": 5.942352914810181,
      "epoch": 4.301342151332426,
      "grad_norm": 1.40625,
      "learning_rate": 0.000327050767256052,
      "loss": 5.063,
      "mean_token_accuracy": 0.20868974924087524,
      "num_tokens": 95892500.0,
      "step": 55285
    },
    {
      "entropy": 5.908770275115967,
      "epoch": 4.301731180704143,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032702380309386714,
      "loss": 4.999,
      "mean_token_accuracy": 0.20895248055458068,
      "num_tokens": 95902154.0,
      "step": 55290
    },
    {
      "entropy": 5.867869567871094,
      "epoch": 4.302120210075861,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032699683814230385,
      "loss": 4.931,
      "mean_token_accuracy": 0.21977607607841493,
      "num_tokens": 95910809.0,
      "step": 55295
    },
    {
      "entropy": 5.8903004169464115,
      "epoch": 4.302509239447578,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003269698724017713,
      "loss": 5.0403,
      "mean_token_accuracy": 0.2052719384431839,
      "num_tokens": 95919194.0,
      "step": 55300
    },
    {
      "entropy": 5.8414942741394045,
      "epoch": 4.302898268819296,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032694290587267867,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22431287169456482,
      "num_tokens": 95927202.0,
      "step": 55305
    },
    {
      "entropy": 5.853916597366333,
      "epoch": 4.303287298191013,
      "grad_norm": 1.375,
      "learning_rate": 0.0003269159385554352,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22306456714868544,
      "num_tokens": 95935821.0,
      "step": 55310
    },
    {
      "entropy": 5.834225225448608,
      "epoch": 4.303676327562731,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003268889704504499,
      "loss": 4.875,
      "mean_token_accuracy": 0.21587986797094344,
      "num_tokens": 95944863.0,
      "step": 55315
    },
    {
      "entropy": 5.890927362442016,
      "epoch": 4.304065356934449,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032686200155813196,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21670316606760026,
      "num_tokens": 95953250.0,
      "step": 55320
    },
    {
      "entropy": 5.849037837982178,
      "epoch": 4.304454386306166,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003268350318788908,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2110270231962204,
      "num_tokens": 95961679.0,
      "step": 55325
    },
    {
      "entropy": 5.842361927032471,
      "epoch": 4.304843415677884,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003268080614131356,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21862058937549592,
      "num_tokens": 95969599.0,
      "step": 55330
    },
    {
      "entropy": 5.806553792953491,
      "epoch": 4.305232445049601,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003267810901612754,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22767717838287355,
      "num_tokens": 95977954.0,
      "step": 55335
    },
    {
      "entropy": 5.857578992843628,
      "epoch": 4.305621474421319,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032675411812371975,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2134447365999222,
      "num_tokens": 95986720.0,
      "step": 55340
    },
    {
      "entropy": 5.777856397628784,
      "epoch": 4.306010503793036,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032672714530087763,
      "loss": 4.8271,
      "mean_token_accuracy": 0.21485936194658278,
      "num_tokens": 95995053.0,
      "step": 55345
    },
    {
      "entropy": 5.859212827682495,
      "epoch": 4.306399533164754,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032670017169315836,
      "loss": 5.0131,
      "mean_token_accuracy": 0.2090982601046562,
      "num_tokens": 96004074.0,
      "step": 55350
    },
    {
      "entropy": 5.8185442924499515,
      "epoch": 4.306788562536472,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003266731973009713,
      "loss": 4.8594,
      "mean_token_accuracy": 0.21759750843048095,
      "num_tokens": 96012845.0,
      "step": 55355
    },
    {
      "entropy": 5.865191221237183,
      "epoch": 4.307177591908189,
      "grad_norm": 1.375,
      "learning_rate": 0.0003266462221247257,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22414237409830093,
      "num_tokens": 96021089.0,
      "step": 55360
    },
    {
      "entropy": 5.862734460830689,
      "epoch": 4.307566621279907,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000326619246164831,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21296327114105223,
      "num_tokens": 96029587.0,
      "step": 55365
    },
    {
      "entropy": 5.829333353042602,
      "epoch": 4.307955650651624,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003265922694216963,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21202111840248108,
      "num_tokens": 96039476.0,
      "step": 55370
    },
    {
      "entropy": 5.829576539993286,
      "epoch": 4.308344680023342,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032656529189573107,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21287769228219985,
      "num_tokens": 96047621.0,
      "step": 55375
    },
    {
      "entropy": 5.771330118179321,
      "epoch": 4.308733709395059,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003265383135873446,
      "loss": 4.7708,
      "mean_token_accuracy": 0.2302372932434082,
      "num_tokens": 96056876.0,
      "step": 55380
    },
    {
      "entropy": 5.831834125518799,
      "epoch": 4.309122738766777,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003265113344969462,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21584170758724214,
      "num_tokens": 96065126.0,
      "step": 55385
    },
    {
      "entropy": 5.745485115051269,
      "epoch": 4.309511768138495,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003264843546249453,
      "loss": 4.869,
      "mean_token_accuracy": 0.21964472234249116,
      "num_tokens": 96074316.0,
      "step": 55390
    },
    {
      "entropy": 5.882855653762817,
      "epoch": 4.309900797510212,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032645737397175135,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21389081627130507,
      "num_tokens": 96082227.0,
      "step": 55395
    },
    {
      "entropy": 5.8174365043640135,
      "epoch": 4.310289826881929,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032643039253777356,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22587417513132096,
      "num_tokens": 96091385.0,
      "step": 55400
    },
    {
      "entropy": 5.859004831314087,
      "epoch": 4.310678856253647,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032640341032342147,
      "loss": 5.0405,
      "mean_token_accuracy": 0.21065380275249482,
      "num_tokens": 96099551.0,
      "step": 55405
    },
    {
      "entropy": 5.726630544662475,
      "epoch": 4.311067885625365,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032637642732910447,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23032439202070237,
      "num_tokens": 96107801.0,
      "step": 55410
    },
    {
      "entropy": 5.8536440372467045,
      "epoch": 4.311456914997082,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032634944355523196,
      "loss": 5.0039,
      "mean_token_accuracy": 0.20518322587013244,
      "num_tokens": 96116917.0,
      "step": 55415
    },
    {
      "entropy": 5.808343935012817,
      "epoch": 4.3118459443688,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003263224590022133,
      "loss": 4.8379,
      "mean_token_accuracy": 0.2260502591729164,
      "num_tokens": 96124970.0,
      "step": 55420
    },
    {
      "entropy": 5.862708806991577,
      "epoch": 4.312234973740518,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032629547367045805,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21923464834690093,
      "num_tokens": 96133956.0,
      "step": 55425
    },
    {
      "entropy": 5.907765054702759,
      "epoch": 4.312624003112235,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032626848756037565,
      "loss": 4.998,
      "mean_token_accuracy": 0.2063562273979187,
      "num_tokens": 96142182.0,
      "step": 55430
    },
    {
      "entropy": 5.8354353427886965,
      "epoch": 4.313013032483952,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00032624150067237553,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22208960205316544,
      "num_tokens": 96150501.0,
      "step": 55435
    },
    {
      "entropy": 5.911330461502075,
      "epoch": 4.31340206185567,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003262145130068672,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21903022527694702,
      "num_tokens": 96159386.0,
      "step": 55440
    },
    {
      "entropy": 5.906987905502319,
      "epoch": 4.313791091227388,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032618752456426017,
      "loss": 5.0147,
      "mean_token_accuracy": 0.20584727376699447,
      "num_tokens": 96167313.0,
      "step": 55445
    },
    {
      "entropy": 5.829735803604126,
      "epoch": 4.314180120599105,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032616053534496387,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22804607450962067,
      "num_tokens": 96176120.0,
      "step": 55450
    },
    {
      "entropy": 5.890226078033447,
      "epoch": 4.314569149970823,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003261335453493879,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20846241265535354,
      "num_tokens": 96185088.0,
      "step": 55455
    },
    {
      "entropy": 5.872335958480835,
      "epoch": 4.314958179342541,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032610655457794184,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21268551796674728,
      "num_tokens": 96193136.0,
      "step": 55460
    },
    {
      "entropy": 5.860608768463135,
      "epoch": 4.3153472087142575,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032607956303103516,
      "loss": 4.969,
      "mean_token_accuracy": 0.21453715413808822,
      "num_tokens": 96202141.0,
      "step": 55465
    },
    {
      "entropy": 5.7723511219024655,
      "epoch": 4.315736238085975,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032605257070907734,
      "loss": 4.8288,
      "mean_token_accuracy": 0.2237903133034706,
      "num_tokens": 96210523.0,
      "step": 55470
    },
    {
      "entropy": 5.753247880935669,
      "epoch": 4.316125267457693,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032602557761247805,
      "loss": 4.8388,
      "mean_token_accuracy": 0.2269977569580078,
      "num_tokens": 96218771.0,
      "step": 55475
    },
    {
      "entropy": 5.824993562698364,
      "epoch": 4.316514296829411,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003259985837416468,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2155532106757164,
      "num_tokens": 96227488.0,
      "step": 55480
    },
    {
      "entropy": 5.849914407730102,
      "epoch": 4.316903326201128,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003259715890969932,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20835191160440444,
      "num_tokens": 96236119.0,
      "step": 55485
    },
    {
      "entropy": 5.874361991882324,
      "epoch": 4.317292355572846,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003259445936789269,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21514023691415787,
      "num_tokens": 96244027.0,
      "step": 55490
    },
    {
      "entropy": 5.792952823638916,
      "epoch": 4.317681384944564,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032591759748785753,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22385378330945968,
      "num_tokens": 96252442.0,
      "step": 55495
    },
    {
      "entropy": 5.808651161193848,
      "epoch": 4.3180704143162805,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003258906005241946,
      "loss": 4.8882,
      "mean_token_accuracy": 0.21964877396821975,
      "num_tokens": 96260818.0,
      "step": 55500
    },
    {
      "entropy": 5.786374568939209,
      "epoch": 4.318459443687998,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003258636027883478,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2192978248000145,
      "num_tokens": 96268914.0,
      "step": 55505
    },
    {
      "entropy": 5.829056453704834,
      "epoch": 4.318848473059716,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003258366042807269,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21526458114385605,
      "num_tokens": 96277438.0,
      "step": 55510
    },
    {
      "entropy": 5.818345212936402,
      "epoch": 4.319237502431434,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003258096050017413,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2207943320274353,
      "num_tokens": 96286088.0,
      "step": 55515
    },
    {
      "entropy": 5.863838529586792,
      "epoch": 4.319626531803151,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003257826049518008,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21066470742225646,
      "num_tokens": 96294741.0,
      "step": 55520
    },
    {
      "entropy": 5.790809535980225,
      "epoch": 4.320015561174869,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003257556041313153,
      "loss": 4.906,
      "mean_token_accuracy": 0.21563355326652528,
      "num_tokens": 96303689.0,
      "step": 55525
    },
    {
      "entropy": 5.771593523025513,
      "epoch": 4.320404590546587,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003257286025406941,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2159102663397789,
      "num_tokens": 96312140.0,
      "step": 55530
    },
    {
      "entropy": 5.879526901245117,
      "epoch": 4.3207936199183035,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00032570160018034717,
      "loss": 4.986,
      "mean_token_accuracy": 0.21532522439956664,
      "num_tokens": 96321161.0,
      "step": 55535
    },
    {
      "entropy": 5.886573123931885,
      "epoch": 4.321182649290021,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032567459705068415,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21237818747758866,
      "num_tokens": 96330290.0,
      "step": 55540
    },
    {
      "entropy": 5.823737335205078,
      "epoch": 4.321571678661739,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032564759315211475,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21561281234025956,
      "num_tokens": 96338998.0,
      "step": 55545
    },
    {
      "entropy": 5.84109582901001,
      "epoch": 4.3219607080334566,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032562058848504877,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2147877961397171,
      "num_tokens": 96348528.0,
      "step": 55550
    },
    {
      "entropy": 5.835260105133057,
      "epoch": 4.322349737405174,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003255935830498959,
      "loss": 4.8967,
      "mean_token_accuracy": 0.2194436490535736,
      "num_tokens": 96357323.0,
      "step": 55555
    },
    {
      "entropy": 5.910776138305664,
      "epoch": 4.322738766776892,
      "grad_norm": 1.421875,
      "learning_rate": 0.000325566576847066,
      "loss": 4.985,
      "mean_token_accuracy": 0.21233998686075212,
      "num_tokens": 96366099.0,
      "step": 55560
    },
    {
      "entropy": 5.890255451202393,
      "epoch": 4.32312779614861,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003255395698769687,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2212217330932617,
      "num_tokens": 96374093.0,
      "step": 55565
    },
    {
      "entropy": 5.859545803070068,
      "epoch": 4.3235168255203265,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003255125621400139,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21254828125238417,
      "num_tokens": 96382641.0,
      "step": 55570
    },
    {
      "entropy": 5.822468185424805,
      "epoch": 4.323905854892044,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003254855536366114,
      "loss": 4.9112,
      "mean_token_accuracy": 0.215688656270504,
      "num_tokens": 96391504.0,
      "step": 55575
    },
    {
      "entropy": 5.791967153549194,
      "epoch": 4.324294884263762,
      "grad_norm": 1.359375,
      "learning_rate": 0.000325458544367171,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21149833351373673,
      "num_tokens": 96400437.0,
      "step": 55580
    },
    {
      "entropy": 5.887205123901367,
      "epoch": 4.3246839136354795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032543153433210243,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20876852571964263,
      "num_tokens": 96408632.0,
      "step": 55585
    },
    {
      "entropy": 5.8072288036346436,
      "epoch": 4.325072943007197,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003254045235318156,
      "loss": 4.8633,
      "mean_token_accuracy": 0.21757478564977645,
      "num_tokens": 96417307.0,
      "step": 55590
    },
    {
      "entropy": 5.825234270095825,
      "epoch": 4.325461972378915,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032537751196672033,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22347909063100815,
      "num_tokens": 96426209.0,
      "step": 55595
    },
    {
      "entropy": 5.844443893432617,
      "epoch": 4.325851001750632,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003253504996372266,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21933119893074035,
      "num_tokens": 96435119.0,
      "step": 55600
    },
    {
      "entropy": 5.8342437744140625,
      "epoch": 4.3262400311223494,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003253234865437442,
      "loss": 4.8845,
      "mean_token_accuracy": 0.2195989653468132,
      "num_tokens": 96443940.0,
      "step": 55605
    },
    {
      "entropy": 5.846450567245483,
      "epoch": 4.326629060494067,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003252964726866829,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21730911135673522,
      "num_tokens": 96452791.0,
      "step": 55610
    },
    {
      "entropy": 5.860593366622925,
      "epoch": 4.327018089865785,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00032526945806645265,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20944087505340575,
      "num_tokens": 96461210.0,
      "step": 55615
    },
    {
      "entropy": 5.869510269165039,
      "epoch": 4.3274071192375025,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003252424426834634,
      "loss": 4.906,
      "mean_token_accuracy": 0.22268157601356506,
      "num_tokens": 96469331.0,
      "step": 55620
    },
    {
      "entropy": 5.871769523620605,
      "epoch": 4.32779614860922,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003252154265381251,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21273259371519088,
      "num_tokens": 96478114.0,
      "step": 55625
    },
    {
      "entropy": 5.881192541122436,
      "epoch": 4.328185177980938,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003251884096308476,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21525080054998397,
      "num_tokens": 96486884.0,
      "step": 55630
    },
    {
      "entropy": 5.8102458953857425,
      "epoch": 4.328574207352655,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032516139196204087,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2091996043920517,
      "num_tokens": 96495584.0,
      "step": 55635
    },
    {
      "entropy": 5.840975666046143,
      "epoch": 4.328963236724372,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032513437353211483,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21153714656829833,
      "num_tokens": 96504241.0,
      "step": 55640
    },
    {
      "entropy": 5.884627962112427,
      "epoch": 4.32935226609609,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003251073543414796,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21262812465429307,
      "num_tokens": 96512859.0,
      "step": 55645
    },
    {
      "entropy": 5.8136622428894045,
      "epoch": 4.329741295467808,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032508033439054493,
      "loss": 4.9396,
      "mean_token_accuracy": 0.21529333293437958,
      "num_tokens": 96521866.0,
      "step": 55650
    },
    {
      "entropy": 5.810681343078613,
      "epoch": 4.3301303248395255,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003250533136797208,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2122107371687889,
      "num_tokens": 96530422.0,
      "step": 55655
    },
    {
      "entropy": 5.732446622848511,
      "epoch": 4.330519354211243,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032502629220941743,
      "loss": 4.7787,
      "mean_token_accuracy": 0.23482239246368408,
      "num_tokens": 96538785.0,
      "step": 55660
    },
    {
      "entropy": 5.826683759689331,
      "epoch": 4.33090838358296,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032499926998004464,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21832953989505768,
      "num_tokens": 96546942.0,
      "step": 55665
    },
    {
      "entropy": 5.88804407119751,
      "epoch": 4.331297412954678,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003249722469920126,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20886122435331345,
      "num_tokens": 96555864.0,
      "step": 55670
    },
    {
      "entropy": 5.8177087783813475,
      "epoch": 4.331686442326395,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003249452232457312,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21929075419902802,
      "num_tokens": 96564493.0,
      "step": 55675
    },
    {
      "entropy": 5.802357912063599,
      "epoch": 4.332075471698113,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032491819874161046,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22094125300645828,
      "num_tokens": 96572580.0,
      "step": 55680
    },
    {
      "entropy": 5.774946880340576,
      "epoch": 4.332464501069831,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003248911734800606,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21621127277612687,
      "num_tokens": 96580837.0,
      "step": 55685
    },
    {
      "entropy": 5.88614821434021,
      "epoch": 4.3328535304415485,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032486414746149154,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20757483094930648,
      "num_tokens": 96590177.0,
      "step": 55690
    },
    {
      "entropy": 5.812265825271607,
      "epoch": 4.333242559813266,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032483712068631345,
      "loss": 4.8436,
      "mean_token_accuracy": 0.2200772926211357,
      "num_tokens": 96597656.0,
      "step": 55695
    },
    {
      "entropy": 5.867917490005493,
      "epoch": 4.333631589184984,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003248100931549364,
      "loss": 4.933,
      "mean_token_accuracy": 0.2085030570626259,
      "num_tokens": 96606118.0,
      "step": 55700
    },
    {
      "entropy": 5.859596872329712,
      "epoch": 4.334020618556701,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00032478306486777037,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21006937175989152,
      "num_tokens": 96615088.0,
      "step": 55705
    },
    {
      "entropy": 5.875917863845825,
      "epoch": 4.334409647928418,
      "grad_norm": 1.375,
      "learning_rate": 0.0003247560358252256,
      "loss": 4.921,
      "mean_token_accuracy": 0.21881041973829268,
      "num_tokens": 96624024.0,
      "step": 55710
    },
    {
      "entropy": 5.834152698516846,
      "epoch": 4.334798677300136,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032472900602771224,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2217002436518669,
      "num_tokens": 96633002.0,
      "step": 55715
    },
    {
      "entropy": 5.839703130722046,
      "epoch": 4.335187706671854,
      "grad_norm": 1.375,
      "learning_rate": 0.00032470197547564036,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2131392017006874,
      "num_tokens": 96641587.0,
      "step": 55720
    },
    {
      "entropy": 5.759793376922607,
      "epoch": 4.3355767360435715,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032467494416942006,
      "loss": 4.8628,
      "mean_token_accuracy": 0.21641114801168443,
      "num_tokens": 96650279.0,
      "step": 55725
    },
    {
      "entropy": 5.8337469577789305,
      "epoch": 4.335965765415289,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003246479121094616,
      "loss": 4.933,
      "mean_token_accuracy": 0.21473925411701203,
      "num_tokens": 96659261.0,
      "step": 55730
    },
    {
      "entropy": 5.841267442703247,
      "epoch": 4.336354794787006,
      "grad_norm": 1.3125,
      "learning_rate": 0.000324620879296175,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2109975442290306,
      "num_tokens": 96668064.0,
      "step": 55735
    },
    {
      "entropy": 5.839495658874512,
      "epoch": 4.336743824158724,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003245938457299706,
      "loss": 4.8955,
      "mean_token_accuracy": 0.21071844398975373,
      "num_tokens": 96676442.0,
      "step": 55740
    },
    {
      "entropy": 5.790213346481323,
      "epoch": 4.337132853530441,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00032456681141125853,
      "loss": 4.8821,
      "mean_token_accuracy": 0.2237157866358757,
      "num_tokens": 96684752.0,
      "step": 55745
    },
    {
      "entropy": 5.824897193908692,
      "epoch": 4.337521882902159,
      "grad_norm": 1.5625,
      "learning_rate": 0.000324539776340449,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21848150342702866,
      "num_tokens": 96693179.0,
      "step": 55750
    },
    {
      "entropy": 5.842472410202026,
      "epoch": 4.337910912273877,
      "grad_norm": 1.375,
      "learning_rate": 0.0003245127405179522,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21417265385389328,
      "num_tokens": 96702139.0,
      "step": 55755
    },
    {
      "entropy": 5.836424827575684,
      "epoch": 4.3382999416455945,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003244857039441784,
      "loss": 4.9315,
      "mean_token_accuracy": 0.22030932009220122,
      "num_tokens": 96710372.0,
      "step": 55760
    },
    {
      "entropy": 5.879325819015503,
      "epoch": 4.338688971017312,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032445866661953774,
      "loss": 4.9714,
      "mean_token_accuracy": 0.2103179484605789,
      "num_tokens": 96718514.0,
      "step": 55765
    },
    {
      "entropy": 5.799934577941895,
      "epoch": 4.339078000389029,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032443162854444064,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21605355441570281,
      "num_tokens": 96727014.0,
      "step": 55770
    },
    {
      "entropy": 5.85666069984436,
      "epoch": 4.339467029760747,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003244045897192972,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21584648340940477,
      "num_tokens": 96735590.0,
      "step": 55775
    },
    {
      "entropy": 5.889454555511475,
      "epoch": 4.339856059132464,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032437755014451777,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20983894169330597,
      "num_tokens": 96745155.0,
      "step": 55780
    },
    {
      "entropy": 5.837863826751709,
      "epoch": 4.340245088504182,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032435050982051246,
      "loss": 4.9101,
      "mean_token_accuracy": 0.21393893510103226,
      "num_tokens": 96754412.0,
      "step": 55785
    },
    {
      "entropy": 5.863303279876709,
      "epoch": 4.3406341178759,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032432346874769187,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22138460576534272,
      "num_tokens": 96763058.0,
      "step": 55790
    },
    {
      "entropy": 5.909774303436279,
      "epoch": 4.3410231472476175,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032429642692646614,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21747536659240724,
      "num_tokens": 96771403.0,
      "step": 55795
    },
    {
      "entropy": 5.857864141464233,
      "epoch": 4.341412176619334,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032426938435724554,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21226586252450944,
      "num_tokens": 96780305.0,
      "step": 55800
    },
    {
      "entropy": 5.820909643173218,
      "epoch": 4.341801205991052,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00032424234104044055,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2312459871172905,
      "num_tokens": 96788536.0,
      "step": 55805
    },
    {
      "entropy": 5.7993567943572994,
      "epoch": 4.34219023536277,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003242152969764613,
      "loss": 4.8487,
      "mean_token_accuracy": 0.2197882816195488,
      "num_tokens": 96797078.0,
      "step": 55810
    },
    {
      "entropy": 5.8345277309417725,
      "epoch": 4.342579264734487,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003241882521657183,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21311198621988298,
      "num_tokens": 96805781.0,
      "step": 55815
    },
    {
      "entropy": 5.918642377853393,
      "epoch": 4.342968294106205,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00032416120660862187,
      "loss": 5.027,
      "mean_token_accuracy": 0.21792485415935517,
      "num_tokens": 96814382.0,
      "step": 55820
    },
    {
      "entropy": 5.948053693771362,
      "epoch": 4.343357323477923,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003241341603055823,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21166738420724868,
      "num_tokens": 96823775.0,
      "step": 55825
    },
    {
      "entropy": 5.806930923461914,
      "epoch": 4.3437463528496405,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003241071132570101,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22760902494192123,
      "num_tokens": 96831952.0,
      "step": 55830
    },
    {
      "entropy": 5.846064662933349,
      "epoch": 4.344135382221357,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003240800654633157,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2155827060341835,
      "num_tokens": 96839848.0,
      "step": 55835
    },
    {
      "entropy": 5.896370077133179,
      "epoch": 4.344524411593075,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003240530169249094,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20939721465110778,
      "num_tokens": 96849282.0,
      "step": 55840
    },
    {
      "entropy": 5.838213300704956,
      "epoch": 4.344913440964793,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003240259676422016,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21659334003925323,
      "num_tokens": 96858917.0,
      "step": 55845
    },
    {
      "entropy": 5.906603908538818,
      "epoch": 4.34530247033651,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032399891761560274,
      "loss": 4.9834,
      "mean_token_accuracy": 0.20922621488571166,
      "num_tokens": 96867921.0,
      "step": 55850
    },
    {
      "entropy": 5.824683475494385,
      "epoch": 4.345691499708228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003239718668455233,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2167795106768608,
      "num_tokens": 96876886.0,
      "step": 55855
    },
    {
      "entropy": 5.836410808563232,
      "epoch": 4.346080529079946,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003239448153323737,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22197705060243605,
      "num_tokens": 96885596.0,
      "step": 55860
    },
    {
      "entropy": 5.877079486846924,
      "epoch": 4.3464695584516635,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032391776307656456,
      "loss": 4.9827,
      "mean_token_accuracy": 0.214811247587204,
      "num_tokens": 96894440.0,
      "step": 55865
    },
    {
      "entropy": 5.870606994628906,
      "epoch": 4.34685858782338,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032389071007850616,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21951773017644882,
      "num_tokens": 96902533.0,
      "step": 55870
    },
    {
      "entropy": 5.880166006088257,
      "epoch": 4.347247617195098,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032386365633860914,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20984472334384918,
      "num_tokens": 96910737.0,
      "step": 55875
    },
    {
      "entropy": 5.800865364074707,
      "epoch": 4.347636646566816,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003238366018572838,
      "loss": 4.8108,
      "mean_token_accuracy": 0.22650467604398727,
      "num_tokens": 96918922.0,
      "step": 55880
    },
    {
      "entropy": 5.826118135452271,
      "epoch": 4.348025675938533,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003238095466349407,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22437690943479538,
      "num_tokens": 96927846.0,
      "step": 55885
    },
    {
      "entropy": 5.913286352157593,
      "epoch": 4.348414705310251,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003237824906719905,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2076628565788269,
      "num_tokens": 96936701.0,
      "step": 55890
    },
    {
      "entropy": 5.9072033882141115,
      "epoch": 4.348803734681969,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00032375543396884367,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2198356032371521,
      "num_tokens": 96944519.0,
      "step": 55895
    },
    {
      "entropy": 5.917604446411133,
      "epoch": 4.3491927640536865,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003237283765259107,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21468027234077453,
      "num_tokens": 96953142.0,
      "step": 55900
    },
    {
      "entropy": 5.836408567428589,
      "epoch": 4.349581793425403,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032370131834360215,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22414413541555406,
      "num_tokens": 96961373.0,
      "step": 55905
    },
    {
      "entropy": 5.904293441772461,
      "epoch": 4.349970822797121,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003236742594223287,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2133316859602928,
      "num_tokens": 96969578.0,
      "step": 55910
    },
    {
      "entropy": 5.8716898441314695,
      "epoch": 4.350359852168839,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003236471997625008,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21582414954900742,
      "num_tokens": 96977990.0,
      "step": 55915
    },
    {
      "entropy": 5.866220140457154,
      "epoch": 4.350748881540556,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003236201393645291,
      "loss": 4.979,
      "mean_token_accuracy": 0.21179738491773606,
      "num_tokens": 96986190.0,
      "step": 55920
    },
    {
      "entropy": 5.93839921951294,
      "epoch": 4.351137910912274,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032359307822882403,
      "loss": 5.0161,
      "mean_token_accuracy": 0.2060386836528778,
      "num_tokens": 96995315.0,
      "step": 55925
    },
    {
      "entropy": 5.8783282279968265,
      "epoch": 4.351526940283992,
      "grad_norm": 1.375,
      "learning_rate": 0.00032356601635579645,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21778611689805985,
      "num_tokens": 97004439.0,
      "step": 55930
    },
    {
      "entropy": 5.915272045135498,
      "epoch": 4.351915969655709,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003235389537458569,
      "loss": 4.9308,
      "mean_token_accuracy": 0.20804882049560547,
      "num_tokens": 97013075.0,
      "step": 55935
    },
    {
      "entropy": 5.891468811035156,
      "epoch": 4.352304999027426,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032351189039941594,
      "loss": 5.1259,
      "mean_token_accuracy": 0.20207825899124146,
      "num_tokens": 97022193.0,
      "step": 55940
    },
    {
      "entropy": 5.83633975982666,
      "epoch": 4.352694028399144,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032348482631688424,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22220561504364014,
      "num_tokens": 97030869.0,
      "step": 55945
    },
    {
      "entropy": 5.936500978469849,
      "epoch": 4.353083057770862,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003234577614986725,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21085729002952575,
      "num_tokens": 97040182.0,
      "step": 55950
    },
    {
      "entropy": 5.949840784072876,
      "epoch": 4.353472087142579,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003234306959451914,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20338288992643355,
      "num_tokens": 97049141.0,
      "step": 55955
    },
    {
      "entropy": 5.924427223205567,
      "epoch": 4.353861116514297,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032340362965685145,
      "loss": 4.9714,
      "mean_token_accuracy": 0.20897796601057053,
      "num_tokens": 97058007.0,
      "step": 55960
    },
    {
      "entropy": 5.907434177398682,
      "epoch": 4.354250145886015,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003233765626340636,
      "loss": 4.9426,
      "mean_token_accuracy": 0.21701187193393706,
      "num_tokens": 97066548.0,
      "step": 55965
    },
    {
      "entropy": 5.945518493652344,
      "epoch": 4.354639175257732,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003233494948772384,
      "loss": 5.022,
      "mean_token_accuracy": 0.21225260347127914,
      "num_tokens": 97075639.0,
      "step": 55970
    },
    {
      "entropy": 5.813122272491455,
      "epoch": 4.355028204629449,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032332242638678647,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22090377062559127,
      "num_tokens": 97084351.0,
      "step": 55975
    },
    {
      "entropy": 5.820841979980469,
      "epoch": 4.355417234001167,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003232953571631187,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22026506066322327,
      "num_tokens": 97093117.0,
      "step": 55980
    },
    {
      "entropy": 5.82263503074646,
      "epoch": 4.355806263372885,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003232682872066457,
      "loss": 4.8704,
      "mean_token_accuracy": 0.21631211340427398,
      "num_tokens": 97101535.0,
      "step": 55985
    },
    {
      "entropy": 5.8013170719146725,
      "epoch": 4.356195292744602,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003232412165177783,
      "loss": 4.913,
      "mean_token_accuracy": 0.21891602873802185,
      "num_tokens": 97109999.0,
      "step": 55990
    },
    {
      "entropy": 5.8168340682983395,
      "epoch": 4.35658432211632,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032321414509692726,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21818098276853562,
      "num_tokens": 97119535.0,
      "step": 55995
    },
    {
      "entropy": 5.9090711116790775,
      "epoch": 4.356973351488037,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032318707294450326,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21734316498041154,
      "num_tokens": 97128357.0,
      "step": 56000
    },
    {
      "entropy": 5.853480482101441,
      "epoch": 4.357362380859755,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003231600000609171,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2131370037794113,
      "num_tokens": 97137317.0,
      "step": 56005
    },
    {
      "entropy": 5.905597352981568,
      "epoch": 4.357751410231472,
      "grad_norm": 1.375,
      "learning_rate": 0.00032313292644657967,
      "loss": 5.001,
      "mean_token_accuracy": 0.21472050249576569,
      "num_tokens": 97146169.0,
      "step": 56010
    },
    {
      "entropy": 5.877661991119385,
      "epoch": 4.35814043960319,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003231058521019017,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2174597918987274,
      "num_tokens": 97155287.0,
      "step": 56015
    },
    {
      "entropy": 5.9289172172546385,
      "epoch": 4.358529468974908,
      "grad_norm": 1.375,
      "learning_rate": 0.0003230787770272939,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21362903714179993,
      "num_tokens": 97163578.0,
      "step": 56020
    },
    {
      "entropy": 5.846210098266601,
      "epoch": 4.358918498346625,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003230517012231673,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21311804056167602,
      "num_tokens": 97172167.0,
      "step": 56025
    },
    {
      "entropy": 5.756001234054565,
      "epoch": 4.359307527718343,
      "grad_norm": 1.53125,
      "learning_rate": 0.00032302462468993253,
      "loss": 4.795,
      "mean_token_accuracy": 0.22805336117744446,
      "num_tokens": 97181464.0,
      "step": 56030
    },
    {
      "entropy": 5.841710710525513,
      "epoch": 4.359696557090061,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003229975474280006,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21403897553682327,
      "num_tokens": 97190820.0,
      "step": 56035
    },
    {
      "entropy": 5.9227015495300295,
      "epoch": 4.360085586461778,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032297046943778216,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20482743233442308,
      "num_tokens": 97200083.0,
      "step": 56040
    },
    {
      "entropy": 5.854891681671143,
      "epoch": 4.360474615833495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032294339071968833,
      "loss": 4.8558,
      "mean_token_accuracy": 0.21929155588150023,
      "num_tokens": 97208703.0,
      "step": 56045
    },
    {
      "entropy": 5.851736307144165,
      "epoch": 4.360863645205213,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032291631127412976,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21540368795394899,
      "num_tokens": 97217887.0,
      "step": 56050
    },
    {
      "entropy": 5.890554714202881,
      "epoch": 4.361252674576931,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003228892311015174,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2004201427102089,
      "num_tokens": 97227858.0,
      "step": 56055
    },
    {
      "entropy": 5.8876978874206545,
      "epoch": 4.361641703948648,
      "grad_norm": 1.484375,
      "learning_rate": 0.00032286215020226224,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21921341568231584,
      "num_tokens": 97235772.0,
      "step": 56060
    },
    {
      "entropy": 5.905483961105347,
      "epoch": 4.362030733320366,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032283506857677517,
      "loss": 4.926,
      "mean_token_accuracy": 0.21788364052772521,
      "num_tokens": 97245068.0,
      "step": 56065
    },
    {
      "entropy": 5.857646560668945,
      "epoch": 4.362419762692083,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003228079862254671,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2192860350012779,
      "num_tokens": 97254595.0,
      "step": 56070
    },
    {
      "entropy": 5.799866962432861,
      "epoch": 4.362808792063801,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032278090314874887,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21131557375192642,
      "num_tokens": 97263133.0,
      "step": 56075
    },
    {
      "entropy": 5.8625412940979,
      "epoch": 4.363197821435518,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00032275381934703147,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21828479319810867,
      "num_tokens": 97271429.0,
      "step": 56080
    },
    {
      "entropy": 5.85198860168457,
      "epoch": 4.363586850807236,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003227267348207258,
      "loss": 4.8769,
      "mean_token_accuracy": 0.21717992424964905,
      "num_tokens": 97280018.0,
      "step": 56085
    },
    {
      "entropy": 5.873661613464355,
      "epoch": 4.363975880178954,
      "grad_norm": 1.375,
      "learning_rate": 0.00032269964957024305,
      "loss": 4.9305,
      "mean_token_accuracy": 0.21584324091672896,
      "num_tokens": 97288552.0,
      "step": 56090
    },
    {
      "entropy": 5.8466979503631595,
      "epoch": 4.364364909550671,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032267256359599396,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21991046518087387,
      "num_tokens": 97297684.0,
      "step": 56095
    },
    {
      "entropy": 5.8576733589172365,
      "epoch": 4.364753938922389,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032264547689838963,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21063765734434128,
      "num_tokens": 97306312.0,
      "step": 56100
    },
    {
      "entropy": 5.974688577651977,
      "epoch": 4.365142968294106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032261838947784103,
      "loss": 5.086,
      "mean_token_accuracy": 0.1932907819747925,
      "num_tokens": 97315566.0,
      "step": 56105
    },
    {
      "entropy": 5.82246618270874,
      "epoch": 4.365531997665824,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003225913013347591,
      "loss": 4.8839,
      "mean_token_accuracy": 0.21617152243852616,
      "num_tokens": 97324359.0,
      "step": 56110
    },
    {
      "entropy": 5.89462513923645,
      "epoch": 4.365921027037541,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032256421246955504,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21327568888664244,
      "num_tokens": 97333377.0,
      "step": 56115
    },
    {
      "entropy": 5.8266651153564455,
      "epoch": 4.366310056409259,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003225371228826396,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22158707082271575,
      "num_tokens": 97342468.0,
      "step": 56120
    },
    {
      "entropy": 5.827818870544434,
      "epoch": 4.366699085780977,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003225100325744241,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22071681916713715,
      "num_tokens": 97350667.0,
      "step": 56125
    },
    {
      "entropy": 5.831667375564575,
      "epoch": 4.367088115152694,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003224829415453194,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2136904314160347,
      "num_tokens": 97359652.0,
      "step": 56130
    },
    {
      "entropy": 5.885065650939941,
      "epoch": 4.367477144524411,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003224558497957367,
      "loss": 4.9877,
      "mean_token_accuracy": 0.20652298182249068,
      "num_tokens": 97368891.0,
      "step": 56135
    },
    {
      "entropy": 5.853997993469238,
      "epoch": 4.367866173896129,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032242875732608697,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21902762055397035,
      "num_tokens": 97377597.0,
      "step": 56140
    },
    {
      "entropy": 5.827123689651489,
      "epoch": 4.3682552032678466,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003224016641367814,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21403760761022567,
      "num_tokens": 97386465.0,
      "step": 56145
    },
    {
      "entropy": 5.860655355453491,
      "epoch": 4.368644232639564,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00032237457022823095,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22551157474517822,
      "num_tokens": 97394780.0,
      "step": 56150
    },
    {
      "entropy": 5.845815134048462,
      "epoch": 4.369033262011282,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032234747560084675,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21327877044677734,
      "num_tokens": 97404276.0,
      "step": 56155
    },
    {
      "entropy": 5.85503830909729,
      "epoch": 4.369422291383,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032232038025504015,
      "loss": 4.933,
      "mean_token_accuracy": 0.22126472294330596,
      "num_tokens": 97412543.0,
      "step": 56160
    },
    {
      "entropy": 5.898508834838867,
      "epoch": 4.369811320754717,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032229328419122196,
      "loss": 5.048,
      "mean_token_accuracy": 0.20882613360881805,
      "num_tokens": 97421886.0,
      "step": 56165
    },
    {
      "entropy": 5.81612639427185,
      "epoch": 4.370200350126434,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003222661874098035,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22832584232091904,
      "num_tokens": 97429910.0,
      "step": 56170
    },
    {
      "entropy": 5.797959804534912,
      "epoch": 4.370589379498152,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003222390899111959,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2189633756875992,
      "num_tokens": 97438658.0,
      "step": 56175
    },
    {
      "entropy": 5.84496374130249,
      "epoch": 4.3709784088698695,
      "grad_norm": 1.5,
      "learning_rate": 0.0003222119916958103,
      "loss": 4.956,
      "mean_token_accuracy": 0.21665065437555314,
      "num_tokens": 97446968.0,
      "step": 56180
    },
    {
      "entropy": 5.876144599914551,
      "epoch": 4.371367438241587,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032218489276405777,
      "loss": 4.9688,
      "mean_token_accuracy": 0.20748453438282013,
      "num_tokens": 97455849.0,
      "step": 56185
    },
    {
      "entropy": 5.908208656311035,
      "epoch": 4.371756467613305,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032215779311634967,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2177588775753975,
      "num_tokens": 97464826.0,
      "step": 56190
    },
    {
      "entropy": 5.881191682815552,
      "epoch": 4.372145496985023,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032213069275309714,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21201578080654143,
      "num_tokens": 97473276.0,
      "step": 56195
    },
    {
      "entropy": 5.910405015945434,
      "epoch": 4.37253452635674,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032210359167471136,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2102323964238167,
      "num_tokens": 97482153.0,
      "step": 56200
    },
    {
      "entropy": 5.86022162437439,
      "epoch": 4.372923555728457,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003220764898816035,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21128607541322708,
      "num_tokens": 97490539.0,
      "step": 56205
    },
    {
      "entropy": 5.8533628463745115,
      "epoch": 4.373312585100175,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00032204938737418495,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22005994766950607,
      "num_tokens": 97500700.0,
      "step": 56210
    },
    {
      "entropy": 5.831777334213257,
      "epoch": 4.3737016144718925,
      "grad_norm": 1.375,
      "learning_rate": 0.00032202228415286664,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21296179592609404,
      "num_tokens": 97509562.0,
      "step": 56215
    },
    {
      "entropy": 5.905174303054809,
      "epoch": 4.37409064384361,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003219951802180602,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20478034466505052,
      "num_tokens": 97518349.0,
      "step": 56220
    },
    {
      "entropy": 5.889245843887329,
      "epoch": 4.374479673215328,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003219680755701766,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21924894750118257,
      "num_tokens": 97526731.0,
      "step": 56225
    },
    {
      "entropy": 5.891793298721313,
      "epoch": 4.374868702587046,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032194097020962726,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2152174636721611,
      "num_tokens": 97535392.0,
      "step": 56230
    },
    {
      "entropy": 5.844496631622315,
      "epoch": 4.375257731958763,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003219138641368234,
      "loss": 4.9642,
      "mean_token_accuracy": 0.21058496981859207,
      "num_tokens": 97544196.0,
      "step": 56235
    },
    {
      "entropy": 5.839735603332519,
      "epoch": 4.37564676133048,
      "grad_norm": 1.515625,
      "learning_rate": 0.00032188675735217627,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22158660590648652,
      "num_tokens": 97551944.0,
      "step": 56240
    },
    {
      "entropy": 5.707260942459106,
      "epoch": 4.376035790702198,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003218596498560973,
      "loss": 4.8369,
      "mean_token_accuracy": 0.22087912559509276,
      "num_tokens": 97560508.0,
      "step": 56245
    },
    {
      "entropy": 5.800538778305054,
      "epoch": 4.3764248200739155,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003218325416489976,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22747107297182084,
      "num_tokens": 97568602.0,
      "step": 56250
    },
    {
      "entropy": 5.854131507873535,
      "epoch": 4.376813849445633,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003218054327312888,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22329197376966475,
      "num_tokens": 97576886.0,
      "step": 56255
    },
    {
      "entropy": 5.911349487304688,
      "epoch": 4.377202878817351,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032177832310338193,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21256973445415497,
      "num_tokens": 97586213.0,
      "step": 56260
    },
    {
      "entropy": 5.789101219177246,
      "epoch": 4.377591908189069,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032175121276568853,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22243961989879607,
      "num_tokens": 97595135.0,
      "step": 56265
    },
    {
      "entropy": 5.847576236724853,
      "epoch": 4.377980937560785,
      "grad_norm": 1.546875,
      "learning_rate": 0.00032172410171861987,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2027269646525383,
      "num_tokens": 97603048.0,
      "step": 56270
    },
    {
      "entropy": 5.841759777069091,
      "epoch": 4.378369966932503,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032169698996258733,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2188049167394638,
      "num_tokens": 97612090.0,
      "step": 56275
    },
    {
      "entropy": 5.892730665206909,
      "epoch": 4.378758996304221,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003216698774980023,
      "loss": 4.894,
      "mean_token_accuracy": 0.21872354447841644,
      "num_tokens": 97621035.0,
      "step": 56280
    },
    {
      "entropy": 5.84887580871582,
      "epoch": 4.3791480256759385,
      "grad_norm": 1.5625,
      "learning_rate": 0.00032164276432527605,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21724647879600525,
      "num_tokens": 97630066.0,
      "step": 56285
    },
    {
      "entropy": 5.862901878356934,
      "epoch": 4.379537055047656,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003216156504448202,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22048344165086747,
      "num_tokens": 97638797.0,
      "step": 56290
    },
    {
      "entropy": 5.931183385848999,
      "epoch": 4.379926084419374,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003215885358570461,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20555678606033326,
      "num_tokens": 97648429.0,
      "step": 56295
    },
    {
      "entropy": 5.844251728057861,
      "epoch": 4.380315113791092,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003215614205623651,
      "loss": 4.8726,
      "mean_token_accuracy": 0.21135499328374863,
      "num_tokens": 97656954.0,
      "step": 56300
    },
    {
      "entropy": 5.844804430007935,
      "epoch": 4.380704143162808,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003215343045611886,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22130008041858673,
      "num_tokens": 97665564.0,
      "step": 56305
    },
    {
      "entropy": 5.849834012985229,
      "epoch": 4.381093172534526,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003215071878539281,
      "loss": 4.956,
      "mean_token_accuracy": 0.21724940687417985,
      "num_tokens": 97674921.0,
      "step": 56310
    },
    {
      "entropy": 5.819569492340088,
      "epoch": 4.381482201906244,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032148007044099507,
      "loss": 4.907,
      "mean_token_accuracy": 0.21422570049762726,
      "num_tokens": 97683564.0,
      "step": 56315
    },
    {
      "entropy": 5.865578985214233,
      "epoch": 4.3818712312779615,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003214529523228009,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21091562360525132,
      "num_tokens": 97693093.0,
      "step": 56320
    },
    {
      "entropy": 5.911924982070923,
      "epoch": 4.382260260649679,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003214258334997572,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2039350226521492,
      "num_tokens": 97701392.0,
      "step": 56325
    },
    {
      "entropy": 5.958255004882813,
      "epoch": 4.382649290021397,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003213987139722753,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21343532353639602,
      "num_tokens": 97710705.0,
      "step": 56330
    },
    {
      "entropy": 5.893885087966919,
      "epoch": 4.383038319393114,
      "grad_norm": 1.46875,
      "learning_rate": 0.00032137159374076693,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2197265714406967,
      "num_tokens": 97719169.0,
      "step": 56335
    },
    {
      "entropy": 5.918248796463013,
      "epoch": 4.383427348764831,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003213444728056432,
      "loss": 5.047,
      "mean_token_accuracy": 0.20918984115123748,
      "num_tokens": 97728444.0,
      "step": 56340
    },
    {
      "entropy": 5.849329137802124,
      "epoch": 4.383816378136549,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00032131735116731604,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21768296658992767,
      "num_tokens": 97736892.0,
      "step": 56345
    },
    {
      "entropy": 5.939564943313599,
      "epoch": 4.384205407508267,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003212902288261966,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21214178055524827,
      "num_tokens": 97746297.0,
      "step": 56350
    },
    {
      "entropy": 5.824271249771118,
      "epoch": 4.3845944368799845,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003212631057826968,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22800871282815932,
      "num_tokens": 97754599.0,
      "step": 56355
    },
    {
      "entropy": 5.909377765655518,
      "epoch": 4.384983466251702,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032123598203722786,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21736262887716293,
      "num_tokens": 97763256.0,
      "step": 56360
    },
    {
      "entropy": 5.883335733413697,
      "epoch": 4.38537249562342,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003212088575902016,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21353209316730498,
      "num_tokens": 97772492.0,
      "step": 56365
    },
    {
      "entropy": 5.832557678222656,
      "epoch": 4.385761524995137,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032118173244202947,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23123684376478196,
      "num_tokens": 97781112.0,
      "step": 56370
    },
    {
      "entropy": 5.87379641532898,
      "epoch": 4.386150554366854,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032115460659312304,
      "loss": 5.0187,
      "mean_token_accuracy": 0.2102784365415573,
      "num_tokens": 97790175.0,
      "step": 56375
    },
    {
      "entropy": 5.843015861511231,
      "epoch": 4.386539583738572,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003211274800438939,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21313754618167877,
      "num_tokens": 97799789.0,
      "step": 56380
    },
    {
      "entropy": 5.891910076141357,
      "epoch": 4.38692861311029,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003211003527947537,
      "loss": 4.8514,
      "mean_token_accuracy": 0.22333042919635773,
      "num_tokens": 97807999.0,
      "step": 56385
    },
    {
      "entropy": 5.777862024307251,
      "epoch": 4.3873176424820075,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032107322484611396,
      "loss": 4.7535,
      "mean_token_accuracy": 0.23310523927211763,
      "num_tokens": 97817133.0,
      "step": 56390
    },
    {
      "entropy": 5.952195167541504,
      "epoch": 4.387706671853725,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032104609619838635,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21147679686546325,
      "num_tokens": 97825427.0,
      "step": 56395
    },
    {
      "entropy": 5.886843824386597,
      "epoch": 4.388095701225443,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003210189668519826,
      "loss": 4.954,
      "mean_token_accuracy": 0.21574897915124894,
      "num_tokens": 97834247.0,
      "step": 56400
    },
    {
      "entropy": 5.880003309249878,
      "epoch": 4.38848473059716,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003209918368073143,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21588183790445328,
      "num_tokens": 97842282.0,
      "step": 56405
    },
    {
      "entropy": 5.832932567596435,
      "epoch": 4.388873759968877,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000320964706064793,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2160721391439438,
      "num_tokens": 97851554.0,
      "step": 56410
    },
    {
      "entropy": 5.851925325393677,
      "epoch": 4.389262789340595,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00032093757462483055,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2203413039445877,
      "num_tokens": 97859941.0,
      "step": 56415
    },
    {
      "entropy": 5.823663234710693,
      "epoch": 4.389651818712313,
      "grad_norm": 1.34375,
      "learning_rate": 0.00032091044248783835,
      "loss": 4.8669,
      "mean_token_accuracy": 0.21934882551431656,
      "num_tokens": 97868265.0,
      "step": 56420
    },
    {
      "entropy": 5.816927766799926,
      "epoch": 4.3900408480840305,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032088330965422836,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21490378677845,
      "num_tokens": 97877750.0,
      "step": 56425
    },
    {
      "entropy": 5.9270195960998535,
      "epoch": 4.390429877455748,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00032085617612441215,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2037125915288925,
      "num_tokens": 97886086.0,
      "step": 56430
    },
    {
      "entropy": 5.8997749328613285,
      "epoch": 4.390818906827466,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003208290418988015,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20419949144124985,
      "num_tokens": 97894761.0,
      "step": 56435
    },
    {
      "entropy": 5.839312744140625,
      "epoch": 4.391207936199183,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000320801906977808,
      "loss": 4.952,
      "mean_token_accuracy": 0.21340329796075821,
      "num_tokens": 97903796.0,
      "step": 56440
    },
    {
      "entropy": 5.789989185333252,
      "epoch": 4.3915969655709,
      "grad_norm": 1.453125,
      "learning_rate": 0.00032077477136184356,
      "loss": 4.9166,
      "mean_token_accuracy": 0.220957413315773,
      "num_tokens": 97912641.0,
      "step": 56445
    },
    {
      "entropy": 5.8472254276275635,
      "epoch": 4.391985994942618,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003207476350513198,
      "loss": 4.921,
      "mean_token_accuracy": 0.21988424509763718,
      "num_tokens": 97921745.0,
      "step": 56450
    },
    {
      "entropy": 5.851008605957031,
      "epoch": 4.392375024314336,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003207204980466485,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2166115790605545,
      "num_tokens": 97930988.0,
      "step": 56455
    },
    {
      "entropy": 5.912194442749024,
      "epoch": 4.3927640536860535,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00032069336034824136,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21005015224218368,
      "num_tokens": 97939907.0,
      "step": 56460
    },
    {
      "entropy": 5.827771139144898,
      "epoch": 4.393153083057771,
      "grad_norm": 1.375,
      "learning_rate": 0.00032066622195651024,
      "loss": 4.8628,
      "mean_token_accuracy": 0.22490894943475723,
      "num_tokens": 97948568.0,
      "step": 56465
    },
    {
      "entropy": 5.788376522064209,
      "epoch": 4.393542112429488,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032063908287186687,
      "loss": 4.9028,
      "mean_token_accuracy": 0.21487530916929246,
      "num_tokens": 97956564.0,
      "step": 56470
    },
    {
      "entropy": 5.897263622283935,
      "epoch": 4.393931141801206,
      "grad_norm": 1.515625,
      "learning_rate": 0.00032061194309472314,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20413736999034882,
      "num_tokens": 97965094.0,
      "step": 56475
    },
    {
      "entropy": 5.787479877471924,
      "epoch": 4.394320171172923,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032058480262549065,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2195443630218506,
      "num_tokens": 97973938.0,
      "step": 56480
    },
    {
      "entropy": 5.876153993606567,
      "epoch": 4.394709200544641,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003205576614645814,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21140616685152053,
      "num_tokens": 97982531.0,
      "step": 56485
    },
    {
      "entropy": 5.86994891166687,
      "epoch": 4.395098229916359,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00032053051961240716,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21883551478385926,
      "num_tokens": 97992147.0,
      "step": 56490
    },
    {
      "entropy": 5.899713516235352,
      "epoch": 4.3954872592880765,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003205033770693797,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21747009009122847,
      "num_tokens": 98000738.0,
      "step": 56495
    },
    {
      "entropy": 5.829110145568848,
      "epoch": 4.395876288659794,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000320476233835911,
      "loss": 4.8581,
      "mean_token_accuracy": 0.2207776740193367,
      "num_tokens": 98009158.0,
      "step": 56500
    },
    {
      "entropy": 5.834641456604004,
      "epoch": 4.396265318031511,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032044908991241274,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21917135864496232,
      "num_tokens": 98017795.0,
      "step": 56505
    },
    {
      "entropy": 5.832931852340698,
      "epoch": 4.396654347403229,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032042194529929694,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2251078560948372,
      "num_tokens": 98026390.0,
      "step": 56510
    },
    {
      "entropy": 5.883548212051392,
      "epoch": 4.397043376774946,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032039479999697545,
      "loss": 4.9656,
      "mean_token_accuracy": 0.21211181581020355,
      "num_tokens": 98035506.0,
      "step": 56515
    },
    {
      "entropy": 5.824854040145874,
      "epoch": 4.397432406146664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032036765400586005,
      "loss": 4.952,
      "mean_token_accuracy": 0.20762148052453994,
      "num_tokens": 98044228.0,
      "step": 56520
    },
    {
      "entropy": 5.897710466384888,
      "epoch": 4.397821435518382,
      "grad_norm": 1.3125,
      "learning_rate": 0.00032034050732636274,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21851904690265656,
      "num_tokens": 98053209.0,
      "step": 56525
    },
    {
      "entropy": 5.899632835388184,
      "epoch": 4.3982104648900995,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003203133599588954,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21904546320438384,
      "num_tokens": 98061892.0,
      "step": 56530
    },
    {
      "entropy": 5.919186353683472,
      "epoch": 4.398599494261817,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00032028621190386997,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21228469908237457,
      "num_tokens": 98071401.0,
      "step": 56535
    },
    {
      "entropy": 5.870324993133545,
      "epoch": 4.398988523633534,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032025906316169835,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22455044388771056,
      "num_tokens": 98080488.0,
      "step": 56540
    },
    {
      "entropy": 5.8134973526000975,
      "epoch": 4.399377553005252,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003202319137327924,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2194547787308693,
      "num_tokens": 98089290.0,
      "step": 56545
    },
    {
      "entropy": 5.8483567237854,
      "epoch": 4.399766582376969,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032020476361756424,
      "loss": 5.0029,
      "mean_token_accuracy": 0.20757050961256027,
      "num_tokens": 98097870.0,
      "step": 56550
    },
    {
      "entropy": 5.862218570709229,
      "epoch": 4.400155611748687,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032017761281642564,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2244698330760002,
      "num_tokens": 98107466.0,
      "step": 56555
    },
    {
      "entropy": 5.88328185081482,
      "epoch": 4.400544641120405,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003201504613297888,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21545010507106782,
      "num_tokens": 98115745.0,
      "step": 56560
    },
    {
      "entropy": 5.803183126449585,
      "epoch": 4.4009336704921225,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032012330915806553,
      "loss": 4.8604,
      "mean_token_accuracy": 0.21407920569181443,
      "num_tokens": 98124516.0,
      "step": 56565
    },
    {
      "entropy": 5.783963632583618,
      "epoch": 4.40132269986384,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032009615630166786,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2261909008026123,
      "num_tokens": 98132966.0,
      "step": 56570
    },
    {
      "entropy": 5.807172060012817,
      "epoch": 4.401711729235557,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003200690027610078,
      "loss": 4.8961,
      "mean_token_accuracy": 0.219764444231987,
      "num_tokens": 98141311.0,
      "step": 56575
    },
    {
      "entropy": 5.932420969009399,
      "epoch": 4.402100758607275,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003200418485364974,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21174367219209672,
      "num_tokens": 98150649.0,
      "step": 56580
    },
    {
      "entropy": 5.930697250366211,
      "epoch": 4.402489787978992,
      "grad_norm": 1.375,
      "learning_rate": 0.0003200146936285485,
      "loss": 5.0443,
      "mean_token_accuracy": 0.2116548776626587,
      "num_tokens": 98159282.0,
      "step": 56585
    },
    {
      "entropy": 5.894499158859253,
      "epoch": 4.40287881735071,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003199875380375734,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22053225040435792,
      "num_tokens": 98167752.0,
      "step": 56590
    },
    {
      "entropy": 5.903307867050171,
      "epoch": 4.403267846722428,
      "grad_norm": 1.5,
      "learning_rate": 0.000319960381763984,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21734940260648727,
      "num_tokens": 98176219.0,
      "step": 56595
    },
    {
      "entropy": 5.829818964004517,
      "epoch": 4.4036568760941455,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003199332248081923,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21353796124458313,
      "num_tokens": 98185210.0,
      "step": 56600
    },
    {
      "entropy": 5.839684915542603,
      "epoch": 4.404045905465862,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031990606717061055,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22315371930599212,
      "num_tokens": 98193512.0,
      "step": 56605
    },
    {
      "entropy": 5.8309722423553465,
      "epoch": 4.40443493483758,
      "grad_norm": 1.5,
      "learning_rate": 0.0003198789088516506,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21393875926733016,
      "num_tokens": 98201714.0,
      "step": 56610
    },
    {
      "entropy": 5.965172147750854,
      "epoch": 4.404823964209298,
      "grad_norm": 1.5625,
      "learning_rate": 0.00031985174985172467,
      "loss": 5.1237,
      "mean_token_accuracy": 0.2046158269047737,
      "num_tokens": 98210464.0,
      "step": 56615
    },
    {
      "entropy": 5.8819815158844,
      "epoch": 4.405212993581015,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003198245901712449,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21772871166467667,
      "num_tokens": 98219142.0,
      "step": 56620
    },
    {
      "entropy": 5.859474945068359,
      "epoch": 4.405602022952733,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003197974298106232,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21113801002502441,
      "num_tokens": 98228424.0,
      "step": 56625
    },
    {
      "entropy": 5.828237295150757,
      "epoch": 4.405991052324451,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031977026877027186,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21250246912240983,
      "num_tokens": 98236919.0,
      "step": 56630
    },
    {
      "entropy": 5.849480485916137,
      "epoch": 4.4063800816961685,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031974310705060297,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21812291145324708,
      "num_tokens": 98245449.0,
      "step": 56635
    },
    {
      "entropy": 5.910617446899414,
      "epoch": 4.406769111067885,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003197159446520286,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21278473138809204,
      "num_tokens": 98254425.0,
      "step": 56640
    },
    {
      "entropy": 5.856750535964966,
      "epoch": 4.407158140439603,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031968878157496105,
      "loss": 4.9424,
      "mean_token_accuracy": 0.21639953553676605,
      "num_tokens": 98262273.0,
      "step": 56645
    },
    {
      "entropy": 5.82481541633606,
      "epoch": 4.407547169811321,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031966161781981224,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21603998839855193,
      "num_tokens": 98271310.0,
      "step": 56650
    },
    {
      "entropy": 5.853538846969604,
      "epoch": 4.407936199183038,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031963445338699446,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2187110036611557,
      "num_tokens": 98280066.0,
      "step": 56655
    },
    {
      "entropy": 5.891941928863526,
      "epoch": 4.408325228554756,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031960728827692,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21383574604988098,
      "num_tokens": 98288067.0,
      "step": 56660
    },
    {
      "entropy": 5.821548318862915,
      "epoch": 4.408714257926474,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003195801224900009,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22011714726686477,
      "num_tokens": 98296872.0,
      "step": 56665
    },
    {
      "entropy": 5.795499610900879,
      "epoch": 4.409103287298191,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031955295602664936,
      "loss": 4.898,
      "mean_token_accuracy": 0.2184193029999733,
      "num_tokens": 98306982.0,
      "step": 56670
    },
    {
      "entropy": 5.894588756561279,
      "epoch": 4.409492316669908,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031952578888727763,
      "loss": 4.9658,
      "mean_token_accuracy": 0.21749274879693986,
      "num_tokens": 98315162.0,
      "step": 56675
    },
    {
      "entropy": 5.887330961227417,
      "epoch": 4.409881346041626,
      "grad_norm": 1.484375,
      "learning_rate": 0.00031949862107229796,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2089128702878952,
      "num_tokens": 98324127.0,
      "step": 56680
    },
    {
      "entropy": 5.908437919616699,
      "epoch": 4.410270375413344,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003194714525821225,
      "loss": 4.909,
      "mean_token_accuracy": 0.2215150073170662,
      "num_tokens": 98332288.0,
      "step": 56685
    },
    {
      "entropy": 5.883612632751465,
      "epoch": 4.410659404785061,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003194442834171636,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21089410781860352,
      "num_tokens": 98340840.0,
      "step": 56690
    },
    {
      "entropy": 5.882678604125976,
      "epoch": 4.411048434156779,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003194171135778334,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2185919240117073,
      "num_tokens": 98349571.0,
      "step": 56695
    },
    {
      "entropy": 5.825513076782227,
      "epoch": 4.411437463528497,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003193899430645442,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2139182761311531,
      "num_tokens": 98358397.0,
      "step": 56700
    },
    {
      "entropy": 5.891409492492675,
      "epoch": 4.411826492900214,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031936277187770824,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2214018478989601,
      "num_tokens": 98366841.0,
      "step": 56705
    },
    {
      "entropy": 5.8951863765716555,
      "epoch": 4.412215522271931,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003193356000177378,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2132929176092148,
      "num_tokens": 98376704.0,
      "step": 56710
    },
    {
      "entropy": 5.846710062026977,
      "epoch": 4.412604551643649,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003193084274850453,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22187179177999497,
      "num_tokens": 98385998.0,
      "step": 56715
    },
    {
      "entropy": 5.856969881057739,
      "epoch": 4.412993581015367,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031928125428004286,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21258797496557236,
      "num_tokens": 98394646.0,
      "step": 56720
    },
    {
      "entropy": 5.8773726463317875,
      "epoch": 4.413382610387084,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003192540804031429,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20959641486406327,
      "num_tokens": 98403245.0,
      "step": 56725
    },
    {
      "entropy": 5.803283548355102,
      "epoch": 4.413771639758802,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003192269058547577,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21666603684425353,
      "num_tokens": 98411258.0,
      "step": 56730
    },
    {
      "entropy": 5.870910120010376,
      "epoch": 4.41416066913052,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003191997306352996,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21379794776439667,
      "num_tokens": 98419792.0,
      "step": 56735
    },
    {
      "entropy": 5.847129487991333,
      "epoch": 4.4145496985022366,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000319172554745181,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21521052718162537,
      "num_tokens": 98428255.0,
      "step": 56740
    },
    {
      "entropy": 5.838669538497925,
      "epoch": 4.414938727873954,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00031914537818481404,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22185811549425125,
      "num_tokens": 98436790.0,
      "step": 56745
    },
    {
      "entropy": 5.819364261627197,
      "epoch": 4.415327757245672,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003191182009546113,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21401439905166625,
      "num_tokens": 98445340.0,
      "step": 56750
    },
    {
      "entropy": 5.792689466476441,
      "epoch": 4.41571678661739,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031909102305498514,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21458850651979447,
      "num_tokens": 98453757.0,
      "step": 56755
    },
    {
      "entropy": 5.85263671875,
      "epoch": 4.416105815989107,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031906384448634784,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21871923804283142,
      "num_tokens": 98462901.0,
      "step": 56760
    },
    {
      "entropy": 5.82844910621643,
      "epoch": 4.416494845360825,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031903666524911185,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21574854999780654,
      "num_tokens": 98470819.0,
      "step": 56765
    },
    {
      "entropy": 5.887957286834717,
      "epoch": 4.416883874732543,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031900948534368954,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20411421358585358,
      "num_tokens": 98479793.0,
      "step": 56770
    },
    {
      "entropy": 5.862755250930786,
      "epoch": 4.4172729041042595,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031898230477049334,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22058097869157792,
      "num_tokens": 98487868.0,
      "step": 56775
    },
    {
      "entropy": 5.905448579788208,
      "epoch": 4.417661933475977,
      "grad_norm": 1.375,
      "learning_rate": 0.0003189551235299357,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21030450016260147,
      "num_tokens": 98496974.0,
      "step": 56780
    },
    {
      "entropy": 5.8414782047271725,
      "epoch": 4.418050962847695,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003189279416224289,
      "loss": 4.9039,
      "mean_token_accuracy": 0.21415306329727174,
      "num_tokens": 98505507.0,
      "step": 56785
    },
    {
      "entropy": 5.858120632171631,
      "epoch": 4.418439992219413,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031890075904838556,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20258676558732985,
      "num_tokens": 98513568.0,
      "step": 56790
    },
    {
      "entropy": 5.867345476150513,
      "epoch": 4.41882902159113,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031887357580821817,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21429697126150132,
      "num_tokens": 98523175.0,
      "step": 56795
    },
    {
      "entropy": 5.945668315887451,
      "epoch": 4.419218050962848,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031884639190233905,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20742809623479844,
      "num_tokens": 98532046.0,
      "step": 56800
    },
    {
      "entropy": 5.878271961212159,
      "epoch": 4.419607080334565,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003188192073311608,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21098416596651076,
      "num_tokens": 98539816.0,
      "step": 56805
    },
    {
      "entropy": 5.893303394317627,
      "epoch": 4.4199961097062825,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031879202209509564,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21036721765995026,
      "num_tokens": 98548364.0,
      "step": 56810
    },
    {
      "entropy": 5.812397146224976,
      "epoch": 4.420385139078,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031876483619455633,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2119004338979721,
      "num_tokens": 98557073.0,
      "step": 56815
    },
    {
      "entropy": 5.818601179122925,
      "epoch": 4.420774168449718,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003187376496299553,
      "loss": 4.9024,
      "mean_token_accuracy": 0.21506871283054352,
      "num_tokens": 98565573.0,
      "step": 56820
    },
    {
      "entropy": 5.821877574920654,
      "epoch": 4.421163197821436,
      "grad_norm": 1.515625,
      "learning_rate": 0.000318710462401705,
      "loss": 4.9404,
      "mean_token_accuracy": 0.22211057543754578,
      "num_tokens": 98573992.0,
      "step": 56825
    },
    {
      "entropy": 5.869712829589844,
      "epoch": 4.421552227193153,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031868327451021816,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21982790380716324,
      "num_tokens": 98582469.0,
      "step": 56830
    },
    {
      "entropy": 5.88722071647644,
      "epoch": 4.421941256564871,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031865608595590706,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21432303190231322,
      "num_tokens": 98590975.0,
      "step": 56835
    },
    {
      "entropy": 5.894414663314819,
      "epoch": 4.422330285936588,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031862889673918437,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21839074790477753,
      "num_tokens": 98600228.0,
      "step": 56840
    },
    {
      "entropy": 5.881117820739746,
      "epoch": 4.4227193153083055,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003186017068604626,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2199365735054016,
      "num_tokens": 98609272.0,
      "step": 56845
    },
    {
      "entropy": 5.911862707138061,
      "epoch": 4.423108344680023,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003185745163201543,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21343216001987458,
      "num_tokens": 98617435.0,
      "step": 56850
    },
    {
      "entropy": 5.916994953155518,
      "epoch": 4.423497374051741,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003185473251186721,
      "loss": 5.0998,
      "mean_token_accuracy": 0.2078026279807091,
      "num_tokens": 98626158.0,
      "step": 56855
    },
    {
      "entropy": 5.828356790542602,
      "epoch": 4.423886403423459,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003185201332564285,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22055112570524216,
      "num_tokens": 98634632.0,
      "step": 56860
    },
    {
      "entropy": 5.913850259780884,
      "epoch": 4.424275432795176,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003184929407338362,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20998370945453643,
      "num_tokens": 98644170.0,
      "step": 56865
    },
    {
      "entropy": 5.924437570571899,
      "epoch": 4.424664462166893,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031846574755130773,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21071734875440598,
      "num_tokens": 98652628.0,
      "step": 56870
    },
    {
      "entropy": 5.923933982849121,
      "epoch": 4.425053491538611,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031843855370925574,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2147386386990547,
      "num_tokens": 98661544.0,
      "step": 56875
    },
    {
      "entropy": 5.932667541503906,
      "epoch": 4.4254425209103285,
      "grad_norm": 1.515625,
      "learning_rate": 0.00031841135920809285,
      "loss": 5.01,
      "mean_token_accuracy": 0.20742432177066802,
      "num_tokens": 98670504.0,
      "step": 56880
    },
    {
      "entropy": 5.8265820980072025,
      "epoch": 4.425831550282046,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003183841640482316,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21314030438661574,
      "num_tokens": 98679001.0,
      "step": 56885
    },
    {
      "entropy": 5.8722243309021,
      "epoch": 4.426220579653764,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031835696823008476,
      "loss": 4.9881,
      "mean_token_accuracy": 0.21235791891813277,
      "num_tokens": 98687461.0,
      "step": 56890
    },
    {
      "entropy": 5.9026154518127445,
      "epoch": 4.426609609025482,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00031832977175406496,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21525291949510575,
      "num_tokens": 98696429.0,
      "step": 56895
    },
    {
      "entropy": 5.852021598815918,
      "epoch": 4.426998638397199,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003183025746205848,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21543249040842055,
      "num_tokens": 98704876.0,
      "step": 56900
    },
    {
      "entropy": 5.802906560897827,
      "epoch": 4.427387667768917,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003182753768300571,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22039939016103743,
      "num_tokens": 98714268.0,
      "step": 56905
    },
    {
      "entropy": 5.912241983413696,
      "epoch": 4.427776697140634,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003182481783828943,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21047123819589614,
      "num_tokens": 98722243.0,
      "step": 56910
    },
    {
      "entropy": 5.851761293411255,
      "epoch": 4.4281657265123515,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031822097927950934,
      "loss": 4.8429,
      "mean_token_accuracy": 0.21856610625982284,
      "num_tokens": 98731212.0,
      "step": 56915
    },
    {
      "entropy": 5.868413925170898,
      "epoch": 4.428554755884069,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031819377952031476,
      "loss": 4.9246,
      "mean_token_accuracy": 0.2150653198361397,
      "num_tokens": 98739982.0,
      "step": 56920
    },
    {
      "entropy": 5.906727743148804,
      "epoch": 4.428943785255787,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031816657910572327,
      "loss": 4.9667,
      "mean_token_accuracy": 0.20531667917966842,
      "num_tokens": 98748413.0,
      "step": 56925
    },
    {
      "entropy": 5.946642684936523,
      "epoch": 4.429332814627505,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031813937803614774,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21544319689273833,
      "num_tokens": 98757813.0,
      "step": 56930
    },
    {
      "entropy": 5.8511651992797855,
      "epoch": 4.429721843999222,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003181121763120008,
      "loss": 4.8891,
      "mean_token_accuracy": 0.21692366451025008,
      "num_tokens": 98766231.0,
      "step": 56935
    },
    {
      "entropy": 5.907589435577393,
      "epoch": 4.430110873370939,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003180849739336952,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2045784994959831,
      "num_tokens": 98774583.0,
      "step": 56940
    },
    {
      "entropy": 5.813856506347657,
      "epoch": 4.430499902742657,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031805777090164364,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22220973670482635,
      "num_tokens": 98782885.0,
      "step": 56945
    },
    {
      "entropy": 5.844394159317017,
      "epoch": 4.4308889321143745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000318030567216259,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21852237582206727,
      "num_tokens": 98792046.0,
      "step": 56950
    },
    {
      "entropy": 5.885601329803467,
      "epoch": 4.431277961486092,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031800336287795395,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21839630901813506,
      "num_tokens": 98800334.0,
      "step": 56955
    },
    {
      "entropy": 5.878577709197998,
      "epoch": 4.43166699085781,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031797615788714133,
      "loss": 4.91,
      "mean_token_accuracy": 0.20704471468925476,
      "num_tokens": 98808647.0,
      "step": 56960
    },
    {
      "entropy": 5.854475116729736,
      "epoch": 4.432056020229528,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031794895224423396,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21235888302326203,
      "num_tokens": 98818112.0,
      "step": 56965
    },
    {
      "entropy": 5.89456992149353,
      "epoch": 4.432445049601245,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003179217459496446,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21699755936861037,
      "num_tokens": 98826893.0,
      "step": 56970
    },
    {
      "entropy": 5.856910753250122,
      "epoch": 4.432834078972962,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000317894539003786,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22007119953632354,
      "num_tokens": 98834484.0,
      "step": 56975
    },
    {
      "entropy": 5.888616418838501,
      "epoch": 4.43322310834468,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000317867331407071,
      "loss": 5.0816,
      "mean_token_accuracy": 0.20982526540756224,
      "num_tokens": 98843661.0,
      "step": 56980
    },
    {
      "entropy": 5.896297788619995,
      "epoch": 4.4336121377163975,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003178401231599126,
      "loss": 4.9938,
      "mean_token_accuracy": 0.20927424728870392,
      "num_tokens": 98852215.0,
      "step": 56985
    },
    {
      "entropy": 5.926761817932129,
      "epoch": 4.434001167088115,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031781291426272347,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21763289123773574,
      "num_tokens": 98862057.0,
      "step": 56990
    },
    {
      "entropy": 5.867588329315185,
      "epoch": 4.434390196459833,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003177857047159165,
      "loss": 4.8279,
      "mean_token_accuracy": 0.22957581728696824,
      "num_tokens": 98870215.0,
      "step": 56995
    },
    {
      "entropy": 5.8454266548156735,
      "epoch": 4.434779225831551,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003177584945199046,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22669149041175843,
      "num_tokens": 98878176.0,
      "step": 57000
    },
    {
      "epoch": 4.434779225831551,
      "eval_entropy": 5.538972745019731,
      "eval_loss": 5.043375492095947,
      "eval_mean_token_accuracy": 0.2201015628365167,
      "eval_num_tokens": 98878176.0,
      "eval_runtime": 21.6483,
      "eval_samples_per_second": 1919.688,
      "eval_steps_per_second": 239.973,
      "step": 57000
    },
    {
      "entropy": 5.865025615692138,
      "epoch": 4.435168255203267,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003177312836751005,
      "loss": 5.013,
      "mean_token_accuracy": 0.20777997374534607,
      "num_tokens": 98886995.0,
      "step": 57005
    },
    {
      "entropy": 5.855606412887573,
      "epoch": 4.435557284574985,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031770407218191727,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2042378544807434,
      "num_tokens": 98895385.0,
      "step": 57010
    },
    {
      "entropy": 5.83737154006958,
      "epoch": 4.435946313946703,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003176768600407677,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22667513638734818,
      "num_tokens": 98903871.0,
      "step": 57015
    },
    {
      "entropy": 5.853383159637451,
      "epoch": 4.4363353433184205,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003176496472520647,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21446166634559632,
      "num_tokens": 98911989.0,
      "step": 57020
    },
    {
      "entropy": 5.859633207321167,
      "epoch": 4.436724372690138,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003176224338162212,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20999007672071457,
      "num_tokens": 98920618.0,
      "step": 57025
    },
    {
      "entropy": 5.802068614959717,
      "epoch": 4.437113402061856,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031759521973365006,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21882838159799575,
      "num_tokens": 98929179.0,
      "step": 57030
    },
    {
      "entropy": 5.9125776290893555,
      "epoch": 4.437502431433574,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031756800500476435,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22074341922998428,
      "num_tokens": 98937900.0,
      "step": 57035
    },
    {
      "entropy": 5.866281270980835,
      "epoch": 4.43789146080529,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031754078962997686,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21960435211658477,
      "num_tokens": 98947310.0,
      "step": 57040
    },
    {
      "entropy": 5.937146997451782,
      "epoch": 4.438280490177008,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003175135736097006,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21295321136713027,
      "num_tokens": 98956203.0,
      "step": 57045
    },
    {
      "entropy": 5.849953508377075,
      "epoch": 4.438669519548726,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031748635694434844,
      "loss": 4.8847,
      "mean_token_accuracy": 0.21855379194021224,
      "num_tokens": 98964690.0,
      "step": 57050
    },
    {
      "entropy": 5.866893243789673,
      "epoch": 4.4390585489204435,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003174591396343336,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2147304967045784,
      "num_tokens": 98974195.0,
      "step": 57055
    },
    {
      "entropy": 5.889646577835083,
      "epoch": 4.439447578292161,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003174319216800688,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21202975511550903,
      "num_tokens": 98982457.0,
      "step": 57060
    },
    {
      "entropy": 5.886439514160156,
      "epoch": 4.439836607663879,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031740470308196724,
      "loss": 4.927,
      "mean_token_accuracy": 0.21616315543651582,
      "num_tokens": 98991449.0,
      "step": 57065
    },
    {
      "entropy": 5.932275867462158,
      "epoch": 4.440225637035597,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003173774838404417,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2162795975804329,
      "num_tokens": 99000472.0,
      "step": 57070
    },
    {
      "entropy": 5.864894866943359,
      "epoch": 4.440614666407313,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031735026395590535,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22521155923604966,
      "num_tokens": 99008642.0,
      "step": 57075
    },
    {
      "entropy": 5.896076250076294,
      "epoch": 4.441003695779031,
      "grad_norm": 1.34375,
      "learning_rate": 0.000317323043428771,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21135891526937484,
      "num_tokens": 99017504.0,
      "step": 57080
    },
    {
      "entropy": 5.887834787368774,
      "epoch": 4.441392725150749,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003172958222594519,
      "loss": 4.9394,
      "mean_token_accuracy": 0.20695120990276336,
      "num_tokens": 99026132.0,
      "step": 57085
    },
    {
      "entropy": 5.903744506835937,
      "epoch": 4.4417817545224665,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003172686004483611,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2157866969704628,
      "num_tokens": 99034535.0,
      "step": 57090
    },
    {
      "entropy": 5.864254474639893,
      "epoch": 4.442170783894184,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031724137799591154,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2082432672381401,
      "num_tokens": 99044591.0,
      "step": 57095
    },
    {
      "entropy": 5.84340615272522,
      "epoch": 4.442559813265902,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031721415490251634,
      "loss": 4.944,
      "mean_token_accuracy": 0.2154535621404648,
      "num_tokens": 99052570.0,
      "step": 57100
    },
    {
      "entropy": 5.8699816226959225,
      "epoch": 4.44294884263762,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003171869311685885,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21432520896196366,
      "num_tokens": 99061533.0,
      "step": 57105
    },
    {
      "entropy": 5.94688458442688,
      "epoch": 4.443337872009336,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003171597067945411,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20566207766532899,
      "num_tokens": 99070214.0,
      "step": 57110
    },
    {
      "entropy": 5.942174959182739,
      "epoch": 4.443726901381054,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031713248178078717,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22151853740215302,
      "num_tokens": 99077968.0,
      "step": 57115
    },
    {
      "entropy": 5.850817346572876,
      "epoch": 4.444115930752772,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031710525612774,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21422065049409866,
      "num_tokens": 99087398.0,
      "step": 57120
    },
    {
      "entropy": 5.950133466720581,
      "epoch": 4.4445049601244895,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003170780298358126,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21400618851184844,
      "num_tokens": 99096057.0,
      "step": 57125
    },
    {
      "entropy": 5.897212123870849,
      "epoch": 4.444893989496207,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031705080290541797,
      "loss": 4.9623,
      "mean_token_accuracy": 0.2148256331682205,
      "num_tokens": 99105034.0,
      "step": 57130
    },
    {
      "entropy": 5.836870241165161,
      "epoch": 4.445283018867925,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031702357533696946,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21777815371751785,
      "num_tokens": 99113992.0,
      "step": 57135
    },
    {
      "entropy": 5.915919494628906,
      "epoch": 4.445672048239642,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031699634713088,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2166342630982399,
      "num_tokens": 99122380.0,
      "step": 57140
    },
    {
      "entropy": 5.845082092285156,
      "epoch": 4.446061077611359,
      "grad_norm": 1.375,
      "learning_rate": 0.0003169691182875628,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20962347984313964,
      "num_tokens": 99131759.0,
      "step": 57145
    },
    {
      "entropy": 5.856068515777588,
      "epoch": 4.446450106983077,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003169418888074311,
      "loss": 4.9054,
      "mean_token_accuracy": 0.21535201817750932,
      "num_tokens": 99140172.0,
      "step": 57150
    },
    {
      "entropy": 5.8785929679870605,
      "epoch": 4.446839136354795,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031691465869089795,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21584688425064086,
      "num_tokens": 99149071.0,
      "step": 57155
    },
    {
      "entropy": 5.889718437194825,
      "epoch": 4.4472281657265125,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003168874279383766,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22119763791561126,
      "num_tokens": 99158266.0,
      "step": 57160
    },
    {
      "entropy": 5.795144748687744,
      "epoch": 4.44761719509823,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003168601965502802,
      "loss": 4.8623,
      "mean_token_accuracy": 0.21960606724023818,
      "num_tokens": 99166707.0,
      "step": 57165
    },
    {
      "entropy": 5.829557609558106,
      "epoch": 4.448006224469948,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003168329645270219,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21763397008180618,
      "num_tokens": 99175694.0,
      "step": 57170
    },
    {
      "entropy": 5.90287561416626,
      "epoch": 4.448395253841665,
      "grad_norm": 1.34375,
      "learning_rate": 0.000316805731869015,
      "loss": 4.9935,
      "mean_token_accuracy": 0.20850178599357605,
      "num_tokens": 99184318.0,
      "step": 57175
    },
    {
      "entropy": 5.830742454528808,
      "epoch": 4.448784283213382,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031677849857667266,
      "loss": 4.867,
      "mean_token_accuracy": 0.2167402520775795,
      "num_tokens": 99192280.0,
      "step": 57180
    },
    {
      "entropy": 5.905914831161499,
      "epoch": 4.4491733125851,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003167512646504081,
      "loss": 5.002,
      "mean_token_accuracy": 0.21043099761009215,
      "num_tokens": 99200067.0,
      "step": 57185
    },
    {
      "entropy": 5.917875862121582,
      "epoch": 4.449562341956818,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003167240300906345,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2167690858244896,
      "num_tokens": 99209130.0,
      "step": 57190
    },
    {
      "entropy": 5.893340921401977,
      "epoch": 4.4499513713285355,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003166967948977652,
      "loss": 4.9144,
      "mean_token_accuracy": 0.21695596277713775,
      "num_tokens": 99218211.0,
      "step": 57195
    },
    {
      "entropy": 5.83864574432373,
      "epoch": 4.450340400700253,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003166695590722134,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22346220165491104,
      "num_tokens": 99226640.0,
      "step": 57200
    },
    {
      "entropy": 5.834119844436645,
      "epoch": 4.45072943007197,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031664232261439235,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21484889686107636,
      "num_tokens": 99235825.0,
      "step": 57205
    },
    {
      "entropy": 5.813468503952026,
      "epoch": 4.451118459443688,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003166150855247153,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2157994493842125,
      "num_tokens": 99244206.0,
      "step": 57210
    },
    {
      "entropy": 5.939279174804687,
      "epoch": 4.451507488815405,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003165878478035956,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21667672246694564,
      "num_tokens": 99252625.0,
      "step": 57215
    },
    {
      "entropy": 5.92590446472168,
      "epoch": 4.451896518187123,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003165606094514465,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21139894425868988,
      "num_tokens": 99261998.0,
      "step": 57220
    },
    {
      "entropy": 5.89979772567749,
      "epoch": 4.452285547558841,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031653337046868137,
      "loss": 5.0373,
      "mean_token_accuracy": 0.20884698629379272,
      "num_tokens": 99271381.0,
      "step": 57225
    },
    {
      "entropy": 5.8654149055480955,
      "epoch": 4.4526745769305585,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031650613085571334,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2065042957663536,
      "num_tokens": 99280198.0,
      "step": 57230
    },
    {
      "entropy": 5.870070171356201,
      "epoch": 4.453063606302276,
      "grad_norm": 1.375,
      "learning_rate": 0.0003164788906129559,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22015132158994674,
      "num_tokens": 99288713.0,
      "step": 57235
    },
    {
      "entropy": 5.867806339263916,
      "epoch": 4.453452635673994,
      "grad_norm": 1.375,
      "learning_rate": 0.0003164516497408223,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2219562217593193,
      "num_tokens": 99297038.0,
      "step": 57240
    },
    {
      "entropy": 5.814713859558106,
      "epoch": 4.453841665045711,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003164244082397259,
      "loss": 4.7863,
      "mean_token_accuracy": 0.21988977789878844,
      "num_tokens": 99305802.0,
      "step": 57245
    },
    {
      "entropy": 5.86818265914917,
      "epoch": 4.454230694417428,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031639716611008,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2170003518462181,
      "num_tokens": 99313841.0,
      "step": 57250
    },
    {
      "entropy": 5.922854900360107,
      "epoch": 4.454619723789146,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003163699233522981,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20481980443000794,
      "num_tokens": 99322002.0,
      "step": 57255
    },
    {
      "entropy": 5.85435266494751,
      "epoch": 4.455008753160864,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003163426799667934,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21833187937736512,
      "num_tokens": 99331299.0,
      "step": 57260
    },
    {
      "entropy": 5.873106908798218,
      "epoch": 4.4553977825325815,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003163154359539793,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21037367433309556,
      "num_tokens": 99340816.0,
      "step": 57265
    },
    {
      "entropy": 5.860009860992432,
      "epoch": 4.455786811904299,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00031628819131426924,
      "loss": 4.9246,
      "mean_token_accuracy": 0.216504068672657,
      "num_tokens": 99350438.0,
      "step": 57270
    },
    {
      "entropy": 5.8602807998657225,
      "epoch": 4.456175841276016,
      "grad_norm": 1.375,
      "learning_rate": 0.0003162609460480766,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2088914096355438,
      "num_tokens": 99358430.0,
      "step": 57275
    },
    {
      "entropy": 5.8631589889526365,
      "epoch": 4.456564870647734,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003162337001558147,
      "loss": 4.9391,
      "mean_token_accuracy": 0.213034550845623,
      "num_tokens": 99366479.0,
      "step": 57280
    },
    {
      "entropy": 5.8362242698669435,
      "epoch": 4.456953900019451,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031620645363789713,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21070931106805801,
      "num_tokens": 99375119.0,
      "step": 57285
    },
    {
      "entropy": 5.858323192596435,
      "epoch": 4.457342929391169,
      "grad_norm": 1.375,
      "learning_rate": 0.0003161792064947372,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21583871692419052,
      "num_tokens": 99385216.0,
      "step": 57290
    },
    {
      "entropy": 5.887515449523926,
      "epoch": 4.457731958762887,
      "grad_norm": 1.53125,
      "learning_rate": 0.00031615195872674836,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21943702101707457,
      "num_tokens": 99393805.0,
      "step": 57295
    },
    {
      "entropy": 5.92804479598999,
      "epoch": 4.4581209881346044,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00031612471033434406,
      "loss": 5.009,
      "mean_token_accuracy": 0.21036478132009506,
      "num_tokens": 99403022.0,
      "step": 57300
    },
    {
      "entropy": 5.846676301956177,
      "epoch": 4.458510017506322,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003160974613179377,
      "loss": 4.8964,
      "mean_token_accuracy": 0.21861130595207215,
      "num_tokens": 99412125.0,
      "step": 57305
    },
    {
      "entropy": 5.884039783477784,
      "epoch": 4.458899046878039,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003160702116779428,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2202650487422943,
      "num_tokens": 99420040.0,
      "step": 57310
    },
    {
      "entropy": 5.854933214187622,
      "epoch": 4.459288076249757,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003160429614147727,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21159075498580932,
      "num_tokens": 99428266.0,
      "step": 57315
    },
    {
      "entropy": 5.853244066238403,
      "epoch": 4.459677105621474,
      "grad_norm": 1.515625,
      "learning_rate": 0.00031601571052884117,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22534742057323456,
      "num_tokens": 99436428.0,
      "step": 57320
    },
    {
      "entropy": 5.842436408996582,
      "epoch": 4.460066134993192,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003159884590205613,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21692524552345277,
      "num_tokens": 99445257.0,
      "step": 57325
    },
    {
      "entropy": 5.926843166351318,
      "epoch": 4.46045516436491,
      "grad_norm": 1.40625,
      "learning_rate": 0.000315961206890347,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21319303214550017,
      "num_tokens": 99453884.0,
      "step": 57330
    },
    {
      "entropy": 5.84315938949585,
      "epoch": 4.460844193736627,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003159339541386115,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21765826940536498,
      "num_tokens": 99462877.0,
      "step": 57335
    },
    {
      "entropy": 5.871442651748657,
      "epoch": 4.461233223108344,
      "grad_norm": 1.375,
      "learning_rate": 0.00031590670076576846,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21479899883270265,
      "num_tokens": 99470955.0,
      "step": 57340
    },
    {
      "entropy": 5.883690738677979,
      "epoch": 4.461622252480062,
      "grad_norm": 1.375,
      "learning_rate": 0.00031587944677223133,
      "loss": 5.0247,
      "mean_token_accuracy": 0.20781624168157578,
      "num_tokens": 99479301.0,
      "step": 57345
    },
    {
      "entropy": 5.9074622631073,
      "epoch": 4.46201128185178,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003158521921584136,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22121800482273102,
      "num_tokens": 99488049.0,
      "step": 57350
    },
    {
      "entropy": 5.91347975730896,
      "epoch": 4.462400311223497,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031582493692472886,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2150060147047043,
      "num_tokens": 99496209.0,
      "step": 57355
    },
    {
      "entropy": 6.0034631252288815,
      "epoch": 4.462789340595215,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003157976810715907,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20692339688539504,
      "num_tokens": 99505474.0,
      "step": 57360
    },
    {
      "entropy": 5.913968420028686,
      "epoch": 4.463178369966933,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031577042459941275,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21789529770612717,
      "num_tokens": 99513708.0,
      "step": 57365
    },
    {
      "entropy": 5.860549449920654,
      "epoch": 4.46356739933865,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031574316750860844,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21666476279497146,
      "num_tokens": 99521954.0,
      "step": 57370
    },
    {
      "entropy": 5.821752595901489,
      "epoch": 4.463956428710367,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003157159097995914,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21571887135505677,
      "num_tokens": 99530559.0,
      "step": 57375
    },
    {
      "entropy": 5.937926197052002,
      "epoch": 4.464345458082085,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031568865147277525,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20940686613321305,
      "num_tokens": 99540568.0,
      "step": 57380
    },
    {
      "entropy": 5.921345281600952,
      "epoch": 4.464734487453803,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031566139252857357,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21525816321372987,
      "num_tokens": 99548868.0,
      "step": 57385
    },
    {
      "entropy": 5.8123210906982425,
      "epoch": 4.46512351682552,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003156341329674,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22649578005075455,
      "num_tokens": 99556817.0,
      "step": 57390
    },
    {
      "entropy": 5.884624767303467,
      "epoch": 4.465512546197238,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003156068727896681,
      "loss": 4.9586,
      "mean_token_accuracy": 0.2171418473124504,
      "num_tokens": 99564962.0,
      "step": 57395
    },
    {
      "entropy": 5.84985032081604,
      "epoch": 4.465901575568956,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003155796119957915,
      "loss": 4.9934,
      "mean_token_accuracy": 0.206262943148613,
      "num_tokens": 99574341.0,
      "step": 57400
    },
    {
      "entropy": 5.943967294692993,
      "epoch": 4.466290604940673,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031555235058618403,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2141242116689682,
      "num_tokens": 99583079.0,
      "step": 57405
    },
    {
      "entropy": 5.877442169189453,
      "epoch": 4.46667963431239,
      "grad_norm": 1.5,
      "learning_rate": 0.00031552508856125906,
      "loss": 4.8939,
      "mean_token_accuracy": 0.21482914984226226,
      "num_tokens": 99591055.0,
      "step": 57410
    },
    {
      "entropy": 5.806022214889526,
      "epoch": 4.467068663684108,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003154978259214304,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2130469262599945,
      "num_tokens": 99600438.0,
      "step": 57415
    },
    {
      "entropy": 5.844544219970703,
      "epoch": 4.467457693055826,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031547056266711173,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21391918808221816,
      "num_tokens": 99608541.0,
      "step": 57420
    },
    {
      "entropy": 5.829662656784057,
      "epoch": 4.467846722427543,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003154432987987166,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21896325945854186,
      "num_tokens": 99617013.0,
      "step": 57425
    },
    {
      "entropy": 5.910103893280029,
      "epoch": 4.468235751799261,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003154160343166589,
      "loss": 4.952,
      "mean_token_accuracy": 0.20986531376838685,
      "num_tokens": 99625957.0,
      "step": 57430
    },
    {
      "entropy": 5.824017143249511,
      "epoch": 4.468624781170979,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031538876922135214,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21520230323076248,
      "num_tokens": 99635783.0,
      "step": 57435
    },
    {
      "entropy": 5.9219707489013675,
      "epoch": 4.469013810542696,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003153615035132101,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21304901242256163,
      "num_tokens": 99644066.0,
      "step": 57440
    },
    {
      "entropy": 5.896878862380982,
      "epoch": 4.469402839914413,
      "grad_norm": 1.375,
      "learning_rate": 0.00031533423719264645,
      "loss": 5.0355,
      "mean_token_accuracy": 0.2126910164952278,
      "num_tokens": 99653194.0,
      "step": 57445
    },
    {
      "entropy": 5.792404460906982,
      "epoch": 4.469791869286131,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000315306970260075,
      "loss": 4.8958,
      "mean_token_accuracy": 0.21507944017648697,
      "num_tokens": 99661332.0,
      "step": 57450
    },
    {
      "entropy": 5.832111120223999,
      "epoch": 4.470180898657849,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003152797027159094,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22227553576231002,
      "num_tokens": 99670163.0,
      "step": 57455
    },
    {
      "entropy": 5.877451467514038,
      "epoch": 4.470569928029566,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031525243456056345,
      "loss": 4.904,
      "mean_token_accuracy": 0.22223387360572816,
      "num_tokens": 99678174.0,
      "step": 57460
    },
    {
      "entropy": 5.8917945384979244,
      "epoch": 4.470958957401284,
      "grad_norm": 1.375,
      "learning_rate": 0.0003152251657944509,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2077690988779068,
      "num_tokens": 99686364.0,
      "step": 57465
    },
    {
      "entropy": 5.798991107940674,
      "epoch": 4.471347986773002,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003151978964179854,
      "loss": 4.843,
      "mean_token_accuracy": 0.21934892237186432,
      "num_tokens": 99694752.0,
      "step": 57470
    },
    {
      "entropy": 5.84685525894165,
      "epoch": 4.4717370161447185,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031517062643158085,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2121386244893074,
      "num_tokens": 99703819.0,
      "step": 57475
    },
    {
      "entropy": 5.836417388916016,
      "epoch": 4.472126045516436,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031514335583565097,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21812244206666948,
      "num_tokens": 99711908.0,
      "step": 57480
    },
    {
      "entropy": 5.853615951538086,
      "epoch": 4.472515074888154,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031511608463060954,
      "loss": 4.8381,
      "mean_token_accuracy": 0.22978344261646272,
      "num_tokens": 99720356.0,
      "step": 57485
    },
    {
      "entropy": 5.936135721206665,
      "epoch": 4.472904104259872,
      "grad_norm": 1.5,
      "learning_rate": 0.00031508881281687044,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2159103512763977,
      "num_tokens": 99729408.0,
      "step": 57490
    },
    {
      "entropy": 5.865327167510986,
      "epoch": 4.473293133631589,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031506154039484736,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22066086828708648,
      "num_tokens": 99738819.0,
      "step": 57495
    },
    {
      "entropy": 5.840186691284179,
      "epoch": 4.473682163003307,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031503426736495424,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22595804929733276,
      "num_tokens": 99747474.0,
      "step": 57500
    },
    {
      "entropy": 5.823077535629272,
      "epoch": 4.474071192375025,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031500699372760476,
      "loss": 4.856,
      "mean_token_accuracy": 0.22064838707447051,
      "num_tokens": 99756541.0,
      "step": 57505
    },
    {
      "entropy": 5.794564199447632,
      "epoch": 4.4744602217467415,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031497971948321287,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21672210991382598,
      "num_tokens": 99765688.0,
      "step": 57510
    },
    {
      "entropy": 5.79774580001831,
      "epoch": 4.474849251118459,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003149524446321923,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2188581258058548,
      "num_tokens": 99774139.0,
      "step": 57515
    },
    {
      "entropy": 5.806543207168579,
      "epoch": 4.475238280490177,
      "grad_norm": 1.46875,
      "learning_rate": 0.000314925169174957,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2278507649898529,
      "num_tokens": 99782429.0,
      "step": 57520
    },
    {
      "entropy": 5.878740215301514,
      "epoch": 4.475627309861895,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031489789311192086,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21509101390838622,
      "num_tokens": 99790901.0,
      "step": 57525
    },
    {
      "entropy": 5.8573215961456295,
      "epoch": 4.476016339233612,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003148706164434977,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2241940677165985,
      "num_tokens": 99799441.0,
      "step": 57530
    },
    {
      "entropy": 5.797119045257569,
      "epoch": 4.47640536860533,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003148433391701014,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2258593961596489,
      "num_tokens": 99808590.0,
      "step": 57535
    },
    {
      "entropy": 5.87029800415039,
      "epoch": 4.476794397977047,
      "grad_norm": 1.5,
      "learning_rate": 0.0003148160612921457,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2087582215666771,
      "num_tokens": 99817098.0,
      "step": 57540
    },
    {
      "entropy": 5.887494516372681,
      "epoch": 4.4771834273487645,
      "grad_norm": 1.59375,
      "learning_rate": 0.00031478878281004474,
      "loss": 4.899,
      "mean_token_accuracy": 0.2140077456831932,
      "num_tokens": 99825282.0,
      "step": 57545
    },
    {
      "entropy": 5.912522459030152,
      "epoch": 4.477572456720482,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031476150372421234,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20480482131242753,
      "num_tokens": 99833883.0,
      "step": 57550
    },
    {
      "entropy": 5.866175365447998,
      "epoch": 4.4779614860922,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003147342240350625,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21857628524303435,
      "num_tokens": 99842533.0,
      "step": 57555
    },
    {
      "entropy": 5.822031784057617,
      "epoch": 4.478350515463918,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003147069437430089,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2231131374835968,
      "num_tokens": 99851212.0,
      "step": 57560
    },
    {
      "entropy": 5.916667890548706,
      "epoch": 4.478739544835635,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031467966284846573,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21020474880933762,
      "num_tokens": 99860135.0,
      "step": 57565
    },
    {
      "entropy": 5.900613403320312,
      "epoch": 4.479128574207353,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003146523813518468,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21814577877521515,
      "num_tokens": 99868387.0,
      "step": 57570
    },
    {
      "entropy": 5.831220865249634,
      "epoch": 4.47951760357907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031462509925356616,
      "loss": 4.9402,
      "mean_token_accuracy": 0.216188283264637,
      "num_tokens": 99877579.0,
      "step": 57575
    },
    {
      "entropy": 5.828810310363769,
      "epoch": 4.4799066329507875,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031459781655403756,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2165072366595268,
      "num_tokens": 99886167.0,
      "step": 57580
    },
    {
      "entropy": 5.8717223644256595,
      "epoch": 4.480295662322505,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003145705332536752,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22159151434898378,
      "num_tokens": 99894914.0,
      "step": 57585
    },
    {
      "entropy": 5.813878536224365,
      "epoch": 4.480684691694223,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000314543249352893,
      "loss": 4.8574,
      "mean_token_accuracy": 0.21957215517759324,
      "num_tokens": 99904275.0,
      "step": 57590
    },
    {
      "entropy": 5.838156652450562,
      "epoch": 4.481073721065941,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031451596485210486,
      "loss": 4.941,
      "mean_token_accuracy": 0.21323951631784438,
      "num_tokens": 99913566.0,
      "step": 57595
    },
    {
      "entropy": 5.875422048568725,
      "epoch": 4.481462750437658,
      "grad_norm": 1.453125,
      "learning_rate": 0.00031448867975172493,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21095221638679504,
      "num_tokens": 99922582.0,
      "step": 57600
    },
    {
      "entropy": 5.827519512176513,
      "epoch": 4.481851779809376,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00031446139405216713,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2075935795903206,
      "num_tokens": 99931426.0,
      "step": 57605
    },
    {
      "entropy": 5.852397537231445,
      "epoch": 4.482240809181093,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003144341077538455,
      "loss": 4.9701,
      "mean_token_accuracy": 0.20843301713466644,
      "num_tokens": 99939726.0,
      "step": 57610
    },
    {
      "entropy": 5.807204532623291,
      "epoch": 4.4826298385528105,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000314406820857174,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2199338734149933,
      "num_tokens": 99948492.0,
      "step": 57615
    },
    {
      "entropy": 5.860810708999634,
      "epoch": 4.483018867924528,
      "grad_norm": 1.5,
      "learning_rate": 0.00031437953336256667,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22088005393743515,
      "num_tokens": 99956577.0,
      "step": 57620
    },
    {
      "entropy": 5.841447496414185,
      "epoch": 4.483407897296246,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003143522452704377,
      "loss": 5.0245,
      "mean_token_accuracy": 0.2072474867105484,
      "num_tokens": 99964896.0,
      "step": 57625
    },
    {
      "entropy": 5.960890388488769,
      "epoch": 4.483796926667964,
      "grad_norm": 1.359375,
      "learning_rate": 0.000314324956581201,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20996006727218627,
      "num_tokens": 99973703.0,
      "step": 57630
    },
    {
      "entropy": 5.815118598937988,
      "epoch": 4.484185956039681,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003142976672952707,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21597819030284882,
      "num_tokens": 99982258.0,
      "step": 57635
    },
    {
      "entropy": 5.890868425369263,
      "epoch": 4.484574985411399,
      "grad_norm": 1.609375,
      "learning_rate": 0.00031427037741306083,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2126392215490341,
      "num_tokens": 99990848.0,
      "step": 57640
    },
    {
      "entropy": 5.768831014633179,
      "epoch": 4.484964014783116,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003142430869349855,
      "loss": 4.9124,
      "mean_token_accuracy": 0.21923135817050934,
      "num_tokens": 99999694.0,
      "step": 57645
    },
    {
      "entropy": 5.797782230377197,
      "epoch": 4.4853530441548335,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003142157958614588,
      "loss": 4.951,
      "mean_token_accuracy": 0.2093740686774254,
      "num_tokens": 100008473.0,
      "step": 57650
    },
    {
      "entropy": 5.921360969543457,
      "epoch": 4.485742073526551,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003141885041928948,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21632418483495713,
      "num_tokens": 100016331.0,
      "step": 57655
    },
    {
      "entropy": 5.79804105758667,
      "epoch": 4.486131102898269,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031416121192970754,
      "loss": 4.8047,
      "mean_token_accuracy": 0.22692911475896835,
      "num_tokens": 100024506.0,
      "step": 57660
    },
    {
      "entropy": 5.799439525604248,
      "epoch": 4.486520132269987,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031413391907231133,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21350111216306686,
      "num_tokens": 100032585.0,
      "step": 57665
    },
    {
      "entropy": 5.770336437225342,
      "epoch": 4.486909161641704,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031410662562112015,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2197784185409546,
      "num_tokens": 100040763.0,
      "step": 57670
    },
    {
      "entropy": 5.905692672729492,
      "epoch": 4.487298191013421,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031407933157654816,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2058960035443306,
      "num_tokens": 100049826.0,
      "step": 57675
    },
    {
      "entropy": 5.910783910751343,
      "epoch": 4.487687220385139,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003140520369390095,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21953096687793733,
      "num_tokens": 100058462.0,
      "step": 57680
    },
    {
      "entropy": 5.865838050842285,
      "epoch": 4.4880762497568565,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003140247417089184,
      "loss": 4.957,
      "mean_token_accuracy": 0.21213608235120773,
      "num_tokens": 100067091.0,
      "step": 57685
    },
    {
      "entropy": 5.871687269210815,
      "epoch": 4.488465279128574,
      "grad_norm": 1.375,
      "learning_rate": 0.0003139974458866889,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20642295926809312,
      "num_tokens": 100076272.0,
      "step": 57690
    },
    {
      "entropy": 5.903164196014404,
      "epoch": 4.488854308500292,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003139701494727353,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20966168940067292,
      "num_tokens": 100085333.0,
      "step": 57695
    },
    {
      "entropy": 5.855655145645142,
      "epoch": 4.48924333787201,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003139428524674716,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2242722675204277,
      "num_tokens": 100093768.0,
      "step": 57700
    },
    {
      "entropy": 5.860012245178223,
      "epoch": 4.489632367243727,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003139155548713122,
      "loss": 5.011,
      "mean_token_accuracy": 0.2080511197447777,
      "num_tokens": 100102452.0,
      "step": 57705
    },
    {
      "entropy": 5.884925842285156,
      "epoch": 4.490021396615444,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031388825668467115,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21219970434904098,
      "num_tokens": 100110864.0,
      "step": 57710
    },
    {
      "entropy": 5.910635328292846,
      "epoch": 4.490410425987162,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031386095790796265,
      "loss": 5.0148,
      "mean_token_accuracy": 0.20813288539648056,
      "num_tokens": 100120772.0,
      "step": 57715
    },
    {
      "entropy": 5.914453172683716,
      "epoch": 4.4907994553588795,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031383365854160104,
      "loss": 4.973,
      "mean_token_accuracy": 0.21915349662303923,
      "num_tokens": 100129100.0,
      "step": 57720
    },
    {
      "entropy": 5.85779767036438,
      "epoch": 4.491188484730597,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003138063585860005,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21512189954519273,
      "num_tokens": 100137992.0,
      "step": 57725
    },
    {
      "entropy": 5.840049743652344,
      "epoch": 4.491577514102315,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003137790580415752,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21135170757770538,
      "num_tokens": 100146608.0,
      "step": 57730
    },
    {
      "entropy": 5.951354694366455,
      "epoch": 4.491966543474033,
      "grad_norm": 1.375,
      "learning_rate": 0.00031375175690873943,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20399457067251206,
      "num_tokens": 100155050.0,
      "step": 57735
    },
    {
      "entropy": 5.857372570037842,
      "epoch": 4.49235557284575,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00031372445518790737,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2183433786034584,
      "num_tokens": 100164496.0,
      "step": 57740
    },
    {
      "entropy": 5.856878089904785,
      "epoch": 4.492744602217467,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031369715287949334,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21057049036026002,
      "num_tokens": 100173076.0,
      "step": 57745
    },
    {
      "entropy": 5.834941577911377,
      "epoch": 4.493133631589185,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031366984998391163,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22243731021881102,
      "num_tokens": 100181972.0,
      "step": 57750
    },
    {
      "entropy": 5.851762056350708,
      "epoch": 4.4935226609609025,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031364254650157646,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21694232672452926,
      "num_tokens": 100191017.0,
      "step": 57755
    },
    {
      "entropy": 5.813315439224243,
      "epoch": 4.49391169033262,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003136152424329023,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23202947229146959,
      "num_tokens": 100198852.0,
      "step": 57760
    },
    {
      "entropy": 5.823428201675415,
      "epoch": 4.494300719704338,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003135879377783032,
      "loss": 4.9751,
      "mean_token_accuracy": 0.20778681784868241,
      "num_tokens": 100207477.0,
      "step": 57765
    },
    {
      "entropy": 5.880980491638184,
      "epoch": 4.494689749076056,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00031356063253819354,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2158778950572014,
      "num_tokens": 100215426.0,
      "step": 57770
    },
    {
      "entropy": 5.814322519302368,
      "epoch": 4.495078778447773,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0003135333267129877,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22320742011070252,
      "num_tokens": 100224511.0,
      "step": 57775
    },
    {
      "entropy": 5.777116870880127,
      "epoch": 4.49546780781949,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003135060203030999,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2197798252105713,
      "num_tokens": 100232925.0,
      "step": 57780
    },
    {
      "entropy": 5.888576507568359,
      "epoch": 4.495856837191208,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031347871330894464,
      "loss": 5.0391,
      "mean_token_accuracy": 0.20857494324445724,
      "num_tokens": 100242140.0,
      "step": 57785
    },
    {
      "entropy": 5.909078788757324,
      "epoch": 4.4962458665629255,
      "grad_norm": 1.4375,
      "learning_rate": 0.000313451405730936,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2109681785106659,
      "num_tokens": 100250431.0,
      "step": 57790
    },
    {
      "entropy": 5.920571422576904,
      "epoch": 4.496634895934643,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003134240975694886,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21439069211483003,
      "num_tokens": 100259117.0,
      "step": 57795
    },
    {
      "entropy": 5.883169174194336,
      "epoch": 4.497023925306361,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031339678882501667,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21548188328742982,
      "num_tokens": 100267273.0,
      "step": 57800
    },
    {
      "entropy": 5.883789873123169,
      "epoch": 4.497412954678079,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00031336947949793455,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2193266823887825,
      "num_tokens": 100276091.0,
      "step": 57805
    },
    {
      "entropy": 5.893096494674682,
      "epoch": 4.497801984049795,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031334216958865663,
      "loss": 4.8374,
      "mean_token_accuracy": 0.22568663954734802,
      "num_tokens": 100284697.0,
      "step": 57810
    },
    {
      "entropy": 5.89234504699707,
      "epoch": 4.498191013421513,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031331485909759734,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21942313611507416,
      "num_tokens": 100292851.0,
      "step": 57815
    },
    {
      "entropy": 5.892862367630005,
      "epoch": 4.498580042793231,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000313287548025171,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2093367725610733,
      "num_tokens": 100301708.0,
      "step": 57820
    },
    {
      "entropy": 5.857749319076538,
      "epoch": 4.4989690721649485,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003132602363717921,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2124334454536438,
      "num_tokens": 100311203.0,
      "step": 57825
    },
    {
      "entropy": 5.845761299133301,
      "epoch": 4.499358101536666,
      "grad_norm": 1.46875,
      "learning_rate": 0.00031323292413787504,
      "loss": 4.918,
      "mean_token_accuracy": 0.22336983382701875,
      "num_tokens": 100320186.0,
      "step": 57830
    },
    {
      "entropy": 5.888300800323487,
      "epoch": 4.499747130908384,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003132056113238341,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21514641046524047,
      "num_tokens": 100329427.0,
      "step": 57835
    },
    {
      "entropy": 5.930241107940674,
      "epoch": 4.5001361602801015,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031317829793008403,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21478995233774184,
      "num_tokens": 100338147.0,
      "step": 57840
    },
    {
      "entropy": 5.768954944610596,
      "epoch": 4.500525189651818,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003131509839570389,
      "loss": 4.8572,
      "mean_token_accuracy": 0.21769253611564637,
      "num_tokens": 100346762.0,
      "step": 57845
    },
    {
      "entropy": 5.85191125869751,
      "epoch": 4.500914219023536,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031312366940511327,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21671951860189437,
      "num_tokens": 100355302.0,
      "step": 57850
    },
    {
      "entropy": 5.923528051376342,
      "epoch": 4.501303248395254,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003130963542747217,
      "loss": 4.947,
      "mean_token_accuracy": 0.21806734204292297,
      "num_tokens": 100363595.0,
      "step": 57855
    },
    {
      "entropy": 5.815790891647339,
      "epoch": 4.5016922777669715,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031306903856627866,
      "loss": 4.8827,
      "mean_token_accuracy": 0.21580542623996735,
      "num_tokens": 100372885.0,
      "step": 57860
    },
    {
      "entropy": 5.863810682296753,
      "epoch": 4.502081307138689,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031304172228019845,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22517270147800444,
      "num_tokens": 100380829.0,
      "step": 57865
    },
    {
      "entropy": 5.842498636245727,
      "epoch": 4.502470336510407,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031301440541689577,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21981606781482696,
      "num_tokens": 100390282.0,
      "step": 57870
    },
    {
      "entropy": 5.81997389793396,
      "epoch": 4.502859365882124,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003129870879767849,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2159394919872284,
      "num_tokens": 100398544.0,
      "step": 57875
    },
    {
      "entropy": 5.907523727416992,
      "epoch": 4.503248395253841,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003129597699602804,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2165328413248062,
      "num_tokens": 100407420.0,
      "step": 57880
    },
    {
      "entropy": 5.7998692989349365,
      "epoch": 4.503637424625559,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003129324513677969,
      "loss": 4.8335,
      "mean_token_accuracy": 0.22802205234766007,
      "num_tokens": 100416691.0,
      "step": 57885
    },
    {
      "entropy": 5.889222192764282,
      "epoch": 4.504026453997277,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00031290513219974875,
      "loss": 4.8931,
      "mean_token_accuracy": 0.21896199136972427,
      "num_tokens": 100424459.0,
      "step": 57890
    },
    {
      "entropy": 5.893585872650147,
      "epoch": 4.504415483368994,
      "grad_norm": 1.375,
      "learning_rate": 0.00031287781245655064,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21434388905763627,
      "num_tokens": 100432437.0,
      "step": 57895
    },
    {
      "entropy": 5.874051189422607,
      "epoch": 4.504804512740712,
      "grad_norm": 1.375,
      "learning_rate": 0.000312850492138617,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21271952390670776,
      "num_tokens": 100441726.0,
      "step": 57900
    },
    {
      "entropy": 5.899984645843506,
      "epoch": 4.50519354211243,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003128231712463623,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21008401811122895,
      "num_tokens": 100450181.0,
      "step": 57905
    },
    {
      "entropy": 5.790151739120484,
      "epoch": 4.5055825714841475,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003127958497802012,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2153690680861473,
      "num_tokens": 100458693.0,
      "step": 57910
    },
    {
      "entropy": 5.904143285751343,
      "epoch": 4.505971600855864,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003127685277405483,
      "loss": 5.0409,
      "mean_token_accuracy": 0.20803707391023635,
      "num_tokens": 100467163.0,
      "step": 57915
    },
    {
      "entropy": 5.925388622283935,
      "epoch": 4.506360630227582,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00031274120512781806,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21088217496871947,
      "num_tokens": 100475546.0,
      "step": 57920
    },
    {
      "entropy": 5.924038553237915,
      "epoch": 4.5067496595993,
      "grad_norm": 1.375,
      "learning_rate": 0.0003127138819424252,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21048192530870438,
      "num_tokens": 100484075.0,
      "step": 57925
    },
    {
      "entropy": 5.977309274673462,
      "epoch": 4.507138688971017,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003126865581847841,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20048573911190032,
      "num_tokens": 100492630.0,
      "step": 57930
    },
    {
      "entropy": 5.898220586776733,
      "epoch": 4.507527718342735,
      "grad_norm": 1.5,
      "learning_rate": 0.00031265923385530943,
      "loss": 4.9748,
      "mean_token_accuracy": 0.213223372399807,
      "num_tokens": 100501389.0,
      "step": 57935
    },
    {
      "entropy": 5.859931516647339,
      "epoch": 4.507916747714452,
      "grad_norm": 1.375,
      "learning_rate": 0.0003126319089544159,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22059128135442735,
      "num_tokens": 100510070.0,
      "step": 57940
    },
    {
      "entropy": 5.926222324371338,
      "epoch": 4.50830577708617,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000312604583482518,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21741771697998047,
      "num_tokens": 100518421.0,
      "step": 57945
    },
    {
      "entropy": 5.856549072265625,
      "epoch": 4.508694806457887,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003125772574400305,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21920562535524368,
      "num_tokens": 100527314.0,
      "step": 57950
    },
    {
      "entropy": 5.854361581802368,
      "epoch": 4.509083835829605,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003125499308273679,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21409669071435927,
      "num_tokens": 100535737.0,
      "step": 57955
    },
    {
      "entropy": 5.855248212814331,
      "epoch": 4.509472865201323,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031252260364494477,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21924584954977036,
      "num_tokens": 100543867.0,
      "step": 57960
    },
    {
      "entropy": 5.871595096588135,
      "epoch": 4.50986189457304,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031249527589317596,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21125365793704987,
      "num_tokens": 100552633.0,
      "step": 57965
    },
    {
      "entropy": 5.8907605648040775,
      "epoch": 4.510250923944758,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003124679475724759,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2145391270518303,
      "num_tokens": 100561842.0,
      "step": 57970
    },
    {
      "entropy": 5.914718246459961,
      "epoch": 4.510639953316476,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003124406186832595,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22656373977661132,
      "num_tokens": 100569988.0,
      "step": 57975
    },
    {
      "entropy": 5.921739387512207,
      "epoch": 4.511028982688193,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003124132892259412,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22258254289627075,
      "num_tokens": 100577795.0,
      "step": 57980
    },
    {
      "entropy": 5.802115869522095,
      "epoch": 4.51141801205991,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031238595920093584,
      "loss": 4.8594,
      "mean_token_accuracy": 0.21999442726373672,
      "num_tokens": 100586359.0,
      "step": 57985
    },
    {
      "entropy": 5.843461799621582,
      "epoch": 4.511807041431628,
      "grad_norm": 1.484375,
      "learning_rate": 0.00031235862860865807,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21094732880592346,
      "num_tokens": 100595062.0,
      "step": 57990
    },
    {
      "entropy": 5.854473304748535,
      "epoch": 4.512196070803346,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003123312974495226,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21433429121971131,
      "num_tokens": 100603606.0,
      "step": 57995
    },
    {
      "entropy": 5.92282509803772,
      "epoch": 4.512585100175063,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00031230396572394406,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2174785777926445,
      "num_tokens": 100611696.0,
      "step": 58000
    },
    {
      "entropy": 5.910829830169678,
      "epoch": 4.512974129546781,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003122766334323372,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22018830925226213,
      "num_tokens": 100620917.0,
      "step": 58005
    },
    {
      "entropy": 5.8818563461303714,
      "epoch": 4.513363158918498,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031224930057511673,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2209864839911461,
      "num_tokens": 100629226.0,
      "step": 58010
    },
    {
      "entropy": 5.779057312011719,
      "epoch": 4.513752188290216,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003122219671526974,
      "loss": 4.8105,
      "mean_token_accuracy": 0.22236605286598204,
      "num_tokens": 100637322.0,
      "step": 58015
    },
    {
      "entropy": 5.827506732940674,
      "epoch": 4.514141217661933,
      "grad_norm": 1.46875,
      "learning_rate": 0.00031219463316549406,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22250427454710006,
      "num_tokens": 100646509.0,
      "step": 58020
    },
    {
      "entropy": 5.883781909942627,
      "epoch": 4.514530247033651,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031216729861392137,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21526462882757186,
      "num_tokens": 100654447.0,
      "step": 58025
    },
    {
      "entropy": 5.9213831424713135,
      "epoch": 4.514919276405369,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003121399634983941,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2094537302851677,
      "num_tokens": 100663228.0,
      "step": 58030
    },
    {
      "entropy": 5.806017160415649,
      "epoch": 4.515308305777086,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031211262781932693,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2223345384001732,
      "num_tokens": 100671533.0,
      "step": 58035
    },
    {
      "entropy": 5.860323190689087,
      "epoch": 4.515697335148804,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031208529157713465,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2130352959036827,
      "num_tokens": 100679749.0,
      "step": 58040
    },
    {
      "entropy": 5.859959983825684,
      "epoch": 4.516086364520521,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031205795477223217,
      "loss": 4.911,
      "mean_token_accuracy": 0.21736120879650117,
      "num_tokens": 100688483.0,
      "step": 58045
    },
    {
      "entropy": 5.8430839538574215,
      "epoch": 4.516475393892239,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031203061740503414,
      "loss": 4.894,
      "mean_token_accuracy": 0.21769499778747559,
      "num_tokens": 100696747.0,
      "step": 58050
    },
    {
      "entropy": 5.885718297958374,
      "epoch": 4.516864423263956,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003120032794759555,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2181092843413353,
      "num_tokens": 100704900.0,
      "step": 58055
    },
    {
      "entropy": 5.87092080116272,
      "epoch": 4.517253452635674,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000311975940985411,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2041421115398407,
      "num_tokens": 100713745.0,
      "step": 58060
    },
    {
      "entropy": 5.85022292137146,
      "epoch": 4.517642482007392,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003119486019338154,
      "loss": 4.9156,
      "mean_token_accuracy": 0.21923898607492448,
      "num_tokens": 100722537.0,
      "step": 58065
    },
    {
      "entropy": 5.905551958084106,
      "epoch": 4.518031511379109,
      "grad_norm": 1.375,
      "learning_rate": 0.00031192126232158367,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21927570104598998,
      "num_tokens": 100730748.0,
      "step": 58070
    },
    {
      "entropy": 5.9061685562133786,
      "epoch": 4.518420540750826,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003118939221491305,
      "loss": 4.9887,
      "mean_token_accuracy": 0.20452854484319688,
      "num_tokens": 100738999.0,
      "step": 58075
    },
    {
      "entropy": 5.9243992328643795,
      "epoch": 4.518809570122544,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003118665814168707,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2107895314693451,
      "num_tokens": 100747729.0,
      "step": 58080
    },
    {
      "entropy": 5.893810701370239,
      "epoch": 4.519198599494262,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031183924012521925,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21545276045799255,
      "num_tokens": 100755779.0,
      "step": 58085
    },
    {
      "entropy": 5.803054618835449,
      "epoch": 4.519587628865979,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000311811898274591,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22326964735984803,
      "num_tokens": 100764372.0,
      "step": 58090
    },
    {
      "entropy": 5.808957719802857,
      "epoch": 4.519976658237697,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031178455586540076,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21837408244609832,
      "num_tokens": 100773884.0,
      "step": 58095
    },
    {
      "entropy": 5.823459625244141,
      "epoch": 4.520365687609415,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031175721289806345,
      "loss": 4.8644,
      "mean_token_accuracy": 0.2216634601354599,
      "num_tokens": 100783175.0,
      "step": 58100
    },
    {
      "entropy": 5.828511095046997,
      "epoch": 4.520754716981132,
      "grad_norm": 1.375,
      "learning_rate": 0.00031172986937299396,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2161809489130974,
      "num_tokens": 100791956.0,
      "step": 58105
    },
    {
      "entropy": 5.881030082702637,
      "epoch": 4.52114374635285,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00031170252529060715,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21496340483427048,
      "num_tokens": 100800197.0,
      "step": 58110
    },
    {
      "entropy": 5.895401811599731,
      "epoch": 4.521532775724567,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003116751806513179,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21107156276702882,
      "num_tokens": 100808658.0,
      "step": 58115
    },
    {
      "entropy": 5.806486654281616,
      "epoch": 4.521921805096285,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003116478354555412,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22433369755744934,
      "num_tokens": 100817640.0,
      "step": 58120
    },
    {
      "entropy": 5.956994295120239,
      "epoch": 4.522310834468002,
      "grad_norm": 1.625,
      "learning_rate": 0.00031162048970369196,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21367547810077667,
      "num_tokens": 100826550.0,
      "step": 58125
    },
    {
      "entropy": 5.92390079498291,
      "epoch": 4.52269986383972,
      "grad_norm": 1.421875,
      "learning_rate": 0.000311593143396185,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2167952075600624,
      "num_tokens": 100835725.0,
      "step": 58130
    },
    {
      "entropy": 5.744177436828613,
      "epoch": 4.523088893211438,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031156579653343543,
      "loss": 4.7396,
      "mean_token_accuracy": 0.22516141533851625,
      "num_tokens": 100843926.0,
      "step": 58135
    },
    {
      "entropy": 5.796554851531982,
      "epoch": 4.523477922583155,
      "grad_norm": 1.484375,
      "learning_rate": 0.000311538449115858,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22058198153972625,
      "num_tokens": 100852580.0,
      "step": 58140
    },
    {
      "entropy": 5.855263614654541,
      "epoch": 4.523866951954872,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003115111011438678,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22500666230916977,
      "num_tokens": 100861026.0,
      "step": 58145
    },
    {
      "entropy": 5.872681951522827,
      "epoch": 4.52425598132659,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003114837526178797,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21567706018686295,
      "num_tokens": 100869959.0,
      "step": 58150
    },
    {
      "entropy": 5.924702453613281,
      "epoch": 4.524645010698308,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003114564035383088,
      "loss": 5.0541,
      "mean_token_accuracy": 0.20988839566707612,
      "num_tokens": 100879448.0,
      "step": 58155
    },
    {
      "entropy": 5.837143516540527,
      "epoch": 4.525034040070025,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031142905390556995,
      "loss": 4.862,
      "mean_token_accuracy": 0.22447603046894074,
      "num_tokens": 100888111.0,
      "step": 58160
    },
    {
      "entropy": 5.785494661331176,
      "epoch": 4.525423069441743,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003114017037200782,
      "loss": 4.8625,
      "mean_token_accuracy": 0.21931740641593933,
      "num_tokens": 100897305.0,
      "step": 58165
    },
    {
      "entropy": 5.793621397018432,
      "epoch": 4.525812098813461,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031137435298224863,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22509240806102754,
      "num_tokens": 100906239.0,
      "step": 58170
    },
    {
      "entropy": 5.8638897895812985,
      "epoch": 4.526201128185178,
      "grad_norm": 1.375,
      "learning_rate": 0.0003113470016924961,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22280639857053758,
      "num_tokens": 100914579.0,
      "step": 58175
    },
    {
      "entropy": 5.882165050506591,
      "epoch": 4.526590157556895,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00031131964985123555,
      "loss": 4.8692,
      "mean_token_accuracy": 0.21803976595401764,
      "num_tokens": 100923291.0,
      "step": 58180
    },
    {
      "entropy": 5.8413878917694095,
      "epoch": 4.526979186928613,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003112922974588822,
      "loss": 4.8776,
      "mean_token_accuracy": 0.21793405413627626,
      "num_tokens": 100932077.0,
      "step": 58185
    },
    {
      "entropy": 5.954155731201172,
      "epoch": 4.527368216300331,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000311264944515851,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20982412695884706,
      "num_tokens": 100941565.0,
      "step": 58190
    },
    {
      "entropy": 5.9046008586883545,
      "epoch": 4.527757245672048,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031123759102255714,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21635877043008805,
      "num_tokens": 100950098.0,
      "step": 58195
    },
    {
      "entropy": 5.829088258743286,
      "epoch": 4.528146275043766,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003112102369794153,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22055758833885192,
      "num_tokens": 100958496.0,
      "step": 58200
    },
    {
      "entropy": 5.762178134918213,
      "epoch": 4.528535304415484,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031118288238684086,
      "loss": 4.8808,
      "mean_token_accuracy": 0.21818768382072448,
      "num_tokens": 100967592.0,
      "step": 58205
    },
    {
      "entropy": 5.9116922378540036,
      "epoch": 4.5289243337872005,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003111555272452486,
      "loss": 5.0417,
      "mean_token_accuracy": 0.2124434843659401,
      "num_tokens": 100976404.0,
      "step": 58210
    },
    {
      "entropy": 5.920695972442627,
      "epoch": 4.529313363158918,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031112817155505394,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2096927985548973,
      "num_tokens": 100984499.0,
      "step": 58215
    },
    {
      "entropy": 5.871499395370483,
      "epoch": 4.529702392530636,
      "grad_norm": 1.375,
      "learning_rate": 0.00031110081531667184,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2221081241965294,
      "num_tokens": 100993420.0,
      "step": 58220
    },
    {
      "entropy": 5.783553838729858,
      "epoch": 4.530091421902354,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031107345853051714,
      "loss": 4.836,
      "mean_token_accuracy": 0.21702542454004287,
      "num_tokens": 101003108.0,
      "step": 58225
    },
    {
      "entropy": 5.860838937759399,
      "epoch": 4.530480451274071,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031104610119700525,
      "loss": 4.9881,
      "mean_token_accuracy": 0.20956325381994248,
      "num_tokens": 101012103.0,
      "step": 58230
    },
    {
      "entropy": 5.913282775878907,
      "epoch": 4.530869480645789,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003110187433165511,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21530283838510514,
      "num_tokens": 101021529.0,
      "step": 58235
    },
    {
      "entropy": 5.947418117523194,
      "epoch": 4.531258510017507,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003109913848895699,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21252587884664537,
      "num_tokens": 101030657.0,
      "step": 58240
    },
    {
      "entropy": 5.779807233810425,
      "epoch": 4.531647539389224,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003109640259164766,
      "loss": 4.8009,
      "mean_token_accuracy": 0.22678810209035874,
      "num_tokens": 101038715.0,
      "step": 58245
    },
    {
      "entropy": 5.758349227905273,
      "epoch": 4.532036568760941,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031093666639768654,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2178892970085144,
      "num_tokens": 101047318.0,
      "step": 58250
    },
    {
      "entropy": 5.817065334320068,
      "epoch": 4.532425598132659,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003109093063336147,
      "loss": 4.7798,
      "mean_token_accuracy": 0.2341475874185562,
      "num_tokens": 101056677.0,
      "step": 58255
    },
    {
      "entropy": 5.911284446716309,
      "epoch": 4.532814627504377,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003108819457246764,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21537545472383499,
      "num_tokens": 101065734.0,
      "step": 58260
    },
    {
      "entropy": 5.863009214401245,
      "epoch": 4.533203656876094,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031085458457128667,
      "loss": 4.8929,
      "mean_token_accuracy": 0.21274728327989578,
      "num_tokens": 101074795.0,
      "step": 58265
    },
    {
      "entropy": 5.855525636672974,
      "epoch": 4.533592686247812,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031082722287386065,
      "loss": 4.9572,
      "mean_token_accuracy": 0.21365399211645125,
      "num_tokens": 101083872.0,
      "step": 58270
    },
    {
      "entropy": 5.8965256690979,
      "epoch": 4.533981715619529,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003107998606328135,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20651938766241074,
      "num_tokens": 101091928.0,
      "step": 58275
    },
    {
      "entropy": 5.81881742477417,
      "epoch": 4.5343707449912465,
      "grad_norm": 1.484375,
      "learning_rate": 0.00031077249784856057,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2182957038283348,
      "num_tokens": 101100740.0,
      "step": 58280
    },
    {
      "entropy": 5.7834474563598635,
      "epoch": 4.534759774362964,
      "grad_norm": 1.390625,
      "learning_rate": 0.00031074513452151683,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21734257489442826,
      "num_tokens": 101108766.0,
      "step": 58285
    },
    {
      "entropy": 5.812333297729492,
      "epoch": 4.535148803734682,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003107177706520975,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22147903442382813,
      "num_tokens": 101118491.0,
      "step": 58290
    },
    {
      "entropy": 5.835483551025391,
      "epoch": 4.5355378331064,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000310690406240718,
      "loss": 4.884,
      "mean_token_accuracy": 0.21727225184440613,
      "num_tokens": 101126550.0,
      "step": 58295
    },
    {
      "entropy": 5.880503463745117,
      "epoch": 4.535926862478117,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031066304128779336,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2176012232899666,
      "num_tokens": 101134812.0,
      "step": 58300
    },
    {
      "entropy": 5.8562592506408695,
      "epoch": 4.536315891849835,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00031063567579373876,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21185854971408843,
      "num_tokens": 101142845.0,
      "step": 58305
    },
    {
      "entropy": 5.884139347076416,
      "epoch": 4.536704921221553,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00031060830975896954,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20779469460248948,
      "num_tokens": 101151540.0,
      "step": 58310
    },
    {
      "entropy": 5.84848985671997,
      "epoch": 4.5370939505932695,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031058094318390085,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21699528396129608,
      "num_tokens": 101159764.0,
      "step": 58315
    },
    {
      "entropy": 5.923594760894775,
      "epoch": 4.537482979964987,
      "grad_norm": 1.46875,
      "learning_rate": 0.00031055357606894806,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21065307408571243,
      "num_tokens": 101168054.0,
      "step": 58320
    },
    {
      "entropy": 5.869775056838989,
      "epoch": 4.537872009336705,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003105262084145263,
      "loss": 4.9883,
      "mean_token_accuracy": 0.20964918434619903,
      "num_tokens": 101176815.0,
      "step": 58325
    },
    {
      "entropy": 5.883952856063843,
      "epoch": 4.538261038708423,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003104988402210509,
      "loss": 4.9029,
      "mean_token_accuracy": 0.21768493205308914,
      "num_tokens": 101185484.0,
      "step": 58330
    },
    {
      "entropy": 5.8145444869995115,
      "epoch": 4.53865006808014,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003104714714889371,
      "loss": 4.8643,
      "mean_token_accuracy": 0.21995907872915268,
      "num_tokens": 101194905.0,
      "step": 58335
    },
    {
      "entropy": 5.899773979187012,
      "epoch": 4.539039097451858,
      "grad_norm": 1.515625,
      "learning_rate": 0.00031044410221860017,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21202611029148102,
      "num_tokens": 101203421.0,
      "step": 58340
    },
    {
      "entropy": 5.83873815536499,
      "epoch": 4.539428126823575,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003104167324104553,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21335465759038924,
      "num_tokens": 101212394.0,
      "step": 58345
    },
    {
      "entropy": 5.903724908828735,
      "epoch": 4.5398171561952925,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031038936206491803,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2073572799563408,
      "num_tokens": 101221824.0,
      "step": 58350
    },
    {
      "entropy": 5.88370156288147,
      "epoch": 4.54020618556701,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003103619911824035,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20669754445552826,
      "num_tokens": 101230585.0,
      "step": 58355
    },
    {
      "entropy": 5.91273865699768,
      "epoch": 4.540595214938728,
      "grad_norm": 1.40625,
      "learning_rate": 0.000310334619763327,
      "loss": 4.9095,
      "mean_token_accuracy": 0.218979874253273,
      "num_tokens": 101239414.0,
      "step": 58360
    },
    {
      "entropy": 5.826918506622315,
      "epoch": 4.540984244310446,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031030724780810385,
      "loss": 4.8559,
      "mean_token_accuracy": 0.21990637481212616,
      "num_tokens": 101248737.0,
      "step": 58365
    },
    {
      "entropy": 5.830650472640992,
      "epoch": 4.541373273682163,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031027987531714944,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21583479195833205,
      "num_tokens": 101258072.0,
      "step": 58370
    },
    {
      "entropy": 5.823217821121216,
      "epoch": 4.541762303053881,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031025250229087915,
      "loss": 4.8503,
      "mean_token_accuracy": 0.21941641718149185,
      "num_tokens": 101266313.0,
      "step": 58375
    },
    {
      "entropy": 5.86814603805542,
      "epoch": 4.542151332425598,
      "grad_norm": 1.4375,
      "learning_rate": 0.00031022512872970823,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21568110287189485,
      "num_tokens": 101276048.0,
      "step": 58380
    },
    {
      "entropy": 5.962202501296997,
      "epoch": 4.5425403617973155,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000310197754634052,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2142393782734871,
      "num_tokens": 101284069.0,
      "step": 58385
    },
    {
      "entropy": 5.776888036727906,
      "epoch": 4.542929391169033,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003101703800043258,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22271857410669327,
      "num_tokens": 101292383.0,
      "step": 58390
    },
    {
      "entropy": 5.8698859214782715,
      "epoch": 4.543318420540751,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00031014300484094516,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22114599943161012,
      "num_tokens": 101300862.0,
      "step": 58395
    },
    {
      "entropy": 5.8967043399810795,
      "epoch": 4.5437074499124686,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003101156291443254,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2017960414290428,
      "num_tokens": 101309583.0,
      "step": 58400
    },
    {
      "entropy": 5.927649784088135,
      "epoch": 4.544096479284186,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003100882529148817,
      "loss": 5.0131,
      "mean_token_accuracy": 0.2143144354224205,
      "num_tokens": 101318568.0,
      "step": 58405
    },
    {
      "entropy": 5.848814487457275,
      "epoch": 4.544485508655903,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00031006087615302976,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21811138838529587,
      "num_tokens": 101326871.0,
      "step": 58410
    },
    {
      "entropy": 5.844626140594483,
      "epoch": 4.544874538027621,
      "grad_norm": 1.421875,
      "learning_rate": 0.00031003349885918473,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2105268955230713,
      "num_tokens": 101334943.0,
      "step": 58415
    },
    {
      "entropy": 5.824334621429443,
      "epoch": 4.5452635673993385,
      "grad_norm": 1.578125,
      "learning_rate": 0.00031000612103376226,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2158386752009392,
      "num_tokens": 101343905.0,
      "step": 58420
    },
    {
      "entropy": 5.823037719726562,
      "epoch": 4.545652596771056,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003099787426771775,
      "loss": 4.8978,
      "mean_token_accuracy": 0.21887197643518447,
      "num_tokens": 101352665.0,
      "step": 58425
    },
    {
      "entropy": 5.871459054946899,
      "epoch": 4.546041626142774,
      "grad_norm": 1.640625,
      "learning_rate": 0.000309951363789846,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2206596091389656,
      "num_tokens": 101360802.0,
      "step": 58430
    },
    {
      "entropy": 5.933388710021973,
      "epoch": 4.5464306555144915,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030992398437218314,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2085282415151596,
      "num_tokens": 101370641.0,
      "step": 58435
    },
    {
      "entropy": 5.924195146560669,
      "epoch": 4.546819684886209,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003098966044246045,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21781107932329177,
      "num_tokens": 101378409.0,
      "step": 58440
    },
    {
      "entropy": 5.879655599594116,
      "epoch": 4.547208714257927,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030986922394752534,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22063378393650054,
      "num_tokens": 101387563.0,
      "step": 58445
    },
    {
      "entropy": 5.816765165328979,
      "epoch": 4.547597743629644,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030984184294136134,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21616096794605255,
      "num_tokens": 101395863.0,
      "step": 58450
    },
    {
      "entropy": 5.831791639328003,
      "epoch": 4.5479867730013614,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030981446140652773,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2230788916349411,
      "num_tokens": 101405409.0,
      "step": 58455
    },
    {
      "entropy": 5.846444511413575,
      "epoch": 4.548375802373079,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00030978707934344014,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22271572798490524,
      "num_tokens": 101412966.0,
      "step": 58460
    },
    {
      "entropy": 5.908572196960449,
      "epoch": 4.548764831744797,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000309759696752514,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21496331095695495,
      "num_tokens": 101421467.0,
      "step": 58465
    },
    {
      "entropy": 5.913785934448242,
      "epoch": 4.5491538611165145,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003097323136341647,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20597818344831467,
      "num_tokens": 101429895.0,
      "step": 58470
    },
    {
      "entropy": 5.818385791778565,
      "epoch": 4.549542890488232,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003097049299888078,
      "loss": 4.8966,
      "mean_token_accuracy": 0.21868670284748076,
      "num_tokens": 101438558.0,
      "step": 58475
    },
    {
      "entropy": 5.898043584823609,
      "epoch": 4.549931919859949,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003096775458168589,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21208728104829788,
      "num_tokens": 101448008.0,
      "step": 58480
    },
    {
      "entropy": 5.8980213642120365,
      "epoch": 4.550320949231667,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030965016111873345,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22232669740915298,
      "num_tokens": 101456341.0,
      "step": 58485
    },
    {
      "entropy": 5.859683179855347,
      "epoch": 4.550709978603384,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030962277589484697,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21831205040216445,
      "num_tokens": 101464511.0,
      "step": 58490
    },
    {
      "entropy": 5.844459676742554,
      "epoch": 4.551099007975102,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030959539014561494,
      "loss": 5.005,
      "mean_token_accuracy": 0.21357445865869523,
      "num_tokens": 101473691.0,
      "step": 58495
    },
    {
      "entropy": 5.9011626720428465,
      "epoch": 4.55148803734682,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030956800387145294,
      "loss": 4.9824,
      "mean_token_accuracy": 0.20818631649017333,
      "num_tokens": 101482450.0,
      "step": 58500
    },
    {
      "entropy": 5.88055272102356,
      "epoch": 4.5518770667185375,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030954061707277647,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21042079478502274,
      "num_tokens": 101490787.0,
      "step": 58505
    },
    {
      "entropy": 5.886212587356567,
      "epoch": 4.552266096090255,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003095132297500011,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2148319289088249,
      "num_tokens": 101499006.0,
      "step": 58510
    },
    {
      "entropy": 5.8386070251464846,
      "epoch": 4.552655125461972,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00030948584190354246,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22053608298301697,
      "num_tokens": 101506893.0,
      "step": 58515
    },
    {
      "entropy": 5.845263576507568,
      "epoch": 4.55304415483369,
      "grad_norm": 1.390625,
      "learning_rate": 0.000309458453533816,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2227756932377815,
      "num_tokens": 101515957.0,
      "step": 58520
    },
    {
      "entropy": 5.9075737476348875,
      "epoch": 4.553433184205407,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030943106464123744,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22161872535943986,
      "num_tokens": 101524835.0,
      "step": 58525
    },
    {
      "entropy": 5.857297801971436,
      "epoch": 4.553822213577125,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030940367522622224,
      "loss": 4.9141,
      "mean_token_accuracy": 0.21938364654779435,
      "num_tokens": 101533991.0,
      "step": 58530
    },
    {
      "entropy": 5.831418991088867,
      "epoch": 4.554211242948843,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000309376285289186,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2176990404725075,
      "num_tokens": 101542523.0,
      "step": 58535
    },
    {
      "entropy": 5.82796277999878,
      "epoch": 4.5546002723205605,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030934889483054433,
      "loss": 4.88,
      "mean_token_accuracy": 0.21640689224004744,
      "num_tokens": 101551030.0,
      "step": 58540
    },
    {
      "entropy": 5.849790096282959,
      "epoch": 4.554989301692277,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003093215038507129,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21625176668167115,
      "num_tokens": 101559585.0,
      "step": 58545
    },
    {
      "entropy": 5.870621824264527,
      "epoch": 4.555378331063995,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003092941123501073,
      "loss": 4.9913,
      "mean_token_accuracy": 0.20585683286190032,
      "num_tokens": 101568463.0,
      "step": 58550
    },
    {
      "entropy": 5.83058123588562,
      "epoch": 4.555767360435713,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030926672032914306,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2217983588576317,
      "num_tokens": 101578634.0,
      "step": 58555
    },
    {
      "entropy": 5.841203069686889,
      "epoch": 4.55615638980743,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003092393277882359,
      "loss": 4.9818,
      "mean_token_accuracy": 0.2127270519733429,
      "num_tokens": 101587824.0,
      "step": 58560
    },
    {
      "entropy": 5.861948871612549,
      "epoch": 4.556545419179148,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030921193472780144,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21374761462211608,
      "num_tokens": 101596282.0,
      "step": 58565
    },
    {
      "entropy": 5.853651762008667,
      "epoch": 4.556934448550866,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003091845411482554,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21788102388381958,
      "num_tokens": 101604948.0,
      "step": 58570
    },
    {
      "entropy": 5.943635511398315,
      "epoch": 4.5573234779225835,
      "grad_norm": 1.578125,
      "learning_rate": 0.00030915714705001323,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20734666138887406,
      "num_tokens": 101612956.0,
      "step": 58575
    },
    {
      "entropy": 5.853701496124268,
      "epoch": 4.557712507294301,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003091297524334908,
      "loss": 4.8699,
      "mean_token_accuracy": 0.21971311122179032,
      "num_tokens": 101621286.0,
      "step": 58580
    },
    {
      "entropy": 5.862766361236572,
      "epoch": 4.558101536666018,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003091023572991036,
      "loss": 4.9373,
      "mean_token_accuracy": 0.20889503955841066,
      "num_tokens": 101630604.0,
      "step": 58585
    },
    {
      "entropy": 5.803920936584473,
      "epoch": 4.558490566037736,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003090749616472675,
      "loss": 4.878,
      "mean_token_accuracy": 0.21854050308465958,
      "num_tokens": 101639565.0,
      "step": 58590
    },
    {
      "entropy": 5.86069769859314,
      "epoch": 4.558879595409453,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003090475654783981,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21249102354049682,
      "num_tokens": 101647865.0,
      "step": 58595
    },
    {
      "entropy": 5.913677930831909,
      "epoch": 4.559268624781171,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003090201687929111,
      "loss": 5.02,
      "mean_token_accuracy": 0.20735414922237397,
      "num_tokens": 101656092.0,
      "step": 58600
    },
    {
      "entropy": 5.9105278015136715,
      "epoch": 4.559657654152889,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00030899277159122213,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21936075538396835,
      "num_tokens": 101664373.0,
      "step": 58605
    },
    {
      "entropy": 5.836515426635742,
      "epoch": 4.560046683524606,
      "grad_norm": 1.5625,
      "learning_rate": 0.00030896537387374703,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22492991238832474,
      "num_tokens": 101672986.0,
      "step": 58610
    },
    {
      "entropy": 5.8848377704620365,
      "epoch": 4.560435712896323,
      "grad_norm": 1.5,
      "learning_rate": 0.00030893797564090136,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20651084035634995,
      "num_tokens": 101682447.0,
      "step": 58615
    },
    {
      "entropy": 5.895258045196533,
      "epoch": 4.560824742268041,
      "grad_norm": 1.375,
      "learning_rate": 0.00030891057689310097,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2162498876452446,
      "num_tokens": 101691511.0,
      "step": 58620
    },
    {
      "entropy": 5.971049690246582,
      "epoch": 4.561213771639759,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003088831776307616,
      "loss": 4.9916,
      "mean_token_accuracy": 0.2121991589665413,
      "num_tokens": 101701148.0,
      "step": 58625
    },
    {
      "entropy": 5.910030889511108,
      "epoch": 4.561602801011476,
      "grad_norm": 1.375,
      "learning_rate": 0.00030885577785429885,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22642982453107835,
      "num_tokens": 101709855.0,
      "step": 58630
    },
    {
      "entropy": 5.785369396209717,
      "epoch": 4.561991830383194,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003088283775641286,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21749065369367598,
      "num_tokens": 101717944.0,
      "step": 58635
    },
    {
      "entropy": 5.833381319046021,
      "epoch": 4.562380859754912,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030880097676066646,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2256601259112358,
      "num_tokens": 101726274.0,
      "step": 58640
    },
    {
      "entropy": 5.956249046325683,
      "epoch": 4.5627698891266295,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003087735754443285,
      "loss": 5.0507,
      "mean_token_accuracy": 0.2046538695693016,
      "num_tokens": 101735321.0,
      "step": 58645
    },
    {
      "entropy": 5.872423791885376,
      "epoch": 4.563158918498346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003087461736155301,
      "loss": 4.927,
      "mean_token_accuracy": 0.2163055270910263,
      "num_tokens": 101744404.0,
      "step": 58650
    },
    {
      "entropy": 5.86171441078186,
      "epoch": 4.563547947870064,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003087187712746873,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2203898772597313,
      "num_tokens": 101752459.0,
      "step": 58655
    },
    {
      "entropy": 5.908582544326782,
      "epoch": 4.563936977241782,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003086913684222159,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22251377254724503,
      "num_tokens": 101760777.0,
      "step": 58660
    },
    {
      "entropy": 5.871193695068359,
      "epoch": 4.564326006613499,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003086639650585315,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21429730504751204,
      "num_tokens": 101769261.0,
      "step": 58665
    },
    {
      "entropy": 5.84931321144104,
      "epoch": 4.564715035985217,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00030863656118405007,
      "loss": 4.941,
      "mean_token_accuracy": 0.20725203901529313,
      "num_tokens": 101777118.0,
      "step": 58670
    },
    {
      "entropy": 5.902321863174438,
      "epoch": 4.565104065356935,
      "grad_norm": 1.53125,
      "learning_rate": 0.00030860915679918724,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21448885947465895,
      "num_tokens": 101785235.0,
      "step": 58675
    },
    {
      "entropy": 5.8392579555511475,
      "epoch": 4.565493094728652,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030858175190435913,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21795457303524018,
      "num_tokens": 101794107.0,
      "step": 58680
    },
    {
      "entropy": 5.907673931121826,
      "epoch": 4.565882124100369,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003085543464999813,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20416492968797684,
      "num_tokens": 101802985.0,
      "step": 58685
    },
    {
      "entropy": 5.903494453430175,
      "epoch": 4.566271153472087,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003085269405864697,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22158308029174806,
      "num_tokens": 101811441.0,
      "step": 58690
    },
    {
      "entropy": 5.856951332092285,
      "epoch": 4.566660182843805,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003084995341642401,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21690763980150224,
      "num_tokens": 101820506.0,
      "step": 58695
    },
    {
      "entropy": 5.885605716705323,
      "epoch": 4.567049212215522,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003084721272337084,
      "loss": 4.896,
      "mean_token_accuracy": 0.21254095882177354,
      "num_tokens": 101828854.0,
      "step": 58700
    },
    {
      "entropy": 5.923724222183227,
      "epoch": 4.56743824158724,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003084447197952904,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20872527360916138,
      "num_tokens": 101837891.0,
      "step": 58705
    },
    {
      "entropy": 5.939090156555176,
      "epoch": 4.567827270958958,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030841731184940205,
      "loss": 4.9468,
      "mean_token_accuracy": 0.20979532301425935,
      "num_tokens": 101847000.0,
      "step": 58710
    },
    {
      "entropy": 5.933237743377686,
      "epoch": 4.568216300330675,
      "grad_norm": 1.5,
      "learning_rate": 0.00030838990339645915,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21373697072267533,
      "num_tokens": 101855029.0,
      "step": 58715
    },
    {
      "entropy": 5.869204568862915,
      "epoch": 4.568605329702392,
      "grad_norm": 1.59375,
      "learning_rate": 0.00030836249443687774,
      "loss": 4.9967,
      "mean_token_accuracy": 0.22325721830129625,
      "num_tokens": 101864025.0,
      "step": 58720
    },
    {
      "entropy": 5.8278093338012695,
      "epoch": 4.56899435907411,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030833508497107344,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22656758278608322,
      "num_tokens": 101873051.0,
      "step": 58725
    },
    {
      "entropy": 5.887861919403076,
      "epoch": 4.569383388445828,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003083076749994623,
      "loss": 4.8708,
      "mean_token_accuracy": 0.2163386896252632,
      "num_tokens": 101880745.0,
      "step": 58730
    },
    {
      "entropy": 5.941604471206665,
      "epoch": 4.569772417817545,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003082802645224603,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2195088595151901,
      "num_tokens": 101889026.0,
      "step": 58735
    },
    {
      "entropy": 5.922458124160767,
      "epoch": 4.570161447189263,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003082528535404832,
      "loss": 5.051,
      "mean_token_accuracy": 0.20273622274398803,
      "num_tokens": 101898511.0,
      "step": 58740
    },
    {
      "entropy": 5.926009750366211,
      "epoch": 4.57055047656098,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030822544205394686,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19863147288560867,
      "num_tokens": 101907416.0,
      "step": 58745
    },
    {
      "entropy": 5.93998155593872,
      "epoch": 4.570939505932698,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030819803006326745,
      "loss": 4.926,
      "mean_token_accuracy": 0.21361775547266007,
      "num_tokens": 101915927.0,
      "step": 58750
    },
    {
      "entropy": 5.9243086814880375,
      "epoch": 4.571328535304415,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030817061756886077,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20720042437314987,
      "num_tokens": 101924289.0,
      "step": 58755
    },
    {
      "entropy": 5.889846467971802,
      "epoch": 4.571717564676133,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030814320457114283,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22068875432014465,
      "num_tokens": 101932718.0,
      "step": 58760
    },
    {
      "entropy": 5.846908950805664,
      "epoch": 4.572106594047851,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030811579107052937,
      "loss": 4.951,
      "mean_token_accuracy": 0.21074649095535278,
      "num_tokens": 101941263.0,
      "step": 58765
    },
    {
      "entropy": 5.858176612854004,
      "epoch": 4.572495623419568,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030808837706743666,
      "loss": 4.9833,
      "mean_token_accuracy": 0.20837116092443467,
      "num_tokens": 101950444.0,
      "step": 58770
    },
    {
      "entropy": 5.958907604217529,
      "epoch": 4.572884652791286,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003080609625622803,
      "loss": 4.999,
      "mean_token_accuracy": 0.21396692991256713,
      "num_tokens": 101958783.0,
      "step": 58775
    },
    {
      "entropy": 5.916350412368774,
      "epoch": 4.573273682163004,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030803354755547653,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2147844448685646,
      "num_tokens": 101967487.0,
      "step": 58780
    },
    {
      "entropy": 5.944613361358643,
      "epoch": 4.573662711534721,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003080061320474412,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2104666382074356,
      "num_tokens": 101976271.0,
      "step": 58785
    },
    {
      "entropy": 5.881683444976806,
      "epoch": 4.574051740906438,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003079787160385905,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2149528980255127,
      "num_tokens": 101985468.0,
      "step": 58790
    },
    {
      "entropy": 5.913859033584595,
      "epoch": 4.574440770278156,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003079512995293402,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21033175736665727,
      "num_tokens": 101995729.0,
      "step": 58795
    },
    {
      "entropy": 5.978483486175537,
      "epoch": 4.574829799649874,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003079238825201064,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2190748929977417,
      "num_tokens": 102005180.0,
      "step": 58800
    },
    {
      "entropy": 5.938677263259888,
      "epoch": 4.575218829021591,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000307896465011305,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2222595825791359,
      "num_tokens": 102013559.0,
      "step": 58805
    },
    {
      "entropy": 5.864674234390259,
      "epoch": 4.575607858393309,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030786904700335217,
      "loss": 4.987,
      "mean_token_accuracy": 0.21454003751277922,
      "num_tokens": 102022108.0,
      "step": 58810
    },
    {
      "entropy": 5.82264666557312,
      "epoch": 4.575996887765026,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003078416284966639,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22037604302167893,
      "num_tokens": 102030658.0,
      "step": 58815
    },
    {
      "entropy": 5.842310857772827,
      "epoch": 4.576385917136744,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00030781420949165607,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21530407369136811,
      "num_tokens": 102038621.0,
      "step": 58820
    },
    {
      "entropy": 5.894441032409668,
      "epoch": 4.576774946508461,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030778678998874494,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2160593420267105,
      "num_tokens": 102047389.0,
      "step": 58825
    },
    {
      "entropy": 5.812646675109863,
      "epoch": 4.577163975880179,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003077593699883465,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2259732738137245,
      "num_tokens": 102056114.0,
      "step": 58830
    },
    {
      "entropy": 5.846849966049194,
      "epoch": 4.577553005251897,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030773194949087665,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22081035524606704,
      "num_tokens": 102063989.0,
      "step": 58835
    },
    {
      "entropy": 5.849257135391236,
      "epoch": 4.577942034623614,
      "grad_norm": 1.375,
      "learning_rate": 0.0003077045284967516,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21161885559558868,
      "num_tokens": 102073348.0,
      "step": 58840
    },
    {
      "entropy": 5.860109615325928,
      "epoch": 4.578331063995332,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003076771070063874,
      "loss": 4.9244,
      "mean_token_accuracy": 0.21194933354854584,
      "num_tokens": 102081909.0,
      "step": 58845
    },
    {
      "entropy": 5.805100584030152,
      "epoch": 4.578720093367049,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003076496850202001,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2220703884959221,
      "num_tokens": 102090577.0,
      "step": 58850
    },
    {
      "entropy": 5.826902675628662,
      "epoch": 4.579109122738767,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030762226253860584,
      "loss": 4.95,
      "mean_token_accuracy": 0.21301692724227905,
      "num_tokens": 102099415.0,
      "step": 58855
    },
    {
      "entropy": 5.890769100189209,
      "epoch": 4.579498152110484,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003075948395620206,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21587304174900054,
      "num_tokens": 102107631.0,
      "step": 58860
    },
    {
      "entropy": 5.88065185546875,
      "epoch": 4.579887181482202,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030756741609086057,
      "loss": 4.9041,
      "mean_token_accuracy": 0.21585996448993683,
      "num_tokens": 102115710.0,
      "step": 58865
    },
    {
      "entropy": 5.871667098999024,
      "epoch": 4.58027621085392,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030753999212554183,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21258566826581954,
      "num_tokens": 102124170.0,
      "step": 58870
    },
    {
      "entropy": 5.869603967666626,
      "epoch": 4.580665240225637,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003075125676664806,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22603644877672197,
      "num_tokens": 102133129.0,
      "step": 58875
    },
    {
      "entropy": 5.867083501815796,
      "epoch": 4.581054269597354,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003074851427140928,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21744220554828644,
      "num_tokens": 102141976.0,
      "step": 58880
    },
    {
      "entropy": 5.849081516265869,
      "epoch": 4.581443298969072,
      "grad_norm": 1.5,
      "learning_rate": 0.00030745771726879467,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21621257811784744,
      "num_tokens": 102150268.0,
      "step": 58885
    },
    {
      "entropy": 5.852190589904785,
      "epoch": 4.58183232834079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003074302913310024,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21267012804746627,
      "num_tokens": 102160689.0,
      "step": 58890
    },
    {
      "entropy": 5.850932693481445,
      "epoch": 4.582221357712507,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000307402864901132,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21642930954694747,
      "num_tokens": 102168975.0,
      "step": 58895
    },
    {
      "entropy": 5.905531549453736,
      "epoch": 4.582610387084225,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003073754379795998,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2070916086435318,
      "num_tokens": 102177802.0,
      "step": 58900
    },
    {
      "entropy": 5.887684869766235,
      "epoch": 4.582999416455943,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030734801056682176,
      "loss": 4.9494,
      "mean_token_accuracy": 0.20766782909631729,
      "num_tokens": 102186735.0,
      "step": 58905
    },
    {
      "entropy": 5.869570350646972,
      "epoch": 4.58338844582766,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030732058266321415,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21863806694746019,
      "num_tokens": 102194420.0,
      "step": 58910
    },
    {
      "entropy": 5.833641624450683,
      "epoch": 4.583777475199377,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003072931542691932,
      "loss": 4.8719,
      "mean_token_accuracy": 0.221488294005394,
      "num_tokens": 102202242.0,
      "step": 58915
    },
    {
      "entropy": 5.907086658477783,
      "epoch": 4.584166504571095,
      "grad_norm": 1.390625,
      "learning_rate": 0.000307265725385175,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2124616503715515,
      "num_tokens": 102210551.0,
      "step": 58920
    },
    {
      "entropy": 5.916331768035889,
      "epoch": 4.584555533942813,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003072382960115758,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21628174930810928,
      "num_tokens": 102219451.0,
      "step": 58925
    },
    {
      "entropy": 5.9182024002075195,
      "epoch": 4.58494456331453,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030721086614881183,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22097554802894592,
      "num_tokens": 102228349.0,
      "step": 58930
    },
    {
      "entropy": 5.898470830917359,
      "epoch": 4.585333592686248,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003071834357972991,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2110875979065895,
      "num_tokens": 102236781.0,
      "step": 58935
    },
    {
      "entropy": 5.851024436950683,
      "epoch": 4.585722622057966,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000307156004957454,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2165671855211258,
      "num_tokens": 102244802.0,
      "step": 58940
    },
    {
      "entropy": 5.8414661407470705,
      "epoch": 4.5861116514296825,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003071285736296928,
      "loss": 4.8531,
      "mean_token_accuracy": 0.2158154532313347,
      "num_tokens": 102253404.0,
      "step": 58945
    },
    {
      "entropy": 5.887237644195556,
      "epoch": 4.5865006808014,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003071011418144315,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2183360457420349,
      "num_tokens": 102261688.0,
      "step": 58950
    },
    {
      "entropy": 5.893262767791748,
      "epoch": 4.586889710173118,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030707370951208655,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20909756273031235,
      "num_tokens": 102270294.0,
      "step": 58955
    },
    {
      "entropy": 5.955899906158447,
      "epoch": 4.587278739544836,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003070462767230741,
      "loss": 5.064,
      "mean_token_accuracy": 0.20277651399374008,
      "num_tokens": 102279475.0,
      "step": 58960
    },
    {
      "entropy": 5.856693506240845,
      "epoch": 4.587667768916553,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003070188434478104,
      "loss": 4.9122,
      "mean_token_accuracy": 0.21719721108675002,
      "num_tokens": 102288630.0,
      "step": 58965
    },
    {
      "entropy": 5.88286771774292,
      "epoch": 4.588056798288271,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003069914096867117,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22452909946441652,
      "num_tokens": 102297701.0,
      "step": 58970
    },
    {
      "entropy": 5.867024230957031,
      "epoch": 4.588445827659989,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003069639754401942,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2168638899922371,
      "num_tokens": 102306665.0,
      "step": 58975
    },
    {
      "entropy": 5.7701804637908936,
      "epoch": 4.588834857031706,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003069365407086744,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22183576971292496,
      "num_tokens": 102315411.0,
      "step": 58980
    },
    {
      "entropy": 5.880029582977295,
      "epoch": 4.589223886403423,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030690910549256836,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2167092427611351,
      "num_tokens": 102324218.0,
      "step": 58985
    },
    {
      "entropy": 5.880406522750855,
      "epoch": 4.589612915775141,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003068816697922925,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21933809518814087,
      "num_tokens": 102332604.0,
      "step": 58990
    },
    {
      "entropy": 5.913788557052612,
      "epoch": 4.5900019451468586,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003068542336082629,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2153335005044937,
      "num_tokens": 102341092.0,
      "step": 58995
    },
    {
      "entropy": 5.900165462493897,
      "epoch": 4.590390974518576,
      "grad_norm": 1.375,
      "learning_rate": 0.00030682679694089615,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22295089662075043,
      "num_tokens": 102349084.0,
      "step": 59000
    },
    {
      "entropy": 5.873656558990478,
      "epoch": 4.590780003890294,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003067993597906083,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21173078417778016,
      "num_tokens": 102357678.0,
      "step": 59005
    },
    {
      "entropy": 5.870819330215454,
      "epoch": 4.591169033262012,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003067719221578158,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2165520265698433,
      "num_tokens": 102365645.0,
      "step": 59010
    },
    {
      "entropy": 5.86516375541687,
      "epoch": 4.5915580626337285,
      "grad_norm": 1.40625,
      "learning_rate": 0.000306744484042935,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21901046633720397,
      "num_tokens": 102373853.0,
      "step": 59015
    },
    {
      "entropy": 5.796182632446289,
      "epoch": 4.591947092005446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003067170454463822,
      "loss": 4.8729,
      "mean_token_accuracy": 0.21695795208215712,
      "num_tokens": 102383027.0,
      "step": 59020
    },
    {
      "entropy": 5.837488508224487,
      "epoch": 4.592336121377164,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030668960636857364,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21978914141654968,
      "num_tokens": 102392114.0,
      "step": 59025
    },
    {
      "entropy": 5.764454889297485,
      "epoch": 4.5927251507488815,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003066621668099258,
      "loss": 4.7553,
      "mean_token_accuracy": 0.2269095927476883,
      "num_tokens": 102400294.0,
      "step": 59030
    },
    {
      "entropy": 5.894205570220947,
      "epoch": 4.593114180120599,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030663472677085494,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2098499581217766,
      "num_tokens": 102408777.0,
      "step": 59035
    },
    {
      "entropy": 5.8207268714904785,
      "epoch": 4.593503209492317,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030660728625177746,
      "loss": 4.866,
      "mean_token_accuracy": 0.2188152179121971,
      "num_tokens": 102417451.0,
      "step": 59040
    },
    {
      "entropy": 5.816992902755738,
      "epoch": 4.593892238864035,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003065798452531097,
      "loss": 4.8858,
      "mean_token_accuracy": 0.2179120421409607,
      "num_tokens": 102425697.0,
      "step": 59045
    },
    {
      "entropy": 5.842831754684449,
      "epoch": 4.5942812682357514,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030655240377526813,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21867853552103042,
      "num_tokens": 102434112.0,
      "step": 59050
    },
    {
      "entropy": 5.877425193786621,
      "epoch": 4.594670297607469,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000306524961818669,
      "loss": 4.9872,
      "mean_token_accuracy": 0.20932792872190475,
      "num_tokens": 102442649.0,
      "step": 59055
    },
    {
      "entropy": 5.93512864112854,
      "epoch": 4.595059326979187,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003064975193837287,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21120850145816802,
      "num_tokens": 102451181.0,
      "step": 59060
    },
    {
      "entropy": 5.836515712738037,
      "epoch": 4.5954483563509045,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003064700764708638,
      "loss": 4.8069,
      "mean_token_accuracy": 0.22665199190378188,
      "num_tokens": 102459453.0,
      "step": 59065
    },
    {
      "entropy": 5.785112762451172,
      "epoch": 4.595837385722622,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030644263308049046,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22357717603445054,
      "num_tokens": 102468677.0,
      "step": 59070
    },
    {
      "entropy": 5.891535949707031,
      "epoch": 4.59622641509434,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003064151892130252,
      "loss": 4.9619,
      "mean_token_accuracy": 0.20587226748466492,
      "num_tokens": 102477821.0,
      "step": 59075
    },
    {
      "entropy": 5.836864376068116,
      "epoch": 4.596615444466057,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003063877448688846,
      "loss": 4.9236,
      "mean_token_accuracy": 0.21815500557422637,
      "num_tokens": 102486054.0,
      "step": 59080
    },
    {
      "entropy": 5.833251476287842,
      "epoch": 4.597004473837774,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003063603000484848,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22125181555747986,
      "num_tokens": 102495587.0,
      "step": 59085
    },
    {
      "entropy": 5.954069471359253,
      "epoch": 4.597393503209492,
      "grad_norm": 1.53125,
      "learning_rate": 0.00030633285475224244,
      "loss": 4.9958,
      "mean_token_accuracy": 0.2236172154545784,
      "num_tokens": 102504335.0,
      "step": 59090
    },
    {
      "entropy": 5.741567325592041,
      "epoch": 4.59778253258121,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030630540898057384,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2208258554339409,
      "num_tokens": 102512550.0,
      "step": 59095
    },
    {
      "entropy": 5.834440755844116,
      "epoch": 4.5981715619529275,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030627796273389546,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2116864249110222,
      "num_tokens": 102521003.0,
      "step": 59100
    },
    {
      "entropy": 5.870408916473389,
      "epoch": 4.598560591324645,
      "grad_norm": 1.5,
      "learning_rate": 0.00030625051601262384,
      "loss": 4.922,
      "mean_token_accuracy": 0.2147119790315628,
      "num_tokens": 102529857.0,
      "step": 59105
    },
    {
      "entropy": 5.9364570617675785,
      "epoch": 4.598949620696363,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003062230688171753,
      "loss": 4.9912,
      "mean_token_accuracy": 0.20949615389108658,
      "num_tokens": 102539965.0,
      "step": 59110
    },
    {
      "entropy": 5.874440145492554,
      "epoch": 4.599338650068081,
      "grad_norm": 1.375,
      "learning_rate": 0.0003061956211479665,
      "loss": 4.8847,
      "mean_token_accuracy": 0.2263513445854187,
      "num_tokens": 102548538.0,
      "step": 59115
    },
    {
      "entropy": 5.848804712295532,
      "epoch": 4.599727679439797,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003061681730054137,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21431005150079727,
      "num_tokens": 102556242.0,
      "step": 59120
    },
    {
      "entropy": 5.741528558731079,
      "epoch": 4.600116708811515,
      "grad_norm": 1.375,
      "learning_rate": 0.0003061407243899335,
      "loss": 4.8302,
      "mean_token_accuracy": 0.2299483373761177,
      "num_tokens": 102565014.0,
      "step": 59125
    },
    {
      "entropy": 5.899964475631714,
      "epoch": 4.600505738183233,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030611327530194243,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2143373653292656,
      "num_tokens": 102573487.0,
      "step": 59130
    },
    {
      "entropy": 5.935739803314209,
      "epoch": 4.6008947675549505,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003060858257418569,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21495617181062698,
      "num_tokens": 102582541.0,
      "step": 59135
    },
    {
      "entropy": 5.947629022598266,
      "epoch": 4.601283796926668,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003060583757100934,
      "loss": 5.0381,
      "mean_token_accuracy": 0.2080756828188896,
      "num_tokens": 102591486.0,
      "step": 59140
    },
    {
      "entropy": 5.864669990539551,
      "epoch": 4.601672826298385,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003060309252070685,
      "loss": 4.9496,
      "mean_token_accuracy": 0.20636269450187683,
      "num_tokens": 102599544.0,
      "step": 59145
    },
    {
      "entropy": 5.93262243270874,
      "epoch": 4.602061855670103,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003060034742331987,
      "loss": 4.994,
      "mean_token_accuracy": 0.20986106097698212,
      "num_tokens": 102608303.0,
      "step": 59150
    },
    {
      "entropy": 5.930326128005982,
      "epoch": 4.60245088504182,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003059760227889006,
      "loss": 4.9877,
      "mean_token_accuracy": 0.214130038022995,
      "num_tokens": 102616541.0,
      "step": 59155
    },
    {
      "entropy": 5.808360147476196,
      "epoch": 4.602839914413538,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003059485708745906,
      "loss": 4.876,
      "mean_token_accuracy": 0.21648477911949157,
      "num_tokens": 102625125.0,
      "step": 59160
    },
    {
      "entropy": 5.8889275074005125,
      "epoch": 4.603228943785256,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003059211184906853,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21307700425386428,
      "num_tokens": 102633826.0,
      "step": 59165
    },
    {
      "entropy": 5.8947694301605225,
      "epoch": 4.6036179731569735,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00030589366563760125,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21525832861661912,
      "num_tokens": 102642155.0,
      "step": 59170
    },
    {
      "entropy": 5.786149692535401,
      "epoch": 4.604007002528691,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030586621231575494,
      "loss": 4.8128,
      "mean_token_accuracy": 0.22814300805330276,
      "num_tokens": 102650884.0,
      "step": 59175
    },
    {
      "entropy": 5.873524522781372,
      "epoch": 4.604396031900409,
      "grad_norm": 1.375,
      "learning_rate": 0.000305838758525563,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22748121619224548,
      "num_tokens": 102659828.0,
      "step": 59180
    },
    {
      "entropy": 5.928497219085694,
      "epoch": 4.604785061272126,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003058113042674421,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20461203902959824,
      "num_tokens": 102668192.0,
      "step": 59185
    },
    {
      "entropy": 5.843136310577393,
      "epoch": 4.605174090643843,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030578384954180866,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21692436635494233,
      "num_tokens": 102677772.0,
      "step": 59190
    },
    {
      "entropy": 5.889706373214722,
      "epoch": 4.605563120015561,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003057563943490792,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2171032801270485,
      "num_tokens": 102687216.0,
      "step": 59195
    },
    {
      "entropy": 5.972531414031982,
      "epoch": 4.605952149387279,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00030572893868967044,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21264603585004807,
      "num_tokens": 102695458.0,
      "step": 59200
    },
    {
      "entropy": 5.896920442581177,
      "epoch": 4.6063411787589965,
      "grad_norm": 1.46875,
      "learning_rate": 0.000305701482563999,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21012648195028305,
      "num_tokens": 102704472.0,
      "step": 59205
    },
    {
      "entropy": 5.843772840499878,
      "epoch": 4.606730208130714,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003056740259724814,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22023798376321793,
      "num_tokens": 102713267.0,
      "step": 59210
    },
    {
      "entropy": 5.798350381851196,
      "epoch": 4.607119237502431,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003056465689155342,
      "loss": 4.7345,
      "mean_token_accuracy": 0.23019516468048096,
      "num_tokens": 102721961.0,
      "step": 59215
    },
    {
      "entropy": 5.87944598197937,
      "epoch": 4.607508266874149,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030561911139357424,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21837567389011384,
      "num_tokens": 102731079.0,
      "step": 59220
    },
    {
      "entropy": 5.8833984375,
      "epoch": 4.607897296245866,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030559165340701797,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21630579233169556,
      "num_tokens": 102739480.0,
      "step": 59225
    },
    {
      "entropy": 5.875898599624634,
      "epoch": 4.608286325617584,
      "grad_norm": 1.375,
      "learning_rate": 0.000305564194956282,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21165303885936737,
      "num_tokens": 102748190.0,
      "step": 59230
    },
    {
      "entropy": 5.846229124069214,
      "epoch": 4.608675354989302,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030553673604178305,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22021922767162322,
      "num_tokens": 102757254.0,
      "step": 59235
    },
    {
      "entropy": 5.832178068161011,
      "epoch": 4.6090643843610195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030550927666393767,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21254366189241408,
      "num_tokens": 102766550.0,
      "step": 59240
    },
    {
      "entropy": 5.906805515289307,
      "epoch": 4.609453413732737,
      "grad_norm": 1.578125,
      "learning_rate": 0.00030548181682316263,
      "loss": 4.9848,
      "mean_token_accuracy": 0.20864721536636352,
      "num_tokens": 102774671.0,
      "step": 59245
    },
    {
      "entropy": 5.903750991821289,
      "epoch": 4.609842443104454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030545435651987453,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21501947194337845,
      "num_tokens": 102783640.0,
      "step": 59250
    },
    {
      "entropy": 5.844103813171387,
      "epoch": 4.610231472476172,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003054268957544901,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2175658717751503,
      "num_tokens": 102792912.0,
      "step": 59255
    },
    {
      "entropy": 5.917260932922363,
      "epoch": 4.610620501847889,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030539943452742585,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21579167544841765,
      "num_tokens": 102801364.0,
      "step": 59260
    },
    {
      "entropy": 5.872364139556884,
      "epoch": 4.611009531219607,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030537197283909875,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21178903728723525,
      "num_tokens": 102810275.0,
      "step": 59265
    },
    {
      "entropy": 5.828303813934326,
      "epoch": 4.611398560591325,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030534451068992506,
      "loss": 4.98,
      "mean_token_accuracy": 0.20569159388542174,
      "num_tokens": 102818724.0,
      "step": 59270
    },
    {
      "entropy": 5.841092729568482,
      "epoch": 4.6117875899630425,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030531704808032183,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21710328310728072,
      "num_tokens": 102826776.0,
      "step": 59275
    },
    {
      "entropy": 5.861355590820312,
      "epoch": 4.612176619334759,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003052895850107057,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2189507082104683,
      "num_tokens": 102835089.0,
      "step": 59280
    },
    {
      "entropy": 5.964153909683228,
      "epoch": 4.612565648706477,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030526212148149326,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2129327431321144,
      "num_tokens": 102843877.0,
      "step": 59285
    },
    {
      "entropy": 5.879856061935425,
      "epoch": 4.612954678078195,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003052346574931013,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2112683594226837,
      "num_tokens": 102852354.0,
      "step": 59290
    },
    {
      "entropy": 5.893152236938477,
      "epoch": 4.613343707449912,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003052071930459465,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21701678037643432,
      "num_tokens": 102860618.0,
      "step": 59295
    },
    {
      "entropy": 5.835953426361084,
      "epoch": 4.61373273682163,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003051797281404457,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21912211179733276,
      "num_tokens": 102868969.0,
      "step": 59300
    },
    {
      "entropy": 5.8892443656921385,
      "epoch": 4.614121766193348,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030515226277701553,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2104501709342003,
      "num_tokens": 102877367.0,
      "step": 59305
    },
    {
      "entropy": 5.896457147598267,
      "epoch": 4.6145107955650655,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003051247969560728,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2165166512131691,
      "num_tokens": 102885606.0,
      "step": 59310
    },
    {
      "entropy": 5.921051454544068,
      "epoch": 4.614899824936783,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003050973306780342,
      "loss": 5.0093,
      "mean_token_accuracy": 0.20715464949607848,
      "num_tokens": 102894110.0,
      "step": 59315
    },
    {
      "entropy": 5.910792589187622,
      "epoch": 4.6152888543085,
      "grad_norm": 1.375,
      "learning_rate": 0.00030506986394331656,
      "loss": 5.004,
      "mean_token_accuracy": 0.21027127504348755,
      "num_tokens": 102903584.0,
      "step": 59320
    },
    {
      "entropy": 5.885996627807617,
      "epoch": 4.615677883680218,
      "grad_norm": 1.375,
      "learning_rate": 0.00030504239675233655,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21193112283945084,
      "num_tokens": 102911651.0,
      "step": 59325
    },
    {
      "entropy": 5.8978838443756105,
      "epoch": 4.616066913051935,
      "grad_norm": 1.53125,
      "learning_rate": 0.00030501492910551094,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21014205664396285,
      "num_tokens": 102920108.0,
      "step": 59330
    },
    {
      "entropy": 5.863290214538575,
      "epoch": 4.616455942423653,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030498746100325656,
      "loss": 4.8932,
      "mean_token_accuracy": 0.21925996243953705,
      "num_tokens": 102928749.0,
      "step": 59335
    },
    {
      "entropy": 5.881446838378906,
      "epoch": 4.616844971795371,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003049599924459903,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21116265207529067,
      "num_tokens": 102936568.0,
      "step": 59340
    },
    {
      "entropy": 5.928707838058472,
      "epoch": 4.6172340011670885,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003049325234341287,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2097833499312401,
      "num_tokens": 102945390.0,
      "step": 59345
    },
    {
      "entropy": 5.972243738174439,
      "epoch": 4.617623030538805,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030490505396808874,
      "loss": 5.0463,
      "mean_token_accuracy": 0.2131298691034317,
      "num_tokens": 102954251.0,
      "step": 59350
    },
    {
      "entropy": 5.926200914382934,
      "epoch": 4.618012059910523,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003048775840482873,
      "loss": 4.9908,
      "mean_token_accuracy": 0.20811918675899505,
      "num_tokens": 102963142.0,
      "step": 59355
    },
    {
      "entropy": 5.8684648990631105,
      "epoch": 4.618401089282241,
      "grad_norm": 1.453125,
      "learning_rate": 0.000304850113675141,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2231759712100029,
      "num_tokens": 102971750.0,
      "step": 59360
    },
    {
      "entropy": 5.945009756088257,
      "epoch": 4.618790118653958,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003048226428490666,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21635287255048752,
      "num_tokens": 102980283.0,
      "step": 59365
    },
    {
      "entropy": 5.867023563385009,
      "epoch": 4.619179148025676,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030479517157048125,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21406133323907853,
      "num_tokens": 102988763.0,
      "step": 59370
    },
    {
      "entropy": 5.818175411224365,
      "epoch": 4.619568177397394,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003047676998398015,
      "loss": 4.881,
      "mean_token_accuracy": 0.22367904633283614,
      "num_tokens": 102997105.0,
      "step": 59375
    },
    {
      "entropy": 5.867564868927002,
      "epoch": 4.6199572067691115,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030474022765744434,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21720581352710724,
      "num_tokens": 103005400.0,
      "step": 59380
    },
    {
      "entropy": 5.909547948837281,
      "epoch": 4.620346236140828,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030471275502382656,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21065307706594466,
      "num_tokens": 103013432.0,
      "step": 59385
    },
    {
      "entropy": 5.89455018043518,
      "epoch": 4.620735265512546,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000304685281939365,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2205229550600052,
      "num_tokens": 103022113.0,
      "step": 59390
    },
    {
      "entropy": 5.7813561916351315,
      "epoch": 4.621124294884264,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003046578084044765,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2236254781484604,
      "num_tokens": 103030417.0,
      "step": 59395
    },
    {
      "entropy": 5.811742782592773,
      "epoch": 4.621513324255981,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000304630334419578,
      "loss": 4.9838,
      "mean_token_accuracy": 0.20927574038505553,
      "num_tokens": 103039652.0,
      "step": 59400
    },
    {
      "entropy": 5.814849853515625,
      "epoch": 4.621902353627699,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003046028599850863,
      "loss": 4.884,
      "mean_token_accuracy": 0.2254044905304909,
      "num_tokens": 103048035.0,
      "step": 59405
    },
    {
      "entropy": 5.85345230102539,
      "epoch": 4.622291382999417,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003045753851014184,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22276206016540528,
      "num_tokens": 103056548.0,
      "step": 59410
    },
    {
      "entropy": 6.000073099136353,
      "epoch": 4.622680412371134,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030454790976899105,
      "loss": 5.0872,
      "mean_token_accuracy": 0.20840395092964173,
      "num_tokens": 103065016.0,
      "step": 59415
    },
    {
      "entropy": 5.926330184936523,
      "epoch": 4.623069441742851,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003045204339882212,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21499303728342056,
      "num_tokens": 103073421.0,
      "step": 59420
    },
    {
      "entropy": 5.849353313446045,
      "epoch": 4.623458471114569,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003044929577595257,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2161320775747299,
      "num_tokens": 103082874.0,
      "step": 59425
    },
    {
      "entropy": 5.887893772125244,
      "epoch": 4.623847500486287,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030446548108332164,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21998862773180008,
      "num_tokens": 103091698.0,
      "step": 59430
    },
    {
      "entropy": 5.908675765991211,
      "epoch": 4.624236529858004,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030443800396002575,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21434952914714814,
      "num_tokens": 103100841.0,
      "step": 59435
    },
    {
      "entropy": 5.831303501129151,
      "epoch": 4.624625559229722,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003044105263900549,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2217041254043579,
      "num_tokens": 103109589.0,
      "step": 59440
    },
    {
      "entropy": 5.861938190460205,
      "epoch": 4.62501458860144,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003043830483738262,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21454712599515915,
      "num_tokens": 103118454.0,
      "step": 59445
    },
    {
      "entropy": 5.839293813705444,
      "epoch": 4.6254036179731575,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003043555699117565,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21116624325513839,
      "num_tokens": 103127144.0,
      "step": 59450
    },
    {
      "entropy": 5.800412178039551,
      "epoch": 4.625792647344874,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030432809100426286,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2210872069001198,
      "num_tokens": 103136130.0,
      "step": 59455
    },
    {
      "entropy": 5.8798499584198,
      "epoch": 4.626181676716592,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003043006116517619,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21400593668222428,
      "num_tokens": 103144147.0,
      "step": 59460
    },
    {
      "entropy": 5.9121373176574705,
      "epoch": 4.62657070608831,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030427313185467104,
      "loss": 4.9794,
      "mean_token_accuracy": 0.20629697293043137,
      "num_tokens": 103153486.0,
      "step": 59465
    },
    {
      "entropy": 5.82157416343689,
      "epoch": 4.626959735460027,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003042456516134068,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2262937232851982,
      "num_tokens": 103162236.0,
      "step": 59470
    },
    {
      "entropy": 5.822551965713501,
      "epoch": 4.627348764831745,
      "grad_norm": 1.5625,
      "learning_rate": 0.00030421817092838644,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22314469665288925,
      "num_tokens": 103169729.0,
      "step": 59475
    },
    {
      "entropy": 5.805689430236816,
      "epoch": 4.627737794203462,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003041906898000268,
      "loss": 4.9809,
      "mean_token_accuracy": 0.213431479036808,
      "num_tokens": 103178429.0,
      "step": 59480
    },
    {
      "entropy": 5.9601325511932375,
      "epoch": 4.62812682357518,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003041632082287449,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20514727830886842,
      "num_tokens": 103186676.0,
      "step": 59485
    },
    {
      "entropy": 5.950713348388672,
      "epoch": 4.628515852946897,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003041357262149577,
      "loss": 4.9942,
      "mean_token_accuracy": 0.21832729279994964,
      "num_tokens": 103195620.0,
      "step": 59490
    },
    {
      "entropy": 5.888232088088989,
      "epoch": 4.628904882318615,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003041082437590824,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2219439461827278,
      "num_tokens": 103204212.0,
      "step": 59495
    },
    {
      "entropy": 5.752078723907471,
      "epoch": 4.629293911690333,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030408076086153575,
      "loss": 4.8507,
      "mean_token_accuracy": 0.21823230981826783,
      "num_tokens": 103212593.0,
      "step": 59500
    },
    {
      "entropy": 5.889556884765625,
      "epoch": 4.62968294106205,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003040532775227347,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21064398288726807,
      "num_tokens": 103220869.0,
      "step": 59505
    },
    {
      "entropy": 5.886502647399903,
      "epoch": 4.630071970433768,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030402579374309656,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2119506523013115,
      "num_tokens": 103229247.0,
      "step": 59510
    },
    {
      "entropy": 5.859588670730591,
      "epoch": 4.630460999805486,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030399830952303815,
      "loss": 4.916,
      "mean_token_accuracy": 0.21957392245531082,
      "num_tokens": 103238190.0,
      "step": 59515
    },
    {
      "entropy": 5.866318035125732,
      "epoch": 4.630850029177203,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030397082486297657,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21058034896850586,
      "num_tokens": 103247284.0,
      "step": 59520
    },
    {
      "entropy": 6.020378589630127,
      "epoch": 4.63123905854892,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003039433397633288,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20766572058200836,
      "num_tokens": 103255015.0,
      "step": 59525
    },
    {
      "entropy": 5.910651063919067,
      "epoch": 4.631628087920638,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030391585422451196,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22275150716304778,
      "num_tokens": 103264129.0,
      "step": 59530
    },
    {
      "entropy": 5.86256594657898,
      "epoch": 4.632017117292356,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030388836824694307,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20690262615680693,
      "num_tokens": 103273268.0,
      "step": 59535
    },
    {
      "entropy": 5.925336599349976,
      "epoch": 4.632406146664073,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003038608818310391,
      "loss": 4.996,
      "mean_token_accuracy": 0.212839475274086,
      "num_tokens": 103281627.0,
      "step": 59540
    },
    {
      "entropy": 5.851765584945679,
      "epoch": 4.632795176035791,
      "grad_norm": 1.34375,
      "learning_rate": 0.00030383339497721726,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21303091943264008,
      "num_tokens": 103290122.0,
      "step": 59545
    },
    {
      "entropy": 5.929911708831787,
      "epoch": 4.633184205407508,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003038059076858944,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2119682639837265,
      "num_tokens": 103298791.0,
      "step": 59550
    },
    {
      "entropy": 5.972873973846435,
      "epoch": 4.633573234779226,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003037784199574879,
      "loss": 5.0341,
      "mean_token_accuracy": 0.20924840569496156,
      "num_tokens": 103307724.0,
      "step": 59555
    },
    {
      "entropy": 5.8803058624267575,
      "epoch": 4.633962264150943,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030375093179241464,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21784429550170897,
      "num_tokens": 103316371.0,
      "step": 59560
    },
    {
      "entropy": 5.891194152832031,
      "epoch": 4.634351293522661,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030372344319109173,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2247490480542183,
      "num_tokens": 103325152.0,
      "step": 59565
    },
    {
      "entropy": 5.885935068130493,
      "epoch": 4.634740322894379,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003036959541539363,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2193233400583267,
      "num_tokens": 103333424.0,
      "step": 59570
    },
    {
      "entropy": 5.900493431091308,
      "epoch": 4.635129352266096,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030366846468136543,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2165011942386627,
      "num_tokens": 103341913.0,
      "step": 59575
    },
    {
      "entropy": 5.946571397781372,
      "epoch": 4.635518381637814,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003036409747737963,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21058931201696396,
      "num_tokens": 103351629.0,
      "step": 59580
    },
    {
      "entropy": 5.913263416290283,
      "epoch": 4.635907411009531,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030361348443164583,
      "loss": 4.945,
      "mean_token_accuracy": 0.21255773454904556,
      "num_tokens": 103359616.0,
      "step": 59585
    },
    {
      "entropy": 5.86098804473877,
      "epoch": 4.6362964403812486,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003035859936553314,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22009864151477815,
      "num_tokens": 103368579.0,
      "step": 59590
    },
    {
      "entropy": 5.898982954025269,
      "epoch": 4.636685469752966,
      "grad_norm": 1.515625,
      "learning_rate": 0.00030355850244527,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21819094717502593,
      "num_tokens": 103377608.0,
      "step": 59595
    },
    {
      "entropy": 5.943006277084351,
      "epoch": 4.637074499124684,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003035310108018787,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21568890064954757,
      "num_tokens": 103386672.0,
      "step": 59600
    },
    {
      "entropy": 5.874470329284668,
      "epoch": 4.637463528496402,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003035035187255748,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21962076127529145,
      "num_tokens": 103395447.0,
      "step": 59605
    },
    {
      "entropy": 5.894077253341675,
      "epoch": 4.637852557868119,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030347602621677534,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2055039197206497,
      "num_tokens": 103404518.0,
      "step": 59610
    },
    {
      "entropy": 5.85761194229126,
      "epoch": 4.638241587239836,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003034485332758975,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2192122980952263,
      "num_tokens": 103412910.0,
      "step": 59615
    },
    {
      "entropy": 5.900339794158936,
      "epoch": 4.638630616611554,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030342103990335846,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21875500828027725,
      "num_tokens": 103420753.0,
      "step": 59620
    },
    {
      "entropy": 5.816905355453491,
      "epoch": 4.6390196459832715,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030339354609957545,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2137668773531914,
      "num_tokens": 103429320.0,
      "step": 59625
    },
    {
      "entropy": 5.76717586517334,
      "epoch": 4.639408675354989,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030336605186496546,
      "loss": 4.7888,
      "mean_token_accuracy": 0.22623176127672195,
      "num_tokens": 103437458.0,
      "step": 59630
    },
    {
      "entropy": 5.865260934829712,
      "epoch": 4.639797704726707,
      "grad_norm": 1.484375,
      "learning_rate": 0.00030333855719994585,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21199748665094376,
      "num_tokens": 103445296.0,
      "step": 59635
    },
    {
      "entropy": 5.808672857284546,
      "epoch": 4.640186734098425,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003033110621049336,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2220495641231537,
      "num_tokens": 103453974.0,
      "step": 59640
    },
    {
      "entropy": 5.748056364059448,
      "epoch": 4.640575763470142,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030328356658034623,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21672202795743942,
      "num_tokens": 103462706.0,
      "step": 59645
    },
    {
      "entropy": 5.810504055023193,
      "epoch": 4.64096479284186,
      "grad_norm": 1.5,
      "learning_rate": 0.00030325607062660066,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22149689942598344,
      "num_tokens": 103471316.0,
      "step": 59650
    },
    {
      "entropy": 5.904333305358887,
      "epoch": 4.641353822213577,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003032285742441143,
      "loss": 4.879,
      "mean_token_accuracy": 0.2216821238398552,
      "num_tokens": 103480198.0,
      "step": 59655
    },
    {
      "entropy": 5.882955265045166,
      "epoch": 4.6417428515852945,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003032010774333041,
      "loss": 4.9226,
      "mean_token_accuracy": 0.21834102720022203,
      "num_tokens": 103489888.0,
      "step": 59660
    },
    {
      "entropy": 5.879792261123657,
      "epoch": 4.642131880957012,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030317358019458753,
      "loss": 5.002,
      "mean_token_accuracy": 0.20893792808055878,
      "num_tokens": 103498722.0,
      "step": 59665
    },
    {
      "entropy": 5.922170925140381,
      "epoch": 4.64252091032873,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003031460825283817,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20604243278503417,
      "num_tokens": 103507790.0,
      "step": 59670
    },
    {
      "entropy": 5.853675270080567,
      "epoch": 4.642909939700448,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00030311858443510383,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21847673505544662,
      "num_tokens": 103516452.0,
      "step": 59675
    },
    {
      "entropy": 5.857326364517212,
      "epoch": 4.643298969072165,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003030910859151712,
      "loss": 4.896,
      "mean_token_accuracy": 0.22030945420265197,
      "num_tokens": 103524330.0,
      "step": 59680
    },
    {
      "entropy": 5.888177299499512,
      "epoch": 4.643687998443882,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030306358696900114,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21369877755641936,
      "num_tokens": 103533202.0,
      "step": 59685
    },
    {
      "entropy": 5.909259223937989,
      "epoch": 4.6440770278156,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030303608759701077,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21647596955299378,
      "num_tokens": 103542036.0,
      "step": 59690
    },
    {
      "entropy": 5.859219217300415,
      "epoch": 4.6444660571873175,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003030085877996174,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21241712868213652,
      "num_tokens": 103551187.0,
      "step": 59695
    },
    {
      "entropy": 5.833377981185913,
      "epoch": 4.644855086559035,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030298108757723823,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21887923926115035,
      "num_tokens": 103559112.0,
      "step": 59700
    },
    {
      "entropy": 5.849860572814942,
      "epoch": 4.645244115930753,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003029535869302906,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22058966159820556,
      "num_tokens": 103566897.0,
      "step": 59705
    },
    {
      "entropy": 5.885828447341919,
      "epoch": 4.645633145302471,
      "grad_norm": 1.375,
      "learning_rate": 0.00030292608585919185,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2080690383911133,
      "num_tokens": 103576638.0,
      "step": 59710
    },
    {
      "entropy": 5.858506679534912,
      "epoch": 4.646022174674188,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030289858436435916,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21105299293994903,
      "num_tokens": 103586086.0,
      "step": 59715
    },
    {
      "entropy": 5.856296491622925,
      "epoch": 4.646411204045905,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030287108244620993,
      "loss": 4.9144,
      "mean_token_accuracy": 0.21640200167894363,
      "num_tokens": 103594686.0,
      "step": 59720
    },
    {
      "entropy": 5.859839391708374,
      "epoch": 4.646800233417623,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030284358010516135,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21841688454151154,
      "num_tokens": 103603657.0,
      "step": 59725
    },
    {
      "entropy": 5.868320512771606,
      "epoch": 4.6471892627893405,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003028160773416307,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21776659339666365,
      "num_tokens": 103612431.0,
      "step": 59730
    },
    {
      "entropy": 5.774597644805908,
      "epoch": 4.647578292161058,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030278857415603534,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22534109801054,
      "num_tokens": 103621552.0,
      "step": 59735
    },
    {
      "entropy": 5.827893161773682,
      "epoch": 4.647967321532776,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030276107054879263,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21908195018768312,
      "num_tokens": 103629751.0,
      "step": 59740
    },
    {
      "entropy": 5.848880863189697,
      "epoch": 4.648356350904494,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030273356652031993,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2170646533370018,
      "num_tokens": 103638759.0,
      "step": 59745
    },
    {
      "entropy": 5.895599365234375,
      "epoch": 4.64874538027621,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003027060620710344,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21032555103302003,
      "num_tokens": 103647234.0,
      "step": 59750
    },
    {
      "entropy": 5.914074182510376,
      "epoch": 4.649134409647928,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030267855720135343,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22329861372709275,
      "num_tokens": 103655494.0,
      "step": 59755
    },
    {
      "entropy": 5.839802598953247,
      "epoch": 4.649523439019646,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003026510519116946,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2151003360748291,
      "num_tokens": 103664704.0,
      "step": 59760
    },
    {
      "entropy": 5.908935022354126,
      "epoch": 4.6499124683913635,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000302623546202475,
      "loss": 5.005,
      "mean_token_accuracy": 0.21299708038568496,
      "num_tokens": 103673324.0,
      "step": 59765
    },
    {
      "entropy": 5.890066051483155,
      "epoch": 4.650301497763081,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003025960400741119,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21138421595096588,
      "num_tokens": 103682047.0,
      "step": 59770
    },
    {
      "entropy": 5.842965936660766,
      "epoch": 4.650690527134799,
      "grad_norm": 1.421875,
      "learning_rate": 0.000302568533527023,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21679498106241227,
      "num_tokens": 103689988.0,
      "step": 59775
    },
    {
      "entropy": 5.801271057128906,
      "epoch": 4.651079556506517,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003025410265616254,
      "loss": 4.9212,
      "mean_token_accuracy": 0.21567331552505492,
      "num_tokens": 103698760.0,
      "step": 59780
    },
    {
      "entropy": 5.840758562088013,
      "epoch": 4.651468585878233,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003025135191783366,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22327022701501847,
      "num_tokens": 103706984.0,
      "step": 59785
    },
    {
      "entropy": 5.895548534393311,
      "epoch": 4.651857615249951,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030248601137757394,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21195153445005416,
      "num_tokens": 103715555.0,
      "step": 59790
    },
    {
      "entropy": 5.93115234375,
      "epoch": 4.652246644621669,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003024585031597547,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22227016985416412,
      "num_tokens": 103723859.0,
      "step": 59795
    },
    {
      "entropy": 5.823552131652832,
      "epoch": 4.6526356739933865,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003024309945252965,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22826244682073593,
      "num_tokens": 103732978.0,
      "step": 59800
    },
    {
      "entropy": 5.88531756401062,
      "epoch": 4.653024703365104,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00030240348547461653,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21223229318857192,
      "num_tokens": 103741148.0,
      "step": 59805
    },
    {
      "entropy": 5.87237958908081,
      "epoch": 4.653413732736822,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030237597600813235,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21004888713359832,
      "num_tokens": 103749997.0,
      "step": 59810
    },
    {
      "entropy": 5.845847129821777,
      "epoch": 4.653802762108539,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003023484661262613,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21808658093214034,
      "num_tokens": 103758177.0,
      "step": 59815
    },
    {
      "entropy": 5.876063966751099,
      "epoch": 4.654191791480256,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003023209558294208,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22040603309869766,
      "num_tokens": 103766319.0,
      "step": 59820
    },
    {
      "entropy": 5.918611764907837,
      "epoch": 4.654580820851974,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00030229344511802824,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2104274719953537,
      "num_tokens": 103776294.0,
      "step": 59825
    },
    {
      "entropy": 5.940850305557251,
      "epoch": 4.654969850223692,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030226593399250114,
      "loss": 5.001,
      "mean_token_accuracy": 0.20811597406864166,
      "num_tokens": 103785456.0,
      "step": 59830
    },
    {
      "entropy": 5.89050407409668,
      "epoch": 4.6553588795954095,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030223842245325687,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22222172021865844,
      "num_tokens": 103793944.0,
      "step": 59835
    },
    {
      "entropy": 5.8901543617248535,
      "epoch": 4.655747908967127,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030221091050071293,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21411545127630233,
      "num_tokens": 103803394.0,
      "step": 59840
    },
    {
      "entropy": 5.874813079833984,
      "epoch": 4.656136938338845,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030218339813528665,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22128608971834182,
      "num_tokens": 103812649.0,
      "step": 59845
    },
    {
      "entropy": 5.839220762252808,
      "epoch": 4.656525967710563,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003021558853573957,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21868777126073838,
      "num_tokens": 103820774.0,
      "step": 59850
    },
    {
      "entropy": 5.794312763214111,
      "epoch": 4.656914997082279,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003021283721674573,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22236402779817582,
      "num_tokens": 103828996.0,
      "step": 59855
    },
    {
      "entropy": 5.8107929706573485,
      "epoch": 4.657304026453997,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003021008585658891,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22290112376213073,
      "num_tokens": 103836833.0,
      "step": 59860
    },
    {
      "entropy": 5.824314928054809,
      "epoch": 4.657693055825715,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00030207334455310846,
      "loss": 4.949,
      "mean_token_accuracy": 0.21553080081939696,
      "num_tokens": 103845433.0,
      "step": 59865
    },
    {
      "entropy": 5.868732643127442,
      "epoch": 4.6580820851974325,
      "grad_norm": 1.5,
      "learning_rate": 0.0003020458301295329,
      "loss": 4.9743,
      "mean_token_accuracy": 0.21500066220760344,
      "num_tokens": 103853613.0,
      "step": 59870
    },
    {
      "entropy": 5.92473578453064,
      "epoch": 4.65847111456915,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030201831529557994,
      "loss": 5.0254,
      "mean_token_accuracy": 0.20314482748508453,
      "num_tokens": 103861773.0,
      "step": 59875
    },
    {
      "entropy": 5.962489461898803,
      "epoch": 4.658860143940868,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003019908000516671,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20798769295215608,
      "num_tokens": 103870432.0,
      "step": 59880
    },
    {
      "entropy": 5.951323175430298,
      "epoch": 4.659249173312585,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003019632843982117,
      "loss": 4.918,
      "mean_token_accuracy": 0.2205438047647476,
      "num_tokens": 103879169.0,
      "step": 59885
    },
    {
      "entropy": 5.848997926712036,
      "epoch": 4.659638202684302,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030193576833563153,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22047965079545975,
      "num_tokens": 103887441.0,
      "step": 59890
    },
    {
      "entropy": 5.871990203857422,
      "epoch": 4.66002723205602,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003019082518643438,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2075105756521225,
      "num_tokens": 103895986.0,
      "step": 59895
    },
    {
      "entropy": 5.78333101272583,
      "epoch": 4.660416261427738,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00030188073498476633,
      "loss": 4.8346,
      "mean_token_accuracy": 0.21951931565999985,
      "num_tokens": 103904623.0,
      "step": 59900
    },
    {
      "entropy": 5.832269382476807,
      "epoch": 4.6608052907994555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003018532176973163,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21452962458133698,
      "num_tokens": 103913553.0,
      "step": 59905
    },
    {
      "entropy": 5.887924242019653,
      "epoch": 4.661194320171173,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00030182570000241154,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21840872019529342,
      "num_tokens": 103922199.0,
      "step": 59910
    },
    {
      "entropy": 5.8273358821868895,
      "epoch": 4.661583349542891,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003017981819004696,
      "loss": 4.8968,
      "mean_token_accuracy": 0.218083156645298,
      "num_tokens": 103930607.0,
      "step": 59915
    },
    {
      "entropy": 5.941971492767334,
      "epoch": 4.661972378914608,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003017706633919078,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21708817034959793,
      "num_tokens": 103938710.0,
      "step": 59920
    },
    {
      "entropy": 5.796772909164429,
      "epoch": 4.662361408286325,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003017431444771437,
      "loss": 4.8996,
      "mean_token_accuracy": 0.21933989226818085,
      "num_tokens": 103947666.0,
      "step": 59925
    },
    {
      "entropy": 5.855378532409668,
      "epoch": 4.662750437658043,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00030171562515659504,
      "loss": 4.904,
      "mean_token_accuracy": 0.21859604865312576,
      "num_tokens": 103955611.0,
      "step": 59930
    },
    {
      "entropy": 5.831744956970215,
      "epoch": 4.663139467029761,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003016881054306793,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22010291814804078,
      "num_tokens": 103964418.0,
      "step": 59935
    },
    {
      "entropy": 5.906565189361572,
      "epoch": 4.6635284964014785,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003016605852998139,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2218442365527153,
      "num_tokens": 103973089.0,
      "step": 59940
    },
    {
      "entropy": 5.946289825439453,
      "epoch": 4.663917525773196,
      "grad_norm": 1.390625,
      "learning_rate": 0.00030163306476441675,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21418512016534805,
      "num_tokens": 103981508.0,
      "step": 59945
    },
    {
      "entropy": 5.847855567932129,
      "epoch": 4.664306555144913,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003016055438249051,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22297352701425552,
      "num_tokens": 103990493.0,
      "step": 59950
    },
    {
      "entropy": 5.881373739242553,
      "epoch": 4.664695584516631,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003015780224816968,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21325118690729142,
      "num_tokens": 103998816.0,
      "step": 59955
    },
    {
      "entropy": 5.901629066467285,
      "epoch": 4.665084613888348,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003015505007352093,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2207338735461235,
      "num_tokens": 104007243.0,
      "step": 59960
    },
    {
      "entropy": 5.947420215606689,
      "epoch": 4.665473643260066,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030152297858586016,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21228749454021453,
      "num_tokens": 104016457.0,
      "step": 59965
    },
    {
      "entropy": 5.92833948135376,
      "epoch": 4.665862672631784,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030149545603406694,
      "loss": 4.9223,
      "mean_token_accuracy": 0.21832142025232315,
      "num_tokens": 104024599.0,
      "step": 59970
    },
    {
      "entropy": 5.837838459014892,
      "epoch": 4.6662517020035015,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003014679330802474,
      "loss": 4.8524,
      "mean_token_accuracy": 0.21554409712553024,
      "num_tokens": 104033142.0,
      "step": 59975
    },
    {
      "entropy": 5.723985815048218,
      "epoch": 4.666640731375219,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030144040972481916,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22220911532640458,
      "num_tokens": 104041843.0,
      "step": 59980
    },
    {
      "entropy": 5.850904226303101,
      "epoch": 4.667029760746937,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030141288596819977,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22250752598047258,
      "num_tokens": 104050085.0,
      "step": 59985
    },
    {
      "entropy": 5.791503143310547,
      "epoch": 4.667418790118654,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030138536181080683,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21352809071540832,
      "num_tokens": 104058199.0,
      "step": 59990
    },
    {
      "entropy": 5.93329005241394,
      "epoch": 4.667807819490371,
      "grad_norm": 1.53125,
      "learning_rate": 0.00030135783725305815,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2120254695415497,
      "num_tokens": 104066829.0,
      "step": 59995
    },
    {
      "entropy": 5.912546014785766,
      "epoch": 4.668196848862089,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003013303122953712,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2081070736050606,
      "num_tokens": 104075253.0,
      "step": 60000
    },
    {
      "epoch": 4.668196848862089,
      "eval_entropy": 5.548210504768672,
      "eval_loss": 5.039699077606201,
      "eval_mean_token_accuracy": 0.22027882334558407,
      "eval_num_tokens": 104075253.0,
      "eval_runtime": 21.9223,
      "eval_samples_per_second": 1895.695,
      "eval_steps_per_second": 236.973,
      "step": 60000
    },
    {
      "entropy": 5.8073512554168705,
      "epoch": 4.668585878233807,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003013027869381635,
      "loss": 4.8123,
      "mean_token_accuracy": 0.22235798686742783,
      "num_tokens": 104083591.0,
      "step": 60005
    },
    {
      "entropy": 5.875702381134033,
      "epoch": 4.6689749076055245,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000301275261181853,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21182302832603456,
      "num_tokens": 104092390.0,
      "step": 60010
    },
    {
      "entropy": 5.885676527023316,
      "epoch": 4.669363936977242,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003012477350268572,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21051882654428483,
      "num_tokens": 104100720.0,
      "step": 60015
    },
    {
      "entropy": 5.825570344924927,
      "epoch": 4.669752966348959,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030122020847359385,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2248877614736557,
      "num_tokens": 104109067.0,
      "step": 60020
    },
    {
      "entropy": 5.911455535888672,
      "epoch": 4.670141995720677,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003011926815224805,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20712292492389678,
      "num_tokens": 104118615.0,
      "step": 60025
    },
    {
      "entropy": 5.8828950881958,
      "epoch": 4.670531025092394,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00030116515417393495,
      "loss": 4.9948,
      "mean_token_accuracy": 0.20596571564674376,
      "num_tokens": 104126491.0,
      "step": 60030
    },
    {
      "entropy": 5.882551479339599,
      "epoch": 4.670920054464112,
      "grad_norm": 1.4375,
      "learning_rate": 0.00030113762642837487,
      "loss": 4.9067,
      "mean_token_accuracy": 0.21406277716159822,
      "num_tokens": 104135116.0,
      "step": 60035
    },
    {
      "entropy": 5.839541912078857,
      "epoch": 4.67130908383583,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030111009828621786,
      "loss": 4.8869,
      "mean_token_accuracy": 0.23054458051919938,
      "num_tokens": 104143581.0,
      "step": 60040
    },
    {
      "entropy": 5.85173716545105,
      "epoch": 4.6716981132075475,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003010825697478817,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21227352023124696,
      "num_tokens": 104152425.0,
      "step": 60045
    },
    {
      "entropy": 5.832494926452637,
      "epoch": 4.672087142579265,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000301055040813784,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2069002076983452,
      "num_tokens": 104160947.0,
      "step": 60050
    },
    {
      "entropy": 5.869179582595825,
      "epoch": 4.672476171950982,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030102751148434255,
      "loss": 4.8987,
      "mean_token_accuracy": 0.208969384431839,
      "num_tokens": 104170012.0,
      "step": 60055
    },
    {
      "entropy": 5.961619329452515,
      "epoch": 4.6728652013227,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000300999981759975,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21854636669158936,
      "num_tokens": 104179331.0,
      "step": 60060
    },
    {
      "entropy": 5.918593215942383,
      "epoch": 4.673254230694417,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003009724516410992,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21699300557374954,
      "num_tokens": 104188411.0,
      "step": 60065
    },
    {
      "entropy": 5.825111150741577,
      "epoch": 4.673643260066135,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003009449211281328,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22157978713512422,
      "num_tokens": 104196805.0,
      "step": 60070
    },
    {
      "entropy": 5.943280172348023,
      "epoch": 4.674032289437853,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003009173902214934,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20735466927289964,
      "num_tokens": 104204877.0,
      "step": 60075
    },
    {
      "entropy": 5.87695574760437,
      "epoch": 4.6744213188095705,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003008898589215989,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2152958557009697,
      "num_tokens": 104213263.0,
      "step": 60080
    },
    {
      "entropy": 5.893117046356201,
      "epoch": 4.674810348181287,
      "grad_norm": 1.421875,
      "learning_rate": 0.00030086232722886704,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22272181510925293,
      "num_tokens": 104221646.0,
      "step": 60085
    },
    {
      "entropy": 5.938587808609009,
      "epoch": 4.675199377553005,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030083479514371545,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2050196722149849,
      "num_tokens": 104230691.0,
      "step": 60090
    },
    {
      "entropy": 5.905352163314819,
      "epoch": 4.675588406924723,
      "grad_norm": 1.375,
      "learning_rate": 0.00030080726266656205,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21153706014156343,
      "num_tokens": 104239718.0,
      "step": 60095
    },
    {
      "entropy": 5.952654886245727,
      "epoch": 4.67597743629644,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003007797297978245,
      "loss": 5.14,
      "mean_token_accuracy": 0.20426367223262787,
      "num_tokens": 104249312.0,
      "step": 60100
    },
    {
      "entropy": 5.9503316402435305,
      "epoch": 4.676366465668158,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003007521965379206,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21429689973592758,
      "num_tokens": 104258499.0,
      "step": 60105
    },
    {
      "entropy": 5.914454317092895,
      "epoch": 4.676755495039876,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000300724662887268,
      "loss": 4.891,
      "mean_token_accuracy": 0.22587245404720308,
      "num_tokens": 104266870.0,
      "step": 60110
    },
    {
      "entropy": 5.904830837249756,
      "epoch": 4.6771445244115935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030069712884628477,
      "loss": 4.9227,
      "mean_token_accuracy": 0.21844279021024704,
      "num_tokens": 104275633.0,
      "step": 60115
    },
    {
      "entropy": 5.868472003936768,
      "epoch": 4.67753355378331,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003006695944153883,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21609628200531006,
      "num_tokens": 104283115.0,
      "step": 60120
    },
    {
      "entropy": 5.834436655044556,
      "epoch": 4.677922583155028,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003006420595949967,
      "loss": 4.8739,
      "mean_token_accuracy": 0.21145954430103303,
      "num_tokens": 104291948.0,
      "step": 60125
    },
    {
      "entropy": 5.863365221023559,
      "epoch": 4.678311612526746,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030061452438552767,
      "loss": 4.8993,
      "mean_token_accuracy": 0.21550666242837907,
      "num_tokens": 104300617.0,
      "step": 60130
    },
    {
      "entropy": 5.924694585800171,
      "epoch": 4.678700641898463,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000300586988787399,
      "loss": 4.981,
      "mean_token_accuracy": 0.21485516726970671,
      "num_tokens": 104310159.0,
      "step": 60135
    },
    {
      "entropy": 5.884992265701294,
      "epoch": 4.679089671270181,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00030055945280102844,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2185334011912346,
      "num_tokens": 104319116.0,
      "step": 60140
    },
    {
      "entropy": 5.815712308883667,
      "epoch": 4.679478700641899,
      "grad_norm": 1.21875,
      "learning_rate": 0.00030053191642683393,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21108656376600266,
      "num_tokens": 104328702.0,
      "step": 60145
    },
    {
      "entropy": 5.915642166137696,
      "epoch": 4.679867730013616,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003005043796652332,
      "loss": 4.9228,
      "mean_token_accuracy": 0.220018470287323,
      "num_tokens": 104337962.0,
      "step": 60150
    },
    {
      "entropy": 5.943922519683838,
      "epoch": 4.680256759385333,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00030047684251664417,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21268443614244462,
      "num_tokens": 104347018.0,
      "step": 60155
    },
    {
      "entropy": 5.738276100158691,
      "epoch": 4.680645788757051,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003004493049814845,
      "loss": 4.8402,
      "mean_token_accuracy": 0.21763478070497513,
      "num_tokens": 104355741.0,
      "step": 60160
    },
    {
      "entropy": 5.8864391326904295,
      "epoch": 4.681034818128769,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030042176706017215,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2185603216290474,
      "num_tokens": 104364003.0,
      "step": 60165
    },
    {
      "entropy": 5.9038321018219,
      "epoch": 4.681423847500486,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030039422875312495,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22179186791181565,
      "num_tokens": 104372611.0,
      "step": 60170
    },
    {
      "entropy": 5.821865177154541,
      "epoch": 4.681812876872204,
      "grad_norm": 1.46875,
      "learning_rate": 0.00030036669006076075,
      "loss": 4.819,
      "mean_token_accuracy": 0.22382308542728424,
      "num_tokens": 104380372.0,
      "step": 60175
    },
    {
      "entropy": 5.827333641052246,
      "epoch": 4.682201906243922,
      "grad_norm": 1.359375,
      "learning_rate": 0.00030033915098349746,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2279147431254387,
      "num_tokens": 104388874.0,
      "step": 60180
    },
    {
      "entropy": 5.8256481170654295,
      "epoch": 4.682590935615639,
      "grad_norm": 1.375,
      "learning_rate": 0.0003003116115217529,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21989302933216096,
      "num_tokens": 104397446.0,
      "step": 60185
    },
    {
      "entropy": 5.918598175048828,
      "epoch": 4.682979964987356,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030028407167594486,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21145930290222167,
      "num_tokens": 104406242.0,
      "step": 60190
    },
    {
      "entropy": 5.849095964431763,
      "epoch": 4.683368994359074,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003002565314464913,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2099806025624275,
      "num_tokens": 104415110.0,
      "step": 60195
    },
    {
      "entropy": 5.855722332000733,
      "epoch": 4.683758023730792,
      "grad_norm": 1.40625,
      "learning_rate": 0.00030022899083381006,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2184421554207802,
      "num_tokens": 104423903.0,
      "step": 60200
    },
    {
      "entropy": 5.841151905059815,
      "epoch": 4.684147053102509,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000300201449838319,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2301768481731415,
      "num_tokens": 104432273.0,
      "step": 60205
    },
    {
      "entropy": 5.92961277961731,
      "epoch": 4.684536082474227,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003001739084604361,
      "loss": 5.0054,
      "mean_token_accuracy": 0.2135171189904213,
      "num_tokens": 104441579.0,
      "step": 60210
    },
    {
      "entropy": 5.8498612403869625,
      "epoch": 4.684925111845945,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00030014636670057913,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22165418416261673,
      "num_tokens": 104450501.0,
      "step": 60215
    },
    {
      "entropy": 5.8545387268066404,
      "epoch": 4.6853141412176615,
      "grad_norm": 1.375,
      "learning_rate": 0.0003001188245591662,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21163256615400314,
      "num_tokens": 104458422.0,
      "step": 60220
    },
    {
      "entropy": 5.9001185417175295,
      "epoch": 4.685703170589379,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003000912820366151,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22065775394439696,
      "num_tokens": 104466567.0,
      "step": 60225
    },
    {
      "entropy": 5.862979888916016,
      "epoch": 4.686092199961097,
      "grad_norm": 1.546875,
      "learning_rate": 0.00030006373913334366,
      "loss": 4.945,
      "mean_token_accuracy": 0.21588771939277648,
      "num_tokens": 104474720.0,
      "step": 60230
    },
    {
      "entropy": 5.919897222518921,
      "epoch": 4.686481229332815,
      "grad_norm": 1.5,
      "learning_rate": 0.0003000361958497699,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21423268169164658,
      "num_tokens": 104483759.0,
      "step": 60235
    },
    {
      "entropy": 5.828099393844605,
      "epoch": 4.686870258704532,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003000086521863116,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2241627112030983,
      "num_tokens": 104492807.0,
      "step": 60240
    },
    {
      "entropy": 5.856683921813965,
      "epoch": 4.68725928807625,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000299981108143387,
      "loss": 5.0408,
      "mean_token_accuracy": 0.20561813563108444,
      "num_tokens": 104500776.0,
      "step": 60245
    },
    {
      "entropy": 5.905717182159424,
      "epoch": 4.687648317447968,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029995356372141367,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21272501200437546,
      "num_tokens": 104509686.0,
      "step": 60250
    },
    {
      "entropy": 5.953820657730103,
      "epoch": 4.6880373468196845,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002999260189208098,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21152054220438005,
      "num_tokens": 104518021.0,
      "step": 60255
    },
    {
      "entropy": 5.888211107254028,
      "epoch": 4.688426376191402,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029989847374199333,
      "loss": 4.941,
      "mean_token_accuracy": 0.22456591874361037,
      "num_tokens": 104527786.0,
      "step": 60260
    },
    {
      "entropy": 5.855815982818603,
      "epoch": 4.68881540556312,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029987092818538217,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2200021490454674,
      "num_tokens": 104536591.0,
      "step": 60265
    },
    {
      "entropy": 5.918423986434936,
      "epoch": 4.689204434934838,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029984338225139405,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2107619822025299,
      "num_tokens": 104544836.0,
      "step": 60270
    },
    {
      "entropy": 5.874045085906983,
      "epoch": 4.689593464306555,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002998158359404473,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2178718402981758,
      "num_tokens": 104554190.0,
      "step": 60275
    },
    {
      "entropy": 5.819086599349975,
      "epoch": 4.689982493678273,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002997882892529597,
      "loss": 4.9219,
      "mean_token_accuracy": 0.21803646683692932,
      "num_tokens": 104562263.0,
      "step": 60280
    },
    {
      "entropy": 5.8563450336456295,
      "epoch": 4.69037152304999,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002997607421893493,
      "loss": 5.015,
      "mean_token_accuracy": 0.2000938430428505,
      "num_tokens": 104570839.0,
      "step": 60285
    },
    {
      "entropy": 5.84790267944336,
      "epoch": 4.6907605524217075,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000299733194750034,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22848785072565078,
      "num_tokens": 104579455.0,
      "step": 60290
    },
    {
      "entropy": 5.936744451522827,
      "epoch": 4.691149581793425,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002997056469354319,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2031017690896988,
      "num_tokens": 104588583.0,
      "step": 60295
    },
    {
      "entropy": 5.9411803722381595,
      "epoch": 4.691538611165143,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002996780987459609,
      "loss": 4.984,
      "mean_token_accuracy": 0.20797375589609146,
      "num_tokens": 104596976.0,
      "step": 60300
    },
    {
      "entropy": 5.933562898635865,
      "epoch": 4.691927640536861,
      "grad_norm": 1.375,
      "learning_rate": 0.0002996505501820391,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21112194508314133,
      "num_tokens": 104606063.0,
      "step": 60305
    },
    {
      "entropy": 5.808654403686523,
      "epoch": 4.692316669908578,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002996230012440843,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2211732879281044,
      "num_tokens": 104614388.0,
      "step": 60310
    },
    {
      "entropy": 5.861419153213501,
      "epoch": 4.692705699280296,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029959545193251473,
      "loss": 4.9447,
      "mean_token_accuracy": 0.21734175682067872,
      "num_tokens": 104622704.0,
      "step": 60315
    },
    {
      "entropy": 5.970580816268921,
      "epoch": 4.693094728652014,
      "grad_norm": 1.5625,
      "learning_rate": 0.00029956790224774834,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21356920897960663,
      "num_tokens": 104630796.0,
      "step": 60320
    },
    {
      "entropy": 5.870044803619384,
      "epoch": 4.6934837580237305,
      "grad_norm": 1.375,
      "learning_rate": 0.00029954035219020305,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22100441455841063,
      "num_tokens": 104638984.0,
      "step": 60325
    },
    {
      "entropy": 5.874324369430542,
      "epoch": 4.693872787395448,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029951280176029703,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21020882576704025,
      "num_tokens": 104647583.0,
      "step": 60330
    },
    {
      "entropy": 5.865918684005737,
      "epoch": 4.694261816767166,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00029948525095844825,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21297832876443862,
      "num_tokens": 104655664.0,
      "step": 60335
    },
    {
      "entropy": 5.8987079620361325,
      "epoch": 4.694650846138884,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002994576997850748,
      "loss": 4.8992,
      "mean_token_accuracy": 0.21728408485651016,
      "num_tokens": 104664081.0,
      "step": 60340
    },
    {
      "entropy": 5.811340999603272,
      "epoch": 4.695039875510601,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002994301482405946,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22392659336328508,
      "num_tokens": 104672751.0,
      "step": 60345
    },
    {
      "entropy": 5.898687076568604,
      "epoch": 4.695428904882318,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002994025963254259,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21644185483455658,
      "num_tokens": 104681957.0,
      "step": 60350
    },
    {
      "entropy": 5.830059385299682,
      "epoch": 4.695817934254036,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002993750440399866,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22976184338331224,
      "num_tokens": 104690563.0,
      "step": 60355
    },
    {
      "entropy": 5.882805109024048,
      "epoch": 4.6962069636257535,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002993474913846948,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21103659719228746,
      "num_tokens": 104699228.0,
      "step": 60360
    },
    {
      "entropy": 5.903840780258179,
      "epoch": 4.696595992997471,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002993199383599685,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21501701921224595,
      "num_tokens": 104708942.0,
      "step": 60365
    },
    {
      "entropy": 5.9156616687774655,
      "epoch": 4.696985022369189,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002992923849662259,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22286065816879272,
      "num_tokens": 104717345.0,
      "step": 60370
    },
    {
      "entropy": 5.877993631362915,
      "epoch": 4.697374051740907,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029926483120388506,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2032250866293907,
      "num_tokens": 104726076.0,
      "step": 60375
    },
    {
      "entropy": 5.856709432601929,
      "epoch": 4.697763081112624,
      "grad_norm": 1.21875,
      "learning_rate": 0.000299237277073364,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2181733474135399,
      "num_tokens": 104735370.0,
      "step": 60380
    },
    {
      "entropy": 5.961450147628784,
      "epoch": 4.698152110484342,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002992097225750808,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21041541844606398,
      "num_tokens": 104744383.0,
      "step": 60385
    },
    {
      "entropy": 5.90102105140686,
      "epoch": 4.698541139856059,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029918216770945363,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21675359606742858,
      "num_tokens": 104752880.0,
      "step": 60390
    },
    {
      "entropy": 5.936983203887939,
      "epoch": 4.6989301692277765,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002991546124769005,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2115202873945236,
      "num_tokens": 104761937.0,
      "step": 60395
    },
    {
      "entropy": 5.923001098632812,
      "epoch": 4.699319198599494,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002991270568778396,
      "loss": 4.9122,
      "mean_token_accuracy": 0.21578485667705535,
      "num_tokens": 104770384.0,
      "step": 60400
    },
    {
      "entropy": 5.868461084365845,
      "epoch": 4.699708227971212,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002990995009126891,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2171441912651062,
      "num_tokens": 104780270.0,
      "step": 60405
    },
    {
      "entropy": 5.903707456588745,
      "epoch": 4.70009725734293,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002990719445818669,
      "loss": 4.8661,
      "mean_token_accuracy": 0.23120139837265014,
      "num_tokens": 104788444.0,
      "step": 60410
    },
    {
      "entropy": 5.900070762634277,
      "epoch": 4.700486286714647,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002990443878857914,
      "loss": 4.962,
      "mean_token_accuracy": 0.21701265424489974,
      "num_tokens": 104798218.0,
      "step": 60415
    },
    {
      "entropy": 5.870543384552002,
      "epoch": 4.700875316086364,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029901683082488045,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22498452365398408,
      "num_tokens": 104806044.0,
      "step": 60420
    },
    {
      "entropy": 5.828421020507813,
      "epoch": 4.701264345458082,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002989892733995524,
      "loss": 4.9002,
      "mean_token_accuracy": 0.21173701733350753,
      "num_tokens": 104814169.0,
      "step": 60425
    },
    {
      "entropy": 5.844472074508667,
      "epoch": 4.7016533748297995,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002989617156102253,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22543649822473527,
      "num_tokens": 104822677.0,
      "step": 60430
    },
    {
      "entropy": 5.8164306640625,
      "epoch": 4.702042404201517,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002989341574573172,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2268856331706047,
      "num_tokens": 104831069.0,
      "step": 60435
    },
    {
      "entropy": 5.771519899368286,
      "epoch": 4.702431433573235,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029890659894124647,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22566627115011215,
      "num_tokens": 104839935.0,
      "step": 60440
    },
    {
      "entropy": 5.851961517333985,
      "epoch": 4.702820462944953,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002988790400624312,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21278622895479202,
      "num_tokens": 104849475.0,
      "step": 60445
    },
    {
      "entropy": 5.838829231262207,
      "epoch": 4.70320949231667,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002988514808212894,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22033341526985167,
      "num_tokens": 104858269.0,
      "step": 60450
    },
    {
      "entropy": 5.923729372024536,
      "epoch": 4.703598521688387,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002988239212182394,
      "loss": 4.9866,
      "mean_token_accuracy": 0.20864265114068986,
      "num_tokens": 104866811.0,
      "step": 60455
    },
    {
      "entropy": 5.976103067398071,
      "epoch": 4.703987551060105,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002987963612536993,
      "loss": 4.9921,
      "mean_token_accuracy": 0.20870812088251114,
      "num_tokens": 104874762.0,
      "step": 60460
    },
    {
      "entropy": 5.79053840637207,
      "epoch": 4.7043765804318225,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029876880092808724,
      "loss": 4.7658,
      "mean_token_accuracy": 0.23350853472948074,
      "num_tokens": 104883594.0,
      "step": 60465
    },
    {
      "entropy": 5.905982589721679,
      "epoch": 4.70476560980354,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002987412402418215,
      "loss": 5.012,
      "mean_token_accuracy": 0.22236628383398055,
      "num_tokens": 104892228.0,
      "step": 60470
    },
    {
      "entropy": 5.810678339004516,
      "epoch": 4.705154639175258,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029871367919532014,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21275557577610016,
      "num_tokens": 104900639.0,
      "step": 60475
    },
    {
      "entropy": 5.803120183944702,
      "epoch": 4.705543668546976,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002986861177890016,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2304033651947975,
      "num_tokens": 104908110.0,
      "step": 60480
    },
    {
      "entropy": 5.869299411773682,
      "epoch": 4.705932697918692,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002986585560232838,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22337280958890915,
      "num_tokens": 104916769.0,
      "step": 60485
    },
    {
      "entropy": 5.917542314529419,
      "epoch": 4.70632172729041,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029863099389858516,
      "loss": 4.952,
      "mean_token_accuracy": 0.2155241310596466,
      "num_tokens": 104925219.0,
      "step": 60490
    },
    {
      "entropy": 5.9612648487091064,
      "epoch": 4.706710756662128,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002986034314153238,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2075718328356743,
      "num_tokens": 104933786.0,
      "step": 60495
    },
    {
      "entropy": 5.896211624145508,
      "epoch": 4.7070997860338455,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029857586857391796,
      "loss": 4.979,
      "mean_token_accuracy": 0.21739242672920228,
      "num_tokens": 104943261.0,
      "step": 60500
    },
    {
      "entropy": 5.884463262557984,
      "epoch": 4.707488815405563,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029854830537478567,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23152933418750762,
      "num_tokens": 104951031.0,
      "step": 60505
    },
    {
      "entropy": 5.917943286895752,
      "epoch": 4.707877844777281,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029852074181834545,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21525523960590362,
      "num_tokens": 104959502.0,
      "step": 60510
    },
    {
      "entropy": 5.8966912746429445,
      "epoch": 4.708266874148999,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029849317790501536,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21382590383291245,
      "num_tokens": 104967753.0,
      "step": 60515
    },
    {
      "entropy": 5.833817863464356,
      "epoch": 4.708655903520716,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029846561363521375,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21644620299339296,
      "num_tokens": 104976546.0,
      "step": 60520
    },
    {
      "entropy": 5.922440385818481,
      "epoch": 4.709044932892433,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002984380490093588,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21363983303308487,
      "num_tokens": 104984895.0,
      "step": 60525
    },
    {
      "entropy": 5.85637845993042,
      "epoch": 4.709433962264151,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002984104840278687,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2120603382587433,
      "num_tokens": 104993994.0,
      "step": 60530
    },
    {
      "entropy": 5.874888753890991,
      "epoch": 4.7098229916358685,
      "grad_norm": 1.375,
      "learning_rate": 0.0002983829186911618,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21609596759080887,
      "num_tokens": 105003174.0,
      "step": 60535
    },
    {
      "entropy": 5.868549633026123,
      "epoch": 4.710212021007586,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029835535299965635,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2201039507985115,
      "num_tokens": 105011734.0,
      "step": 60540
    },
    {
      "entropy": 5.906525182723999,
      "epoch": 4.710601050379304,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029832778695377057,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21255347579717637,
      "num_tokens": 105021656.0,
      "step": 60545
    },
    {
      "entropy": 5.872804784774781,
      "epoch": 4.710990079751022,
      "grad_norm": 1.375,
      "learning_rate": 0.00029830022055392277,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2164387285709381,
      "num_tokens": 105030061.0,
      "step": 60550
    },
    {
      "entropy": 5.880702829360962,
      "epoch": 4.711379109122738,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00029827265380053116,
      "loss": 4.984,
      "mean_token_accuracy": 0.20973785370588302,
      "num_tokens": 105038047.0,
      "step": 60555
    },
    {
      "entropy": 5.926590442657471,
      "epoch": 4.711768138494456,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002982450866940141,
      "loss": 5.0745,
      "mean_token_accuracy": 0.2051724448800087,
      "num_tokens": 105046691.0,
      "step": 60560
    },
    {
      "entropy": 5.77610592842102,
      "epoch": 4.712157167866174,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002982175192347899,
      "loss": 4.8,
      "mean_token_accuracy": 0.21941739618778228,
      "num_tokens": 105054897.0,
      "step": 60565
    },
    {
      "entropy": 5.8982954025268555,
      "epoch": 4.7125461972378915,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002981899514232767,
      "loss": 4.9566,
      "mean_token_accuracy": 0.21923065483570098,
      "num_tokens": 105063554.0,
      "step": 60570
    },
    {
      "entropy": 5.974019479751587,
      "epoch": 4.712935226609609,
      "grad_norm": 1.40625,
      "learning_rate": 0.000298162383259893,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20524168759584427,
      "num_tokens": 105072555.0,
      "step": 60575
    },
    {
      "entropy": 5.83348536491394,
      "epoch": 4.713324255981327,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029813481474505693,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21284767240285873,
      "num_tokens": 105081712.0,
      "step": 60580
    },
    {
      "entropy": 5.925499773025512,
      "epoch": 4.713713285353045,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029810724587918684,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2083306059241295,
      "num_tokens": 105089963.0,
      "step": 60585
    },
    {
      "entropy": 5.906161069869995,
      "epoch": 4.714102314724761,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002980796766627012,
      "loss": 4.981,
      "mean_token_accuracy": 0.21710223108530044,
      "num_tokens": 105098584.0,
      "step": 60590
    },
    {
      "entropy": 5.8503645896911625,
      "epoch": 4.714491344096479,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002980521070960181,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21807724833488465,
      "num_tokens": 105107042.0,
      "step": 60595
    },
    {
      "entropy": 5.890273332595825,
      "epoch": 4.714880373468197,
      "grad_norm": 1.5,
      "learning_rate": 0.00029802453717955597,
      "loss": 4.976,
      "mean_token_accuracy": 0.2159767359495163,
      "num_tokens": 105115323.0,
      "step": 60600
    },
    {
      "entropy": 5.928406715393066,
      "epoch": 4.7152694028399145,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002979969669137332,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20628589987754822,
      "num_tokens": 105124838.0,
      "step": 60605
    },
    {
      "entropy": 5.938727807998657,
      "epoch": 4.715658432211632,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000297969396298968,
      "loss": 5.0011,
      "mean_token_accuracy": 0.20809753239154816,
      "num_tokens": 105133407.0,
      "step": 60610
    },
    {
      "entropy": 5.926307535171508,
      "epoch": 4.71604746158335,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002979418253356788,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21411479264497757,
      "num_tokens": 105142310.0,
      "step": 60615
    },
    {
      "entropy": 5.864467144012451,
      "epoch": 4.716436490955067,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002979142540242839,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21570830345153807,
      "num_tokens": 105150536.0,
      "step": 60620
    },
    {
      "entropy": 5.798748922348023,
      "epoch": 4.716825520326784,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029788668236520164,
      "loss": 4.8295,
      "mean_token_accuracy": 0.22521100342273712,
      "num_tokens": 105159543.0,
      "step": 60625
    },
    {
      "entropy": 5.854570579528809,
      "epoch": 4.717214549698502,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029785911035885046,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2218244731426239,
      "num_tokens": 105168204.0,
      "step": 60630
    },
    {
      "entropy": 5.8912147045135494,
      "epoch": 4.71760357907022,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029783153800564863,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22351886034011842,
      "num_tokens": 105177523.0,
      "step": 60635
    },
    {
      "entropy": 5.899677467346192,
      "epoch": 4.7179926084419375,
      "grad_norm": 1.671875,
      "learning_rate": 0.00029780396530601456,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22080015987157822,
      "num_tokens": 105186087.0,
      "step": 60640
    },
    {
      "entropy": 5.890599584579467,
      "epoch": 4.718381637813655,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029777639226036663,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21785981208086014,
      "num_tokens": 105195423.0,
      "step": 60645
    },
    {
      "entropy": 5.924311256408691,
      "epoch": 4.718770667185373,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002977488188691232,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2132745146751404,
      "num_tokens": 105204275.0,
      "step": 60650
    },
    {
      "entropy": 5.89634952545166,
      "epoch": 4.7191596965570906,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002977212451327026,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21226166635751725,
      "num_tokens": 105212685.0,
      "step": 60655
    },
    {
      "entropy": 5.841070938110351,
      "epoch": 4.719548725928807,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002976936710515233,
      "loss": 4.891,
      "mean_token_accuracy": 0.2180558279156685,
      "num_tokens": 105221645.0,
      "step": 60660
    },
    {
      "entropy": 5.833380794525146,
      "epoch": 4.719937755300525,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029766609662600367,
      "loss": 4.9475,
      "mean_token_accuracy": 0.22205691784620285,
      "num_tokens": 105229681.0,
      "step": 60665
    },
    {
      "entropy": 5.8945780277252195,
      "epoch": 4.720326784672243,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029763852185656206,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2138862729072571,
      "num_tokens": 105238150.0,
      "step": 60670
    },
    {
      "entropy": 5.90069842338562,
      "epoch": 4.7207158140439605,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029761094674361695,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21595190167427064,
      "num_tokens": 105246334.0,
      "step": 60675
    },
    {
      "entropy": 5.892250108718872,
      "epoch": 4.721104843415678,
      "grad_norm": 1.375,
      "learning_rate": 0.00029758337128758675,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21278634816408157,
      "num_tokens": 105254796.0,
      "step": 60680
    },
    {
      "entropy": 5.87995867729187,
      "epoch": 4.721493872787395,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002975557954888898,
      "loss": 4.931,
      "mean_token_accuracy": 0.22121565490961076,
      "num_tokens": 105263718.0,
      "step": 60685
    },
    {
      "entropy": 5.906523656845093,
      "epoch": 4.721882902159113,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029752821934794457,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21669235825538635,
      "num_tokens": 105271947.0,
      "step": 60690
    },
    {
      "entropy": 5.870176506042481,
      "epoch": 4.72227193153083,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029750064286516943,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2086670830845833,
      "num_tokens": 105280660.0,
      "step": 60695
    },
    {
      "entropy": 5.896756505966186,
      "epoch": 4.722660960902548,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002974730660409828,
      "loss": 4.9167,
      "mean_token_accuracy": 0.21602464020252227,
      "num_tokens": 105289604.0,
      "step": 60700
    },
    {
      "entropy": 5.940533113479614,
      "epoch": 4.723049990274266,
      "grad_norm": 1.515625,
      "learning_rate": 0.00029744548887580325,
      "loss": 4.9944,
      "mean_token_accuracy": 0.21379899382591247,
      "num_tokens": 105298112.0,
      "step": 60705
    },
    {
      "entropy": 5.904570865631103,
      "epoch": 4.7234390196459834,
      "grad_norm": 1.546875,
      "learning_rate": 0.00029741791137004916,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2108463779091835,
      "num_tokens": 105306065.0,
      "step": 60710
    },
    {
      "entropy": 5.887718439102173,
      "epoch": 4.723828049017701,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029739033352413883,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21735145300626754,
      "num_tokens": 105315107.0,
      "step": 60715
    },
    {
      "entropy": 5.841065502166748,
      "epoch": 4.724217078389419,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029736275533849094,
      "loss": 4.9287,
      "mean_token_accuracy": 0.216402131319046,
      "num_tokens": 105323470.0,
      "step": 60720
    },
    {
      "entropy": 5.831031560897827,
      "epoch": 4.724606107761136,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002973351768135239,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22297154366970062,
      "num_tokens": 105332644.0,
      "step": 60725
    },
    {
      "entropy": 5.870612335205078,
      "epoch": 4.724995137132853,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029730759794965595,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22400619685649872,
      "num_tokens": 105341672.0,
      "step": 60730
    },
    {
      "entropy": 5.819486904144287,
      "epoch": 4.725384166504571,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002972800187473057,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2211390644311905,
      "num_tokens": 105349958.0,
      "step": 60735
    },
    {
      "entropy": 5.855258178710938,
      "epoch": 4.725773195876289,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029725243920689163,
      "loss": 4.944,
      "mean_token_accuracy": 0.21145863085985184,
      "num_tokens": 105359069.0,
      "step": 60740
    },
    {
      "entropy": 5.900996685028076,
      "epoch": 4.726162225248006,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029722485932883233,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21387088894844056,
      "num_tokens": 105367779.0,
      "step": 60745
    },
    {
      "entropy": 5.842031049728393,
      "epoch": 4.726551254619724,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029719727911354604,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21801753342151642,
      "num_tokens": 105376135.0,
      "step": 60750
    },
    {
      "entropy": 5.885424089431763,
      "epoch": 4.726940283991441,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029716969856145143,
      "loss": 4.9819,
      "mean_token_accuracy": 0.217241770029068,
      "num_tokens": 105384934.0,
      "step": 60755
    },
    {
      "entropy": 5.891575145721435,
      "epoch": 4.727329313363159,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029714211767296693,
      "loss": 5.0151,
      "mean_token_accuracy": 0.2055464580655098,
      "num_tokens": 105393238.0,
      "step": 60760
    },
    {
      "entropy": 5.858783817291259,
      "epoch": 4.727718342734876,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002971145364485111,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2132525160908699,
      "num_tokens": 105402772.0,
      "step": 60765
    },
    {
      "entropy": 5.902745056152344,
      "epoch": 4.728107372106594,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002970869548885023,
      "loss": 4.96,
      "mean_token_accuracy": 0.21192507743835448,
      "num_tokens": 105411578.0,
      "step": 60770
    },
    {
      "entropy": 5.8758580684661865,
      "epoch": 4.728496401478312,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002970593729933591,
      "loss": 4.9254,
      "mean_token_accuracy": 0.21759480237960815,
      "num_tokens": 105420397.0,
      "step": 60775
    },
    {
      "entropy": 5.875942850112915,
      "epoch": 4.728885430850029,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029703179076350005,
      "loss": 4.9007,
      "mean_token_accuracy": 0.2245413213968277,
      "num_tokens": 105429087.0,
      "step": 60780
    },
    {
      "entropy": 5.860939407348633,
      "epoch": 4.729274460221747,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029700420819934363,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21735534220933914,
      "num_tokens": 105436900.0,
      "step": 60785
    },
    {
      "entropy": 5.866411828994751,
      "epoch": 4.729663489593464,
      "grad_norm": 1.375,
      "learning_rate": 0.00029697662530130836,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21231728941202163,
      "num_tokens": 105445740.0,
      "step": 60790
    },
    {
      "entropy": 5.8576117038726805,
      "epoch": 4.730052518965182,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002969490420698129,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21322918683290482,
      "num_tokens": 105454754.0,
      "step": 60795
    },
    {
      "entropy": 5.798538255691528,
      "epoch": 4.730441548336899,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029692145850527553,
      "loss": 4.8304,
      "mean_token_accuracy": 0.22737499177455903,
      "num_tokens": 105463237.0,
      "step": 60800
    },
    {
      "entropy": 5.845492172241211,
      "epoch": 4.730830577708617,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029689387460811494,
      "loss": 4.912,
      "mean_token_accuracy": 0.2192944496870041,
      "num_tokens": 105472459.0,
      "step": 60805
    },
    {
      "entropy": 5.922378253936768,
      "epoch": 4.731219607080335,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002968662903787497,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22024778574705123,
      "num_tokens": 105481528.0,
      "step": 60810
    },
    {
      "entropy": 5.852470350265503,
      "epoch": 4.731608636452052,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002968387058175983,
      "loss": 4.8445,
      "mean_token_accuracy": 0.22382921129465103,
      "num_tokens": 105490317.0,
      "step": 60815
    },
    {
      "entropy": 5.833373403549194,
      "epoch": 4.731997665823769,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002968111209250792,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22285645008087157,
      "num_tokens": 105499039.0,
      "step": 60820
    },
    {
      "entropy": 5.813837385177612,
      "epoch": 4.732386695195487,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029678353570161117,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22093296200037002,
      "num_tokens": 105508233.0,
      "step": 60825
    },
    {
      "entropy": 5.869218873977661,
      "epoch": 4.732775724567205,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002967559501476126,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21013854295015336,
      "num_tokens": 105516867.0,
      "step": 60830
    },
    {
      "entropy": 5.836968660354614,
      "epoch": 4.733164753938922,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029672836426350213,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2240413397550583,
      "num_tokens": 105525305.0,
      "step": 60835
    },
    {
      "entropy": 5.888692474365234,
      "epoch": 4.73355378331064,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029670077804969835,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21696747690439225,
      "num_tokens": 105533488.0,
      "step": 60840
    },
    {
      "entropy": 5.835903549194336,
      "epoch": 4.733942812682358,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029667319150661976,
      "loss": 4.9043,
      "mean_token_accuracy": 0.21555159538984298,
      "num_tokens": 105542768.0,
      "step": 60845
    },
    {
      "entropy": 5.818588972091675,
      "epoch": 4.734331842054075,
      "grad_norm": 1.421875,
      "learning_rate": 0.000296645604634685,
      "loss": 4.8265,
      "mean_token_accuracy": 0.2333204284310341,
      "num_tokens": 105551097.0,
      "step": 60850
    },
    {
      "entropy": 5.855669832229614,
      "epoch": 4.734720871425793,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002966180174343127,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2197147622704506,
      "num_tokens": 105559846.0,
      "step": 60855
    },
    {
      "entropy": 5.854354572296143,
      "epoch": 4.73510990079751,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029659042990592125,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2156075954437256,
      "num_tokens": 105568697.0,
      "step": 60860
    },
    {
      "entropy": 5.9596349716186525,
      "epoch": 4.735498930169228,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002965628420499295,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2051233634352684,
      "num_tokens": 105577139.0,
      "step": 60865
    },
    {
      "entropy": 5.928594923019409,
      "epoch": 4.735887959540945,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002965352538667559,
      "loss": 4.894,
      "mean_token_accuracy": 0.22407492101192475,
      "num_tokens": 105585557.0,
      "step": 60870
    },
    {
      "entropy": 5.798439168930054,
      "epoch": 4.736276988912663,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00029650766535681917,
      "loss": 4.8091,
      "mean_token_accuracy": 0.22160778641700746,
      "num_tokens": 105593671.0,
      "step": 60875
    },
    {
      "entropy": 5.81412148475647,
      "epoch": 4.736666018284381,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029648007652053774,
      "loss": 4.8639,
      "mean_token_accuracy": 0.219798743724823,
      "num_tokens": 105602668.0,
      "step": 60880
    },
    {
      "entropy": 5.82104640007019,
      "epoch": 4.737055047656098,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029645248735833045,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22374660074710845,
      "num_tokens": 105610727.0,
      "step": 60885
    },
    {
      "entropy": 5.878356313705444,
      "epoch": 4.737444077027815,
      "grad_norm": 1.375,
      "learning_rate": 0.00029642489787061574,
      "loss": 4.9559,
      "mean_token_accuracy": 0.217888905107975,
      "num_tokens": 105619050.0,
      "step": 60890
    },
    {
      "entropy": 5.877041959762574,
      "epoch": 4.737833106399533,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002963973080578123,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21296107172966003,
      "num_tokens": 105628065.0,
      "step": 60895
    },
    {
      "entropy": 5.826031637191773,
      "epoch": 4.738222135771251,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029636971792033873,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2174631893634796,
      "num_tokens": 105636838.0,
      "step": 60900
    },
    {
      "entropy": 5.88527512550354,
      "epoch": 4.738611165142968,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002963421274586138,
      "loss": 4.9687,
      "mean_token_accuracy": 0.213816037774086,
      "num_tokens": 105645739.0,
      "step": 60905
    },
    {
      "entropy": 5.954970264434815,
      "epoch": 4.739000194514686,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029631453667305603,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21654482632875444,
      "num_tokens": 105654077.0,
      "step": 60910
    },
    {
      "entropy": 5.875184440612793,
      "epoch": 4.739389223886404,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002962869455640841,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2197732463479042,
      "num_tokens": 105662274.0,
      "step": 60915
    },
    {
      "entropy": 5.933502244949341,
      "epoch": 4.739778253258121,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002962593541321167,
      "loss": 4.9638,
      "mean_token_accuracy": 0.20897582173347473,
      "num_tokens": 105670153.0,
      "step": 60920
    },
    {
      "entropy": 5.919316577911377,
      "epoch": 4.740167282629838,
      "grad_norm": 1.5,
      "learning_rate": 0.0002962317623775723,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21574748158454896,
      "num_tokens": 105679801.0,
      "step": 60925
    },
    {
      "entropy": 5.876159715652466,
      "epoch": 4.740556312001556,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029620417030086976,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21289939284324647,
      "num_tokens": 105688944.0,
      "step": 60930
    },
    {
      "entropy": 5.914960241317749,
      "epoch": 4.740945341373274,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029617657790242767,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2156111091375351,
      "num_tokens": 105697368.0,
      "step": 60935
    },
    {
      "entropy": 5.844474363327026,
      "epoch": 4.741334370744991,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029614898518266476,
      "loss": 4.8668,
      "mean_token_accuracy": 0.21742608398199081,
      "num_tokens": 105706068.0,
      "step": 60940
    },
    {
      "entropy": 5.815180540084839,
      "epoch": 4.741723400116709,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002961213921419997,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21509056091308593,
      "num_tokens": 105714178.0,
      "step": 60945
    },
    {
      "entropy": 5.944170331954956,
      "epoch": 4.742112429488427,
      "grad_norm": 1.40625,
      "learning_rate": 0.000296093798780851,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21122508496046066,
      "num_tokens": 105722628.0,
      "step": 60950
    },
    {
      "entropy": 5.92220630645752,
      "epoch": 4.7425014588601435,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002960662050996377,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21475653797388078,
      "num_tokens": 105732024.0,
      "step": 60955
    },
    {
      "entropy": 5.824908971786499,
      "epoch": 4.742890488231861,
      "grad_norm": 1.375,
      "learning_rate": 0.00029603861109877807,
      "loss": 4.8744,
      "mean_token_accuracy": 0.21619491875171662,
      "num_tokens": 105740612.0,
      "step": 60960
    },
    {
      "entropy": 5.9146082401275635,
      "epoch": 4.743279517603579,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029601101677869106,
      "loss": 4.888,
      "mean_token_accuracy": 0.2171686753630638,
      "num_tokens": 105750014.0,
      "step": 60965
    },
    {
      "entropy": 5.887810802459716,
      "epoch": 4.743668546975297,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002959834221397952,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21673284769058226,
      "num_tokens": 105759272.0,
      "step": 60970
    },
    {
      "entropy": 5.790692472457886,
      "epoch": 4.744057576347014,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029595582718250954,
      "loss": 4.9265,
      "mean_token_accuracy": 0.21436749249696732,
      "num_tokens": 105767630.0,
      "step": 60975
    },
    {
      "entropy": 5.903926086425781,
      "epoch": 4.744446605718732,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002959282319072524,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20204175859689713,
      "num_tokens": 105775823.0,
      "step": 60980
    },
    {
      "entropy": 5.906321620941162,
      "epoch": 4.74483563509045,
      "grad_norm": 1.5,
      "learning_rate": 0.0002959006363144427,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2183700829744339,
      "num_tokens": 105784565.0,
      "step": 60985
    },
    {
      "entropy": 5.900314664840698,
      "epoch": 4.7452246644621665,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002958730404044991,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21297525465488434,
      "num_tokens": 105793944.0,
      "step": 60990
    },
    {
      "entropy": 5.846686172485351,
      "epoch": 4.745613693833884,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002958454441778403,
      "loss": 4.9146,
      "mean_token_accuracy": 0.21707866936922074,
      "num_tokens": 105802416.0,
      "step": 60995
    },
    {
      "entropy": 5.896412372589111,
      "epoch": 4.746002723205602,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029581784763488506,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21247471123933792,
      "num_tokens": 105811648.0,
      "step": 61000
    },
    {
      "entropy": 5.95093355178833,
      "epoch": 4.74639175257732,
      "grad_norm": 1.625,
      "learning_rate": 0.0002957902507760522,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2138322576880455,
      "num_tokens": 105819458.0,
      "step": 61005
    },
    {
      "entropy": 5.885247898101807,
      "epoch": 4.746780781949037,
      "grad_norm": 1.5,
      "learning_rate": 0.0002957626536017603,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21853520423173906,
      "num_tokens": 105828178.0,
      "step": 61010
    },
    {
      "entropy": 5.844181871414184,
      "epoch": 4.747169811320755,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002957350561124282,
      "loss": 4.877,
      "mean_token_accuracy": 0.21989412009716033,
      "num_tokens": 105836548.0,
      "step": 61015
    },
    {
      "entropy": 5.882378959655762,
      "epoch": 4.747558840692472,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002957074583084747,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21192325949668883,
      "num_tokens": 105845374.0,
      "step": 61020
    },
    {
      "entropy": 5.810941076278686,
      "epoch": 4.7479478700641895,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002956798601903184,
      "loss": 4.8262,
      "mean_token_accuracy": 0.22525006532669067,
      "num_tokens": 105854294.0,
      "step": 61025
    },
    {
      "entropy": 5.903068161010742,
      "epoch": 4.748336899435907,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00029565226175837814,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22202681601047516,
      "num_tokens": 105863019.0,
      "step": 61030
    },
    {
      "entropy": 5.853875780105591,
      "epoch": 4.748725928807625,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002956246630130727,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2194087728857994,
      "num_tokens": 105871233.0,
      "step": 61035
    },
    {
      "entropy": 5.893203639984131,
      "epoch": 4.749114958179343,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029559706395482075,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21711246222257613,
      "num_tokens": 105880005.0,
      "step": 61040
    },
    {
      "entropy": 5.8818943977355955,
      "epoch": 4.74950398755106,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00029556946458404124,
      "loss": 4.96,
      "mean_token_accuracy": 0.21537002176046371,
      "num_tokens": 105888578.0,
      "step": 61045
    },
    {
      "entropy": 5.876488065719604,
      "epoch": 4.749893016922778,
      "grad_norm": 1.5,
      "learning_rate": 0.00029554186490115284,
      "loss": 4.957,
      "mean_token_accuracy": 0.21126961410045625,
      "num_tokens": 105896842.0,
      "step": 61050
    },
    {
      "entropy": 5.934459972381592,
      "epoch": 4.750282046294496,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002955142649065743,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21262956857681276,
      "num_tokens": 105904722.0,
      "step": 61055
    },
    {
      "entropy": 5.91472430229187,
      "epoch": 4.7506710756662125,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002954866646007244,
      "loss": 5.006,
      "mean_token_accuracy": 0.21256323903799057,
      "num_tokens": 105913118.0,
      "step": 61060
    },
    {
      "entropy": 5.912500667572021,
      "epoch": 4.75106010503793,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029545906398402205,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2017812982201576,
      "num_tokens": 105921798.0,
      "step": 61065
    },
    {
      "entropy": 5.881893348693848,
      "epoch": 4.751449134409648,
      "grad_norm": 1.375,
      "learning_rate": 0.00029543146305688594,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21557374447584152,
      "num_tokens": 105930914.0,
      "step": 61070
    },
    {
      "entropy": 5.882312774658203,
      "epoch": 4.751838163781366,
      "grad_norm": 1.375,
      "learning_rate": 0.0002954038618197349,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21585414707660674,
      "num_tokens": 105940266.0,
      "step": 61075
    },
    {
      "entropy": 5.922711801528931,
      "epoch": 4.752227193153083,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002953762602729877,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22216977775096894,
      "num_tokens": 105948221.0,
      "step": 61080
    },
    {
      "entropy": 5.843612098693848,
      "epoch": 4.752616222524801,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029534865841706324,
      "loss": 4.9285,
      "mean_token_accuracy": 0.21978115886449814,
      "num_tokens": 105956864.0,
      "step": 61085
    },
    {
      "entropy": 5.926437187194824,
      "epoch": 4.753005251896518,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002953210562523803,
      "loss": 4.9989,
      "mean_token_accuracy": 0.22348757982254028,
      "num_tokens": 105965926.0,
      "step": 61090
    },
    {
      "entropy": 5.898807048797607,
      "epoch": 4.7533942812682355,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002952934537793576,
      "loss": 4.969,
      "mean_token_accuracy": 0.21146332323551179,
      "num_tokens": 105975510.0,
      "step": 61095
    },
    {
      "entropy": 5.8703193187713625,
      "epoch": 4.753783310639953,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029526585099841395,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2104182481765747,
      "num_tokens": 105984408.0,
      "step": 61100
    },
    {
      "entropy": 5.844713973999023,
      "epoch": 4.754172340011671,
      "grad_norm": 1.25,
      "learning_rate": 0.00029523824790996845,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22372340708971022,
      "num_tokens": 105993496.0,
      "step": 61105
    },
    {
      "entropy": 5.900966358184815,
      "epoch": 4.754561369383389,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029521064451443966,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22201010435819626,
      "num_tokens": 106002040.0,
      "step": 61110
    },
    {
      "entropy": 5.889397668838501,
      "epoch": 4.754950398755106,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029518304081224656,
      "loss": 4.95,
      "mean_token_accuracy": 0.20875708609819413,
      "num_tokens": 106010510.0,
      "step": 61115
    },
    {
      "entropy": 5.921111154556274,
      "epoch": 4.755339428126824,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029515543680380793,
      "loss": 4.9238,
      "mean_token_accuracy": 0.221443372964859,
      "num_tokens": 106019258.0,
      "step": 61120
    },
    {
      "entropy": 5.907075500488281,
      "epoch": 4.755728457498541,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029512783248954253,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21693806648254393,
      "num_tokens": 106027947.0,
      "step": 61125
    },
    {
      "entropy": 5.889995861053467,
      "epoch": 4.7561174868702585,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002951002278698695,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21608293652534485,
      "num_tokens": 106036141.0,
      "step": 61130
    },
    {
      "entropy": 5.852833032608032,
      "epoch": 4.756506516241976,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029507262294520726,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22119842916727067,
      "num_tokens": 106044821.0,
      "step": 61135
    },
    {
      "entropy": 5.835148906707763,
      "epoch": 4.756895545613694,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002950450177159751,
      "loss": 4.9556,
      "mean_token_accuracy": 0.2211356058716774,
      "num_tokens": 106052910.0,
      "step": 61140
    },
    {
      "entropy": 5.8114612102508545,
      "epoch": 4.757284574985412,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029501741218259167,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22332770079374314,
      "num_tokens": 106062025.0,
      "step": 61145
    },
    {
      "entropy": 5.902966547012329,
      "epoch": 4.757673604357129,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029498980634547585,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21825567185878753,
      "num_tokens": 106070075.0,
      "step": 61150
    },
    {
      "entropy": 5.8145452499389645,
      "epoch": 4.758062633728846,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029496220020504656,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2161939799785614,
      "num_tokens": 106079225.0,
      "step": 61155
    },
    {
      "entropy": 5.7966302871704105,
      "epoch": 4.758451663100564,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029493459376172263,
      "loss": 4.7493,
      "mean_token_accuracy": 0.2272977724671364,
      "num_tokens": 106087767.0,
      "step": 61160
    },
    {
      "entropy": 5.896937036514283,
      "epoch": 4.7588406924722815,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002949069870159229,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22175297141075134,
      "num_tokens": 106096670.0,
      "step": 61165
    },
    {
      "entropy": 5.860300159454345,
      "epoch": 4.759229721843999,
      "grad_norm": 1.40625,
      "learning_rate": 0.00029487937996806635,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2171230733394623,
      "num_tokens": 106105768.0,
      "step": 61170
    },
    {
      "entropy": 5.8539710521698,
      "epoch": 4.759618751215717,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029485177261857193,
      "loss": 4.843,
      "mean_token_accuracy": 0.22562188357114793,
      "num_tokens": 106113831.0,
      "step": 61175
    },
    {
      "entropy": 5.832643127441406,
      "epoch": 4.760007780587435,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002948241649678584,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22861818075180054,
      "num_tokens": 106121814.0,
      "step": 61180
    },
    {
      "entropy": 5.928608083724976,
      "epoch": 4.760396809959152,
      "grad_norm": 1.5,
      "learning_rate": 0.0002947965570163448,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21545332968235015,
      "num_tokens": 106130983.0,
      "step": 61185
    },
    {
      "entropy": 5.892875099182129,
      "epoch": 4.76078583933087,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002947689487644499,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21343118101358413,
      "num_tokens": 106139983.0,
      "step": 61190
    },
    {
      "entropy": 5.837586879730225,
      "epoch": 4.761174868702587,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002947413402125927,
      "loss": 4.8853,
      "mean_token_accuracy": 0.21965064257383346,
      "num_tokens": 106148132.0,
      "step": 61195
    },
    {
      "entropy": 5.859208774566651,
      "epoch": 4.7615638980743045,
      "grad_norm": 1.375,
      "learning_rate": 0.0002947137313611919,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21253736466169357,
      "num_tokens": 106156481.0,
      "step": 61200
    },
    {
      "entropy": 5.870739603042603,
      "epoch": 4.761952927446022,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029468612221066683,
      "loss": 5.0031,
      "mean_token_accuracy": 0.20898440182209016,
      "num_tokens": 106165815.0,
      "step": 61205
    },
    {
      "entropy": 5.874511480331421,
      "epoch": 4.76234195681774,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002946585127614361,
      "loss": 4.8604,
      "mean_token_accuracy": 0.2184752941131592,
      "num_tokens": 106174073.0,
      "step": 61210
    },
    {
      "entropy": 5.883990669250489,
      "epoch": 4.762730986189458,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002946309030139188,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21315957009792327,
      "num_tokens": 106182788.0,
      "step": 61215
    },
    {
      "entropy": 5.9313554763793945,
      "epoch": 4.763120015561174,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002946032929685337,
      "loss": 4.9535,
      "mean_token_accuracy": 0.20816124230623245,
      "num_tokens": 106191640.0,
      "step": 61220
    },
    {
      "entropy": 5.851598787307739,
      "epoch": 4.763509044932892,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002945756826256999,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2108073502779007,
      "num_tokens": 106200341.0,
      "step": 61225
    },
    {
      "entropy": 5.8882745742797855,
      "epoch": 4.76389807430461,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002945480719858363,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2329368844628334,
      "num_tokens": 106208246.0,
      "step": 61230
    },
    {
      "entropy": 5.972222948074341,
      "epoch": 4.7642871036763275,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002945204610493617,
      "loss": 5.0993,
      "mean_token_accuracy": 0.20270868092775346,
      "num_tokens": 106216917.0,
      "step": 61235
    },
    {
      "entropy": 5.8756755828857425,
      "epoch": 4.764676133048045,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029449284981669524,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21971877813339233,
      "num_tokens": 106225049.0,
      "step": 61240
    },
    {
      "entropy": 5.854177093505859,
      "epoch": 4.765065162419763,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029446523828825577,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22476483285427093,
      "num_tokens": 106233066.0,
      "step": 61245
    },
    {
      "entropy": 5.871186065673828,
      "epoch": 4.7654541917914806,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002944376264644624,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21867943853139876,
      "num_tokens": 106242107.0,
      "step": 61250
    },
    {
      "entropy": 5.9037737369537355,
      "epoch": 4.765843221163198,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002944100143457339,
      "loss": 4.9083,
      "mean_token_accuracy": 0.21939605474472046,
      "num_tokens": 106250571.0,
      "step": 61255
    },
    {
      "entropy": 5.963947486877442,
      "epoch": 4.766232250534915,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002943824019324893,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20453084707260133,
      "num_tokens": 106259069.0,
      "step": 61260
    },
    {
      "entropy": 5.926756048202515,
      "epoch": 4.766621279906633,
      "grad_norm": 1.375,
      "learning_rate": 0.0002943547892251476,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2270295113325119,
      "num_tokens": 106267971.0,
      "step": 61265
    },
    {
      "entropy": 5.915213584899902,
      "epoch": 4.7670103092783505,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029432717622412784,
      "loss": 4.9004,
      "mean_token_accuracy": 0.21815563142299652,
      "num_tokens": 106275972.0,
      "step": 61270
    },
    {
      "entropy": 5.8746179103851315,
      "epoch": 4.767399338650068,
      "grad_norm": 1.515625,
      "learning_rate": 0.00029429956292984893,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21695407778024672,
      "num_tokens": 106284292.0,
      "step": 61275
    },
    {
      "entropy": 5.806115579605103,
      "epoch": 4.767788368021786,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002942719493427299,
      "loss": 4.8801,
      "mean_token_accuracy": 0.21529111713171006,
      "num_tokens": 106292909.0,
      "step": 61280
    },
    {
      "entropy": 5.8449324607849125,
      "epoch": 4.7681773973935035,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002942443354631897,
      "loss": 4.8688,
      "mean_token_accuracy": 0.21771973967552186,
      "num_tokens": 106301724.0,
      "step": 61285
    },
    {
      "entropy": 5.903324270248413,
      "epoch": 4.76856642676522,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029421672129164726,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21339768320322036,
      "num_tokens": 106310818.0,
      "step": 61290
    },
    {
      "entropy": 5.9232923030853275,
      "epoch": 4.768955456136938,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029418910682852176,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21534694284200667,
      "num_tokens": 106318929.0,
      "step": 61295
    },
    {
      "entropy": 5.9295577049255375,
      "epoch": 4.769344485508656,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029416149207423203,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2113225743174553,
      "num_tokens": 106327502.0,
      "step": 61300
    },
    {
      "entropy": 5.870498418807983,
      "epoch": 4.7697335148803734,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029413387702919723,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21916650980710983,
      "num_tokens": 106335882.0,
      "step": 61305
    },
    {
      "entropy": 5.96025333404541,
      "epoch": 4.770122544252091,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029410626169383623,
      "loss": 5.0731,
      "mean_token_accuracy": 0.20890649408102036,
      "num_tokens": 106344593.0,
      "step": 61310
    },
    {
      "entropy": 5.8693946361541744,
      "epoch": 4.770511573623809,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029407864606856814,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22215885072946548,
      "num_tokens": 106352930.0,
      "step": 61315
    },
    {
      "entropy": 5.849253797531128,
      "epoch": 4.7709006029955265,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000294051030153812,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21666816920042037,
      "num_tokens": 106360826.0,
      "step": 61320
    },
    {
      "entropy": 5.877149248123169,
      "epoch": 4.771289632367243,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002940234139499868,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2249186173081398,
      "num_tokens": 106369752.0,
      "step": 61325
    },
    {
      "entropy": 5.8540667533874515,
      "epoch": 4.771678661738961,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029399579745751156,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22689461410045625,
      "num_tokens": 106377632.0,
      "step": 61330
    },
    {
      "entropy": 5.826297426223755,
      "epoch": 4.772067691110679,
      "grad_norm": 1.375,
      "learning_rate": 0.00029396818067680525,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21397999823093414,
      "num_tokens": 106386159.0,
      "step": 61335
    },
    {
      "entropy": 5.804233169555664,
      "epoch": 4.772456720482396,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029394056360828707,
      "loss": 4.852,
      "mean_token_accuracy": 0.21680935323238373,
      "num_tokens": 106395529.0,
      "step": 61340
    },
    {
      "entropy": 5.92691559791565,
      "epoch": 4.772845749854114,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00029391294625237595,
      "loss": 4.992,
      "mean_token_accuracy": 0.2062061011791229,
      "num_tokens": 106404178.0,
      "step": 61345
    },
    {
      "entropy": 5.847654914855957,
      "epoch": 4.773234779225832,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002938853286094911,
      "loss": 4.9163,
      "mean_token_accuracy": 0.220000958442688,
      "num_tokens": 106412750.0,
      "step": 61350
    },
    {
      "entropy": 5.83144850730896,
      "epoch": 4.773623808597549,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029385771068005123,
      "loss": 4.898,
      "mean_token_accuracy": 0.21834155470132827,
      "num_tokens": 106421972.0,
      "step": 61355
    },
    {
      "entropy": 5.821387815475464,
      "epoch": 4.774012837969266,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002938300924644757,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2178368330001831,
      "num_tokens": 106430702.0,
      "step": 61360
    },
    {
      "entropy": 5.906731176376343,
      "epoch": 4.774401867340984,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029380247396318355,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21817118376493455,
      "num_tokens": 106439543.0,
      "step": 61365
    },
    {
      "entropy": 5.827312231063843,
      "epoch": 4.774790896712702,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002937748551765937,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2267797663807869,
      "num_tokens": 106448361.0,
      "step": 61370
    },
    {
      "entropy": 5.914904022216797,
      "epoch": 4.775179926084419,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002937472361051254,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21219406574964522,
      "num_tokens": 106456448.0,
      "step": 61375
    },
    {
      "entropy": 5.8572954654693605,
      "epoch": 4.775568955456137,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002937196167491976,
      "loss": 4.8202,
      "mean_token_accuracy": 0.22672727108001708,
      "num_tokens": 106464347.0,
      "step": 61380
    },
    {
      "entropy": 5.796436500549317,
      "epoch": 4.775957984827855,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029369199710922944,
      "loss": 4.8835,
      "mean_token_accuracy": 0.21994436234235765,
      "num_tokens": 106472755.0,
      "step": 61385
    },
    {
      "entropy": 5.839485120773316,
      "epoch": 4.7763470141995725,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029366437718563994,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21521196514368057,
      "num_tokens": 106481382.0,
      "step": 61390
    },
    {
      "entropy": 5.930896759033203,
      "epoch": 4.776736043571289,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002936367569788482,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21474696397781373,
      "num_tokens": 106490230.0,
      "step": 61395
    },
    {
      "entropy": 5.923521709442139,
      "epoch": 4.777125072943007,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002936091364892733,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21415612399578093,
      "num_tokens": 106498242.0,
      "step": 61400
    },
    {
      "entropy": 5.9067620754241945,
      "epoch": 4.777514102314725,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00029358151571733446,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21870145499706267,
      "num_tokens": 106506362.0,
      "step": 61405
    },
    {
      "entropy": 5.831527853012085,
      "epoch": 4.777903131686442,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029355389466345064,
      "loss": 4.8602,
      "mean_token_accuracy": 0.21716355234384538,
      "num_tokens": 106515022.0,
      "step": 61410
    },
    {
      "entropy": 5.883844423294067,
      "epoch": 4.77829216105816,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029352627332804104,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22303903698921204,
      "num_tokens": 106522932.0,
      "step": 61415
    },
    {
      "entropy": 5.883833217620849,
      "epoch": 4.778681190429878,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029349865171152475,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21399435102939607,
      "num_tokens": 106532123.0,
      "step": 61420
    },
    {
      "entropy": 5.85845160484314,
      "epoch": 4.779070219801595,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00029347102981432084,
      "loss": 4.88,
      "mean_token_accuracy": 0.22085824459791184,
      "num_tokens": 106540222.0,
      "step": 61425
    },
    {
      "entropy": 5.895369338989258,
      "epoch": 4.779459249173312,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002934434076368484,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22770921289920806,
      "num_tokens": 106548530.0,
      "step": 61430
    },
    {
      "entropy": 5.830578851699829,
      "epoch": 4.77984827854503,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002934157851795266,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21685124188661575,
      "num_tokens": 106557032.0,
      "step": 61435
    },
    {
      "entropy": 5.83912467956543,
      "epoch": 4.780237307916748,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002933881624427746,
      "loss": 4.908,
      "mean_token_accuracy": 0.21519740521907807,
      "num_tokens": 106565323.0,
      "step": 61440
    },
    {
      "entropy": 5.896981048583984,
      "epoch": 4.780626337288465,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002933605394270116,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21188769787549971,
      "num_tokens": 106574018.0,
      "step": 61445
    },
    {
      "entropy": 5.897304534912109,
      "epoch": 4.781015366660183,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002933329161326565,
      "loss": 4.984,
      "mean_token_accuracy": 0.2137632817029953,
      "num_tokens": 106583050.0,
      "step": 61450
    },
    {
      "entropy": 5.917555809020996,
      "epoch": 4.781404396031901,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029330529256012863,
      "loss": 4.992,
      "mean_token_accuracy": 0.2110005423426628,
      "num_tokens": 106592013.0,
      "step": 61455
    },
    {
      "entropy": 5.883837556838989,
      "epoch": 4.781793425403618,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029327766870984714,
      "loss": 4.8613,
      "mean_token_accuracy": 0.224580579996109,
      "num_tokens": 106600497.0,
      "step": 61460
    },
    {
      "entropy": 5.867414331436157,
      "epoch": 4.782182454775335,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029325004458223097,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2242203563451767,
      "num_tokens": 106609615.0,
      "step": 61465
    },
    {
      "entropy": 5.961280727386475,
      "epoch": 4.782571484147053,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00029322242017769954,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20489207953214644,
      "num_tokens": 106618777.0,
      "step": 61470
    },
    {
      "entropy": 5.932095003128052,
      "epoch": 4.782960513518771,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029319479549667183,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21638449281454086,
      "num_tokens": 106626672.0,
      "step": 61475
    },
    {
      "entropy": 5.930276107788086,
      "epoch": 4.783349542890488,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002931671705395671,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2155462995171547,
      "num_tokens": 106635469.0,
      "step": 61480
    },
    {
      "entropy": 5.836870765686035,
      "epoch": 4.783738572262206,
      "grad_norm": 1.515625,
      "learning_rate": 0.00029313954530680447,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2218319967389107,
      "num_tokens": 106644042.0,
      "step": 61485
    },
    {
      "entropy": 5.853882646560669,
      "epoch": 4.784127601633923,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029311191979880313,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20881106555461884,
      "num_tokens": 106653187.0,
      "step": 61490
    },
    {
      "entropy": 5.936031293869019,
      "epoch": 4.784516631005641,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002930842940159822,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21924409568309783,
      "num_tokens": 106662120.0,
      "step": 61495
    },
    {
      "entropy": 5.955626344680786,
      "epoch": 4.784905660377358,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029305666795876093,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2199181228876114,
      "num_tokens": 106670964.0,
      "step": 61500
    },
    {
      "entropy": 5.919579792022705,
      "epoch": 4.785294689749076,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00029302904162755843,
      "loss": 4.9922,
      "mean_token_accuracy": 0.20902244448661805,
      "num_tokens": 106679264.0,
      "step": 61505
    },
    {
      "entropy": 5.914316034317016,
      "epoch": 4.785683719120794,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029300141502279395,
      "loss": 5.0628,
      "mean_token_accuracy": 0.21200062781572343,
      "num_tokens": 106687842.0,
      "step": 61510
    },
    {
      "entropy": 5.920604610443116,
      "epoch": 4.786072748492511,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029297378814488657,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2144693613052368,
      "num_tokens": 106696585.0,
      "step": 61515
    },
    {
      "entropy": 5.907338857650757,
      "epoch": 4.786461777864229,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029294616099425556,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21135954111814498,
      "num_tokens": 106705841.0,
      "step": 61520
    },
    {
      "entropy": 5.881907272338867,
      "epoch": 4.786850807235947,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002929185335713202,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22044856250286102,
      "num_tokens": 106714674.0,
      "step": 61525
    },
    {
      "entropy": 5.87093596458435,
      "epoch": 4.787239836607664,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029289090587649965,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2189219117164612,
      "num_tokens": 106723498.0,
      "step": 61530
    },
    {
      "entropy": 5.9294881343841555,
      "epoch": 4.787628865979381,
      "grad_norm": 1.34375,
      "learning_rate": 0.000292863277910213,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21484384536743165,
      "num_tokens": 106732502.0,
      "step": 61535
    },
    {
      "entropy": 5.920049142837525,
      "epoch": 4.788017895351099,
      "grad_norm": 1.5,
      "learning_rate": 0.0002928356496728796,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21666008681058885,
      "num_tokens": 106741217.0,
      "step": 61540
    },
    {
      "entropy": 5.88206672668457,
      "epoch": 4.788406924722817,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029280802116491854,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21823852360248566,
      "num_tokens": 106749753.0,
      "step": 61545
    },
    {
      "entropy": 5.857629060745239,
      "epoch": 4.788795954094534,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029278039238674915,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21169043928384781,
      "num_tokens": 106758699.0,
      "step": 61550
    },
    {
      "entropy": 5.940484046936035,
      "epoch": 4.789184983466251,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002927527633387906,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20772104561328888,
      "num_tokens": 106767796.0,
      "step": 61555
    },
    {
      "entropy": 5.931923770904541,
      "epoch": 4.789574012837969,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029272513402146203,
      "loss": 4.9618,
      "mean_token_accuracy": 0.20842180997133256,
      "num_tokens": 106776969.0,
      "step": 61560
    },
    {
      "entropy": 5.833198118209839,
      "epoch": 4.789963042209687,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002926975044351828,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21857596039772034,
      "num_tokens": 106786679.0,
      "step": 61565
    },
    {
      "entropy": 5.93925838470459,
      "epoch": 4.790352071581404,
      "grad_norm": 1.375,
      "learning_rate": 0.0002926698745803723,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21423836350440978,
      "num_tokens": 106794964.0,
      "step": 61570
    },
    {
      "entropy": 5.908381557464599,
      "epoch": 4.790741100953122,
      "grad_norm": 1.375,
      "learning_rate": 0.00029264224445744937,
      "loss": 4.9095,
      "mean_token_accuracy": 0.21793148964643477,
      "num_tokens": 106803823.0,
      "step": 61575
    },
    {
      "entropy": 5.926155996322632,
      "epoch": 4.79113013032484,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029261461406683353,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21640522927045822,
      "num_tokens": 106812380.0,
      "step": 61580
    },
    {
      "entropy": 5.845181846618653,
      "epoch": 4.791519159696557,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029258698340894397,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22322425246238708,
      "num_tokens": 106821528.0,
      "step": 61585
    },
    {
      "entropy": 5.887009239196777,
      "epoch": 4.791908189068275,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00029255935248419983,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21669257581233978,
      "num_tokens": 106829925.0,
      "step": 61590
    },
    {
      "entropy": 5.854901552200317,
      "epoch": 4.792297218439992,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002925317212930206,
      "loss": 4.9482,
      "mean_token_accuracy": 0.213443423807621,
      "num_tokens": 106838555.0,
      "step": 61595
    },
    {
      "entropy": 5.839565277099609,
      "epoch": 4.79268624781171,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029250408983582527,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2237298846244812,
      "num_tokens": 106847047.0,
      "step": 61600
    },
    {
      "entropy": 5.862899684906006,
      "epoch": 4.793075277183427,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002924764581130333,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23150792568922043,
      "num_tokens": 106855111.0,
      "step": 61605
    },
    {
      "entropy": 5.907291221618652,
      "epoch": 4.793464306555145,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000292448826125064,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21615599989891052,
      "num_tokens": 106864170.0,
      "step": 61610
    },
    {
      "entropy": 5.850291967391968,
      "epoch": 4.793853335926863,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002924211938723364,
      "loss": 4.8629,
      "mean_token_accuracy": 0.21678697913885117,
      "num_tokens": 106873482.0,
      "step": 61615
    },
    {
      "entropy": 5.88385329246521,
      "epoch": 4.79424236529858,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029239356135526994,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21475729942321778,
      "num_tokens": 106881776.0,
      "step": 61620
    },
    {
      "entropy": 5.90189938545227,
      "epoch": 4.794631394670297,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000292365928574284,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20489224642515183,
      "num_tokens": 106890405.0,
      "step": 61625
    },
    {
      "entropy": 5.914997673034668,
      "epoch": 4.795020424042015,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002923382955297975,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2186702683568001,
      "num_tokens": 106899611.0,
      "step": 61630
    },
    {
      "entropy": 5.869638967514038,
      "epoch": 4.795409453413733,
      "grad_norm": 1.375,
      "learning_rate": 0.0002923106622222301,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22128242552280425,
      "num_tokens": 106909273.0,
      "step": 61635
    },
    {
      "entropy": 5.782846450805664,
      "epoch": 4.79579848278545,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002922830286520009,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22206541746854783,
      "num_tokens": 106917531.0,
      "step": 61640
    },
    {
      "entropy": 5.896895551681519,
      "epoch": 4.796187512157168,
      "grad_norm": 1.5625,
      "learning_rate": 0.00029225539481952934,
      "loss": 4.971,
      "mean_token_accuracy": 0.220217502117157,
      "num_tokens": 106926355.0,
      "step": 61645
    },
    {
      "entropy": 5.8039137840271,
      "epoch": 4.796576541528886,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002922277607252346,
      "loss": 4.8239,
      "mean_token_accuracy": 0.22635396122932433,
      "num_tokens": 106934836.0,
      "step": 61650
    },
    {
      "entropy": 5.942328929901123,
      "epoch": 4.796965570900603,
      "grad_norm": 1.359375,
      "learning_rate": 0.000292200126369536,
      "loss": 5.008,
      "mean_token_accuracy": 0.20458555817604065,
      "num_tokens": 106944078.0,
      "step": 61655
    },
    {
      "entropy": 5.8109241962432865,
      "epoch": 4.79735460027232,
      "grad_norm": 1.390625,
      "learning_rate": 0.00029217249175285284,
      "loss": 4.8112,
      "mean_token_accuracy": 0.22997771799564362,
      "num_tokens": 106952545.0,
      "step": 61660
    },
    {
      "entropy": 5.857165670394897,
      "epoch": 4.797743629644038,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00029214485687560437,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22011952400207518,
      "num_tokens": 106962181.0,
      "step": 61665
    },
    {
      "entropy": 5.75851674079895,
      "epoch": 4.798132659015756,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002921172217382101,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2260972961783409,
      "num_tokens": 106970780.0,
      "step": 61670
    },
    {
      "entropy": 5.870467567443848,
      "epoch": 4.798521688387473,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029208958634108917,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21760610789060592,
      "num_tokens": 106980037.0,
      "step": 61675
    },
    {
      "entropy": 5.861391830444336,
      "epoch": 4.798910717759191,
      "grad_norm": 1.453125,
      "learning_rate": 0.000292061950684661,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2221352756023407,
      "num_tokens": 106988822.0,
      "step": 61680
    },
    {
      "entropy": 5.905781555175781,
      "epoch": 4.799299747130909,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00029203431476934496,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22204350531101227,
      "num_tokens": 106996696.0,
      "step": 61685
    },
    {
      "entropy": 5.935604572296143,
      "epoch": 4.7996887765026255,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029200667859556026,
      "loss": 5.061,
      "mean_token_accuracy": 0.20652461796998978,
      "num_tokens": 107005530.0,
      "step": 61690
    },
    {
      "entropy": 5.828819656372071,
      "epoch": 4.800077805874343,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002919790421637262,
      "loss": 4.8198,
      "mean_token_accuracy": 0.22885854542255402,
      "num_tokens": 107013779.0,
      "step": 61695
    },
    {
      "entropy": 5.8279221534729,
      "epoch": 4.800466835246061,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002919514054742622,
      "loss": 4.8475,
      "mean_token_accuracy": 0.2204282596707344,
      "num_tokens": 107022397.0,
      "step": 61700
    },
    {
      "entropy": 5.774351596832275,
      "epoch": 4.800855864617779,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029192376852758774,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21601344048976898,
      "num_tokens": 107031213.0,
      "step": 61705
    },
    {
      "entropy": 5.9311291694641115,
      "epoch": 4.801244893989496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002918961313241219,
      "loss": 5.051,
      "mean_token_accuracy": 0.2075706511735916,
      "num_tokens": 107040369.0,
      "step": 61710
    },
    {
      "entropy": 6.012514877319336,
      "epoch": 4.801633923361214,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002918684938642842,
      "loss": 5.089,
      "mean_token_accuracy": 0.20581374168395997,
      "num_tokens": 107049303.0,
      "step": 61715
    },
    {
      "entropy": 5.846624231338501,
      "epoch": 4.802022952732932,
      "grad_norm": 1.375,
      "learning_rate": 0.00029184085614849403,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21580785512924194,
      "num_tokens": 107057860.0,
      "step": 61720
    },
    {
      "entropy": 5.853810501098633,
      "epoch": 4.802411982104649,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002918132181771706,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20705979764461518,
      "num_tokens": 107066318.0,
      "step": 61725
    },
    {
      "entropy": 5.869130706787109,
      "epoch": 4.802801011476366,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029178557995073335,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23005930036306382,
      "num_tokens": 107075183.0,
      "step": 61730
    },
    {
      "entropy": 5.802637434005737,
      "epoch": 4.803190040848084,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002917579414696017,
      "loss": 4.8278,
      "mean_token_accuracy": 0.22055953443050386,
      "num_tokens": 107083375.0,
      "step": 61735
    },
    {
      "entropy": 5.8412333011627195,
      "epoch": 4.803579070219802,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029173030273419484,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22183843553066254,
      "num_tokens": 107091990.0,
      "step": 61740
    },
    {
      "entropy": 5.882032775878907,
      "epoch": 4.803968099591519,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002917026637449323,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2129744365811348,
      "num_tokens": 107101196.0,
      "step": 61745
    },
    {
      "entropy": 5.8227927684783936,
      "epoch": 4.804357128963237,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002916750245022335,
      "loss": 4.8417,
      "mean_token_accuracy": 0.21931996643543245,
      "num_tokens": 107109379.0,
      "step": 61750
    },
    {
      "entropy": 5.837630367279052,
      "epoch": 4.804746158334955,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002916473850065177,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2231481820344925,
      "num_tokens": 107118635.0,
      "step": 61755
    },
    {
      "entropy": 5.827294969558716,
      "epoch": 4.8051351877066715,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00029161974525820434,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2166684478521347,
      "num_tokens": 107126170.0,
      "step": 61760
    },
    {
      "entropy": 5.836792850494385,
      "epoch": 4.805524217078389,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002915921052577127,
      "loss": 4.9629,
      "mean_token_accuracy": 0.20916857868432998,
      "num_tokens": 107134406.0,
      "step": 61765
    },
    {
      "entropy": 5.9114978313446045,
      "epoch": 4.805913246450107,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029156446500546243,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21086731255054475,
      "num_tokens": 107143577.0,
      "step": 61770
    },
    {
      "entropy": 5.910972023010254,
      "epoch": 4.806302275821825,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029153682450187265,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21004211157560349,
      "num_tokens": 107151908.0,
      "step": 61775
    },
    {
      "entropy": 5.916540241241455,
      "epoch": 4.806691305193542,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029150918374736294,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21713487654924393,
      "num_tokens": 107159924.0,
      "step": 61780
    },
    {
      "entropy": 5.9685547828674315,
      "epoch": 4.80708033456526,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002914815427423526,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21994968205690385,
      "num_tokens": 107167962.0,
      "step": 61785
    },
    {
      "entropy": 5.875788593292237,
      "epoch": 4.807469363936978,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002914539014872612,
      "loss": 4.9123,
      "mean_token_accuracy": 0.21743139028549194,
      "num_tokens": 107177760.0,
      "step": 61790
    },
    {
      "entropy": 5.85589632987976,
      "epoch": 4.8078583933086945,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029142625998250786,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21955206096172333,
      "num_tokens": 107186369.0,
      "step": 61795
    },
    {
      "entropy": 5.909440326690674,
      "epoch": 4.808247422680412,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029139861822851233,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21461800187826158,
      "num_tokens": 107194526.0,
      "step": 61800
    },
    {
      "entropy": 5.915717697143554,
      "epoch": 4.80863645205213,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029137097622569376,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20820139050483705,
      "num_tokens": 107202350.0,
      "step": 61805
    },
    {
      "entropy": 5.904705333709717,
      "epoch": 4.809025481423848,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002913433339744718,
      "loss": 4.9767,
      "mean_token_accuracy": 0.21579258739948273,
      "num_tokens": 107211227.0,
      "step": 61810
    },
    {
      "entropy": 5.852851867675781,
      "epoch": 4.809414510795565,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002913156914752656,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21095031201839448,
      "num_tokens": 107219909.0,
      "step": 61815
    },
    {
      "entropy": 5.8806154251098635,
      "epoch": 4.809803540167283,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002912880487284948,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2206324964761734,
      "num_tokens": 107228065.0,
      "step": 61820
    },
    {
      "entropy": 5.838554716110229,
      "epoch": 4.810192569539,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029126040573457886,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2244905337691307,
      "num_tokens": 107236730.0,
      "step": 61825
    },
    {
      "entropy": 5.9198075294494625,
      "epoch": 4.8105815989107175,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00029123276249393707,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2163254514336586,
      "num_tokens": 107245491.0,
      "step": 61830
    },
    {
      "entropy": 5.849360990524292,
      "epoch": 4.810970628282435,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029120511900698894,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21540557146072387,
      "num_tokens": 107254428.0,
      "step": 61835
    },
    {
      "entropy": 5.965359830856324,
      "epoch": 4.811359657654153,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000291177475274154,
      "loss": 5.0839,
      "mean_token_accuracy": 0.2058124929666519,
      "num_tokens": 107263070.0,
      "step": 61840
    },
    {
      "entropy": 5.96596131324768,
      "epoch": 4.8117486870258706,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002911498312958516,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21580142974853517,
      "num_tokens": 107271779.0,
      "step": 61845
    },
    {
      "entropy": 5.8652587890625,
      "epoch": 4.812137716397588,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029112218707250115,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2264339178800583,
      "num_tokens": 107280489.0,
      "step": 61850
    },
    {
      "entropy": 5.847153425216675,
      "epoch": 4.812526745769306,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029109454260452223,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2139160230755806,
      "num_tokens": 107289536.0,
      "step": 61855
    },
    {
      "entropy": 5.895623350143433,
      "epoch": 4.812915775141024,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029106689789233425,
      "loss": 4.9337,
      "mean_token_accuracy": 0.2089979588985443,
      "num_tokens": 107298082.0,
      "step": 61860
    },
    {
      "entropy": 5.9282997131347654,
      "epoch": 4.8133048045127405,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002910392529363566,
      "loss": 4.9899,
      "mean_token_accuracy": 0.20948233753442763,
      "num_tokens": 107306769.0,
      "step": 61865
    },
    {
      "entropy": 5.947876596450806,
      "epoch": 4.813693833884458,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029101160773700887,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21442489624023436,
      "num_tokens": 107315122.0,
      "step": 61870
    },
    {
      "entropy": 5.910910940170288,
      "epoch": 4.814082863256176,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002909839622947104,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21395664364099504,
      "num_tokens": 107323182.0,
      "step": 61875
    },
    {
      "entropy": 5.893836450576782,
      "epoch": 4.8144718926278935,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002909563166098809,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22578376829624175,
      "num_tokens": 107331502.0,
      "step": 61880
    },
    {
      "entropy": 5.842210292816162,
      "epoch": 4.814860921999611,
      "grad_norm": 1.5,
      "learning_rate": 0.00029092867068293966,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21833933740854264,
      "num_tokens": 107340440.0,
      "step": 61885
    },
    {
      "entropy": 5.9327116966247555,
      "epoch": 4.815249951371328,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002909010245143061,
      "loss": 4.9901,
      "mean_token_accuracy": 0.218713741004467,
      "num_tokens": 107348983.0,
      "step": 61890
    },
    {
      "entropy": 5.935223054885864,
      "epoch": 4.815638980743046,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002908733781043998,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21698195040225982,
      "num_tokens": 107358209.0,
      "step": 61895
    },
    {
      "entropy": 5.80257363319397,
      "epoch": 4.8160280101147634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002908457314536403,
      "loss": 4.8188,
      "mean_token_accuracy": 0.21979471296072006,
      "num_tokens": 107366901.0,
      "step": 61900
    },
    {
      "entropy": 5.881351661682129,
      "epoch": 4.816417039486481,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00029081808456244703,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2138473644852638,
      "num_tokens": 107374742.0,
      "step": 61905
    },
    {
      "entropy": 5.865342998504639,
      "epoch": 4.816806068858199,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00029079043743123953,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21983760893344878,
      "num_tokens": 107383608.0,
      "step": 61910
    },
    {
      "entropy": 5.864350700378418,
      "epoch": 4.8171950982299165,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00029076279006043723,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21005658507347108,
      "num_tokens": 107392453.0,
      "step": 61915
    },
    {
      "entropy": 5.918496370315552,
      "epoch": 4.817584127601634,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002907351424504597,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21348755061626434,
      "num_tokens": 107401114.0,
      "step": 61920
    },
    {
      "entropy": 5.893487405776978,
      "epoch": 4.817973156973352,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002907074946017265,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21582500487565995,
      "num_tokens": 107409630.0,
      "step": 61925
    },
    {
      "entropy": 5.875457620620727,
      "epoch": 4.818362186345069,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000290679846514657,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2117021158337593,
      "num_tokens": 107418404.0,
      "step": 61930
    },
    {
      "entropy": 5.861436128616333,
      "epoch": 4.818751215716786,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029065219818967075,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2218944564461708,
      "num_tokens": 107427094.0,
      "step": 61935
    },
    {
      "entropy": 5.884388065338134,
      "epoch": 4.819140245088504,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029062454962718736,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21704270243644713,
      "num_tokens": 107436407.0,
      "step": 61940
    },
    {
      "entropy": 5.93908724784851,
      "epoch": 4.819529274460222,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002905969008276263,
      "loss": 5.0393,
      "mean_token_accuracy": 0.20738417506217957,
      "num_tokens": 107444647.0,
      "step": 61945
    },
    {
      "entropy": 5.917088031768799,
      "epoch": 4.8199183038319395,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002905692517914071,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2218498021364212,
      "num_tokens": 107453212.0,
      "step": 61950
    },
    {
      "entropy": 5.909564018249512,
      "epoch": 4.820307333203657,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029054160251894924,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21248814165592195,
      "num_tokens": 107462870.0,
      "step": 61955
    },
    {
      "entropy": 5.898371696472168,
      "epoch": 4.820696362575374,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002905139530106724,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21813525706529618,
      "num_tokens": 107470483.0,
      "step": 61960
    },
    {
      "entropy": 5.899685001373291,
      "epoch": 4.821085391947092,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029048630326699594,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22235707193613052,
      "num_tokens": 107478720.0,
      "step": 61965
    },
    {
      "entropy": 5.830667734146118,
      "epoch": 4.821474421318809,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029045865328833944,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22177153676748276,
      "num_tokens": 107487304.0,
      "step": 61970
    },
    {
      "entropy": 5.917517948150635,
      "epoch": 4.821863450690527,
      "grad_norm": 1.421875,
      "learning_rate": 0.00029043100307512247,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21476486027240754,
      "num_tokens": 107495716.0,
      "step": 61975
    },
    {
      "entropy": 5.9309529781341555,
      "epoch": 4.822252480062245,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00029040335262776467,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22576336413621903,
      "num_tokens": 107504884.0,
      "step": 61980
    },
    {
      "entropy": 5.879701709747314,
      "epoch": 4.8226415094339625,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002903757019466855,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22229569107294084,
      "num_tokens": 107513756.0,
      "step": 61985
    },
    {
      "entropy": 5.869576168060303,
      "epoch": 4.82303053880568,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029034805103230453,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2164277881383896,
      "num_tokens": 107521768.0,
      "step": 61990
    },
    {
      "entropy": 5.94304461479187,
      "epoch": 4.823419568177397,
      "grad_norm": 1.375,
      "learning_rate": 0.00029032039988504126,
      "loss": 5.0502,
      "mean_token_accuracy": 0.20650026351213455,
      "num_tokens": 107530454.0,
      "step": 61995
    },
    {
      "entropy": 5.8508758544921875,
      "epoch": 4.823808597549115,
      "grad_norm": 1.484375,
      "learning_rate": 0.00029029274850531534,
      "loss": 4.84,
      "mean_token_accuracy": 0.2179330751299858,
      "num_tokens": 107539114.0,
      "step": 62000
    },
    {
      "entropy": 5.839717197418213,
      "epoch": 4.824197626920832,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029026509689354635,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22520221024751663,
      "num_tokens": 107548698.0,
      "step": 62005
    },
    {
      "entropy": 5.8696778297424315,
      "epoch": 4.82458665629255,
      "grad_norm": 1.34375,
      "learning_rate": 0.00029023744505015373,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21642776280641557,
      "num_tokens": 107558391.0,
      "step": 62010
    },
    {
      "entropy": 5.96497163772583,
      "epoch": 4.824975685664268,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00029020979297555717,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21327578276395798,
      "num_tokens": 107567640.0,
      "step": 62015
    },
    {
      "entropy": 5.906280183792115,
      "epoch": 4.8253647150359855,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029018214067017624,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22210168540477754,
      "num_tokens": 107576394.0,
      "step": 62020
    },
    {
      "entropy": 5.871028280258178,
      "epoch": 4.825753744407702,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002901544881344305,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2147579684853554,
      "num_tokens": 107584957.0,
      "step": 62025
    },
    {
      "entropy": 5.935571813583374,
      "epoch": 4.82614277377942,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002901268353687394,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21331010162830352,
      "num_tokens": 107593712.0,
      "step": 62030
    },
    {
      "entropy": 5.837939119338989,
      "epoch": 4.826531803151138,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029009918237352274,
      "loss": 4.9058,
      "mean_token_accuracy": 0.21637653261423112,
      "num_tokens": 107602169.0,
      "step": 62035
    },
    {
      "entropy": 5.931995868682861,
      "epoch": 4.826920832522855,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002900715291492,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2193202033638954,
      "num_tokens": 107610610.0,
      "step": 62040
    },
    {
      "entropy": 5.86958794593811,
      "epoch": 4.827309861894573,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002900438756961908,
      "loss": 4.8537,
      "mean_token_accuracy": 0.21837698817253112,
      "num_tokens": 107619181.0,
      "step": 62045
    },
    {
      "entropy": 5.84639892578125,
      "epoch": 4.827698891266291,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002900162220149148,
      "loss": 4.9221,
      "mean_token_accuracy": 0.22707315236330033,
      "num_tokens": 107627577.0,
      "step": 62050
    },
    {
      "entropy": 5.829078149795532,
      "epoch": 4.8280879206380085,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028998856810579144,
      "loss": 4.911,
      "mean_token_accuracy": 0.21625610440969467,
      "num_tokens": 107635967.0,
      "step": 62055
    },
    {
      "entropy": 5.844146728515625,
      "epoch": 4.828476950009726,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002899609139692405,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21824893206357956,
      "num_tokens": 107644736.0,
      "step": 62060
    },
    {
      "entropy": 5.8452691555023195,
      "epoch": 4.828865979381443,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028993325960568133,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21232314258813859,
      "num_tokens": 107654182.0,
      "step": 62065
    },
    {
      "entropy": 5.905589818954468,
      "epoch": 4.829255008753161,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002899056050155339,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22382637113332748,
      "num_tokens": 107662261.0,
      "step": 62070
    },
    {
      "entropy": 5.861131191253662,
      "epoch": 4.829644038124878,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002898779501992176,
      "loss": 4.9597,
      "mean_token_accuracy": 0.2151722565293312,
      "num_tokens": 107670633.0,
      "step": 62075
    },
    {
      "entropy": 5.851321506500244,
      "epoch": 4.830033067496596,
      "grad_norm": 1.296875,
      "learning_rate": 0.000289850295157152,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2192945286631584,
      "num_tokens": 107679627.0,
      "step": 62080
    },
    {
      "entropy": 5.939016580581665,
      "epoch": 4.830422096868314,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028982263988975687,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22275057733058928,
      "num_tokens": 107688051.0,
      "step": 62085
    },
    {
      "entropy": 5.886831331253052,
      "epoch": 4.8308111262400315,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002897949843974518,
      "loss": 4.8907,
      "mean_token_accuracy": 0.21942337453365326,
      "num_tokens": 107696611.0,
      "step": 62090
    },
    {
      "entropy": 5.863000106811524,
      "epoch": 4.831200155611748,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002897673286806562,
      "loss": 4.8384,
      "mean_token_accuracy": 0.2208089277148247,
      "num_tokens": 107705236.0,
      "step": 62095
    },
    {
      "entropy": 5.877387380599975,
      "epoch": 4.831589184983466,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002897396727397901,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22393605709075928,
      "num_tokens": 107713500.0,
      "step": 62100
    },
    {
      "entropy": 5.846330165863037,
      "epoch": 4.831978214355184,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002897120165752729,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22097325325012207,
      "num_tokens": 107722855.0,
      "step": 62105
    },
    {
      "entropy": 5.910072708129883,
      "epoch": 4.832367243726901,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002896843601875243,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21377041488885878,
      "num_tokens": 107731818.0,
      "step": 62110
    },
    {
      "entropy": 5.89212007522583,
      "epoch": 4.832756273098619,
      "grad_norm": 1.375,
      "learning_rate": 0.00028965670357696384,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2159130722284317,
      "num_tokens": 107741008.0,
      "step": 62115
    },
    {
      "entropy": 5.885447549819946,
      "epoch": 4.833145302470337,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028962904674401124,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2212527334690094,
      "num_tokens": 107750172.0,
      "step": 62120
    },
    {
      "entropy": 5.874864006042481,
      "epoch": 4.8335343318420545,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002896013896890862,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21130937486886978,
      "num_tokens": 107759152.0,
      "step": 62125
    },
    {
      "entropy": 5.874926233291626,
      "epoch": 4.833923361213771,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028957373241260816,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22224509119987487,
      "num_tokens": 107767636.0,
      "step": 62130
    },
    {
      "entropy": 5.849129486083984,
      "epoch": 4.834312390585489,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000289546074914997,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2185625210404396,
      "num_tokens": 107776771.0,
      "step": 62135
    },
    {
      "entropy": 5.8795229434967045,
      "epoch": 4.834701419957207,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002895184171966724,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20924233496189118,
      "num_tokens": 107785498.0,
      "step": 62140
    },
    {
      "entropy": 5.895925521850586,
      "epoch": 4.835090449328924,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002894907592580538,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2157543942332268,
      "num_tokens": 107794407.0,
      "step": 62145
    },
    {
      "entropy": 5.870306348800659,
      "epoch": 4.835479478700642,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028946310109956107,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22420860081911087,
      "num_tokens": 107803570.0,
      "step": 62150
    },
    {
      "entropy": 5.91444845199585,
      "epoch": 4.83586850807236,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002894354427216138,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21051058173179626,
      "num_tokens": 107811908.0,
      "step": 62155
    },
    {
      "entropy": 5.81254768371582,
      "epoch": 4.836257537444077,
      "grad_norm": 1.25,
      "learning_rate": 0.00028940778412463164,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22054685652256012,
      "num_tokens": 107820819.0,
      "step": 62160
    },
    {
      "entropy": 5.869008541107178,
      "epoch": 4.836646566815794,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028938012530903425,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22547455877065659,
      "num_tokens": 107829583.0,
      "step": 62165
    },
    {
      "entropy": 5.9023009777069095,
      "epoch": 4.837035596187512,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002893524662752414,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2111619770526886,
      "num_tokens": 107838880.0,
      "step": 62170
    },
    {
      "entropy": 5.981964874267578,
      "epoch": 4.83742462555923,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028932480702367264,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21153555661439896,
      "num_tokens": 107848646.0,
      "step": 62175
    },
    {
      "entropy": 5.875278520584106,
      "epoch": 4.837813654930947,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002892971475547478,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21869334429502488,
      "num_tokens": 107856629.0,
      "step": 62180
    },
    {
      "entropy": 5.901328420639038,
      "epoch": 4.838202684302665,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028926948786888655,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21563210636377333,
      "num_tokens": 107865219.0,
      "step": 62185
    },
    {
      "entropy": 5.866179275512695,
      "epoch": 4.838591713674383,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028924182796650854,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22186527103185655,
      "num_tokens": 107873448.0,
      "step": 62190
    },
    {
      "entropy": 5.861680412292481,
      "epoch": 4.8389807430461,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028921416784803336,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2125310942530632,
      "num_tokens": 107881355.0,
      "step": 62195
    },
    {
      "entropy": 5.8479033470153805,
      "epoch": 4.839369772417817,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002891865075138809,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21401943117380143,
      "num_tokens": 107890112.0,
      "step": 62200
    },
    {
      "entropy": 5.917454433441162,
      "epoch": 4.839758801789535,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028915884696447064,
      "loss": 4.959,
      "mean_token_accuracy": 0.21557158827781678,
      "num_tokens": 107898764.0,
      "step": 62205
    },
    {
      "entropy": 5.882563066482544,
      "epoch": 4.840147831161253,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002891311862002225,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21420456767082213,
      "num_tokens": 107907895.0,
      "step": 62210
    },
    {
      "entropy": 5.9254903316497805,
      "epoch": 4.84053686053297,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002891035252215561,
      "loss": 4.954,
      "mean_token_accuracy": 0.21492969840765,
      "num_tokens": 107916083.0,
      "step": 62215
    },
    {
      "entropy": 5.941675806045533,
      "epoch": 4.840925889904688,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028907586402889115,
      "loss": 5.0056,
      "mean_token_accuracy": 0.209830379486084,
      "num_tokens": 107925783.0,
      "step": 62220
    },
    {
      "entropy": 5.952735042572021,
      "epoch": 4.841314919276405,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002890482026226473,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21599783301353453,
      "num_tokens": 107934579.0,
      "step": 62225
    },
    {
      "entropy": 5.813460826873779,
      "epoch": 4.841703948648123,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028902054100324445,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2246412843465805,
      "num_tokens": 107944067.0,
      "step": 62230
    },
    {
      "entropy": 5.925602626800537,
      "epoch": 4.84209297801984,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00028899287917110214,
      "loss": 4.9817,
      "mean_token_accuracy": 0.20910331159830092,
      "num_tokens": 107952787.0,
      "step": 62235
    },
    {
      "entropy": 5.856427478790283,
      "epoch": 4.842482007391558,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002889652171266402,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22313415110111237,
      "num_tokens": 107961628.0,
      "step": 62240
    },
    {
      "entropy": 5.853416967391968,
      "epoch": 4.842871036763276,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00028893755487027827,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22663459330797195,
      "num_tokens": 107969885.0,
      "step": 62245
    },
    {
      "entropy": 5.887207508087158,
      "epoch": 4.843260066134993,
      "grad_norm": 1.375,
      "learning_rate": 0.00028890989240243614,
      "loss": 4.8676,
      "mean_token_accuracy": 0.21991741955280303,
      "num_tokens": 107978018.0,
      "step": 62250
    },
    {
      "entropy": 5.914603614807129,
      "epoch": 4.843649095506711,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028888222972353356,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2128503367304802,
      "num_tokens": 107986267.0,
      "step": 62255
    },
    {
      "entropy": 5.900496578216552,
      "epoch": 4.844038124878429,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028885456683399024,
      "loss": 4.99,
      "mean_token_accuracy": 0.21534063816070556,
      "num_tokens": 107995083.0,
      "step": 62260
    },
    {
      "entropy": 5.81304030418396,
      "epoch": 4.844427154250146,
      "grad_norm": 1.5625,
      "learning_rate": 0.00028882690373422586,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22105390727519988,
      "num_tokens": 108003616.0,
      "step": 62265
    },
    {
      "entropy": 5.815735864639282,
      "epoch": 4.844816183621863,
      "grad_norm": 1.5,
      "learning_rate": 0.0002887992404246603,
      "loss": 4.876,
      "mean_token_accuracy": 0.2197629228234291,
      "num_tokens": 108012429.0,
      "step": 62270
    },
    {
      "entropy": 5.922047805786133,
      "epoch": 4.845205212993581,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002887715769057132,
      "loss": 4.963,
      "mean_token_accuracy": 0.20835772901773453,
      "num_tokens": 108021436.0,
      "step": 62275
    },
    {
      "entropy": 5.876093435287475,
      "epoch": 4.845594242365299,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028874391317780437,
      "loss": 4.8938,
      "mean_token_accuracy": 0.21242813915014266,
      "num_tokens": 108030677.0,
      "step": 62280
    },
    {
      "entropy": 5.882072877883911,
      "epoch": 4.845983271737016,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002887162492413534,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2163059875369072,
      "num_tokens": 108038922.0,
      "step": 62285
    },
    {
      "entropy": 5.910499477386475,
      "epoch": 4.846372301108734,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002886885850967803,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21437685936689377,
      "num_tokens": 108048279.0,
      "step": 62290
    },
    {
      "entropy": 5.887196159362793,
      "epoch": 4.846761330480451,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028866092074450467,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2174646556377411,
      "num_tokens": 108057875.0,
      "step": 62295
    },
    {
      "entropy": 5.822545433044434,
      "epoch": 4.847150359852169,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002886332561849464,
      "loss": 4.7654,
      "mean_token_accuracy": 0.2337066188454628,
      "num_tokens": 108066056.0,
      "step": 62300
    },
    {
      "entropy": 5.803968811035157,
      "epoch": 4.847539389223886,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028860559141852504,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22822553813457488,
      "num_tokens": 108074641.0,
      "step": 62305
    },
    {
      "entropy": 5.7953619956970215,
      "epoch": 4.847928418595604,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028857792644566057,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22146357148885726,
      "num_tokens": 108083433.0,
      "step": 62310
    },
    {
      "entropy": 5.840663242340088,
      "epoch": 4.848317447967322,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002885502612667727,
      "loss": 4.8686,
      "mean_token_accuracy": 0.21989228278398515,
      "num_tokens": 108092191.0,
      "step": 62315
    },
    {
      "entropy": 5.849950981140137,
      "epoch": 4.848706477339039,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028852259588228106,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22538453787565232,
      "num_tokens": 108100141.0,
      "step": 62320
    },
    {
      "entropy": 5.882545518875122,
      "epoch": 4.849095506710757,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028849493029260573,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21220564544200898,
      "num_tokens": 108108863.0,
      "step": 62325
    },
    {
      "entropy": 5.91484751701355,
      "epoch": 4.849484536082474,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002884672644981661,
      "loss": 4.9607,
      "mean_token_accuracy": 0.20958262383937837,
      "num_tokens": 108117708.0,
      "step": 62330
    },
    {
      "entropy": 5.92452712059021,
      "epoch": 4.849873565454192,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002884395984993823,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21214203238487245,
      "num_tokens": 108126739.0,
      "step": 62335
    },
    {
      "entropy": 5.844645929336548,
      "epoch": 4.850262594825909,
      "grad_norm": 1.421875,
      "learning_rate": 0.000288411932296674,
      "loss": 4.8237,
      "mean_token_accuracy": 0.22675587832927704,
      "num_tokens": 108135173.0,
      "step": 62340
    },
    {
      "entropy": 5.85316481590271,
      "epoch": 4.850651624197627,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002883842658904609,
      "loss": 4.812,
      "mean_token_accuracy": 0.24071256667375565,
      "num_tokens": 108144225.0,
      "step": 62345
    },
    {
      "entropy": 5.793784236907959,
      "epoch": 4.851040653569345,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002883565992811629,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2242927297949791,
      "num_tokens": 108152701.0,
      "step": 62350
    },
    {
      "entropy": 5.846787691116333,
      "epoch": 4.851429682941062,
      "grad_norm": 1.375,
      "learning_rate": 0.0002883289324691998,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2189866080880165,
      "num_tokens": 108161467.0,
      "step": 62355
    },
    {
      "entropy": 5.904319477081299,
      "epoch": 4.851818712312779,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002883012654549913,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21888498067855836,
      "num_tokens": 108170725.0,
      "step": 62360
    },
    {
      "entropy": 5.919288396835327,
      "epoch": 4.852207741684497,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028827359823895726,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21134620159864426,
      "num_tokens": 108179173.0,
      "step": 62365
    },
    {
      "entropy": 5.925215387344361,
      "epoch": 4.852596771056215,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028824593082151754,
      "loss": 4.9115,
      "mean_token_accuracy": 0.22222490906715392,
      "num_tokens": 108187976.0,
      "step": 62370
    },
    {
      "entropy": 5.916512823104858,
      "epoch": 4.852985800427932,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002882182632030919,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21128493249416352,
      "num_tokens": 108196272.0,
      "step": 62375
    },
    {
      "entropy": 5.852432107925415,
      "epoch": 4.85337482979965,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002881905953841002,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21789026260375977,
      "num_tokens": 108204715.0,
      "step": 62380
    },
    {
      "entropy": 5.91992678642273,
      "epoch": 4.853763859171368,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002881629273649621,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21823257803916932,
      "num_tokens": 108213122.0,
      "step": 62385
    },
    {
      "entropy": 5.866619634628296,
      "epoch": 4.854152888543085,
      "grad_norm": 1.5,
      "learning_rate": 0.00028813525914609767,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21817003488540648,
      "num_tokens": 108221792.0,
      "step": 62390
    },
    {
      "entropy": 5.913182544708252,
      "epoch": 4.854541917914803,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028810759072792644,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21674841940402984,
      "num_tokens": 108231542.0,
      "step": 62395
    },
    {
      "entropy": 5.900798654556274,
      "epoch": 4.85493094728652,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002880799221108684,
      "loss": 5.0144,
      "mean_token_accuracy": 0.20791709572076797,
      "num_tokens": 108240565.0,
      "step": 62400
    },
    {
      "entropy": 5.951684856414795,
      "epoch": 4.855319976658238,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002880522532953435,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20978060662746428,
      "num_tokens": 108249058.0,
      "step": 62405
    },
    {
      "entropy": 5.896510744094849,
      "epoch": 4.855709006029955,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002880245842817713,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22688811123371125,
      "num_tokens": 108256659.0,
      "step": 62410
    },
    {
      "entropy": 5.859241056442261,
      "epoch": 4.856098035401673,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002879969150705718,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22180895507335663,
      "num_tokens": 108265230.0,
      "step": 62415
    },
    {
      "entropy": 5.9231500148773195,
      "epoch": 4.856487064773391,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002879692456621649,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20798333436250688,
      "num_tokens": 108273563.0,
      "step": 62420
    },
    {
      "entropy": 5.845493125915527,
      "epoch": 4.8568760941451075,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002879415760569702,
      "loss": 4.8982,
      "mean_token_accuracy": 0.2185686230659485,
      "num_tokens": 108281746.0,
      "step": 62425
    },
    {
      "entropy": 5.91968002319336,
      "epoch": 4.857265123516825,
      "grad_norm": 1.375,
      "learning_rate": 0.0002879139062554077,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2107301712036133,
      "num_tokens": 108290066.0,
      "step": 62430
    },
    {
      "entropy": 5.938291072845459,
      "epoch": 4.857654152888543,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028788623625789725,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2107479214668274,
      "num_tokens": 108298984.0,
      "step": 62435
    },
    {
      "entropy": 5.920580101013184,
      "epoch": 4.8580431822602606,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002878585660648587,
      "loss": 4.9159,
      "mean_token_accuracy": 0.21830906271934508,
      "num_tokens": 108308101.0,
      "step": 62440
    },
    {
      "entropy": 5.932555055618286,
      "epoch": 4.858432211631978,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002878308956767118,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2188437357544899,
      "num_tokens": 108316679.0,
      "step": 62445
    },
    {
      "entropy": 5.871972560882568,
      "epoch": 4.858821241003696,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002878032250938766,
      "loss": 4.9082,
      "mean_token_accuracy": 0.225163072347641,
      "num_tokens": 108325751.0,
      "step": 62450
    },
    {
      "entropy": 5.90297498703003,
      "epoch": 4.859210270375414,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028777555431677276,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2088163435459137,
      "num_tokens": 108335000.0,
      "step": 62455
    },
    {
      "entropy": 5.814658689498901,
      "epoch": 4.859599299747131,
      "grad_norm": 1.375,
      "learning_rate": 0.00028774788334582025,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21545420140028,
      "num_tokens": 108343862.0,
      "step": 62460
    },
    {
      "entropy": 5.900244331359863,
      "epoch": 4.859988329118848,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002877202121814389,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20971643477678298,
      "num_tokens": 108353521.0,
      "step": 62465
    },
    {
      "entropy": 5.820962142944336,
      "epoch": 4.860377358490566,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002876925408240485,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22710434049367906,
      "num_tokens": 108362633.0,
      "step": 62470
    },
    {
      "entropy": 5.8281372547149655,
      "epoch": 4.8607663878622835,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002876648692740691,
      "loss": 4.8014,
      "mean_token_accuracy": 0.2258273482322693,
      "num_tokens": 108371376.0,
      "step": 62475
    },
    {
      "entropy": 5.847288322448731,
      "epoch": 4.861155417234001,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028763719753192044,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2301265612244606,
      "num_tokens": 108379107.0,
      "step": 62480
    },
    {
      "entropy": 5.9092645168304445,
      "epoch": 4.861544446605719,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002876095255980224,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20798783153295516,
      "num_tokens": 108387001.0,
      "step": 62485
    },
    {
      "entropy": 5.927778720855713,
      "epoch": 4.861933475977437,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028758185347279485,
      "loss": 4.9452,
      "mean_token_accuracy": 0.219222030043602,
      "num_tokens": 108395513.0,
      "step": 62490
    },
    {
      "entropy": 5.903514719009399,
      "epoch": 4.8623225053491534,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002875541811566578,
      "loss": 4.928,
      "mean_token_accuracy": 0.21933302879333497,
      "num_tokens": 108404304.0,
      "step": 62495
    },
    {
      "entropy": 5.872148609161377,
      "epoch": 4.862711534720871,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002875265086500309,
      "loss": 4.929,
      "mean_token_accuracy": 0.21544578224420546,
      "num_tokens": 108412610.0,
      "step": 62500
    },
    {
      "entropy": 5.844034910202026,
      "epoch": 4.863100564092589,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028749883595333424,
      "loss": 4.8393,
      "mean_token_accuracy": 0.2292870834469795,
      "num_tokens": 108421007.0,
      "step": 62505
    },
    {
      "entropy": 5.792699527740479,
      "epoch": 4.8634895934643065,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002874711630669876,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2186249554157257,
      "num_tokens": 108429574.0,
      "step": 62510
    },
    {
      "entropy": 5.840561819076538,
      "epoch": 4.863878622836024,
      "grad_norm": 1.546875,
      "learning_rate": 0.000287443489991411,
      "loss": 4.825,
      "mean_token_accuracy": 0.22579430490732194,
      "num_tokens": 108438097.0,
      "step": 62515
    },
    {
      "entropy": 5.861608505249023,
      "epoch": 4.864267652207742,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002874158167270242,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2143292471766472,
      "num_tokens": 108447014.0,
      "step": 62520
    },
    {
      "entropy": 5.870207262039185,
      "epoch": 4.86465668157946,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028738814327424714,
      "loss": 4.9529,
      "mean_token_accuracy": 0.219376839697361,
      "num_tokens": 108456094.0,
      "step": 62525
    },
    {
      "entropy": 5.91980390548706,
      "epoch": 4.865045710951176,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002873604696334997,
      "loss": 4.9705,
      "mean_token_accuracy": 0.2123442217707634,
      "num_tokens": 108464799.0,
      "step": 62530
    },
    {
      "entropy": 5.878115320205689,
      "epoch": 4.865434740322894,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028733279580520185,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2252853259444237,
      "num_tokens": 108472677.0,
      "step": 62535
    },
    {
      "entropy": 5.8735191822052,
      "epoch": 4.865823769694612,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028730512178977343,
      "loss": 4.988,
      "mean_token_accuracy": 0.21094561517238616,
      "num_tokens": 108481589.0,
      "step": 62540
    },
    {
      "entropy": 5.9461023807525635,
      "epoch": 4.8662127990663295,
      "grad_norm": 1.375,
      "learning_rate": 0.00028727744758763434,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21543501764535905,
      "num_tokens": 108490485.0,
      "step": 62545
    },
    {
      "entropy": 5.969494009017945,
      "epoch": 4.866601828438047,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028724977319920453,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21681982576847075,
      "num_tokens": 108499154.0,
      "step": 62550
    },
    {
      "entropy": 5.932147884368897,
      "epoch": 4.866990857809765,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00028722209862490393,
      "loss": 4.9918,
      "mean_token_accuracy": 0.20581575781106948,
      "num_tokens": 108507210.0,
      "step": 62555
    },
    {
      "entropy": 5.943839168548584,
      "epoch": 4.867379887181482,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028719442386515245,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21154083013534547,
      "num_tokens": 108515696.0,
      "step": 62560
    },
    {
      "entropy": 5.842122554779053,
      "epoch": 4.867768916553199,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028716674892036994,
      "loss": 4.8509,
      "mean_token_accuracy": 0.2205411598086357,
      "num_tokens": 108524554.0,
      "step": 62565
    },
    {
      "entropy": 5.907657814025879,
      "epoch": 4.868157945924917,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002871390737909765,
      "loss": 4.885,
      "mean_token_accuracy": 0.21514279246330262,
      "num_tokens": 108533284.0,
      "step": 62570
    },
    {
      "entropy": 5.942115449905396,
      "epoch": 4.868546975296635,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028711139847739187,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21199230402708052,
      "num_tokens": 108541629.0,
      "step": 62575
    },
    {
      "entropy": 5.877481079101562,
      "epoch": 4.8689360046683525,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00028708372298003607,
      "loss": 4.9362,
      "mean_token_accuracy": 0.2135868862271309,
      "num_tokens": 108549879.0,
      "step": 62580
    },
    {
      "entropy": 5.92747745513916,
      "epoch": 4.86932503404007,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028705604729932895,
      "loss": 5.0239,
      "mean_token_accuracy": 0.211185921728611,
      "num_tokens": 108557902.0,
      "step": 62585
    },
    {
      "entropy": 5.870274257659912,
      "epoch": 4.869714063411788,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028702837143569057,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22543398886919022,
      "num_tokens": 108566354.0,
      "step": 62590
    },
    {
      "entropy": 5.8789654731750485,
      "epoch": 4.870103092783506,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028700069538954064,
      "loss": 4.971,
      "mean_token_accuracy": 0.21002316176891328,
      "num_tokens": 108574844.0,
      "step": 62595
    },
    {
      "entropy": 5.8980919361114506,
      "epoch": 4.870492122155222,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028697301916129944,
      "loss": 4.893,
      "mean_token_accuracy": 0.21198370009660722,
      "num_tokens": 108583284.0,
      "step": 62600
    },
    {
      "entropy": 5.835699987411499,
      "epoch": 4.87088115152694,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002869453427513866,
      "loss": 4.9332,
      "mean_token_accuracy": 0.21970461308956146,
      "num_tokens": 108592005.0,
      "step": 62605
    },
    {
      "entropy": 5.809485197067261,
      "epoch": 4.871270180898658,
      "grad_norm": 1.53125,
      "learning_rate": 0.00028691766616022234,
      "loss": 4.8088,
      "mean_token_accuracy": 0.2306772992014885,
      "num_tokens": 108601516.0,
      "step": 62610
    },
    {
      "entropy": 5.911069059371949,
      "epoch": 4.8716592102703755,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002868899893882264,
      "loss": 4.9375,
      "mean_token_accuracy": 0.214309224486351,
      "num_tokens": 108610794.0,
      "step": 62615
    },
    {
      "entropy": 5.914481258392334,
      "epoch": 4.872048239642093,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028686231243581873,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22024333477020264,
      "num_tokens": 108619508.0,
      "step": 62620
    },
    {
      "entropy": 5.858764934539795,
      "epoch": 4.872437269013811,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028683463530341944,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22517129480838777,
      "num_tokens": 108627667.0,
      "step": 62625
    },
    {
      "entropy": 5.9410676002502445,
      "epoch": 4.872826298385528,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002868069579914483,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21013107448816298,
      "num_tokens": 108636756.0,
      "step": 62630
    },
    {
      "entropy": 5.864767074584961,
      "epoch": 4.873215327757245,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028677928050032547,
      "loss": 4.8759,
      "mean_token_accuracy": 0.22185056507587433,
      "num_tokens": 108645388.0,
      "step": 62635
    },
    {
      "entropy": 5.844821405410767,
      "epoch": 4.873604357128963,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028675160283047073,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22035716325044633,
      "num_tokens": 108655406.0,
      "step": 62640
    },
    {
      "entropy": 5.888237762451172,
      "epoch": 4.873993386500681,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00028672392498230414,
      "loss": 4.9096,
      "mean_token_accuracy": 0.21594117432832718,
      "num_tokens": 108663733.0,
      "step": 62645
    },
    {
      "entropy": 5.8950825214385985,
      "epoch": 4.8743824158723985,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00028669624695624573,
      "loss": 4.9013,
      "mean_token_accuracy": 0.21335013210773468,
      "num_tokens": 108671914.0,
      "step": 62650
    },
    {
      "entropy": 5.861316728591919,
      "epoch": 4.874771445244116,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00028666856875271535,
      "loss": 4.966,
      "mean_token_accuracy": 0.21195148676633835,
      "num_tokens": 108679940.0,
      "step": 62655
    },
    {
      "entropy": 5.871832656860351,
      "epoch": 4.875160474615834,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002866408903721329,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21383584290742874,
      "num_tokens": 108688297.0,
      "step": 62660
    },
    {
      "entropy": 5.943823051452637,
      "epoch": 4.875549503987551,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028661321181491854,
      "loss": 5.0531,
      "mean_token_accuracy": 0.21138908714056015,
      "num_tokens": 108697422.0,
      "step": 62665
    },
    {
      "entropy": 5.888514804840088,
      "epoch": 4.875938533359268,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002865855330814922,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2148138016462326,
      "num_tokens": 108706342.0,
      "step": 62670
    },
    {
      "entropy": 5.856881523132325,
      "epoch": 4.876327562730986,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002865578541722738,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2264973357319832,
      "num_tokens": 108715406.0,
      "step": 62675
    },
    {
      "entropy": 5.91471791267395,
      "epoch": 4.876716592102704,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002865301750876834,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21009822487831115,
      "num_tokens": 108725137.0,
      "step": 62680
    },
    {
      "entropy": 5.924348592758179,
      "epoch": 4.8771056214744215,
      "grad_norm": 1.375,
      "learning_rate": 0.0002865024958281409,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21461839675903321,
      "num_tokens": 108733243.0,
      "step": 62685
    },
    {
      "entropy": 5.925912666320801,
      "epoch": 4.877494650846139,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028647481639406644,
      "loss": 5.0546,
      "mean_token_accuracy": 0.20848291218280793,
      "num_tokens": 108742242.0,
      "step": 62690
    },
    {
      "entropy": 5.940831995010376,
      "epoch": 4.877883680217856,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028644713678587983,
      "loss": 4.98,
      "mean_token_accuracy": 0.21553827822208405,
      "num_tokens": 108750736.0,
      "step": 62695
    },
    {
      "entropy": 5.905335569381714,
      "epoch": 4.878272709589574,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028641945700400104,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22155216038227082,
      "num_tokens": 108759229.0,
      "step": 62700
    },
    {
      "entropy": 5.861728858947754,
      "epoch": 4.878661738961291,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028639177704885033,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21700639724731446,
      "num_tokens": 108767805.0,
      "step": 62705
    },
    {
      "entropy": 5.920863962173462,
      "epoch": 4.879050768333009,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028636409692084743,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21157653480768204,
      "num_tokens": 108776529.0,
      "step": 62710
    },
    {
      "entropy": 5.849791193008423,
      "epoch": 4.879439797704727,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002863364166204125,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2203465700149536,
      "num_tokens": 108784944.0,
      "step": 62715
    },
    {
      "entropy": 5.8636078357696535,
      "epoch": 4.8798288270764445,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028630873614796543,
      "loss": 4.8544,
      "mean_token_accuracy": 0.21947544813156128,
      "num_tokens": 108793225.0,
      "step": 62720
    },
    {
      "entropy": 5.858047103881836,
      "epoch": 4.880217856448162,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028628105550392637,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21337395608425141,
      "num_tokens": 108801794.0,
      "step": 62725
    },
    {
      "entropy": 5.830312776565552,
      "epoch": 4.88060688581988,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002862533746887152,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21703822314739227,
      "num_tokens": 108810938.0,
      "step": 62730
    },
    {
      "entropy": 5.887650823593139,
      "epoch": 4.880995915191597,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028622569370275196,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22171899676322937,
      "num_tokens": 108820160.0,
      "step": 62735
    },
    {
      "entropy": 5.889841079711914,
      "epoch": 4.881384944563314,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00028619801254645675,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2191880688071251,
      "num_tokens": 108828929.0,
      "step": 62740
    },
    {
      "entropy": 5.828569221496582,
      "epoch": 4.881773973935032,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002861703312202495,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21420105397701264,
      "num_tokens": 108837543.0,
      "step": 62745
    },
    {
      "entropy": 5.824690675735473,
      "epoch": 4.88216300330675,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028614264972455023,
      "loss": 4.8918,
      "mean_token_accuracy": 0.21821432262659074,
      "num_tokens": 108847109.0,
      "step": 62750
    },
    {
      "entropy": 5.909644985198975,
      "epoch": 4.8825520326784675,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028611496805977906,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22138262838125228,
      "num_tokens": 108856350.0,
      "step": 62755
    },
    {
      "entropy": 5.805631113052368,
      "epoch": 4.882941062050184,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028608728622635585,
      "loss": 4.7841,
      "mean_token_accuracy": 0.22793570756912232,
      "num_tokens": 108864709.0,
      "step": 62760
    },
    {
      "entropy": 5.87542576789856,
      "epoch": 4.883330091421902,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028605960422470073,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21415722668170928,
      "num_tokens": 108873876.0,
      "step": 62765
    },
    {
      "entropy": 5.8209922313690186,
      "epoch": 4.88371912079362,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002860319220552338,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22071389555931092,
      "num_tokens": 108882408.0,
      "step": 62770
    },
    {
      "entropy": 5.823239850997925,
      "epoch": 4.884108150165337,
      "grad_norm": 1.5625,
      "learning_rate": 0.00028600423971837484,
      "loss": 4.849,
      "mean_token_accuracy": 0.2264445975422859,
      "num_tokens": 108890548.0,
      "step": 62775
    },
    {
      "entropy": 5.882374000549317,
      "epoch": 4.884497179537055,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028597655721454427,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21895410269498825,
      "num_tokens": 108898690.0,
      "step": 62780
    },
    {
      "entropy": 5.850329065322876,
      "epoch": 4.884886208908773,
      "grad_norm": 1.484375,
      "learning_rate": 0.00028594887454416176,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2171856239438057,
      "num_tokens": 108907198.0,
      "step": 62785
    },
    {
      "entropy": 5.883897304534912,
      "epoch": 4.8852752382804905,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002859211917076476,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2156778246164322,
      "num_tokens": 108916451.0,
      "step": 62790
    },
    {
      "entropy": 5.89300856590271,
      "epoch": 4.885664267652208,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028589350870542165,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21836104542016982,
      "num_tokens": 108924975.0,
      "step": 62795
    },
    {
      "entropy": 5.872204732894898,
      "epoch": 4.886053297023925,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002858658255379041,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21131481230258942,
      "num_tokens": 108933114.0,
      "step": 62800
    },
    {
      "entropy": 5.786011171340943,
      "epoch": 4.886442326395643,
      "grad_norm": 1.375,
      "learning_rate": 0.00028583814220551495,
      "loss": 4.812,
      "mean_token_accuracy": 0.2269524872303009,
      "num_tokens": 108941429.0,
      "step": 62805
    },
    {
      "entropy": 5.931719589233398,
      "epoch": 4.88683135576736,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002858104587086742,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21086300015449524,
      "num_tokens": 108949762.0,
      "step": 62810
    },
    {
      "entropy": 5.8741038799285885,
      "epoch": 4.887220385139078,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000285782775047802,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21982812881469727,
      "num_tokens": 108958488.0,
      "step": 62815
    },
    {
      "entropy": 5.839327335357666,
      "epoch": 4.887609414510796,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002857550912233183,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2202258974313736,
      "num_tokens": 108966976.0,
      "step": 62820
    },
    {
      "entropy": 5.793496131896973,
      "epoch": 4.8879984438825135,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002857274072356432,
      "loss": 4.8581,
      "mean_token_accuracy": 0.21851739883422852,
      "num_tokens": 108975435.0,
      "step": 62825
    },
    {
      "entropy": 5.843374538421631,
      "epoch": 4.88838747325423,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028569972308519674,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22015581876039506,
      "num_tokens": 108983721.0,
      "step": 62830
    },
    {
      "entropy": 5.894990301132202,
      "epoch": 4.888776502625948,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028567203877239914,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21833908706903457,
      "num_tokens": 108992376.0,
      "step": 62835
    },
    {
      "entropy": 5.89663872718811,
      "epoch": 4.889165531997666,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028564435429767035,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21532341837882996,
      "num_tokens": 109001049.0,
      "step": 62840
    },
    {
      "entropy": 5.926431035995483,
      "epoch": 4.889554561369383,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002856166696614303,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2161971688270569,
      "num_tokens": 109009859.0,
      "step": 62845
    },
    {
      "entropy": 5.923861694335938,
      "epoch": 4.889943590741101,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002855889848640992,
      "loss": 5.0266,
      "mean_token_accuracy": 0.21285324841737746,
      "num_tokens": 109018491.0,
      "step": 62850
    },
    {
      "entropy": 5.834464120864868,
      "epoch": 4.890332620112819,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002855612999060971,
      "loss": 4.911,
      "mean_token_accuracy": 0.21825989484786987,
      "num_tokens": 109026702.0,
      "step": 62855
    },
    {
      "entropy": 5.880823802947998,
      "epoch": 4.8907216494845365,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002855336147878442,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21308285892009735,
      "num_tokens": 109036382.0,
      "step": 62860
    },
    {
      "entropy": 5.898878145217895,
      "epoch": 4.891110678856253,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028550592950976036,
      "loss": 4.9508,
      "mean_token_accuracy": 0.20910721719264985,
      "num_tokens": 109045149.0,
      "step": 62865
    },
    {
      "entropy": 5.913369846343994,
      "epoch": 4.891499708227971,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028547824407226574,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21461388021707534,
      "num_tokens": 109053937.0,
      "step": 62870
    },
    {
      "entropy": 5.891812181472778,
      "epoch": 4.891888737599689,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002854505584757805,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22122907638549805,
      "num_tokens": 109062812.0,
      "step": 62875
    },
    {
      "entropy": 5.91003189086914,
      "epoch": 4.892277766971406,
      "grad_norm": 1.5,
      "learning_rate": 0.00028542287272072475,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22089508920907974,
      "num_tokens": 109072335.0,
      "step": 62880
    },
    {
      "entropy": 5.890327262878418,
      "epoch": 4.892666796343124,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002853951868075184,
      "loss": 4.9568,
      "mean_token_accuracy": 0.20583431720733641,
      "num_tokens": 109081334.0,
      "step": 62885
    },
    {
      "entropy": 5.91676959991455,
      "epoch": 4.893055825714842,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002853675007365816,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21861480176448822,
      "num_tokens": 109089995.0,
      "step": 62890
    },
    {
      "entropy": 5.963656425476074,
      "epoch": 4.893444855086559,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002853398145083345,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20463999807834626,
      "num_tokens": 109097791.0,
      "step": 62895
    },
    {
      "entropy": 5.839676713943481,
      "epoch": 4.893833884458276,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028531212812319724,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2175234168767929,
      "num_tokens": 109106256.0,
      "step": 62900
    },
    {
      "entropy": 5.891665363311768,
      "epoch": 4.894222913829994,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002852844415815898,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21448959559202194,
      "num_tokens": 109115862.0,
      "step": 62905
    },
    {
      "entropy": 5.84844217300415,
      "epoch": 4.894611943201712,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028525675488393234,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21901315003633498,
      "num_tokens": 109124127.0,
      "step": 62910
    },
    {
      "entropy": 5.9595317363739015,
      "epoch": 4.895000972573429,
      "grad_norm": 1.609375,
      "learning_rate": 0.00028522906803064494,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20333793759346008,
      "num_tokens": 109132975.0,
      "step": 62915
    },
    {
      "entropy": 5.950059700012207,
      "epoch": 4.895390001945147,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002852013810221477,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21391311883926392,
      "num_tokens": 109142398.0,
      "step": 62920
    },
    {
      "entropy": 5.888796710968018,
      "epoch": 4.895779031316865,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002851736938588607,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22038516104221345,
      "num_tokens": 109151043.0,
      "step": 62925
    },
    {
      "entropy": 5.88486270904541,
      "epoch": 4.8961680606885825,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002851460065412042,
      "loss": 4.903,
      "mean_token_accuracy": 0.21586000621318818,
      "num_tokens": 109159671.0,
      "step": 62930
    },
    {
      "entropy": 5.867138004302978,
      "epoch": 4.896557090060299,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002851183190695981,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22196243405342103,
      "num_tokens": 109168358.0,
      "step": 62935
    },
    {
      "entropy": 5.909192276000977,
      "epoch": 4.896946119432017,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002850906314444627,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21271082907915115,
      "num_tokens": 109176698.0,
      "step": 62940
    },
    {
      "entropy": 5.829588842391968,
      "epoch": 4.897335148803735,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028506294366621796,
      "loss": 4.864,
      "mean_token_accuracy": 0.2202983409166336,
      "num_tokens": 109185362.0,
      "step": 62945
    },
    {
      "entropy": 5.941349935531616,
      "epoch": 4.897724178175452,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028503525573528413,
      "loss": 5.0275,
      "mean_token_accuracy": 0.20749955326318742,
      "num_tokens": 109193719.0,
      "step": 62950
    },
    {
      "entropy": 5.968068075180054,
      "epoch": 4.89811320754717,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002850075676520812,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21175200194120408,
      "num_tokens": 109202138.0,
      "step": 62955
    },
    {
      "entropy": 5.898190927505493,
      "epoch": 4.898502236918888,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002849798794170294,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2161297544836998,
      "num_tokens": 109211065.0,
      "step": 62960
    },
    {
      "entropy": 5.943305397033692,
      "epoch": 4.898891266290605,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002849521910305488,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2180306151509285,
      "num_tokens": 109219292.0,
      "step": 62965
    },
    {
      "entropy": 5.882408285140992,
      "epoch": 4.899280295662322,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028492450249305957,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22348613291978836,
      "num_tokens": 109228332.0,
      "step": 62970
    },
    {
      "entropy": 5.845983409881592,
      "epoch": 4.89966932503404,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028489681380498177,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22201534807682038,
      "num_tokens": 109237299.0,
      "step": 62975
    },
    {
      "entropy": 5.895946931838989,
      "epoch": 4.900058354405758,
      "grad_norm": 1.59375,
      "learning_rate": 0.00028486912496673557,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22163930386304856,
      "num_tokens": 109245301.0,
      "step": 62980
    },
    {
      "entropy": 5.824560880661011,
      "epoch": 4.900447383777475,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028484143597874115,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21273715198040008,
      "num_tokens": 109253695.0,
      "step": 62985
    },
    {
      "entropy": 5.84009485244751,
      "epoch": 4.900836413149193,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002848137468414186,
      "loss": 4.9393,
      "mean_token_accuracy": 0.211474047601223,
      "num_tokens": 109262381.0,
      "step": 62990
    },
    {
      "entropy": 5.886327314376831,
      "epoch": 4.901225442520911,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000284786057555188,
      "loss": 4.9212,
      "mean_token_accuracy": 0.21678577363491058,
      "num_tokens": 109270765.0,
      "step": 62995
    },
    {
      "entropy": 5.915979814529419,
      "epoch": 4.9016144718926276,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002847583681204696,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2226930856704712,
      "num_tokens": 109278719.0,
      "step": 63000
    },
    {
      "epoch": 4.9016144718926276,
      "eval_entropy": 5.592337388781198,
      "eval_loss": 5.036297798156738,
      "eval_mean_token_accuracy": 0.2212241230559418,
      "eval_num_tokens": 109278719.0,
      "eval_runtime": 22.0294,
      "eval_samples_per_second": 1886.482,
      "eval_steps_per_second": 235.822,
      "step": 63000
    },
    {
      "entropy": 5.845660972595215,
      "epoch": 4.902003501264345,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028473067853768347,
      "loss": 4.8916,
      "mean_token_accuracy": 0.21708036661148072,
      "num_tokens": 109287000.0,
      "step": 63005
    },
    {
      "entropy": 5.947008657455444,
      "epoch": 4.902392530636063,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028470298880724985,
      "loss": 5.0595,
      "mean_token_accuracy": 0.2071488693356514,
      "num_tokens": 109296029.0,
      "step": 63010
    },
    {
      "entropy": 5.861974382400513,
      "epoch": 4.902781560007781,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002846752989295888,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21617467552423478,
      "num_tokens": 109303990.0,
      "step": 63015
    },
    {
      "entropy": 5.922281980514526,
      "epoch": 4.903170589379498,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002846476089051204,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21947211176156997,
      "num_tokens": 109312922.0,
      "step": 63020
    },
    {
      "entropy": 5.831495952606201,
      "epoch": 4.903559618751216,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028461991873426494,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2157473534345627,
      "num_tokens": 109321296.0,
      "step": 63025
    },
    {
      "entropy": 5.933815956115723,
      "epoch": 4.903948648122933,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002845922284174426,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2108047276735306,
      "num_tokens": 109329541.0,
      "step": 63030
    },
    {
      "entropy": 5.900624561309814,
      "epoch": 4.9043376774946505,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028456453795507335,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2303169399499893,
      "num_tokens": 109338054.0,
      "step": 63035
    },
    {
      "entropy": 5.882947683334351,
      "epoch": 4.904726706866368,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002845368473475776,
      "loss": 4.8945,
      "mean_token_accuracy": 0.225942961871624,
      "num_tokens": 109346216.0,
      "step": 63040
    },
    {
      "entropy": 5.8657355308532715,
      "epoch": 4.905115736238086,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028450915659537524,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21595780402421952,
      "num_tokens": 109355033.0,
      "step": 63045
    },
    {
      "entropy": 5.864536285400391,
      "epoch": 4.905504765609804,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002844814656988866,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2164296329021454,
      "num_tokens": 109363515.0,
      "step": 63050
    },
    {
      "entropy": 5.975519323348999,
      "epoch": 4.905893794981521,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028445377465853176,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20222811847925187,
      "num_tokens": 109373063.0,
      "step": 63055
    },
    {
      "entropy": 5.861296272277832,
      "epoch": 4.906282824353239,
      "grad_norm": 1.625,
      "learning_rate": 0.00028442608347473093,
      "loss": 4.8597,
      "mean_token_accuracy": 0.21710907518863679,
      "num_tokens": 109380781.0,
      "step": 63060
    },
    {
      "entropy": 5.787471628189087,
      "epoch": 4.906671853724957,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028439839214790437,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2307184636592865,
      "num_tokens": 109389222.0,
      "step": 63065
    },
    {
      "entropy": 5.92974820137024,
      "epoch": 4.9070608830966735,
      "grad_norm": 1.5,
      "learning_rate": 0.0002843707006784722,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21363822668790816,
      "num_tokens": 109398101.0,
      "step": 63070
    },
    {
      "entropy": 5.906003093719482,
      "epoch": 4.907449912468391,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002843430090668545,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2221415624022484,
      "num_tokens": 109406508.0,
      "step": 63075
    },
    {
      "entropy": 5.955307722091675,
      "epoch": 4.907838941840109,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028431531731347155,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20807292610406875,
      "num_tokens": 109414943.0,
      "step": 63080
    },
    {
      "entropy": 5.856168031692505,
      "epoch": 4.908227971211827,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002842876254187434,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2138863116502762,
      "num_tokens": 109424438.0,
      "step": 63085
    },
    {
      "entropy": 5.892143487930298,
      "epoch": 4.908617000583544,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028425993338309036,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21412809789180756,
      "num_tokens": 109433108.0,
      "step": 63090
    },
    {
      "entropy": 5.917541027069092,
      "epoch": 4.909006029955261,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002842322412069325,
      "loss": 4.9857,
      "mean_token_accuracy": 0.20951674431562423,
      "num_tokens": 109442019.0,
      "step": 63095
    },
    {
      "entropy": 5.926276397705078,
      "epoch": 4.909395059326979,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002842045488906902,
      "loss": 4.9763,
      "mean_token_accuracy": 0.217173233628273,
      "num_tokens": 109450819.0,
      "step": 63100
    },
    {
      "entropy": 5.943784236907959,
      "epoch": 4.9097840886986965,
      "grad_norm": 1.515625,
      "learning_rate": 0.00028417685643478347,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2140013337135315,
      "num_tokens": 109459551.0,
      "step": 63105
    },
    {
      "entropy": 5.902006053924561,
      "epoch": 4.910173118070414,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028414916383963265,
      "loss": 4.944,
      "mean_token_accuracy": 0.2118055507540703,
      "num_tokens": 109468351.0,
      "step": 63110
    },
    {
      "entropy": 5.859724617004394,
      "epoch": 4.910562147442132,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002841214711056577,
      "loss": 4.9004,
      "mean_token_accuracy": 0.2193540558218956,
      "num_tokens": 109477951.0,
      "step": 63115
    },
    {
      "entropy": 5.843268775939942,
      "epoch": 4.91095117681385,
      "grad_norm": 1.421875,
      "learning_rate": 0.000284093778233279,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22145945578813553,
      "num_tokens": 109486113.0,
      "step": 63120
    },
    {
      "entropy": 5.892044973373413,
      "epoch": 4.911340206185567,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002840660852229166,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21553106009960174,
      "num_tokens": 109494566.0,
      "step": 63125
    },
    {
      "entropy": 5.869701814651489,
      "epoch": 4.911729235557285,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002840383920749908,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21697404533624648,
      "num_tokens": 109503738.0,
      "step": 63130
    },
    {
      "entropy": 5.853439426422119,
      "epoch": 4.912118264929002,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002840106987899219,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22182758450508117,
      "num_tokens": 109512155.0,
      "step": 63135
    },
    {
      "entropy": 5.955921268463134,
      "epoch": 4.9125072943007195,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002839830053681299,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20780492275953294,
      "num_tokens": 109521206.0,
      "step": 63140
    },
    {
      "entropy": 5.8957914352417,
      "epoch": 4.912896323672437,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028395531181003516,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22184519916772844,
      "num_tokens": 109529486.0,
      "step": 63145
    },
    {
      "entropy": 5.876897954940796,
      "epoch": 4.913285353044155,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002839276181160577,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21965682357549668,
      "num_tokens": 109538152.0,
      "step": 63150
    },
    {
      "entropy": 5.852418088912964,
      "epoch": 4.913674382415873,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000283899924286618,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21349960565567017,
      "num_tokens": 109546828.0,
      "step": 63155
    },
    {
      "entropy": 5.874808931350708,
      "epoch": 4.91406341178759,
      "grad_norm": 1.421875,
      "learning_rate": 0.000283872230322136,
      "loss": 4.89,
      "mean_token_accuracy": 0.21371839195489883,
      "num_tokens": 109555378.0,
      "step": 63160
    },
    {
      "entropy": 5.88665189743042,
      "epoch": 4.914452441159307,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028384453622303207,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21079155057668686,
      "num_tokens": 109564290.0,
      "step": 63165
    },
    {
      "entropy": 5.81317629814148,
      "epoch": 4.914841470531025,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028381684198972636,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21897746175527572,
      "num_tokens": 109572958.0,
      "step": 63170
    },
    {
      "entropy": 5.888412618637085,
      "epoch": 4.9152304999027425,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028378914762263905,
      "loss": 4.9037,
      "mean_token_accuracy": 0.21928623914718628,
      "num_tokens": 109581669.0,
      "step": 63175
    },
    {
      "entropy": 5.766787719726563,
      "epoch": 4.91561952927446,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028376145312219054,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22201623767614365,
      "num_tokens": 109589890.0,
      "step": 63180
    },
    {
      "entropy": 5.937801933288574,
      "epoch": 4.916008558646178,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002837337584888009,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1991310551762581,
      "num_tokens": 109598522.0,
      "step": 63185
    },
    {
      "entropy": 5.882436895370484,
      "epoch": 4.916397588017896,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002837060637228903,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22643204480409623,
      "num_tokens": 109606527.0,
      "step": 63190
    },
    {
      "entropy": 5.930106925964355,
      "epoch": 4.916786617389613,
      "grad_norm": 1.375,
      "learning_rate": 0.00028367836882487905,
      "loss": 4.9809,
      "mean_token_accuracy": 0.20324242860078812,
      "num_tokens": 109615498.0,
      "step": 63195
    },
    {
      "entropy": 5.893737125396728,
      "epoch": 4.91717564676133,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002836506737951875,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20879970639944076,
      "num_tokens": 109624624.0,
      "step": 63200
    },
    {
      "entropy": 5.880253648757934,
      "epoch": 4.917564676133048,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002836229786342356,
      "loss": 4.901,
      "mean_token_accuracy": 0.216366009414196,
      "num_tokens": 109633076.0,
      "step": 63205
    },
    {
      "entropy": 5.933247661590576,
      "epoch": 4.9179537055047655,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002835952833424438,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22866974025964737,
      "num_tokens": 109640869.0,
      "step": 63210
    },
    {
      "entropy": 5.843147325515747,
      "epoch": 4.918342734876483,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028356758792023217,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21429307609796525,
      "num_tokens": 109649485.0,
      "step": 63215
    },
    {
      "entropy": 5.848204898834228,
      "epoch": 4.918731764248201,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028353989236802114,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2213117867708206,
      "num_tokens": 109658440.0,
      "step": 63220
    },
    {
      "entropy": 5.8321432113647464,
      "epoch": 4.919120793619919,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028351219668623076,
      "loss": 4.899,
      "mean_token_accuracy": 0.2190529078245163,
      "num_tokens": 109667325.0,
      "step": 63225
    },
    {
      "entropy": 5.925417041778564,
      "epoch": 4.919509822991635,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002834845008752813,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2156813845038414,
      "num_tokens": 109675348.0,
      "step": 63230
    },
    {
      "entropy": 5.870559978485107,
      "epoch": 4.919898852363353,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028345680493559313,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22022017389535903,
      "num_tokens": 109683884.0,
      "step": 63235
    },
    {
      "entropy": 5.856965637207031,
      "epoch": 4.920287881735071,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002834291088675864,
      "loss": 4.9643,
      "mean_token_accuracy": 0.21265001296997071,
      "num_tokens": 109692499.0,
      "step": 63240
    },
    {
      "entropy": 5.889193487167359,
      "epoch": 4.9206769111067885,
      "grad_norm": 1.5,
      "learning_rate": 0.00028340141267168135,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2175105705857277,
      "num_tokens": 109700700.0,
      "step": 63245
    },
    {
      "entropy": 5.92943844795227,
      "epoch": 4.921065940478506,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028337371634829824,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21137484312057495,
      "num_tokens": 109709062.0,
      "step": 63250
    },
    {
      "entropy": 5.888901472091675,
      "epoch": 4.921454969850224,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002833460198978573,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2244207516312599,
      "num_tokens": 109718229.0,
      "step": 63255
    },
    {
      "entropy": 5.854289865493774,
      "epoch": 4.921843999221942,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002833183233207788,
      "loss": 4.8716,
      "mean_token_accuracy": 0.21974745988845826,
      "num_tokens": 109726449.0,
      "step": 63260
    },
    {
      "entropy": 5.829619312286377,
      "epoch": 4.922233028593659,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028329062661748295,
      "loss": 4.8328,
      "mean_token_accuracy": 0.21937422454357147,
      "num_tokens": 109734279.0,
      "step": 63265
    },
    {
      "entropy": 5.77800726890564,
      "epoch": 4.922622057965376,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00028326292978839005,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22635478973388673,
      "num_tokens": 109742795.0,
      "step": 63270
    },
    {
      "entropy": 5.9006123542785645,
      "epoch": 4.923011087337094,
      "grad_norm": 1.375,
      "learning_rate": 0.00028323523283392034,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2135780856013298,
      "num_tokens": 109752009.0,
      "step": 63275
    },
    {
      "entropy": 5.891555690765381,
      "epoch": 4.9234001167088115,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00028320753575449415,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21826127171516418,
      "num_tokens": 109760716.0,
      "step": 63280
    },
    {
      "entropy": 5.894096612930298,
      "epoch": 4.923789146080529,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002831798385505316,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22058100551366805,
      "num_tokens": 109769649.0,
      "step": 63285
    },
    {
      "entropy": 5.9327469825744625,
      "epoch": 4.924178175452247,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00028315214122245295,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2197945550084114,
      "num_tokens": 109778058.0,
      "step": 63290
    },
    {
      "entropy": 5.903349828720093,
      "epoch": 4.924567204823965,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002831244437706786,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21896747797727584,
      "num_tokens": 109786100.0,
      "step": 63295
    },
    {
      "entropy": 5.893560600280762,
      "epoch": 4.924956234195681,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002830967461956288,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21065692454576493,
      "num_tokens": 109794736.0,
      "step": 63300
    },
    {
      "entropy": 5.852028703689575,
      "epoch": 4.925345263567399,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028306904849772374,
      "loss": 4.8621,
      "mean_token_accuracy": 0.21240765750408172,
      "num_tokens": 109803293.0,
      "step": 63305
    },
    {
      "entropy": 5.904577350616455,
      "epoch": 4.925734292939117,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002830413506773837,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21251896470785142,
      "num_tokens": 109811639.0,
      "step": 63310
    },
    {
      "entropy": 5.8953687191009525,
      "epoch": 4.9261233223108345,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000283013652735029,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22485243380069733,
      "num_tokens": 109820048.0,
      "step": 63315
    },
    {
      "entropy": 5.95179328918457,
      "epoch": 4.926512351682552,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028298595467107985,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22567200660705566,
      "num_tokens": 109828600.0,
      "step": 63320
    },
    {
      "entropy": 5.842519903182984,
      "epoch": 4.92690138105427,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002829582564859565,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2174249142408371,
      "num_tokens": 109837774.0,
      "step": 63325
    },
    {
      "entropy": 5.868889951705933,
      "epoch": 4.927290410425988,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002829305581800793,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22109757363796234,
      "num_tokens": 109846506.0,
      "step": 63330
    },
    {
      "entropy": 5.861656284332275,
      "epoch": 4.927679439797704,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002829028597538685,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2178611069917679,
      "num_tokens": 109855314.0,
      "step": 63335
    },
    {
      "entropy": 5.942794275283814,
      "epoch": 4.928068469169422,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028287516120774445,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20927295833826065,
      "num_tokens": 109863768.0,
      "step": 63340
    },
    {
      "entropy": 5.878686857223511,
      "epoch": 4.92845749854114,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002828474625421273,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21473852246999742,
      "num_tokens": 109871634.0,
      "step": 63345
    },
    {
      "entropy": 5.87903208732605,
      "epoch": 4.9288465279128575,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002828197637574374,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21939936876296998,
      "num_tokens": 109880846.0,
      "step": 63350
    },
    {
      "entropy": 5.796025371551513,
      "epoch": 4.929235557284575,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028279206485409505,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22162344604730605,
      "num_tokens": 109889614.0,
      "step": 63355
    },
    {
      "entropy": 5.9180442810058596,
      "epoch": 4.929624586656293,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002827643658325205,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21627540290355682,
      "num_tokens": 109897846.0,
      "step": 63360
    },
    {
      "entropy": 5.87094259262085,
      "epoch": 4.93001361602801,
      "grad_norm": 1.421875,
      "learning_rate": 0.000282736666693134,
      "loss": 4.8162,
      "mean_token_accuracy": 0.22752740532159804,
      "num_tokens": 109905893.0,
      "step": 63365
    },
    {
      "entropy": 5.847710752487183,
      "epoch": 4.930402645399727,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002827089674363559,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21025339066982268,
      "num_tokens": 109914137.0,
      "step": 63370
    },
    {
      "entropy": 5.863405323028564,
      "epoch": 4.930791674771445,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002826812680626066,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21870004534721374,
      "num_tokens": 109923113.0,
      "step": 63375
    },
    {
      "entropy": 5.925350284576416,
      "epoch": 4.931180704143163,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002826535685723062,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21588694006204606,
      "num_tokens": 109932165.0,
      "step": 63380
    },
    {
      "entropy": 5.938417577743531,
      "epoch": 4.9315697335148805,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028262586896587517,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21277597397565842,
      "num_tokens": 109941069.0,
      "step": 63385
    },
    {
      "entropy": 5.89747953414917,
      "epoch": 4.931958762886598,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002825981692437336,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2161966383457184,
      "num_tokens": 109950096.0,
      "step": 63390
    },
    {
      "entropy": 5.821556282043457,
      "epoch": 4.932347792258316,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028257046940630197,
      "loss": 4.8358,
      "mean_token_accuracy": 0.2238988146185875,
      "num_tokens": 109959217.0,
      "step": 63395
    },
    {
      "entropy": 5.899184799194336,
      "epoch": 4.932736821630033,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00028254276945400056,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2146111771464348,
      "num_tokens": 109967242.0,
      "step": 63400
    },
    {
      "entropy": 5.865417194366455,
      "epoch": 4.93312585100175,
      "grad_norm": 1.546875,
      "learning_rate": 0.00028251506938724953,
      "loss": 4.912,
      "mean_token_accuracy": 0.2243200033903122,
      "num_tokens": 109976243.0,
      "step": 63405
    },
    {
      "entropy": 5.954142713546753,
      "epoch": 4.933514880373468,
      "grad_norm": 1.375,
      "learning_rate": 0.00028248736920646927,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2075088992714882,
      "num_tokens": 109984980.0,
      "step": 63410
    },
    {
      "entropy": 5.907082653045654,
      "epoch": 4.933903909745186,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028245966891208017,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21756117641925812,
      "num_tokens": 109994070.0,
      "step": 63415
    },
    {
      "entropy": 5.858321905136108,
      "epoch": 4.9342929391169035,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002824319685045025,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21705792397260665,
      "num_tokens": 110002689.0,
      "step": 63420
    },
    {
      "entropy": 5.824409008026123,
      "epoch": 4.934681968488621,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028240426798415645,
      "loss": 4.8097,
      "mean_token_accuracy": 0.22267137467861176,
      "num_tokens": 110011615.0,
      "step": 63425
    },
    {
      "entropy": 5.857821273803711,
      "epoch": 4.935070997860338,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002823765673514625,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21611316204071046,
      "num_tokens": 110020556.0,
      "step": 63430
    },
    {
      "entropy": 5.880559682846069,
      "epoch": 4.935460027232056,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028234886660684084,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2092095583677292,
      "num_tokens": 110029629.0,
      "step": 63435
    },
    {
      "entropy": 5.93054428100586,
      "epoch": 4.935849056603773,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002823211657507118,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2157832056283951,
      "num_tokens": 110039476.0,
      "step": 63440
    },
    {
      "entropy": 5.906149530410767,
      "epoch": 4.936238085975491,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002822934647834958,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21256107985973358,
      "num_tokens": 110048235.0,
      "step": 63445
    },
    {
      "entropy": 5.859096527099609,
      "epoch": 4.936627115347209,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00028226576370561304,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22193579375743866,
      "num_tokens": 110056703.0,
      "step": 63450
    },
    {
      "entropy": 5.832886219024658,
      "epoch": 4.9370161447189265,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028223806251748384,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21847161501646042,
      "num_tokens": 110064932.0,
      "step": 63455
    },
    {
      "entropy": 5.904750871658325,
      "epoch": 4.937405174090644,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002822103612195286,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21442256569862367,
      "num_tokens": 110073715.0,
      "step": 63460
    },
    {
      "entropy": 5.901986694335937,
      "epoch": 4.937794203462362,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002821826598121676,
      "loss": 5.0334,
      "mean_token_accuracy": 0.20750907063484192,
      "num_tokens": 110082626.0,
      "step": 63465
    },
    {
      "entropy": 5.890605354309082,
      "epoch": 4.938183232834079,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002821549582958212,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21418662816286088,
      "num_tokens": 110091768.0,
      "step": 63470
    },
    {
      "entropy": 5.9705322265625,
      "epoch": 4.938572262205796,
      "grad_norm": 1.5,
      "learning_rate": 0.0002821272566709096,
      "loss": 5.049,
      "mean_token_accuracy": 0.21606716960668565,
      "num_tokens": 110100372.0,
      "step": 63475
    },
    {
      "entropy": 5.9737039566040036,
      "epoch": 4.938961291577514,
      "grad_norm": 1.359375,
      "learning_rate": 0.00028209955493785333,
      "loss": 5.0414,
      "mean_token_accuracy": 0.21236948668956757,
      "num_tokens": 110109045.0,
      "step": 63480
    },
    {
      "entropy": 5.839304780960083,
      "epoch": 4.939350320949232,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002820718530970726,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21627112627029418,
      "num_tokens": 110117612.0,
      "step": 63485
    },
    {
      "entropy": 5.838115119934082,
      "epoch": 4.9397393503209495,
      "grad_norm": 1.484375,
      "learning_rate": 0.00028204415114898765,
      "loss": 4.8599,
      "mean_token_accuracy": 0.21942494213581085,
      "num_tokens": 110126068.0,
      "step": 63490
    },
    {
      "entropy": 5.938824796676636,
      "epoch": 4.940128379692667,
      "grad_norm": 1.484375,
      "learning_rate": 0.00028201644909401903,
      "loss": 5.0546,
      "mean_token_accuracy": 0.2052411526441574,
      "num_tokens": 110134742.0,
      "step": 63495
    },
    {
      "entropy": 5.977343559265137,
      "epoch": 4.940517409064384,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028198874693258696,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2094852313399315,
      "num_tokens": 110143065.0,
      "step": 63500
    },
    {
      "entropy": 5.854104995727539,
      "epoch": 4.940906438436102,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002819610446651117,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22552470862865448,
      "num_tokens": 110151605.0,
      "step": 63505
    },
    {
      "entropy": 5.869261837005615,
      "epoch": 4.941295467807819,
      "grad_norm": 1.421875,
      "learning_rate": 0.00028193334229201374,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21594999432563783,
      "num_tokens": 110160645.0,
      "step": 63510
    },
    {
      "entropy": 5.722666692733765,
      "epoch": 4.941684497179537,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00028190563981371334,
      "loss": 4.7748,
      "mean_token_accuracy": 0.2226237028837204,
      "num_tokens": 110169058.0,
      "step": 63515
    },
    {
      "entropy": 5.8366954803466795,
      "epoch": 4.942073526551255,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002818779372306308,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22498802095651627,
      "num_tokens": 110177274.0,
      "step": 63520
    },
    {
      "entropy": 5.81015944480896,
      "epoch": 4.9424625559229725,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00028185023454318646,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22174597680568695,
      "num_tokens": 110186254.0,
      "step": 63525
    },
    {
      "entropy": 5.8742513179779055,
      "epoch": 4.94285158529469,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028182253175180084,
      "loss": 4.8762,
      "mean_token_accuracy": 0.21548761874437333,
      "num_tokens": 110194993.0,
      "step": 63530
    },
    {
      "entropy": 5.856391954421997,
      "epoch": 4.943240614666407,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002817948288568942,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22460591197013854,
      "num_tokens": 110203330.0,
      "step": 63535
    },
    {
      "entropy": 5.849644517898559,
      "epoch": 4.943629644038125,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002817671258588868,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2171579509973526,
      "num_tokens": 110213092.0,
      "step": 63540
    },
    {
      "entropy": 5.905259609222412,
      "epoch": 4.944018673409842,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002817394227581989,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21379088908433913,
      "num_tokens": 110221575.0,
      "step": 63545
    },
    {
      "entropy": 5.986494016647339,
      "epoch": 4.94440770278156,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002817117195552511,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2128536194562912,
      "num_tokens": 110231010.0,
      "step": 63550
    },
    {
      "entropy": 5.903822040557861,
      "epoch": 4.944796732153278,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028168401625046363,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21209092289209366,
      "num_tokens": 110239681.0,
      "step": 63555
    },
    {
      "entropy": 5.857723140716553,
      "epoch": 4.9451857615249954,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002816563128442568,
      "loss": 4.8909,
      "mean_token_accuracy": 0.21939959973096848,
      "num_tokens": 110248190.0,
      "step": 63560
    },
    {
      "entropy": 5.851400423049927,
      "epoch": 4.945574790896712,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028162860933705105,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22191061079502106,
      "num_tokens": 110257651.0,
      "step": 63565
    },
    {
      "entropy": 5.877963018417359,
      "epoch": 4.94596382026843,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002816009057292667,
      "loss": 4.927,
      "mean_token_accuracy": 0.22086296677589418,
      "num_tokens": 110265539.0,
      "step": 63570
    },
    {
      "entropy": 5.96783766746521,
      "epoch": 4.946352849640148,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028157320202132417,
      "loss": 5.0825,
      "mean_token_accuracy": 0.21048882901668547,
      "num_tokens": 110273727.0,
      "step": 63575
    },
    {
      "entropy": 5.87131667137146,
      "epoch": 4.946741879011865,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028154549821364375,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2216145411133766,
      "num_tokens": 110282451.0,
      "step": 63580
    },
    {
      "entropy": 5.7974076747894285,
      "epoch": 4.947130908383583,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00028151779430664575,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21319199353456497,
      "num_tokens": 110291447.0,
      "step": 63585
    },
    {
      "entropy": 5.8686034202575685,
      "epoch": 4.947519937755301,
      "grad_norm": 1.53125,
      "learning_rate": 0.00028149009030075056,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21626957207918168,
      "num_tokens": 110299422.0,
      "step": 63590
    },
    {
      "entropy": 5.802679395675659,
      "epoch": 4.947908967127018,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028146238619637865,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22541553825139998,
      "num_tokens": 110308811.0,
      "step": 63595
    },
    {
      "entropy": 5.90958948135376,
      "epoch": 4.948297996498736,
      "grad_norm": 1.46875,
      "learning_rate": 0.00028143468199395036,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2138850599527359,
      "num_tokens": 110317970.0,
      "step": 63600
    },
    {
      "entropy": 5.90444049835205,
      "epoch": 4.948687025870453,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028140697769388597,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22285108864307404,
      "num_tokens": 110326932.0,
      "step": 63605
    },
    {
      "entropy": 5.913523387908936,
      "epoch": 4.949076055242171,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002813792732966059,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20508617162704468,
      "num_tokens": 110335407.0,
      "step": 63610
    },
    {
      "entropy": 5.923509120941162,
      "epoch": 4.949465084613888,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002813515688025305,
      "loss": 4.9611,
      "mean_token_accuracy": 0.20967944711446762,
      "num_tokens": 110343118.0,
      "step": 63615
    },
    {
      "entropy": 5.891967058181763,
      "epoch": 4.949854113985606,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002813238642120802,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21591049879789354,
      "num_tokens": 110351922.0,
      "step": 63620
    },
    {
      "entropy": 5.950574922561645,
      "epoch": 4.950243143357324,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028129615952567517,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2166236624121666,
      "num_tokens": 110360893.0,
      "step": 63625
    },
    {
      "entropy": 5.842494344711303,
      "epoch": 4.9506321727290405,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002812684547437361,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22448484003543853,
      "num_tokens": 110368947.0,
      "step": 63630
    },
    {
      "entropy": 5.845922470092773,
      "epoch": 4.951021202100758,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002812407498666831,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21805597692728043,
      "num_tokens": 110377710.0,
      "step": 63635
    },
    {
      "entropy": 5.849162197113037,
      "epoch": 4.951410231472476,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002812130448949367,
      "loss": 4.912,
      "mean_token_accuracy": 0.216579432785511,
      "num_tokens": 110385969.0,
      "step": 63640
    },
    {
      "entropy": 5.92461142539978,
      "epoch": 4.951799260844194,
      "grad_norm": 1.484375,
      "learning_rate": 0.00028118533982891727,
      "loss": 4.935,
      "mean_token_accuracy": 0.21646480709314347,
      "num_tokens": 110394935.0,
      "step": 63645
    },
    {
      "entropy": 5.8492515087127686,
      "epoch": 4.952188290215911,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002811576346690451,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22742342650890351,
      "num_tokens": 110403459.0,
      "step": 63650
    },
    {
      "entropy": 5.797526550292969,
      "epoch": 4.952577319587629,
      "grad_norm": 1.609375,
      "learning_rate": 0.00028112992941574074,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22081702053546906,
      "num_tokens": 110411908.0,
      "step": 63655
    },
    {
      "entropy": 5.842470121383667,
      "epoch": 4.952966348959347,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028110222406942424,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21674100458621978,
      "num_tokens": 110419831.0,
      "step": 63660
    },
    {
      "entropy": 5.857442426681518,
      "epoch": 4.953355378331064,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002810745186305164,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22223664075136185,
      "num_tokens": 110428111.0,
      "step": 63665
    },
    {
      "entropy": 5.820807266235351,
      "epoch": 4.953744407702781,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002810468130994373,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22213053703308105,
      "num_tokens": 110437478.0,
      "step": 63670
    },
    {
      "entropy": 5.893518400192261,
      "epoch": 4.954133437074499,
      "grad_norm": 1.53125,
      "learning_rate": 0.00028101910747660745,
      "loss": 4.9069,
      "mean_token_accuracy": 0.21988317370414734,
      "num_tokens": 110445196.0,
      "step": 63675
    },
    {
      "entropy": 5.836811923980713,
      "epoch": 4.954522466446217,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002809914017624472,
      "loss": 4.9314,
      "mean_token_accuracy": 0.21533963978290557,
      "num_tokens": 110453633.0,
      "step": 63680
    },
    {
      "entropy": 5.867154264450074,
      "epoch": 4.954911495817934,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002809636959573771,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22303626090288162,
      "num_tokens": 110462607.0,
      "step": 63685
    },
    {
      "entropy": 5.887663888931274,
      "epoch": 4.955300525189652,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002809359900618173,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21666764169931413,
      "num_tokens": 110471738.0,
      "step": 63690
    },
    {
      "entropy": 5.862070989608765,
      "epoch": 4.95568955456137,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00028090828407618837,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2147664099931717,
      "num_tokens": 110480133.0,
      "step": 63695
    },
    {
      "entropy": 5.877325487136841,
      "epoch": 4.9560785839330865,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00028088057800091054,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21376505494117737,
      "num_tokens": 110489331.0,
      "step": 63700
    },
    {
      "entropy": 5.858159685134888,
      "epoch": 4.956467613304804,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002808528718364043,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21441205441951752,
      "num_tokens": 110497872.0,
      "step": 63705
    },
    {
      "entropy": 5.857069206237793,
      "epoch": 4.956856642676522,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028082516558309005,
      "loss": 4.871,
      "mean_token_accuracy": 0.22673405408859254,
      "num_tokens": 110506602.0,
      "step": 63710
    },
    {
      "entropy": 5.887046957015992,
      "epoch": 4.95724567204824,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002807974592413883,
      "loss": 4.9363,
      "mean_token_accuracy": 0.21468460857868193,
      "num_tokens": 110514714.0,
      "step": 63715
    },
    {
      "entropy": 5.9019293785095215,
      "epoch": 4.957634701419957,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028076975281171926,
      "loss": 4.9544,
      "mean_token_accuracy": 0.2109535276889801,
      "num_tokens": 110523103.0,
      "step": 63720
    },
    {
      "entropy": 5.869557857513428,
      "epoch": 4.958023730791675,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00028074204629450334,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22498640418052673,
      "num_tokens": 110532123.0,
      "step": 63725
    },
    {
      "entropy": 5.817840909957885,
      "epoch": 4.958412760163393,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002807143396901611,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21906392723321916,
      "num_tokens": 110540507.0,
      "step": 63730
    },
    {
      "entropy": 5.901196622848511,
      "epoch": 4.9588017895351095,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00028068663299911286,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21583758443593978,
      "num_tokens": 110549254.0,
      "step": 63735
    },
    {
      "entropy": 5.85047721862793,
      "epoch": 4.959190818906827,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002806589262217789,
      "loss": 4.865,
      "mean_token_accuracy": 0.22236533761024474,
      "num_tokens": 110557192.0,
      "step": 63740
    },
    {
      "entropy": 5.853070449829102,
      "epoch": 4.959579848278545,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00028063121935857985,
      "loss": 4.9189,
      "mean_token_accuracy": 0.219621379673481,
      "num_tokens": 110565888.0,
      "step": 63745
    },
    {
      "entropy": 5.897838306427002,
      "epoch": 4.959968877650263,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002806035124099359,
      "loss": 4.932,
      "mean_token_accuracy": 0.21701975762844086,
      "num_tokens": 110575366.0,
      "step": 63750
    },
    {
      "entropy": 5.969503688812256,
      "epoch": 4.96035790702198,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028057580537626766,
      "loss": 5.0497,
      "mean_token_accuracy": 0.20820772051811218,
      "num_tokens": 110585223.0,
      "step": 63755
    },
    {
      "entropy": 5.926539516448974,
      "epoch": 4.960746936393698,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00028054809825799537,
      "loss": 4.9663,
      "mean_token_accuracy": 0.2113854169845581,
      "num_tokens": 110593799.0,
      "step": 63760
    },
    {
      "entropy": 5.911235094070435,
      "epoch": 4.961135965765415,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028052039105553965,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2192264422774315,
      "num_tokens": 110603853.0,
      "step": 63765
    },
    {
      "entropy": 5.8950474739074705,
      "epoch": 4.9615249951371325,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002804926837693207,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20988713651895524,
      "num_tokens": 110612243.0,
      "step": 63770
    },
    {
      "entropy": 5.897330856323242,
      "epoch": 4.96191402450885,
      "grad_norm": 1.40625,
      "learning_rate": 0.000280464976399759,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21213668584823608,
      "num_tokens": 110621570.0,
      "step": 63775
    },
    {
      "entropy": 5.922954130172729,
      "epoch": 4.962303053880568,
      "grad_norm": 1.390625,
      "learning_rate": 0.000280437268947275,
      "loss": 4.983,
      "mean_token_accuracy": 0.20959360748529435,
      "num_tokens": 110630508.0,
      "step": 63780
    },
    {
      "entropy": 5.900032663345337,
      "epoch": 4.962692083252286,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002804095614122891,
      "loss": 4.984,
      "mean_token_accuracy": 0.21195309460163117,
      "num_tokens": 110639028.0,
      "step": 63785
    },
    {
      "entropy": 5.903289794921875,
      "epoch": 4.963081112624003,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00028038185379522176,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21936021447181703,
      "num_tokens": 110647467.0,
      "step": 63790
    },
    {
      "entropy": 5.852658939361572,
      "epoch": 4.963470141995721,
      "grad_norm": 1.4375,
      "learning_rate": 0.00028035414609649327,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21426823139190673,
      "num_tokens": 110656074.0,
      "step": 63795
    },
    {
      "entropy": 5.873316431045533,
      "epoch": 4.963859171367439,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028032643831652415,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21776635348796844,
      "num_tokens": 110665156.0,
      "step": 63800
    },
    {
      "entropy": 5.925295400619507,
      "epoch": 4.9642482007391555,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002802987304557349,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2102704167366028,
      "num_tokens": 110673784.0,
      "step": 63805
    },
    {
      "entropy": 5.909707069396973,
      "epoch": 4.964637230110873,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002802710225145458,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21142228692770004,
      "num_tokens": 110682745.0,
      "step": 63810
    },
    {
      "entropy": 5.9285131931304935,
      "epoch": 4.965026259482591,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028024331449337726,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21819272339344026,
      "num_tokens": 110691350.0,
      "step": 63815
    },
    {
      "entropy": 5.826099920272827,
      "epoch": 4.965415288854309,
      "grad_norm": 1.453125,
      "learning_rate": 0.00028021560639264985,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21952420175075532,
      "num_tokens": 110700569.0,
      "step": 63820
    },
    {
      "entropy": 5.853891658782959,
      "epoch": 4.965804318226026,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002801878982127838,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21637140214443207,
      "num_tokens": 110709251.0,
      "step": 63825
    },
    {
      "entropy": 5.810581398010254,
      "epoch": 4.966193347597744,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002801601899541998,
      "loss": 4.823,
      "mean_token_accuracy": 0.2268988698720932,
      "num_tokens": 110717416.0,
      "step": 63830
    },
    {
      "entropy": 5.900151777267456,
      "epoch": 4.966582376969461,
      "grad_norm": 1.453125,
      "learning_rate": 0.000280132481617318,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21254925429821014,
      "num_tokens": 110725887.0,
      "step": 63835
    },
    {
      "entropy": 5.909772491455078,
      "epoch": 4.9669714063411785,
      "grad_norm": 1.5,
      "learning_rate": 0.00028010477320255904,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22225174605846404,
      "num_tokens": 110734292.0,
      "step": 63840
    },
    {
      "entropy": 5.925844430923462,
      "epoch": 4.967360435712896,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002800770647103433,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2203982502222061,
      "num_tokens": 110743414.0,
      "step": 63845
    },
    {
      "entropy": 5.961088085174561,
      "epoch": 4.967749465084614,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002800493561410911,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20802550613880158,
      "num_tokens": 110752009.0,
      "step": 63850
    },
    {
      "entropy": 5.86592435836792,
      "epoch": 4.968138494456332,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000280021647495223,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21360279768705367,
      "num_tokens": 110760685.0,
      "step": 63855
    },
    {
      "entropy": 5.89526596069336,
      "epoch": 4.968527523828049,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027999393877315944,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21093704253435136,
      "num_tokens": 110770134.0,
      "step": 63860
    },
    {
      "entropy": 5.865665483474731,
      "epoch": 4.968916553199767,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002799662299753207,
      "loss": 4.9147,
      "mean_token_accuracy": 0.21159274131059647,
      "num_tokens": 110778337.0,
      "step": 63865
    },
    {
      "entropy": 5.771656465530396,
      "epoch": 4.969305582571484,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027993852110212735,
      "loss": 4.7746,
      "mean_token_accuracy": 0.22371368259191513,
      "num_tokens": 110786509.0,
      "step": 63870
    },
    {
      "entropy": 5.857584142684937,
      "epoch": 4.9696946119432015,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002799108121539999,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2185181826353073,
      "num_tokens": 110795093.0,
      "step": 63875
    },
    {
      "entropy": 5.877140188217163,
      "epoch": 4.970083641314919,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027988310313135857,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22066117227077484,
      "num_tokens": 110803959.0,
      "step": 63880
    },
    {
      "entropy": 5.899916124343872,
      "epoch": 4.970472670686637,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027985539403462407,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22463965713977813,
      "num_tokens": 110812567.0,
      "step": 63885
    },
    {
      "entropy": 5.9445148468017575,
      "epoch": 4.970861700058355,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027982768486421657,
      "loss": 5.0562,
      "mean_token_accuracy": 0.20831960886716844,
      "num_tokens": 110820944.0,
      "step": 63890
    },
    {
      "entropy": 5.77889723777771,
      "epoch": 4.971250729430072,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027979997562055663,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2226160928606987,
      "num_tokens": 110830361.0,
      "step": 63895
    },
    {
      "entropy": 5.922188758850098,
      "epoch": 4.971639758801789,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00027977226630406476,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20399985313415528,
      "num_tokens": 110839832.0,
      "step": 63900
    },
    {
      "entropy": 5.940261173248291,
      "epoch": 4.972028788173507,
      "grad_norm": 1.5,
      "learning_rate": 0.0002797445569151613,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21473704725503923,
      "num_tokens": 110848693.0,
      "step": 63905
    },
    {
      "entropy": 5.954789876937866,
      "epoch": 4.9724178175452245,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002797168474542668,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20703244507312774,
      "num_tokens": 110857347.0,
      "step": 63910
    },
    {
      "entropy": 5.815934705734253,
      "epoch": 4.972806846916942,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027968913792180154,
      "loss": 4.8685,
      "mean_token_accuracy": 0.21862866580486298,
      "num_tokens": 110866141.0,
      "step": 63915
    },
    {
      "entropy": 5.8908562660217285,
      "epoch": 4.97319587628866,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002796614283181862,
      "loss": 4.923,
      "mean_token_accuracy": 0.22083892524242402,
      "num_tokens": 110874875.0,
      "step": 63920
    },
    {
      "entropy": 5.857335996627808,
      "epoch": 4.973584905660378,
      "grad_norm": 1.453125,
      "learning_rate": 0.000279633718643841,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21625324040651323,
      "num_tokens": 110884045.0,
      "step": 63925
    },
    {
      "entropy": 5.847949552536011,
      "epoch": 4.973973935032095,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027960600889918657,
      "loss": 4.882,
      "mean_token_accuracy": 0.22922610342502595,
      "num_tokens": 110892751.0,
      "step": 63930
    },
    {
      "entropy": 5.786898183822632,
      "epoch": 4.974362964403813,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002795782990846433,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22101684808731079,
      "num_tokens": 110901929.0,
      "step": 63935
    },
    {
      "entropy": 5.891020250320435,
      "epoch": 4.97475199377553,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002795505892006316,
      "loss": 4.9193,
      "mean_token_accuracy": 0.21833216398954391,
      "num_tokens": 110911162.0,
      "step": 63940
    },
    {
      "entropy": 5.870971250534057,
      "epoch": 4.9751410231472475,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002795228792475719,
      "loss": 4.9085,
      "mean_token_accuracy": 0.21584829837083816,
      "num_tokens": 110919811.0,
      "step": 63945
    },
    {
      "entropy": 5.8591306686401365,
      "epoch": 4.975530052518965,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002794951692258848,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2223811626434326,
      "num_tokens": 110928452.0,
      "step": 63950
    },
    {
      "entropy": 5.9334807872772215,
      "epoch": 4.975919081890683,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002794674591359906,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20930658131837845,
      "num_tokens": 110937911.0,
      "step": 63955
    },
    {
      "entropy": 5.831922960281372,
      "epoch": 4.976308111262401,
      "grad_norm": 1.5,
      "learning_rate": 0.0002794397489783098,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2260296165943146,
      "num_tokens": 110946960.0,
      "step": 63960
    },
    {
      "entropy": 5.8928529739379885,
      "epoch": 4.976697140634117,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002794120387532629,
      "loss": 4.9536,
      "mean_token_accuracy": 0.20918354243040085,
      "num_tokens": 110955512.0,
      "step": 63965
    },
    {
      "entropy": 5.895640468597412,
      "epoch": 4.977086170005835,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002793843284612703,
      "loss": 5.0293,
      "mean_token_accuracy": 0.20898516327142716,
      "num_tokens": 110964372.0,
      "step": 63970
    },
    {
      "entropy": 5.899881362915039,
      "epoch": 4.977475199377553,
      "grad_norm": 1.375,
      "learning_rate": 0.0002793566181027526,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21969883590936662,
      "num_tokens": 110972707.0,
      "step": 63975
    },
    {
      "entropy": 5.904075384140015,
      "epoch": 4.9778642287492705,
      "grad_norm": 1.5,
      "learning_rate": 0.00027932890767813005,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2184451162815094,
      "num_tokens": 110981943.0,
      "step": 63980
    },
    {
      "entropy": 5.909880685806274,
      "epoch": 4.978253258120988,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002793011971878233,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21889750510454178,
      "num_tokens": 110990826.0,
      "step": 63985
    },
    {
      "entropy": 5.936341285705566,
      "epoch": 4.978642287492706,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002792734866322526,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20118037462234498,
      "num_tokens": 111000580.0,
      "step": 63990
    },
    {
      "entropy": 5.924557876586914,
      "epoch": 4.979031316864424,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002792457760118387,
      "loss": 4.9644,
      "mean_token_accuracy": 0.21293821781873704,
      "num_tokens": 111009587.0,
      "step": 63995
    },
    {
      "entropy": 5.89310507774353,
      "epoch": 4.979420346236141,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027921806532700186,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21938625276088713,
      "num_tokens": 111017654.0,
      "step": 64000
    },
    {
      "entropy": 5.826642799377441,
      "epoch": 4.979809375607858,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002791903545781625,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23218424171209334,
      "num_tokens": 111026629.0,
      "step": 64005
    },
    {
      "entropy": 5.881484317779541,
      "epoch": 4.980198404979576,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027916264376574124,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21525414735078813,
      "num_tokens": 111035627.0,
      "step": 64010
    },
    {
      "entropy": 5.866163682937622,
      "epoch": 4.9805874343512935,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027913493289015857,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21798885017633438,
      "num_tokens": 111043802.0,
      "step": 64015
    },
    {
      "entropy": 5.9055314540863035,
      "epoch": 4.980976463723011,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027910722195183474,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21158768832683564,
      "num_tokens": 111052814.0,
      "step": 64020
    },
    {
      "entropy": 5.907416486740113,
      "epoch": 4.981365493094729,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002790795109511904,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21356550306081773,
      "num_tokens": 111061437.0,
      "step": 64025
    },
    {
      "entropy": 5.990430068969727,
      "epoch": 4.981754522466447,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027905179988864593,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2072521209716797,
      "num_tokens": 111069877.0,
      "step": 64030
    },
    {
      "entropy": 5.952732086181641,
      "epoch": 4.982143551838163,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002790240887646219,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21191806793212892,
      "num_tokens": 111078501.0,
      "step": 64035
    },
    {
      "entropy": 5.907199239730835,
      "epoch": 4.982532581209881,
      "grad_norm": 1.515625,
      "learning_rate": 0.00027899637757953876,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22508083879947663,
      "num_tokens": 111086461.0,
      "step": 64040
    },
    {
      "entropy": 5.876067113876343,
      "epoch": 4.982921610581599,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027896866633381694,
      "loss": 4.9076,
      "mean_token_accuracy": 0.22683394700288773,
      "num_tokens": 111095277.0,
      "step": 64045
    },
    {
      "entropy": 5.808910751342774,
      "epoch": 4.9833106399533165,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002789409550278769,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2256898507475853,
      "num_tokens": 111104304.0,
      "step": 64050
    },
    {
      "entropy": 5.812211751937866,
      "epoch": 4.983699669325034,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002789132436621391,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22071948349475862,
      "num_tokens": 111113110.0,
      "step": 64055
    },
    {
      "entropy": 5.883360910415649,
      "epoch": 4.984088698696752,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002788855322370241,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21941233277320862,
      "num_tokens": 111120856.0,
      "step": 64060
    },
    {
      "entropy": 5.986951446533203,
      "epoch": 4.98447772806847,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027885782075295234,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20592739135026933,
      "num_tokens": 111130061.0,
      "step": 64065
    },
    {
      "entropy": 5.860374784469604,
      "epoch": 4.984866757440186,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002788301092103443,
      "loss": 4.8359,
      "mean_token_accuracy": 0.22496977895498277,
      "num_tokens": 111139596.0,
      "step": 64070
    },
    {
      "entropy": 5.972198724746704,
      "epoch": 4.985255786811904,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027880239760962044,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21518083661794662,
      "num_tokens": 111147288.0,
      "step": 64075
    },
    {
      "entropy": 5.873819017410279,
      "epoch": 4.985644816183622,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027877468595120125,
      "loss": 4.8955,
      "mean_token_accuracy": 0.2185990557074547,
      "num_tokens": 111155940.0,
      "step": 64080
    },
    {
      "entropy": 5.917134761810303,
      "epoch": 4.9860338455553395,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002787469742355072,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21209343075752257,
      "num_tokens": 111164165.0,
      "step": 64085
    },
    {
      "entropy": 5.8869551658630375,
      "epoch": 4.986422874927057,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002787192624629587,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2152844175696373,
      "num_tokens": 111171979.0,
      "step": 64090
    },
    {
      "entropy": 5.881446075439453,
      "epoch": 4.986811904298775,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002786915506339764,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2103766083717346,
      "num_tokens": 111180816.0,
      "step": 64095
    },
    {
      "entropy": 5.845865392684937,
      "epoch": 4.987200933670492,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002786638387489807,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2204076200723648,
      "num_tokens": 111189294.0,
      "step": 64100
    },
    {
      "entropy": 5.914604711532593,
      "epoch": 4.987589963042209,
      "grad_norm": 1.5,
      "learning_rate": 0.00027863612680839204,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21569467335939407,
      "num_tokens": 111197741.0,
      "step": 64105
    },
    {
      "entropy": 5.906512928009033,
      "epoch": 4.987978992413927,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000278608414812631,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22186311930418015,
      "num_tokens": 111206080.0,
      "step": 64110
    },
    {
      "entropy": 5.920551872253418,
      "epoch": 4.988368021785645,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027858070276211804,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21341196447610855,
      "num_tokens": 111214589.0,
      "step": 64115
    },
    {
      "entropy": 5.870954513549805,
      "epoch": 4.9887570511573625,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002785529906572735,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21427757740020753,
      "num_tokens": 111223956.0,
      "step": 64120
    },
    {
      "entropy": 5.9437885761260985,
      "epoch": 4.98914608052908,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000278525278498518,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20386583507061004,
      "num_tokens": 111232985.0,
      "step": 64125
    },
    {
      "entropy": 5.96984372138977,
      "epoch": 4.989535109900798,
      "grad_norm": 1.515625,
      "learning_rate": 0.00027849756628627205,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21732331216335296,
      "num_tokens": 111241557.0,
      "step": 64130
    },
    {
      "entropy": 5.903753709793091,
      "epoch": 4.9899241392725155,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002784698540209561,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22474494576454163,
      "num_tokens": 111249832.0,
      "step": 64135
    },
    {
      "entropy": 5.835906219482422,
      "epoch": 4.990313168644232,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027844214170299065,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21527326107025146,
      "num_tokens": 111258017.0,
      "step": 64140
    },
    {
      "entropy": 5.769734239578247,
      "epoch": 4.99070219801595,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027841442933279614,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22111235857009887,
      "num_tokens": 111266774.0,
      "step": 64145
    },
    {
      "entropy": 5.875031614303589,
      "epoch": 4.991091227387668,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027838671691079314,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21523204296827317,
      "num_tokens": 111275994.0,
      "step": 64150
    },
    {
      "entropy": 5.8610505104064945,
      "epoch": 4.9914802567593854,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002783590044374021,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21947066038846968,
      "num_tokens": 111283752.0,
      "step": 64155
    },
    {
      "entropy": 5.826385879516602,
      "epoch": 4.991869286131103,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002783312919130435,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22135344594717027,
      "num_tokens": 111292217.0,
      "step": 64160
    },
    {
      "entropy": 5.854293918609619,
      "epoch": 4.992258315502821,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002783035793381378,
      "loss": 5.0118,
      "mean_token_accuracy": 0.2152214005589485,
      "num_tokens": 111300649.0,
      "step": 64165
    },
    {
      "entropy": 5.941743993759156,
      "epoch": 4.992647344874538,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027827586671310556,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22137117385864258,
      "num_tokens": 111309132.0,
      "step": 64170
    },
    {
      "entropy": 5.81886773109436,
      "epoch": 4.993036374246255,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002782481540383673,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23413552343845367,
      "num_tokens": 111317089.0,
      "step": 64175
    },
    {
      "entropy": 5.821245288848877,
      "epoch": 4.993425403617973,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027822044131434344,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21556792557239532,
      "num_tokens": 111326273.0,
      "step": 64180
    },
    {
      "entropy": 5.822574090957642,
      "epoch": 4.993814432989691,
      "grad_norm": 1.5,
      "learning_rate": 0.0002781927285414545,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21645500808954238,
      "num_tokens": 111334815.0,
      "step": 64185
    },
    {
      "entropy": 5.887893438339233,
      "epoch": 4.994203462361408,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027816501572012104,
      "loss": 4.9636,
      "mean_token_accuracy": 0.218571275472641,
      "num_tokens": 111343563.0,
      "step": 64190
    },
    {
      "entropy": 5.850993394851685,
      "epoch": 4.994592491733126,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002781373028507634,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22337769269943236,
      "num_tokens": 111352310.0,
      "step": 64195
    },
    {
      "entropy": 5.903102207183838,
      "epoch": 4.994981521104844,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002781095899338023,
      "loss": 4.9505,
      "mean_token_accuracy": 0.20949295461177825,
      "num_tokens": 111361276.0,
      "step": 64200
    },
    {
      "entropy": 5.879453897476196,
      "epoch": 4.995370550476561,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002780818769696581,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2148409366607666,
      "num_tokens": 111369870.0,
      "step": 64205
    },
    {
      "entropy": 5.844564437866211,
      "epoch": 4.995759579848278,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027805416395875127,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2153105244040489,
      "num_tokens": 111378967.0,
      "step": 64210
    },
    {
      "entropy": 5.879635000228882,
      "epoch": 4.996148609219996,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002780264509015024,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2325071409344673,
      "num_tokens": 111387534.0,
      "step": 64215
    },
    {
      "entropy": 5.946760177612305,
      "epoch": 4.996537638591714,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027799873779833193,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2219708725810051,
      "num_tokens": 111395802.0,
      "step": 64220
    },
    {
      "entropy": 5.870682859420777,
      "epoch": 4.996926667963431,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002779710246496604,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20953334420919417,
      "num_tokens": 111404980.0,
      "step": 64225
    },
    {
      "entropy": 5.846844816207886,
      "epoch": 4.997315697335149,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002779433114559083,
      "loss": 4.838,
      "mean_token_accuracy": 0.22834526747465134,
      "num_tokens": 111412690.0,
      "step": 64230
    },
    {
      "entropy": 5.843465328216553,
      "epoch": 4.997704726706866,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027791559821749615,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22586438953876495,
      "num_tokens": 111420831.0,
      "step": 64235
    },
    {
      "entropy": 5.894962739944458,
      "epoch": 4.998093756078584,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027788788493484447,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22111471742391586,
      "num_tokens": 111429141.0,
      "step": 64240
    },
    {
      "entropy": 5.857069492340088,
      "epoch": 4.998482785450301,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002778601716083736,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2204052224755287,
      "num_tokens": 111437851.0,
      "step": 64245
    },
    {
      "entropy": 5.938580465316773,
      "epoch": 4.998871814822019,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002778324582385043,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21551517099142076,
      "num_tokens": 111446208.0,
      "step": 64250
    },
    {
      "entropy": 5.898428010940552,
      "epoch": 4.999260844193737,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027780474482565687,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21395378261804582,
      "num_tokens": 111455099.0,
      "step": 64255
    },
    {
      "entropy": 5.884850025177002,
      "epoch": 4.999649873565454,
      "grad_norm": 1.375,
      "learning_rate": 0.00027777703137025193,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22067487239837646,
      "num_tokens": 111464383.0,
      "step": 64260
    },
    {
      "entropy": 5.973176903194851,
      "epoch": 5.0,
      "grad_norm": 2.8125,
      "learning_rate": 0.00027774931787271,
      "loss": 5.0698,
      "mean_token_accuracy": 0.21202192703882852,
      "num_tokens": 111471835.0,
      "step": 64265
    },
    {
      "entropy": 5.909226369857788,
      "epoch": 5.000389029371718,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002777216043334516,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2119364395737648,
      "num_tokens": 111480304.0,
      "step": 64270
    },
    {
      "entropy": 5.906182289123535,
      "epoch": 5.000778058743435,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027769389075289714,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21264814585447311,
      "num_tokens": 111488837.0,
      "step": 64275
    },
    {
      "entropy": 5.874850559234619,
      "epoch": 5.001167088115153,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002776661771314671,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22183497697114946,
      "num_tokens": 111497055.0,
      "step": 64280
    },
    {
      "entropy": 5.821565437316894,
      "epoch": 5.00155611748687,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002776384634695821,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22468568235635758,
      "num_tokens": 111506345.0,
      "step": 64285
    },
    {
      "entropy": 5.828790950775146,
      "epoch": 5.001945146858588,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002776107497676626,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2215821549296379,
      "num_tokens": 111515138.0,
      "step": 64290
    },
    {
      "entropy": 5.863262510299682,
      "epoch": 5.002334176230305,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002775830360261291,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2301979124546051,
      "num_tokens": 111524487.0,
      "step": 64295
    },
    {
      "entropy": 5.84428038597107,
      "epoch": 5.002723205602023,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027755532224540226,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22167651057243348,
      "num_tokens": 111533455.0,
      "step": 64300
    },
    {
      "entropy": 5.88657579421997,
      "epoch": 5.003112234973741,
      "grad_norm": 1.5625,
      "learning_rate": 0.00027752760842590245,
      "loss": 4.9454,
      "mean_token_accuracy": 0.216737699508667,
      "num_tokens": 111541922.0,
      "step": 64305
    },
    {
      "entropy": 5.878822612762451,
      "epoch": 5.003501264345458,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002774998945680502,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2233734220266342,
      "num_tokens": 111550154.0,
      "step": 64310
    },
    {
      "entropy": 5.9098132133483885,
      "epoch": 5.003890293717176,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000277472180672266,
      "loss": 4.8055,
      "mean_token_accuracy": 0.22785235047340394,
      "num_tokens": 111559007.0,
      "step": 64315
    },
    {
      "entropy": 5.855810213088989,
      "epoch": 5.004279323088893,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027744446673897034,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22538969814777374,
      "num_tokens": 111567839.0,
      "step": 64320
    },
    {
      "entropy": 5.840738010406494,
      "epoch": 5.004668352460611,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027741675276858374,
      "loss": 4.8129,
      "mean_token_accuracy": 0.22136402428150176,
      "num_tokens": 111575902.0,
      "step": 64325
    },
    {
      "entropy": 5.8498505592346195,
      "epoch": 5.005057381832328,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027738903876152693,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2221868798136711,
      "num_tokens": 111585543.0,
      "step": 64330
    },
    {
      "entropy": 5.949696493148804,
      "epoch": 5.005446411204046,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002773613247182202,
      "loss": 5.0054,
      "mean_token_accuracy": 0.20731735974550247,
      "num_tokens": 111594144.0,
      "step": 64335
    },
    {
      "entropy": 5.887598752975464,
      "epoch": 5.005835440575764,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027733361063908405,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22005427479743958,
      "num_tokens": 111602633.0,
      "step": 64340
    },
    {
      "entropy": 5.926147556304931,
      "epoch": 5.006224469947481,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002773058965245393,
      "loss": 4.9701,
      "mean_token_accuracy": 0.2121441513299942,
      "num_tokens": 111610847.0,
      "step": 64345
    },
    {
      "entropy": 5.88891921043396,
      "epoch": 5.006613499319198,
      "grad_norm": 1.484375,
      "learning_rate": 0.000277278182375006,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2319459468126297,
      "num_tokens": 111619462.0,
      "step": 64350
    },
    {
      "entropy": 5.845765113830566,
      "epoch": 5.007002528690916,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002772504681909049,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2150522366166115,
      "num_tokens": 111627762.0,
      "step": 64355
    },
    {
      "entropy": 5.882256603240966,
      "epoch": 5.007391558062634,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002772227539726566,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21489426791667937,
      "num_tokens": 111636983.0,
      "step": 64360
    },
    {
      "entropy": 5.897719192504883,
      "epoch": 5.007780587434351,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002771950397206816,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21634673774242402,
      "num_tokens": 111646167.0,
      "step": 64365
    },
    {
      "entropy": 5.947975111007691,
      "epoch": 5.008169616806069,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002771673254354003,
      "loss": 4.997,
      "mean_token_accuracy": 0.21402380466461182,
      "num_tokens": 111655213.0,
      "step": 64370
    },
    {
      "entropy": 5.847512912750244,
      "epoch": 5.008558646177787,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027713961111723335,
      "loss": 4.8925,
      "mean_token_accuracy": 0.21850191950798034,
      "num_tokens": 111664477.0,
      "step": 64375
    },
    {
      "entropy": 5.85119595527649,
      "epoch": 5.008947675549504,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00027711189676660116,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2314981698989868,
      "num_tokens": 111674001.0,
      "step": 64380
    },
    {
      "entropy": 5.950858449935913,
      "epoch": 5.009336704921221,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027708418238392435,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20829713791608812,
      "num_tokens": 111682428.0,
      "step": 64385
    },
    {
      "entropy": 5.925779151916504,
      "epoch": 5.009725734292939,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002770564679696233,
      "loss": 4.9133,
      "mean_token_accuracy": 0.219101157784462,
      "num_tokens": 111691688.0,
      "step": 64390
    },
    {
      "entropy": 5.885391187667847,
      "epoch": 5.010114763664657,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027702875352411866,
      "loss": 4.871,
      "mean_token_accuracy": 0.22441920638084412,
      "num_tokens": 111699485.0,
      "step": 64395
    },
    {
      "entropy": 5.907472562789917,
      "epoch": 5.010503793036374,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002770010390478309,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21735292822122573,
      "num_tokens": 111708293.0,
      "step": 64400
    },
    {
      "entropy": 5.8503913402557375,
      "epoch": 5.010892822408092,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027697332454118057,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22068884521722792,
      "num_tokens": 111716920.0,
      "step": 64405
    },
    {
      "entropy": 5.876481008529663,
      "epoch": 5.01128185177981,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027694561000458825,
      "loss": 4.9153,
      "mean_token_accuracy": 0.21625497490167617,
      "num_tokens": 111725746.0,
      "step": 64410
    },
    {
      "entropy": 5.839608240127563,
      "epoch": 5.011670881151527,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002769178954384743,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22555433809757233,
      "num_tokens": 111734337.0,
      "step": 64415
    },
    {
      "entropy": 5.807862710952759,
      "epoch": 5.012059910523244,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002768901808432594,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2221294179558754,
      "num_tokens": 111742842.0,
      "step": 64420
    },
    {
      "entropy": 5.782453155517578,
      "epoch": 5.012448939894962,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000276862466219364,
      "loss": 4.8042,
      "mean_token_accuracy": 0.23139130622148513,
      "num_tokens": 111751506.0,
      "step": 64425
    },
    {
      "entropy": 5.83946681022644,
      "epoch": 5.01283796926668,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002768347515672086,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21661840826272966,
      "num_tokens": 111760110.0,
      "step": 64430
    },
    {
      "entropy": 5.838093996047974,
      "epoch": 5.013226998638397,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027680703688721383,
      "loss": 4.887,
      "mean_token_accuracy": 0.22454843521118165,
      "num_tokens": 111768781.0,
      "step": 64435
    },
    {
      "entropy": 5.864094161987305,
      "epoch": 5.013616028010115,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027677932217980004,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22545695453882217,
      "num_tokens": 111776830.0,
      "step": 64440
    },
    {
      "entropy": 5.897670793533325,
      "epoch": 5.014005057381833,
      "grad_norm": 1.4375,
      "learning_rate": 0.000276751607445388,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21299912929534912,
      "num_tokens": 111785366.0,
      "step": 64445
    },
    {
      "entropy": 5.877817678451538,
      "epoch": 5.0143940867535495,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000276723892684398,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2155481532216072,
      "num_tokens": 111793724.0,
      "step": 64450
    },
    {
      "entropy": 5.86826376914978,
      "epoch": 5.014783116125267,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027669617789725074,
      "loss": 4.852,
      "mean_token_accuracy": 0.22119883298873902,
      "num_tokens": 111802892.0,
      "step": 64455
    },
    {
      "entropy": 5.867434310913086,
      "epoch": 5.015172145496985,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002766684630843667,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2152613341808319,
      "num_tokens": 111811768.0,
      "step": 64460
    },
    {
      "entropy": 5.914344120025635,
      "epoch": 5.015561174868703,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002766407482461663,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21625232249498366,
      "num_tokens": 111821344.0,
      "step": 64465
    },
    {
      "entropy": 5.877723646163941,
      "epoch": 5.01595020424042,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002766130333830702,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22738804519176484,
      "num_tokens": 111830731.0,
      "step": 64470
    },
    {
      "entropy": 5.823696517944336,
      "epoch": 5.016339233612138,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027658531849549894,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22013920545578003,
      "num_tokens": 111838637.0,
      "step": 64475
    },
    {
      "entropy": 5.881544733047486,
      "epoch": 5.016728262983856,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002765576035838729,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21595467031002044,
      "num_tokens": 111846646.0,
      "step": 64480
    },
    {
      "entropy": 5.8424018859863285,
      "epoch": 5.0171172923555725,
      "grad_norm": 1.375,
      "learning_rate": 0.0002765298886486128,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22106529027223587,
      "num_tokens": 111855717.0,
      "step": 64485
    },
    {
      "entropy": 5.838183736801147,
      "epoch": 5.01750632172729,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002765021736901389,
      "loss": 4.922,
      "mean_token_accuracy": 0.21723163723945618,
      "num_tokens": 111865351.0,
      "step": 64490
    },
    {
      "entropy": 5.838608837127685,
      "epoch": 5.017895351099008,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002764744587088721,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22504724711179733,
      "num_tokens": 111873246.0,
      "step": 64495
    },
    {
      "entropy": 5.88311767578125,
      "epoch": 5.018284380470726,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002764467437052327,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21749533265829085,
      "num_tokens": 111882035.0,
      "step": 64500
    },
    {
      "entropy": 5.815058183670044,
      "epoch": 5.018673409842443,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027641902867964126,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22354974150657653,
      "num_tokens": 111890505.0,
      "step": 64505
    },
    {
      "entropy": 5.898570537567139,
      "epoch": 5.019062439214161,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027639131363251825,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22015136778354644,
      "num_tokens": 111899662.0,
      "step": 64510
    },
    {
      "entropy": 5.856615257263184,
      "epoch": 5.019451468585879,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00027636359856428436,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20836184173822403,
      "num_tokens": 111908425.0,
      "step": 64515
    },
    {
      "entropy": 5.817131996154785,
      "epoch": 5.0198404979575955,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002763358834753599,
      "loss": 4.862,
      "mean_token_accuracy": 0.22499207705259322,
      "num_tokens": 111917070.0,
      "step": 64520
    },
    {
      "entropy": 5.902232360839844,
      "epoch": 5.020229527329313,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027630816836616564,
      "loss": 4.918,
      "mean_token_accuracy": 0.21824053972959517,
      "num_tokens": 111925747.0,
      "step": 64525
    },
    {
      "entropy": 5.90600619316101,
      "epoch": 5.020618556701031,
      "grad_norm": 1.53125,
      "learning_rate": 0.00027628045323712194,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21482586860656738,
      "num_tokens": 111934277.0,
      "step": 64530
    },
    {
      "entropy": 6.003481388092041,
      "epoch": 5.0210075860727486,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027625273808864955,
      "loss": 5.016,
      "mean_token_accuracy": 0.2088635340332985,
      "num_tokens": 111942911.0,
      "step": 64535
    },
    {
      "entropy": 5.899829626083374,
      "epoch": 5.021396615444466,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002762250229211687,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22460928708314895,
      "num_tokens": 111952000.0,
      "step": 64540
    },
    {
      "entropy": 5.844606876373291,
      "epoch": 5.021785644816184,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027619730773510015,
      "loss": 4.9549,
      "mean_token_accuracy": 0.21338469088077544,
      "num_tokens": 111960878.0,
      "step": 64545
    },
    {
      "entropy": 5.913014125823975,
      "epoch": 5.022174674187901,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027616959253086437,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22320106774568557,
      "num_tokens": 111969025.0,
      "step": 64550
    },
    {
      "entropy": 5.911015796661377,
      "epoch": 5.0225637035596185,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002761418773088818,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2378765180706978,
      "num_tokens": 111977470.0,
      "step": 64555
    },
    {
      "entropy": 5.925042057037354,
      "epoch": 5.022952732931336,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002761141620695731,
      "loss": 4.979,
      "mean_token_accuracy": 0.21598169952630997,
      "num_tokens": 111986305.0,
      "step": 64560
    },
    {
      "entropy": 5.902093696594238,
      "epoch": 5.023341762303054,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027608644681335876,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2112624913454056,
      "num_tokens": 111994842.0,
      "step": 64565
    },
    {
      "entropy": 5.915882349014282,
      "epoch": 5.0237307916747715,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027605873154065935,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22018698751926422,
      "num_tokens": 112003495.0,
      "step": 64570
    },
    {
      "entropy": 5.862339782714844,
      "epoch": 5.024119821046489,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027603101625189543,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22215728014707564,
      "num_tokens": 112012235.0,
      "step": 64575
    },
    {
      "entropy": 5.9120502948760985,
      "epoch": 5.024508850418207,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002760033009474874,
      "loss": 4.9685,
      "mean_token_accuracy": 0.20994774252176285,
      "num_tokens": 112021002.0,
      "step": 64580
    },
    {
      "entropy": 5.914652347564697,
      "epoch": 5.024897879789924,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002759755856278558,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22479916214942933,
      "num_tokens": 112029188.0,
      "step": 64585
    },
    {
      "entropy": 5.851552534103393,
      "epoch": 5.0252869091616414,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002759478702934213,
      "loss": 4.8234,
      "mean_token_accuracy": 0.22999429404735566,
      "num_tokens": 112038007.0,
      "step": 64590
    },
    {
      "entropy": 5.857545185089111,
      "epoch": 5.025675938533359,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027592015494460443,
      "loss": 4.907,
      "mean_token_accuracy": 0.2208712339401245,
      "num_tokens": 112046759.0,
      "step": 64595
    },
    {
      "entropy": 5.875380182266236,
      "epoch": 5.026064967905077,
      "grad_norm": 1.5625,
      "learning_rate": 0.00027589243958182566,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2218281462788582,
      "num_tokens": 112054615.0,
      "step": 64600
    },
    {
      "entropy": 5.862544965744019,
      "epoch": 5.0264539972767945,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027586472420550553,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2270636036992073,
      "num_tokens": 112063376.0,
      "step": 64605
    },
    {
      "entropy": 5.877909183502197,
      "epoch": 5.026843026648512,
      "grad_norm": 1.5,
      "learning_rate": 0.0002758370088160646,
      "loss": 4.9457,
      "mean_token_accuracy": 0.21472433507442473,
      "num_tokens": 112071585.0,
      "step": 64610
    },
    {
      "entropy": 5.849989223480224,
      "epoch": 5.02723205602023,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002758092934139234,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2321088880300522,
      "num_tokens": 112079453.0,
      "step": 64615
    },
    {
      "entropy": 5.826113891601563,
      "epoch": 5.027621085391947,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002757815779995023,
      "loss": 4.8517,
      "mean_token_accuracy": 0.225997656583786,
      "num_tokens": 112088746.0,
      "step": 64620
    },
    {
      "entropy": 5.859411001205444,
      "epoch": 5.028010114763664,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002757538625732222,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21667447686195374,
      "num_tokens": 112097157.0,
      "step": 64625
    },
    {
      "entropy": 5.811166477203369,
      "epoch": 5.028399144135382,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002757261471355034,
      "loss": 4.8363,
      "mean_token_accuracy": 0.22257663458585739,
      "num_tokens": 112105436.0,
      "step": 64630
    },
    {
      "entropy": 5.887614965438843,
      "epoch": 5.0287881735071,
      "grad_norm": 1.625,
      "learning_rate": 0.00027569843168676647,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21997662782669067,
      "num_tokens": 112113701.0,
      "step": 64635
    },
    {
      "entropy": 5.848079442977905,
      "epoch": 5.0291772028788175,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002756707162274319,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22155907601118088,
      "num_tokens": 112122436.0,
      "step": 64640
    },
    {
      "entropy": 5.839847040176392,
      "epoch": 5.029566232250535,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027564300075792026,
      "loss": 4.8555,
      "mean_token_accuracy": 0.21865054070949555,
      "num_tokens": 112131491.0,
      "step": 64645
    },
    {
      "entropy": 5.926815414428711,
      "epoch": 5.029955261622252,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027561528527865215,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2216126799583435,
      "num_tokens": 112140458.0,
      "step": 64650
    },
    {
      "entropy": 5.886229562759399,
      "epoch": 5.03034429099397,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002755875697900481,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2206069514155388,
      "num_tokens": 112149571.0,
      "step": 64655
    },
    {
      "entropy": 5.894085168838501,
      "epoch": 5.030733320365687,
      "grad_norm": 1.5,
      "learning_rate": 0.00027555985429252857,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21675858199596404,
      "num_tokens": 112158470.0,
      "step": 64660
    },
    {
      "entropy": 5.866319370269776,
      "epoch": 5.031122349737405,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002755321387865141,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22225476056337357,
      "num_tokens": 112166395.0,
      "step": 64665
    },
    {
      "entropy": 5.8216753005981445,
      "epoch": 5.031511379109123,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002755044232724253,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22428909689188004,
      "num_tokens": 112174990.0,
      "step": 64670
    },
    {
      "entropy": 5.919644021987915,
      "epoch": 5.0319004084808405,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027547670775068274,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2129521608352661,
      "num_tokens": 112184117.0,
      "step": 64675
    },
    {
      "entropy": 6.0109185695648195,
      "epoch": 5.032289437852558,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027544899222170684,
      "loss": 5.0867,
      "mean_token_accuracy": 0.2167431116104126,
      "num_tokens": 112193747.0,
      "step": 64680
    },
    {
      "entropy": 5.903535509109497,
      "epoch": 5.032678467224275,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027542127668591824,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21410429626703262,
      "num_tokens": 112202485.0,
      "step": 64685
    },
    {
      "entropy": 5.807920789718628,
      "epoch": 5.033067496595993,
      "grad_norm": 1.859375,
      "learning_rate": 0.0002753935611437374,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23694253116846084,
      "num_tokens": 112211844.0,
      "step": 64690
    },
    {
      "entropy": 5.892452335357666,
      "epoch": 5.03345652596771,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027536584559558494,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2132840037345886,
      "num_tokens": 112220494.0,
      "step": 64695
    },
    {
      "entropy": 5.868812322616577,
      "epoch": 5.033845555339428,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002753381300418813,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21918901950120925,
      "num_tokens": 112228804.0,
      "step": 64700
    },
    {
      "entropy": 5.942786169052124,
      "epoch": 5.034234584711146,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002753104144830471,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2168677791953087,
      "num_tokens": 112237707.0,
      "step": 64705
    },
    {
      "entropy": 5.839637517929077,
      "epoch": 5.0346236140828635,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027528269891950285,
      "loss": 4.8672,
      "mean_token_accuracy": 0.21986932456493377,
      "num_tokens": 112245365.0,
      "step": 64710
    },
    {
      "entropy": 5.833718204498291,
      "epoch": 5.035012643454581,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002752549833516691,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21476005166769027,
      "num_tokens": 112253710.0,
      "step": 64715
    },
    {
      "entropy": 5.863571691513061,
      "epoch": 5.035401672826298,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002752272677799664,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21887447983026503,
      "num_tokens": 112261990.0,
      "step": 64720
    },
    {
      "entropy": 5.905410814285278,
      "epoch": 5.035790702198016,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027519955220481525,
      "loss": 4.929,
      "mean_token_accuracy": 0.22362429648637772,
      "num_tokens": 112271555.0,
      "step": 64725
    },
    {
      "entropy": 5.876361942291259,
      "epoch": 5.036179731569733,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002751718366266362,
      "loss": 4.8501,
      "mean_token_accuracy": 0.21785062402486802,
      "num_tokens": 112280365.0,
      "step": 64730
    },
    {
      "entropy": 5.898961210250855,
      "epoch": 5.036568760941451,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002751441210458498,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21646856814622878,
      "num_tokens": 112288240.0,
      "step": 64735
    },
    {
      "entropy": 5.96207218170166,
      "epoch": 5.036957790313169,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027511640546287665,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21324578076601028,
      "num_tokens": 112297769.0,
      "step": 64740
    },
    {
      "entropy": 5.907105493545532,
      "epoch": 5.0373468196848865,
      "grad_norm": 1.375,
      "learning_rate": 0.0002750886898781372,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22068289667367935,
      "num_tokens": 112306669.0,
      "step": 64745
    },
    {
      "entropy": 5.827813005447387,
      "epoch": 5.037735849056604,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000275060974292052,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22098648101091384,
      "num_tokens": 112315377.0,
      "step": 64750
    },
    {
      "entropy": 5.78583254814148,
      "epoch": 5.038124878428321,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002750332587050416,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2223580524325371,
      "num_tokens": 112323964.0,
      "step": 64755
    },
    {
      "entropy": 5.888118267059326,
      "epoch": 5.038513907800039,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002750055431175266,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2230849876999855,
      "num_tokens": 112331965.0,
      "step": 64760
    },
    {
      "entropy": 5.86181788444519,
      "epoch": 5.038902937171756,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027497782752992743,
      "loss": 4.9515,
      "mean_token_accuracy": 0.20751924365758895,
      "num_tokens": 112340780.0,
      "step": 64765
    },
    {
      "entropy": 5.9145880222320555,
      "epoch": 5.039291966543474,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002749501119426647,
      "loss": 4.9742,
      "mean_token_accuracy": 0.20553751587867736,
      "num_tokens": 112349604.0,
      "step": 64770
    },
    {
      "entropy": 5.817593050003052,
      "epoch": 5.039680995915192,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027492239635615903,
      "loss": 4.7891,
      "mean_token_accuracy": 0.22874020785093307,
      "num_tokens": 112358754.0,
      "step": 64775
    },
    {
      "entropy": 5.977034711837769,
      "epoch": 5.0400700252869095,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002748946807708308,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20542526692152024,
      "num_tokens": 112367371.0,
      "step": 64780
    },
    {
      "entropy": 5.969960260391235,
      "epoch": 5.040459054658626,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002748669651871006,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2175255998969078,
      "num_tokens": 112376498.0,
      "step": 64785
    },
    {
      "entropy": 5.862517976760865,
      "epoch": 5.040848084030344,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00027483924960538907,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2166166126728058,
      "num_tokens": 112385609.0,
      "step": 64790
    },
    {
      "entropy": 5.868311738967895,
      "epoch": 5.041237113402062,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027481153402611664,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21647924482822417,
      "num_tokens": 112394571.0,
      "step": 64795
    },
    {
      "entropy": 5.820017385482788,
      "epoch": 5.041626142773779,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027478381844970386,
      "loss": 4.856,
      "mean_token_accuracy": 0.2283809095621109,
      "num_tokens": 112404187.0,
      "step": 64800
    },
    {
      "entropy": 5.873910427093506,
      "epoch": 5.042015172145497,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002747561028765713,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21526931822299958,
      "num_tokens": 112413301.0,
      "step": 64805
    },
    {
      "entropy": 5.860207796096802,
      "epoch": 5.042404201517215,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00027472838730713956,
      "loss": 4.891,
      "mean_token_accuracy": 0.22126493155956267,
      "num_tokens": 112422914.0,
      "step": 64810
    },
    {
      "entropy": 5.873942279815674,
      "epoch": 5.0427932308889325,
      "grad_norm": 1.375,
      "learning_rate": 0.00027470067174182905,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22297078520059585,
      "num_tokens": 112431808.0,
      "step": 64815
    },
    {
      "entropy": 5.869992351531982,
      "epoch": 5.043182260260649,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002746729561810604,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22310978323221206,
      "num_tokens": 112441098.0,
      "step": 64820
    },
    {
      "entropy": 5.976985931396484,
      "epoch": 5.043571289632367,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027464524062525405,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21123768538236617,
      "num_tokens": 112449921.0,
      "step": 64825
    },
    {
      "entropy": 5.925273513793945,
      "epoch": 5.043960319004085,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002746175250748306,
      "loss": 4.959,
      "mean_token_accuracy": 0.21295267045497895,
      "num_tokens": 112458296.0,
      "step": 64830
    },
    {
      "entropy": 5.89515905380249,
      "epoch": 5.044349348375802,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002745898095302106,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22082732170820235,
      "num_tokens": 112466374.0,
      "step": 64835
    },
    {
      "entropy": 5.807960510253906,
      "epoch": 5.04473837774752,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027456209399181466,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22418073564767838,
      "num_tokens": 112474795.0,
      "step": 64840
    },
    {
      "entropy": 5.889541244506836,
      "epoch": 5.045127407119238,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027453437846006325,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2250228703022003,
      "num_tokens": 112483843.0,
      "step": 64845
    },
    {
      "entropy": 5.897929668426514,
      "epoch": 5.0455164364909555,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027450666293537683,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2148088589310646,
      "num_tokens": 112493156.0,
      "step": 64850
    },
    {
      "entropy": 5.911912965774536,
      "epoch": 5.045905465862672,
      "grad_norm": 1.546875,
      "learning_rate": 0.000274478947418176,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22611821442842484,
      "num_tokens": 112502112.0,
      "step": 64855
    },
    {
      "entropy": 5.872188568115234,
      "epoch": 5.04629449523439,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002744512319088814,
      "loss": 4.8298,
      "mean_token_accuracy": 0.2223617374897003,
      "num_tokens": 112510565.0,
      "step": 64860
    },
    {
      "entropy": 5.902511978149414,
      "epoch": 5.046683524606108,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002744235164079135,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2180855542421341,
      "num_tokens": 112519677.0,
      "step": 64865
    },
    {
      "entropy": 5.9034675598144535,
      "epoch": 5.047072553977825,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002743958009156928,
      "loss": 4.9347,
      "mean_token_accuracy": 0.21795149892568588,
      "num_tokens": 112527865.0,
      "step": 64870
    },
    {
      "entropy": 5.863388109207153,
      "epoch": 5.047461583349543,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002743680854326398,
      "loss": 4.92,
      "mean_token_accuracy": 0.21361682415008545,
      "num_tokens": 112536738.0,
      "step": 64875
    },
    {
      "entropy": 5.8773950099945065,
      "epoch": 5.047850612721261,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002743403699591751,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21387017965316774,
      "num_tokens": 112545055.0,
      "step": 64880
    },
    {
      "entropy": 5.95924015045166,
      "epoch": 5.048239642092978,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027431265449571937,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21599520593881608,
      "num_tokens": 112553986.0,
      "step": 64885
    },
    {
      "entropy": 5.839030647277832,
      "epoch": 5.048628671464695,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027428493904269285,
      "loss": 4.8555,
      "mean_token_accuracy": 0.21672775149345397,
      "num_tokens": 112562980.0,
      "step": 64890
    },
    {
      "entropy": 5.8693924903869625,
      "epoch": 5.049017700836413,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002742572236005164,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22095745354890822,
      "num_tokens": 112571884.0,
      "step": 64895
    },
    {
      "entropy": 5.802287244796753,
      "epoch": 5.049406730208131,
      "grad_norm": 1.53125,
      "learning_rate": 0.00027422950816961025,
      "loss": 4.8307,
      "mean_token_accuracy": 0.22675658464431764,
      "num_tokens": 112580120.0,
      "step": 64900
    },
    {
      "entropy": 5.853997039794922,
      "epoch": 5.049795759579848,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000274201792750395,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2200113609433174,
      "num_tokens": 112588603.0,
      "step": 64905
    },
    {
      "entropy": 5.893977546691895,
      "epoch": 5.050184788951566,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002741740773432914,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22451066374778747,
      "num_tokens": 112596528.0,
      "step": 64910
    },
    {
      "entropy": 5.800830221176147,
      "epoch": 5.050573818323284,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002741463619487199,
      "loss": 4.764,
      "mean_token_accuracy": 0.23505058288574218,
      "num_tokens": 112606080.0,
      "step": 64915
    },
    {
      "entropy": 5.798304080963135,
      "epoch": 5.050962847695001,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002741186465671011,
      "loss": 4.9136,
      "mean_token_accuracy": 0.21558415591716767,
      "num_tokens": 112614403.0,
      "step": 64920
    },
    {
      "entropy": 5.892044401168823,
      "epoch": 5.051351877066718,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002740909311988553,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21385367959737778,
      "num_tokens": 112622402.0,
      "step": 64925
    },
    {
      "entropy": 5.893374109268189,
      "epoch": 5.051740906438436,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002740632158444031,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2216365799307823,
      "num_tokens": 112631556.0,
      "step": 64930
    },
    {
      "entropy": 5.869762325286866,
      "epoch": 5.052129935810154,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002740355005041652,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23569984436035157,
      "num_tokens": 112640328.0,
      "step": 64935
    },
    {
      "entropy": 5.855741834640503,
      "epoch": 5.052518965181871,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002740077851785621,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21776291131973266,
      "num_tokens": 112648302.0,
      "step": 64940
    },
    {
      "entropy": 5.9218274593353275,
      "epoch": 5.052907994553589,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027398006986801415,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21496013700962066,
      "num_tokens": 112657202.0,
      "step": 64945
    },
    {
      "entropy": 5.902160596847534,
      "epoch": 5.053297023925307,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002739523545729421,
      "loss": 4.966,
      "mean_token_accuracy": 0.2142154946923256,
      "num_tokens": 112665869.0,
      "step": 64950
    },
    {
      "entropy": 5.917887020111084,
      "epoch": 5.053686053297024,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002739246392937664,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21729104369878768,
      "num_tokens": 112675141.0,
      "step": 64955
    },
    {
      "entropy": 5.888119125366211,
      "epoch": 5.054075082668741,
      "grad_norm": 1.390625,
      "learning_rate": 0.00027389692403090754,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22151168137788774,
      "num_tokens": 112684735.0,
      "step": 64960
    },
    {
      "entropy": 5.793762302398681,
      "epoch": 5.054464112040459,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00027386920878478616,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22848789691925048,
      "num_tokens": 112693383.0,
      "step": 64965
    },
    {
      "entropy": 5.915420150756836,
      "epoch": 5.054853141412177,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027384149355582266,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21750899404287338,
      "num_tokens": 112702211.0,
      "step": 64970
    },
    {
      "entropy": 5.874306869506836,
      "epoch": 5.055242170783894,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002738137783444377,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23560174107551574,
      "num_tokens": 112710398.0,
      "step": 64975
    },
    {
      "entropy": 5.9192358493804935,
      "epoch": 5.055631200155612,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027378606315105175,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21411571949720382,
      "num_tokens": 112719278.0,
      "step": 64980
    },
    {
      "entropy": 5.817488336563111,
      "epoch": 5.056020229527329,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002737583479760854,
      "loss": 4.854,
      "mean_token_accuracy": 0.22727371007204056,
      "num_tokens": 112728164.0,
      "step": 64985
    },
    {
      "entropy": 5.859387588500977,
      "epoch": 5.056409258899047,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002737306328199591,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2227487489581108,
      "num_tokens": 112737036.0,
      "step": 64990
    },
    {
      "entropy": 5.935918855667114,
      "epoch": 5.056798288270764,
      "grad_norm": 1.5625,
      "learning_rate": 0.00027370291768309354,
      "loss": 4.9122,
      "mean_token_accuracy": 0.21944176256656647,
      "num_tokens": 112745113.0,
      "step": 64995
    },
    {
      "entropy": 5.852061319351196,
      "epoch": 5.057187317642482,
      "grad_norm": 1.5,
      "learning_rate": 0.000273675202565909,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22317281663417815,
      "num_tokens": 112753667.0,
      "step": 65000
    },
    {
      "entropy": 5.914431095123291,
      "epoch": 5.0575763470142,
      "grad_norm": 1.5,
      "learning_rate": 0.00027364748746882623,
      "loss": 4.9038,
      "mean_token_accuracy": 0.21511419713497162,
      "num_tokens": 112761409.0,
      "step": 65005
    },
    {
      "entropy": 5.854408645629883,
      "epoch": 5.057965376385917,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027361977239226566,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21941624879837035,
      "num_tokens": 112770235.0,
      "step": 65010
    },
    {
      "entropy": 5.90045804977417,
      "epoch": 5.058354405757635,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002735920573366478,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21663012951612473,
      "num_tokens": 112778300.0,
      "step": 65015
    },
    {
      "entropy": 5.908638143539429,
      "epoch": 5.058743435129352,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002735643423023933,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21123852878808974,
      "num_tokens": 112787518.0,
      "step": 65020
    },
    {
      "entropy": 5.891512250900268,
      "epoch": 5.05913246450107,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002735366272899226,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22273272573947905,
      "num_tokens": 112796100.0,
      "step": 65025
    },
    {
      "entropy": 5.798455858230591,
      "epoch": 5.059521493872787,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027350891229965623,
      "loss": 4.8079,
      "mean_token_accuracy": 0.23303077667951583,
      "num_tokens": 112804756.0,
      "step": 65030
    },
    {
      "entropy": 5.890562582015991,
      "epoch": 5.059910523244505,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002734811973320147,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21506002992391587,
      "num_tokens": 112813020.0,
      "step": 65035
    },
    {
      "entropy": 5.884553241729736,
      "epoch": 5.060299552616223,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027345348238741864,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21912122070789336,
      "num_tokens": 112822364.0,
      "step": 65040
    },
    {
      "entropy": 5.936220264434814,
      "epoch": 5.06068858198794,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027342576746628854,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22500730454921722,
      "num_tokens": 112830709.0,
      "step": 65045
    },
    {
      "entropy": 5.853667736053467,
      "epoch": 5.061077611359658,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002733980525690449,
      "loss": 4.8307,
      "mean_token_accuracy": 0.21764699816703797,
      "num_tokens": 112839108.0,
      "step": 65050
    },
    {
      "entropy": 5.797406482696533,
      "epoch": 5.061466640731375,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027337033769610827,
      "loss": 4.8278,
      "mean_token_accuracy": 0.21902662962675096,
      "num_tokens": 112847521.0,
      "step": 65055
    },
    {
      "entropy": 5.863193988800049,
      "epoch": 5.061855670103093,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002733426228478991,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22062177658081056,
      "num_tokens": 112856758.0,
      "step": 65060
    },
    {
      "entropy": 5.864192342758178,
      "epoch": 5.06224469947481,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002733149080248381,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22126346081495285,
      "num_tokens": 112864628.0,
      "step": 65065
    },
    {
      "entropy": 5.883143615722656,
      "epoch": 5.062633728846528,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002732871932273456,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2171648398041725,
      "num_tokens": 112873407.0,
      "step": 65070
    },
    {
      "entropy": 5.885685777664184,
      "epoch": 5.063022758218246,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002732594784558423,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21406883746385574,
      "num_tokens": 112881577.0,
      "step": 65075
    },
    {
      "entropy": 5.839009714126587,
      "epoch": 5.063411787589963,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002732317637107486,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22209562510251998,
      "num_tokens": 112889985.0,
      "step": 65080
    },
    {
      "entropy": 5.894692611694336,
      "epoch": 5.063800816961681,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027320404899248507,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21197282522916794,
      "num_tokens": 112897506.0,
      "step": 65085
    },
    {
      "entropy": 5.980425119400024,
      "epoch": 5.064189846333398,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002731763343014722,
      "loss": 5.052,
      "mean_token_accuracy": 0.20575623363256454,
      "num_tokens": 112906433.0,
      "step": 65090
    },
    {
      "entropy": 5.921547079086304,
      "epoch": 5.064578875705116,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027314861963813065,
      "loss": 4.9728,
      "mean_token_accuracy": 0.20803272128105163,
      "num_tokens": 112915006.0,
      "step": 65095
    },
    {
      "entropy": 5.837696361541748,
      "epoch": 5.064967905076833,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027312090500288065,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2220611333847046,
      "num_tokens": 112923056.0,
      "step": 65100
    },
    {
      "entropy": 5.854428052902222,
      "epoch": 5.065356934448551,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002730931903961431,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21267687380313874,
      "num_tokens": 112931345.0,
      "step": 65105
    },
    {
      "entropy": 5.862637186050415,
      "epoch": 5.065745963820269,
      "grad_norm": 1.484375,
      "learning_rate": 0.00027306547581833836,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22649916559457778,
      "num_tokens": 112939983.0,
      "step": 65110
    },
    {
      "entropy": 5.802713346481323,
      "epoch": 5.066134993191986,
      "grad_norm": 1.46875,
      "learning_rate": 0.00027303776126988687,
      "loss": 4.7651,
      "mean_token_accuracy": 0.22732220143079757,
      "num_tokens": 112948760.0,
      "step": 65115
    },
    {
      "entropy": 5.828072357177734,
      "epoch": 5.066524022563703,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027301004675120926,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21431052684783936,
      "num_tokens": 112957861.0,
      "step": 65120
    },
    {
      "entropy": 5.886091184616089,
      "epoch": 5.066913051935421,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002729823322627261,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22917547821998596,
      "num_tokens": 112966027.0,
      "step": 65125
    },
    {
      "entropy": 5.905043125152588,
      "epoch": 5.0673020813071386,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002729546178048577,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21229495257139205,
      "num_tokens": 112974256.0,
      "step": 65130
    },
    {
      "entropy": 5.844264984130859,
      "epoch": 5.067691110678856,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002729269033780248,
      "loss": 4.893,
      "mean_token_accuracy": 0.2265363499522209,
      "num_tokens": 112982611.0,
      "step": 65135
    },
    {
      "entropy": 5.872763729095459,
      "epoch": 5.068080140050574,
      "grad_norm": 1.46875,
      "learning_rate": 0.00027289918898264784,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21456132382154464,
      "num_tokens": 112991258.0,
      "step": 65140
    },
    {
      "entropy": 5.847701215744019,
      "epoch": 5.068469169422292,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002728714746191473,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21162209808826446,
      "num_tokens": 113000354.0,
      "step": 65145
    },
    {
      "entropy": 5.854842710494995,
      "epoch": 5.068858198794009,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027284376028794376,
      "loss": 4.7918,
      "mean_token_accuracy": 0.22677891254425048,
      "num_tokens": 113008858.0,
      "step": 65150
    },
    {
      "entropy": 5.915000343322754,
      "epoch": 5.069247228165726,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027281604598945774,
      "loss": 4.817,
      "mean_token_accuracy": 0.2210051581263542,
      "num_tokens": 113017073.0,
      "step": 65155
    },
    {
      "entropy": 5.906723737716675,
      "epoch": 5.069636257537444,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002727883317241098,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2243797540664673,
      "num_tokens": 113025782.0,
      "step": 65160
    },
    {
      "entropy": 5.868117141723633,
      "epoch": 5.0700252869091615,
      "grad_norm": 1.5625,
      "learning_rate": 0.00027276061749232034,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21898618191480637,
      "num_tokens": 113034478.0,
      "step": 65165
    },
    {
      "entropy": 5.897947692871094,
      "epoch": 5.070414316280879,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002727329032945099,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21958808600902557,
      "num_tokens": 113043474.0,
      "step": 65170
    },
    {
      "entropy": 5.923863315582276,
      "epoch": 5.070803345652597,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002727051891310991,
      "loss": 4.943,
      "mean_token_accuracy": 0.2204459384083748,
      "num_tokens": 113051600.0,
      "step": 65175
    },
    {
      "entropy": 5.833640289306641,
      "epoch": 5.071192375024315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002726774750025084,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22142624706029893,
      "num_tokens": 113061062.0,
      "step": 65180
    },
    {
      "entropy": 5.870120477676392,
      "epoch": 5.071581404396032,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027264976090915835,
      "loss": 4.942,
      "mean_token_accuracy": 0.2112548068165779,
      "num_tokens": 113069795.0,
      "step": 65185
    },
    {
      "entropy": 5.844617080688477,
      "epoch": 5.071970433767749,
      "grad_norm": 1.5,
      "learning_rate": 0.0002726220468514694,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22756513506174086,
      "num_tokens": 113079037.0,
      "step": 65190
    },
    {
      "entropy": 5.90989499092102,
      "epoch": 5.072359463139467,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002725943328298621,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21848677694797516,
      "num_tokens": 113087431.0,
      "step": 65195
    },
    {
      "entropy": 5.787341594696045,
      "epoch": 5.0727484925111845,
      "grad_norm": 1.484375,
      "learning_rate": 0.000272566618844757,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22504399716854095,
      "num_tokens": 113095985.0,
      "step": 65200
    },
    {
      "entropy": 5.862154579162597,
      "epoch": 5.073137521882902,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027253890489657457,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2184295192360878,
      "num_tokens": 113104687.0,
      "step": 65205
    },
    {
      "entropy": 5.7922694206237795,
      "epoch": 5.07352655125462,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027251119098573533,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22839094698429108,
      "num_tokens": 113112799.0,
      "step": 65210
    },
    {
      "entropy": 5.951018953323365,
      "epoch": 5.073915580626338,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027248347711265983,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21086585372686387,
      "num_tokens": 113121421.0,
      "step": 65215
    },
    {
      "entropy": 5.88949556350708,
      "epoch": 5.074304609998054,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002724557632777686,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22497667372226715,
      "num_tokens": 113130708.0,
      "step": 65220
    },
    {
      "entropy": 5.854046821594238,
      "epoch": 5.074693639369772,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027242804948148197,
      "loss": 4.8807,
      "mean_token_accuracy": 0.21787459999322892,
      "num_tokens": 113139034.0,
      "step": 65225
    },
    {
      "entropy": 5.923110628128052,
      "epoch": 5.07508266874149,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002724003357242207,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2221601888537407,
      "num_tokens": 113148304.0,
      "step": 65230
    },
    {
      "entropy": 5.88196063041687,
      "epoch": 5.0754716981132075,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002723726220064051,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2134503334760666,
      "num_tokens": 113156984.0,
      "step": 65235
    },
    {
      "entropy": 5.821055507659912,
      "epoch": 5.075860727484925,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002723449083284559,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2156686916947365,
      "num_tokens": 113165885.0,
      "step": 65240
    },
    {
      "entropy": 5.916143846511841,
      "epoch": 5.076249756856643,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027231719469079344,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21719736456871033,
      "num_tokens": 113175196.0,
      "step": 65245
    },
    {
      "entropy": 5.900101184844971,
      "epoch": 5.076638786228361,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027228948109383826,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2231779471039772,
      "num_tokens": 113183811.0,
      "step": 65250
    },
    {
      "entropy": 5.913023376464844,
      "epoch": 5.077027815600077,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027226176753801095,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2208199679851532,
      "num_tokens": 113192846.0,
      "step": 65255
    },
    {
      "entropy": 5.937267732620239,
      "epoch": 5.077416844971795,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00027223405402373194,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21527650356292724,
      "num_tokens": 113201311.0,
      "step": 65260
    },
    {
      "entropy": 5.872148609161377,
      "epoch": 5.077805874343513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027220634055142174,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21491804718971252,
      "num_tokens": 113209799.0,
      "step": 65265
    },
    {
      "entropy": 5.902370595932007,
      "epoch": 5.0781949037152305,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002721786271215009,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2167784631252289,
      "num_tokens": 113219107.0,
      "step": 65270
    },
    {
      "entropy": 5.914782571792602,
      "epoch": 5.078583933086948,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002721509137343899,
      "loss": 4.949,
      "mean_token_accuracy": 0.2099096730351448,
      "num_tokens": 113228395.0,
      "step": 65275
    },
    {
      "entropy": 5.970580005645752,
      "epoch": 5.078972962458666,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027212320039050924,
      "loss": 5.0805,
      "mean_token_accuracy": 0.20246295183897017,
      "num_tokens": 113237934.0,
      "step": 65280
    },
    {
      "entropy": 5.901640176773071,
      "epoch": 5.079361991830384,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027209548709027944,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2157839596271515,
      "num_tokens": 113246184.0,
      "step": 65285
    },
    {
      "entropy": 5.868247079849243,
      "epoch": 5.0797510212021,
      "grad_norm": 1.4375,
      "learning_rate": 0.000272067773834121,
      "loss": 4.8152,
      "mean_token_accuracy": 0.2276090681552887,
      "num_tokens": 113254741.0,
      "step": 65290
    },
    {
      "entropy": 5.904131507873535,
      "epoch": 5.080140050573818,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002720400606224544,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2172112762928009,
      "num_tokens": 113262516.0,
      "step": 65295
    },
    {
      "entropy": 5.847368383407593,
      "epoch": 5.080529079945536,
      "grad_norm": 1.53125,
      "learning_rate": 0.00027201234745570014,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22352254390716553,
      "num_tokens": 113270714.0,
      "step": 65300
    },
    {
      "entropy": 5.787071990966797,
      "epoch": 5.0809181093172535,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002719846343342788,
      "loss": 4.8334,
      "mean_token_accuracy": 0.22568640410900115,
      "num_tokens": 113279244.0,
      "step": 65305
    },
    {
      "entropy": 5.828607082366943,
      "epoch": 5.081307138688971,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002719569212586108,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2223309949040413,
      "num_tokens": 113287429.0,
      "step": 65310
    },
    {
      "entropy": 5.908442401885987,
      "epoch": 5.081696168060689,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002719292082291167,
      "loss": 5.0136,
      "mean_token_accuracy": 0.20804882794618607,
      "num_tokens": 113296050.0,
      "step": 65315
    },
    {
      "entropy": 5.911419010162353,
      "epoch": 5.082085197432406,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000271901495246217,
      "loss": 4.9612,
      "mean_token_accuracy": 0.21422354876995087,
      "num_tokens": 113304153.0,
      "step": 65320
    },
    {
      "entropy": 5.856732320785523,
      "epoch": 5.082474226804123,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002718737823103322,
      "loss": 4.8601,
      "mean_token_accuracy": 0.2237371787428856,
      "num_tokens": 113312561.0,
      "step": 65325
    },
    {
      "entropy": 5.874054765701294,
      "epoch": 5.082863256175841,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027184606942188275,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22814837247133254,
      "num_tokens": 113321480.0,
      "step": 65330
    },
    {
      "entropy": 5.895652341842651,
      "epoch": 5.083252285547559,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027181835658128913,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21708130985498428,
      "num_tokens": 113330510.0,
      "step": 65335
    },
    {
      "entropy": 5.82962737083435,
      "epoch": 5.0836413149192765,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002717906437889719,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22329147756099701,
      "num_tokens": 113340273.0,
      "step": 65340
    },
    {
      "entropy": 5.8773986339569095,
      "epoch": 5.084030344290994,
      "grad_norm": 1.59375,
      "learning_rate": 0.00027176293104535154,
      "loss": 4.8517,
      "mean_token_accuracy": 0.22134459763765335,
      "num_tokens": 113348963.0,
      "step": 65345
    },
    {
      "entropy": 5.814590549468994,
      "epoch": 5.084419373662712,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027173521835084863,
      "loss": 4.814,
      "mean_token_accuracy": 0.23178064674139023,
      "num_tokens": 113357678.0,
      "step": 65350
    },
    {
      "entropy": 5.847152233123779,
      "epoch": 5.084808403034429,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002717075057058835,
      "loss": 4.8191,
      "mean_token_accuracy": 0.22073301523923874,
      "num_tokens": 113366631.0,
      "step": 65355
    },
    {
      "entropy": 5.8026128768920895,
      "epoch": 5.085197432406146,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002716797931108768,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2252953380346298,
      "num_tokens": 113374591.0,
      "step": 65360
    },
    {
      "entropy": 5.826650381088257,
      "epoch": 5.085586461777864,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002716520805662489,
      "loss": 4.8943,
      "mean_token_accuracy": 0.21870972514152526,
      "num_tokens": 113384208.0,
      "step": 65365
    },
    {
      "entropy": 5.819668960571289,
      "epoch": 5.085975491149582,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027162436807242024,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21797839552164078,
      "num_tokens": 113393113.0,
      "step": 65370
    },
    {
      "entropy": 5.836445093154907,
      "epoch": 5.0863645205212995,
      "grad_norm": 1.515625,
      "learning_rate": 0.00027159665562981154,
      "loss": 4.925,
      "mean_token_accuracy": 0.22273023724555968,
      "num_tokens": 113401703.0,
      "step": 65375
    },
    {
      "entropy": 6.037199831008911,
      "epoch": 5.086753549893017,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002715689432388433,
      "loss": 5.0714,
      "mean_token_accuracy": 0.20666299015283585,
      "num_tokens": 113410636.0,
      "step": 65380
    },
    {
      "entropy": 5.912101984024048,
      "epoch": 5.087142579264735,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002715412308999357,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2101287141442299,
      "num_tokens": 113420443.0,
      "step": 65385
    },
    {
      "entropy": 5.876407623291016,
      "epoch": 5.087531608636452,
      "grad_norm": 1.46875,
      "learning_rate": 0.00027151351861350953,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22003291696310043,
      "num_tokens": 113428646.0,
      "step": 65390
    },
    {
      "entropy": 5.832454872131348,
      "epoch": 5.087920638008169,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002714858063799851,
      "loss": 4.8625,
      "mean_token_accuracy": 0.21711015999317168,
      "num_tokens": 113438526.0,
      "step": 65395
    },
    {
      "entropy": 5.861706733703613,
      "epoch": 5.088309667379887,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027145809419978293,
      "loss": 4.8694,
      "mean_token_accuracy": 0.21587102115154266,
      "num_tokens": 113447081.0,
      "step": 65400
    },
    {
      "entropy": 5.913576269149781,
      "epoch": 5.088698696751605,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00027143038207332363,
      "loss": 4.965,
      "mean_token_accuracy": 0.21326445639133454,
      "num_tokens": 113456309.0,
      "step": 65405
    },
    {
      "entropy": 5.911952066421509,
      "epoch": 5.0890877261233225,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027140267000102755,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2093920886516571,
      "num_tokens": 113464874.0,
      "step": 65410
    },
    {
      "entropy": 5.89169282913208,
      "epoch": 5.08947675549504,
      "grad_norm": 1.609375,
      "learning_rate": 0.00027137495798331524,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21414714455604553,
      "num_tokens": 113473841.0,
      "step": 65415
    },
    {
      "entropy": 5.823452043533325,
      "epoch": 5.089865784866757,
      "grad_norm": 1.75,
      "learning_rate": 0.00027134724602060717,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2303641378879547,
      "num_tokens": 113482594.0,
      "step": 65420
    },
    {
      "entropy": 5.864932918548584,
      "epoch": 5.090254814238475,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027131953411332384,
      "loss": 4.952,
      "mean_token_accuracy": 0.21977291256189346,
      "num_tokens": 113491256.0,
      "step": 65425
    },
    {
      "entropy": 5.844240331649781,
      "epoch": 5.090643843610192,
      "grad_norm": 1.53125,
      "learning_rate": 0.00027129182226188575,
      "loss": 4.8852,
      "mean_token_accuracy": 0.21609150916337966,
      "num_tokens": 113500221.0,
      "step": 65430
    },
    {
      "entropy": 5.759958982467651,
      "epoch": 5.09103287298191,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002712641104667132,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2331371396780014,
      "num_tokens": 113509060.0,
      "step": 65435
    },
    {
      "entropy": 5.924038362503052,
      "epoch": 5.091421902353628,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000271236398728227,
      "loss": 4.983,
      "mean_token_accuracy": 0.21313105821609496,
      "num_tokens": 113517953.0,
      "step": 65440
    },
    {
      "entropy": 5.93276743888855,
      "epoch": 5.0918109317253455,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027120868704684736,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21829409897327423,
      "num_tokens": 113526520.0,
      "step": 65445
    },
    {
      "entropy": 5.887671995162964,
      "epoch": 5.092199961097063,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002711809754229949,
      "loss": 4.941,
      "mean_token_accuracy": 0.21588577032089235,
      "num_tokens": 113535403.0,
      "step": 65450
    },
    {
      "entropy": 5.833799266815186,
      "epoch": 5.09258899046878,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027115326385709004,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22690628170967103,
      "num_tokens": 113543763.0,
      "step": 65455
    },
    {
      "entropy": 5.911220932006836,
      "epoch": 5.092978019840498,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002711255523495534,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2185896009206772,
      "num_tokens": 113551558.0,
      "step": 65460
    },
    {
      "entropy": 5.942658710479736,
      "epoch": 5.093367049212215,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002710978409008052,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22096379846334457,
      "num_tokens": 113559603.0,
      "step": 65465
    },
    {
      "entropy": 5.855960273742676,
      "epoch": 5.093756078583933,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027107012951126607,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23044936805963517,
      "num_tokens": 113568116.0,
      "step": 65470
    },
    {
      "entropy": 5.870174884796143,
      "epoch": 5.094145107955651,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027104241818135655,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2154032915830612,
      "num_tokens": 113576778.0,
      "step": 65475
    },
    {
      "entropy": 5.796998119354248,
      "epoch": 5.0945341373273685,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027101470691149693,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23330315351486205,
      "num_tokens": 113585453.0,
      "step": 65480
    },
    {
      "entropy": 5.891551399230957,
      "epoch": 5.094923166699086,
      "grad_norm": 1.515625,
      "learning_rate": 0.00027098699570210787,
      "loss": 4.8596,
      "mean_token_accuracy": 0.22142219990491868,
      "num_tokens": 113593727.0,
      "step": 65485
    },
    {
      "entropy": 5.888092517852783,
      "epoch": 5.095312196070803,
      "grad_norm": 1.421875,
      "learning_rate": 0.00027095928455360977,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21614392548799516,
      "num_tokens": 113601584.0,
      "step": 65490
    },
    {
      "entropy": 5.868610811233521,
      "epoch": 5.095701225442521,
      "grad_norm": 1.546875,
      "learning_rate": 0.000270931573466423,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21638845801353454,
      "num_tokens": 113610509.0,
      "step": 65495
    },
    {
      "entropy": 5.911000394821167,
      "epoch": 5.096090254814238,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002709038624409681,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2081785038113594,
      "num_tokens": 113618642.0,
      "step": 65500
    },
    {
      "entropy": 5.9551338195800785,
      "epoch": 5.096479284185956,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027087615147766567,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21395208686590195,
      "num_tokens": 113627385.0,
      "step": 65505
    },
    {
      "entropy": 5.969971370697022,
      "epoch": 5.096868313557674,
      "grad_norm": 1.484375,
      "learning_rate": 0.000270848440576936,
      "loss": 4.973,
      "mean_token_accuracy": 0.21126056760549544,
      "num_tokens": 113636581.0,
      "step": 65510
    },
    {
      "entropy": 5.869914293289185,
      "epoch": 5.0972573429293915,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00027082072973919974,
      "loss": 4.8711,
      "mean_token_accuracy": 0.22226247042417527,
      "num_tokens": 113645825.0,
      "step": 65515
    },
    {
      "entropy": 5.8987444877624515,
      "epoch": 5.097646372301108,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00027079301896487716,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21859501749277116,
      "num_tokens": 113654709.0,
      "step": 65520
    },
    {
      "entropy": 5.941103267669678,
      "epoch": 5.098035401672826,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002707653082543889,
      "loss": 5.0135,
      "mean_token_accuracy": 0.2200246348977089,
      "num_tokens": 113663536.0,
      "step": 65525
    },
    {
      "entropy": 5.842590284347534,
      "epoch": 5.098424431044544,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002707375976081553,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21426498144865036,
      "num_tokens": 113671597.0,
      "step": 65530
    },
    {
      "entropy": 5.855186700820923,
      "epoch": 5.098813460416261,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027070988702659687,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2239503711462021,
      "num_tokens": 113679913.0,
      "step": 65535
    },
    {
      "entropy": 5.866179370880127,
      "epoch": 5.099202489787979,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00027068217651013415,
      "loss": 4.899,
      "mean_token_accuracy": 0.21892814934253693,
      "num_tokens": 113688263.0,
      "step": 65540
    },
    {
      "entropy": 5.925367403030395,
      "epoch": 5.099591519159697,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002706544660591875,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2114056646823883,
      "num_tokens": 113698391.0,
      "step": 65545
    },
    {
      "entropy": 5.8344817638397215,
      "epoch": 5.0999805485314145,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00027062675567417736,
      "loss": 4.8168,
      "mean_token_accuracy": 0.2248559519648552,
      "num_tokens": 113705949.0,
      "step": 65550
    },
    {
      "entropy": 5.798208236694336,
      "epoch": 5.100369577903131,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027059904535552424,
      "loss": 4.8812,
      "mean_token_accuracy": 0.21425600796937944,
      "num_tokens": 113714104.0,
      "step": 65555
    },
    {
      "entropy": 5.9197744846344,
      "epoch": 5.100758607274849,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027057133510364866,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21435917317867278,
      "num_tokens": 113723099.0,
      "step": 65560
    },
    {
      "entropy": 5.908179473876953,
      "epoch": 5.101147636646567,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002705436249189709,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21976491063833237,
      "num_tokens": 113731102.0,
      "step": 65565
    },
    {
      "entropy": 5.951141738891602,
      "epoch": 5.101536666018284,
      "grad_norm": 1.5,
      "learning_rate": 0.00027051591480191167,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21568616181612016,
      "num_tokens": 113739328.0,
      "step": 65570
    },
    {
      "entropy": 5.850593852996826,
      "epoch": 5.101925695390002,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002704882047528912,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21705861538648605,
      "num_tokens": 113748592.0,
      "step": 65575
    },
    {
      "entropy": 5.862431573867798,
      "epoch": 5.10231472476172,
      "grad_norm": 1.546875,
      "learning_rate": 0.00027046049477233013,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23088606595993041,
      "num_tokens": 113756779.0,
      "step": 65580
    },
    {
      "entropy": 5.913510704040528,
      "epoch": 5.1027037541334375,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002704327848606488,
      "loss": 5.0266,
      "mean_token_accuracy": 0.21351926326751708,
      "num_tokens": 113765315.0,
      "step": 65585
    },
    {
      "entropy": 5.861938524246216,
      "epoch": 5.103092783505154,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002704050750182677,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2186524122953415,
      "num_tokens": 113773589.0,
      "step": 65590
    },
    {
      "entropy": 5.7822352886199955,
      "epoch": 5.103481812876872,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00027037736524560717,
      "loss": 4.8185,
      "mean_token_accuracy": 0.23314427584409714,
      "num_tokens": 113782477.0,
      "step": 65595
    },
    {
      "entropy": 5.829677486419678,
      "epoch": 5.10387084224859,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002703496555430879,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21592042446136475,
      "num_tokens": 113791092.0,
      "step": 65600
    },
    {
      "entropy": 5.888465738296508,
      "epoch": 5.104259871620307,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00027032194591113015,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22251616567373275,
      "num_tokens": 113799770.0,
      "step": 65605
    },
    {
      "entropy": 5.888960170745849,
      "epoch": 5.104648900992025,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002702942363501544,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22633911520242692,
      "num_tokens": 113809062.0,
      "step": 65610
    },
    {
      "entropy": 5.846899843215942,
      "epoch": 5.105037930363743,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027026652686058117,
      "loss": 4.8545,
      "mean_token_accuracy": 0.2220997393131256,
      "num_tokens": 113817944.0,
      "step": 65615
    },
    {
      "entropy": 5.8574799537658695,
      "epoch": 5.1054269597354605,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00027023881744283083,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21502801775932312,
      "num_tokens": 113826486.0,
      "step": 65620
    },
    {
      "entropy": 6.026064348220825,
      "epoch": 5.105815989107177,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002702111080973239,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2053128033876419,
      "num_tokens": 113834982.0,
      "step": 65625
    },
    {
      "entropy": 5.8985813617706295,
      "epoch": 5.106205018478895,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002701833988244808,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2196221485733986,
      "num_tokens": 113844521.0,
      "step": 65630
    },
    {
      "entropy": 5.850252342224121,
      "epoch": 5.106594047850613,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00027015568962472175,
      "loss": 4.8391,
      "mean_token_accuracy": 0.22889325320720671,
      "num_tokens": 113853332.0,
      "step": 65635
    },
    {
      "entropy": 5.840620660781861,
      "epoch": 5.10698307722233,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00027012798049846757,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21579495817422867,
      "num_tokens": 113861959.0,
      "step": 65640
    },
    {
      "entropy": 5.970750236511231,
      "epoch": 5.107372106594048,
      "grad_norm": 1.625,
      "learning_rate": 0.0002701002714461385,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21270469427108765,
      "num_tokens": 113869589.0,
      "step": 65645
    },
    {
      "entropy": 5.880407333374023,
      "epoch": 5.107761135965766,
      "grad_norm": 1.375,
      "learning_rate": 0.0002700725624681551,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2121308773756027,
      "num_tokens": 113878610.0,
      "step": 65650
    },
    {
      "entropy": 5.879530954360962,
      "epoch": 5.108150165337483,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027004485356493763,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22617058902978898,
      "num_tokens": 113887670.0,
      "step": 65655
    },
    {
      "entropy": 5.906849670410156,
      "epoch": 5.1085391947092,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027001714473690665,
      "loss": 4.9547,
      "mean_token_accuracy": 0.21259665936231614,
      "num_tokens": 113896282.0,
      "step": 65660
    },
    {
      "entropy": 5.900206708908081,
      "epoch": 5.108928224080918,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026998943598448256,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21903114914894103,
      "num_tokens": 113905144.0,
      "step": 65665
    },
    {
      "entropy": 5.86332516670227,
      "epoch": 5.109317253452636,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002699617273080858,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22082848995923995,
      "num_tokens": 113913418.0,
      "step": 65670
    },
    {
      "entropy": 5.859861516952515,
      "epoch": 5.109706282824353,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002699340187081368,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21316328942775725,
      "num_tokens": 113923135.0,
      "step": 65675
    },
    {
      "entropy": 5.9240306377410885,
      "epoch": 5.110095312196071,
      "grad_norm": 1.59375,
      "learning_rate": 0.000269906310185056,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22368897944688798,
      "num_tokens": 113932015.0,
      "step": 65680
    },
    {
      "entropy": 5.974324083328247,
      "epoch": 5.110484341567789,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026987860173926387,
      "loss": 5.0292,
      "mean_token_accuracy": 0.20854596197605133,
      "num_tokens": 113940273.0,
      "step": 65685
    },
    {
      "entropy": 5.938892078399658,
      "epoch": 5.1108733709395056,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026985089337118077,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21817571818828582,
      "num_tokens": 113948607.0,
      "step": 65690
    },
    {
      "entropy": 5.858677864074707,
      "epoch": 5.111262400311223,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026982318508122724,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22024273574352266,
      "num_tokens": 113956957.0,
      "step": 65695
    },
    {
      "entropy": 5.916488552093506,
      "epoch": 5.111651429682941,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002697954768698235,
      "loss": 5.006,
      "mean_token_accuracy": 0.20754637569189072,
      "num_tokens": 113966320.0,
      "step": 65700
    },
    {
      "entropy": 5.871480941772461,
      "epoch": 5.112040459054659,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026976776873739017,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2208838254213333,
      "num_tokens": 113974719.0,
      "step": 65705
    },
    {
      "entropy": 5.919399404525757,
      "epoch": 5.112429488426376,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002697400606843477,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21417030394077302,
      "num_tokens": 113983224.0,
      "step": 65710
    },
    {
      "entropy": 5.905768394470215,
      "epoch": 5.112818517798094,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026971235271111636,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21973634660243987,
      "num_tokens": 113992209.0,
      "step": 65715
    },
    {
      "entropy": 5.858517599105835,
      "epoch": 5.113207547169812,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026968464481811667,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22360142916440964,
      "num_tokens": 114000896.0,
      "step": 65720
    },
    {
      "entropy": 5.879826021194458,
      "epoch": 5.1135965765415285,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026965693700576904,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22303366959095,
      "num_tokens": 114010015.0,
      "step": 65725
    },
    {
      "entropy": 5.818126106262207,
      "epoch": 5.113985605913246,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026962922927449394,
      "loss": 4.8123,
      "mean_token_accuracy": 0.22186270505189895,
      "num_tokens": 114018229.0,
      "step": 65730
    },
    {
      "entropy": 5.8122330665588375,
      "epoch": 5.114374635284964,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002696015216247117,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22155270725488663,
      "num_tokens": 114027039.0,
      "step": 65735
    },
    {
      "entropy": 5.873907423019409,
      "epoch": 5.114763664656682,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002695738140568427,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20710396766662598,
      "num_tokens": 114035137.0,
      "step": 65740
    },
    {
      "entropy": 5.906149530410767,
      "epoch": 5.115152694028399,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002695461065713075,
      "loss": 4.9145,
      "mean_token_accuracy": 0.2211265593767166,
      "num_tokens": 114044105.0,
      "step": 65745
    },
    {
      "entropy": 5.868191957473755,
      "epoch": 5.115541723400117,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026951839916852647,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22507392466068268,
      "num_tokens": 114052535.0,
      "step": 65750
    },
    {
      "entropy": 5.915162992477417,
      "epoch": 5.115930752771834,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00026949069184892,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22042029350996017,
      "num_tokens": 114060620.0,
      "step": 65755
    },
    {
      "entropy": 5.860546731948853,
      "epoch": 5.1163197821435515,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026946298461290845,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2275201916694641,
      "num_tokens": 114070748.0,
      "step": 65760
    },
    {
      "entropy": 5.945041656494141,
      "epoch": 5.116708811515269,
      "grad_norm": 1.5,
      "learning_rate": 0.0002694352774609123,
      "loss": 4.9367,
      "mean_token_accuracy": 0.21616561561822892,
      "num_tokens": 114079525.0,
      "step": 65765
    },
    {
      "entropy": 5.893023347854614,
      "epoch": 5.117097840886987,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000269407570393352,
      "loss": 4.8906,
      "mean_token_accuracy": 0.21943559497594833,
      "num_tokens": 114088074.0,
      "step": 65770
    },
    {
      "entropy": 5.946559143066406,
      "epoch": 5.117486870258705,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000269379863410648,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2176995947957039,
      "num_tokens": 114097203.0,
      "step": 65775
    },
    {
      "entropy": 5.814003896713257,
      "epoch": 5.117875899630422,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002693521565132205,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22075533866882324,
      "num_tokens": 114105123.0,
      "step": 65780
    },
    {
      "entropy": 5.922133541107177,
      "epoch": 5.11826492900214,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026932444970149006,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22361107915639877,
      "num_tokens": 114114656.0,
      "step": 65785
    },
    {
      "entropy": 5.908982372283935,
      "epoch": 5.118653958373857,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002692967429758771,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22133497297763824,
      "num_tokens": 114122810.0,
      "step": 65790
    },
    {
      "entropy": 5.914796638488769,
      "epoch": 5.1190429877455745,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026926903633680193,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22167709171772004,
      "num_tokens": 114131809.0,
      "step": 65795
    },
    {
      "entropy": 5.870538854598999,
      "epoch": 5.119432017117292,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026924132978468505,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21987212151288987,
      "num_tokens": 114139534.0,
      "step": 65800
    },
    {
      "entropy": 5.883299255371094,
      "epoch": 5.11982104648901,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026921362331994684,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22485719472169877,
      "num_tokens": 114148248.0,
      "step": 65805
    },
    {
      "entropy": 5.893091154098511,
      "epoch": 5.120210075860728,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026918591694300776,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22348200380802155,
      "num_tokens": 114157130.0,
      "step": 65810
    },
    {
      "entropy": 5.82591986656189,
      "epoch": 5.120599105232445,
      "grad_norm": 1.5,
      "learning_rate": 0.00026915821065428804,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22988693416118622,
      "num_tokens": 114165544.0,
      "step": 65815
    },
    {
      "entropy": 5.949701166152954,
      "epoch": 5.120988134604163,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002691305044542082,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20471552312374114,
      "num_tokens": 114173829.0,
      "step": 65820
    },
    {
      "entropy": 5.850264930725098,
      "epoch": 5.12137716397588,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026910279834318855,
      "loss": 4.8699,
      "mean_token_accuracy": 0.21821389645338057,
      "num_tokens": 114182380.0,
      "step": 65825
    },
    {
      "entropy": 5.9048666000366214,
      "epoch": 5.1217661933475975,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026907509232164964,
      "loss": 5.001,
      "mean_token_accuracy": 0.20319158732891082,
      "num_tokens": 114191071.0,
      "step": 65830
    },
    {
      "entropy": 5.865254497528076,
      "epoch": 5.122155222719315,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026904738639001163,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21703410297632217,
      "num_tokens": 114199476.0,
      "step": 65835
    },
    {
      "entropy": 5.809226369857788,
      "epoch": 5.122544252091033,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002690196805486952,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22425133734941483,
      "num_tokens": 114208373.0,
      "step": 65840
    },
    {
      "entropy": 5.831390810012818,
      "epoch": 5.122933281462751,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026899197479812053,
      "loss": 4.947,
      "mean_token_accuracy": 0.21973471641540526,
      "num_tokens": 114216649.0,
      "step": 65845
    },
    {
      "entropy": 5.790037441253662,
      "epoch": 5.123322310834468,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002689642691387081,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22262126952409744,
      "num_tokens": 114225431.0,
      "step": 65850
    },
    {
      "entropy": 5.845479679107666,
      "epoch": 5.123711340206185,
      "grad_norm": 1.578125,
      "learning_rate": 0.00026893656357087833,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22621013671159745,
      "num_tokens": 114233857.0,
      "step": 65855
    },
    {
      "entropy": 5.849991273880005,
      "epoch": 5.124100369577903,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002689088580950515,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21286094337701797,
      "num_tokens": 114242741.0,
      "step": 65860
    },
    {
      "entropy": 5.844541025161743,
      "epoch": 5.1244893989496205,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000268881152711648,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23468927592039107,
      "num_tokens": 114250666.0,
      "step": 65865
    },
    {
      "entropy": 5.858936834335327,
      "epoch": 5.124878428321338,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002688534474210884,
      "loss": 4.9464,
      "mean_token_accuracy": 0.20674145966768265,
      "num_tokens": 114258870.0,
      "step": 65870
    },
    {
      "entropy": 5.824692153930664,
      "epoch": 5.125267457693056,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026882574222379283,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22415810823440552,
      "num_tokens": 114268700.0,
      "step": 65875
    },
    {
      "entropy": 5.9452191352844235,
      "epoch": 5.125656487064774,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026879803712018187,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2125726342201233,
      "num_tokens": 114277482.0,
      "step": 65880
    },
    {
      "entropy": 5.915705347061158,
      "epoch": 5.126045516436491,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002687703321106758,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22471041232347488,
      "num_tokens": 114286182.0,
      "step": 65885
    },
    {
      "entropy": 5.926186180114746,
      "epoch": 5.126434545808208,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002687426271956951,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2176584228873253,
      "num_tokens": 114295556.0,
      "step": 65890
    },
    {
      "entropy": 5.84512734413147,
      "epoch": 5.126823575179926,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026871492237566,
      "loss": 4.7851,
      "mean_token_accuracy": 0.21994955092668533,
      "num_tokens": 114304342.0,
      "step": 65895
    },
    {
      "entropy": 5.876171064376831,
      "epoch": 5.1272126045516435,
      "grad_norm": 1.5,
      "learning_rate": 0.00026868721765099086,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23264408111572266,
      "num_tokens": 114312653.0,
      "step": 65900
    },
    {
      "entropy": 5.771732521057129,
      "epoch": 5.127601633923361,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002686595130221082,
      "loss": 4.8363,
      "mean_token_accuracy": 0.22190264165401458,
      "num_tokens": 114321081.0,
      "step": 65905
    },
    {
      "entropy": 5.804191541671753,
      "epoch": 5.127990663295079,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026863180848943244,
      "loss": 4.919,
      "mean_token_accuracy": 0.21626996397972106,
      "num_tokens": 114330208.0,
      "step": 65910
    },
    {
      "entropy": 5.849740982055664,
      "epoch": 5.128379692666797,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002686041040533838,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23210158199071884,
      "num_tokens": 114338201.0,
      "step": 65915
    },
    {
      "entropy": 5.836041212081909,
      "epoch": 5.128768722038514,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002685763997143827,
      "loss": 4.8449,
      "mean_token_accuracy": 0.22335275262594223,
      "num_tokens": 114346877.0,
      "step": 65920
    },
    {
      "entropy": 5.829005336761474,
      "epoch": 5.129157751410231,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026854869547284947,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22081327438354492,
      "num_tokens": 114355506.0,
      "step": 65925
    },
    {
      "entropy": 5.882661247253418,
      "epoch": 5.129546780781949,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026852099132920455,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21788639277219773,
      "num_tokens": 114364372.0,
      "step": 65930
    },
    {
      "entropy": 5.92300271987915,
      "epoch": 5.1299358101536665,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002684932872838683,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2142435386776924,
      "num_tokens": 114372568.0,
      "step": 65935
    },
    {
      "entropy": 5.948053359985352,
      "epoch": 5.130324839525384,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000268465583337261,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20777359753847122,
      "num_tokens": 114381388.0,
      "step": 65940
    },
    {
      "entropy": 5.943252277374268,
      "epoch": 5.130713868897102,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002684378794898031,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21952307522296904,
      "num_tokens": 114390209.0,
      "step": 65945
    },
    {
      "entropy": 5.88835563659668,
      "epoch": 5.13110289826882,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026841017574191496,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22517536580562592,
      "num_tokens": 114398860.0,
      "step": 65950
    },
    {
      "entropy": 5.873006296157837,
      "epoch": 5.131491927640537,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002683824720940168,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21143682599067687,
      "num_tokens": 114407553.0,
      "step": 65955
    },
    {
      "entropy": 5.877561950683594,
      "epoch": 5.131880957012254,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026835476854652926,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2240485891699791,
      "num_tokens": 114416807.0,
      "step": 65960
    },
    {
      "entropy": 5.847120475769043,
      "epoch": 5.132269986383972,
      "grad_norm": 1.5,
      "learning_rate": 0.0002683270650998724,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21624185740947724,
      "num_tokens": 114425250.0,
      "step": 65965
    },
    {
      "entropy": 5.850070858001709,
      "epoch": 5.1326590157556895,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002682993617544667,
      "loss": 4.9434,
      "mean_token_accuracy": 0.20776990056037903,
      "num_tokens": 114434530.0,
      "step": 65970
    },
    {
      "entropy": 5.9022375583648685,
      "epoch": 5.133048045127407,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002682716585107326,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21379211246967317,
      "num_tokens": 114442964.0,
      "step": 65975
    },
    {
      "entropy": 5.912150812149048,
      "epoch": 5.133437074499125,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026824395536909036,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21660968661308289,
      "num_tokens": 114452381.0,
      "step": 65980
    },
    {
      "entropy": 5.813210153579712,
      "epoch": 5.133826103870843,
      "grad_norm": 1.5,
      "learning_rate": 0.0002682162523299603,
      "loss": 4.87,
      "mean_token_accuracy": 0.22334378957748413,
      "num_tokens": 114460590.0,
      "step": 65985
    },
    {
      "entropy": 5.921731805801391,
      "epoch": 5.134215133242559,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026818854939376284,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21537007987499238,
      "num_tokens": 114469616.0,
      "step": 65990
    },
    {
      "entropy": 5.89310393333435,
      "epoch": 5.134604162614277,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026816084656091835,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22286620438098909,
      "num_tokens": 114478200.0,
      "step": 65995
    },
    {
      "entropy": 5.873892402648925,
      "epoch": 5.134993191985995,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026813314383184714,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22126066237688063,
      "num_tokens": 114486389.0,
      "step": 66000
    },
    {
      "epoch": 5.134993191985995,
      "eval_entropy": 5.563581949937102,
      "eval_loss": 5.036537170410156,
      "eval_mean_token_accuracy": 0.2218062281665949,
      "eval_num_tokens": 114486389.0,
      "eval_runtime": 21.8285,
      "eval_samples_per_second": 1903.842,
      "eval_steps_per_second": 237.992,
      "step": 66000
    },
    {
      "entropy": 5.872282600402832,
      "epoch": 5.1353822213577125,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026810544120696943,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2182406008243561,
      "num_tokens": 114495338.0,
      "step": 66005
    },
    {
      "entropy": 5.917290544509887,
      "epoch": 5.13577125072943,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026807773868670575,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21585901379585265,
      "num_tokens": 114503744.0,
      "step": 66010
    },
    {
      "entropy": 5.911306953430175,
      "epoch": 5.136160280101148,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002680500362714764,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21984894275665284,
      "num_tokens": 114512991.0,
      "step": 66015
    },
    {
      "entropy": 5.9440789222717285,
      "epoch": 5.136549309472866,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00026802233396170164,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21140967160463334,
      "num_tokens": 114522067.0,
      "step": 66020
    },
    {
      "entropy": 5.911148357391357,
      "epoch": 5.136938338844582,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002679946317578018,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22086064517498016,
      "num_tokens": 114530569.0,
      "step": 66025
    },
    {
      "entropy": 5.854669666290283,
      "epoch": 5.1373273682163,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002679669296601973,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2238999590277672,
      "num_tokens": 114538716.0,
      "step": 66030
    },
    {
      "entropy": 5.813155794143677,
      "epoch": 5.137716397588018,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026793922766930845,
      "loss": 4.799,
      "mean_token_accuracy": 0.22434172183275222,
      "num_tokens": 114547572.0,
      "step": 66035
    },
    {
      "entropy": 5.823918247222901,
      "epoch": 5.1381054269597355,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026791152578555563,
      "loss": 4.882,
      "mean_token_accuracy": 0.21725434064865112,
      "num_tokens": 114555502.0,
      "step": 66040
    },
    {
      "entropy": 5.836768341064453,
      "epoch": 5.138494456331453,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026788382400935913,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21897660195827484,
      "num_tokens": 114564190.0,
      "step": 66045
    },
    {
      "entropy": 5.910693025588989,
      "epoch": 5.138883485703171,
      "grad_norm": 1.375,
      "learning_rate": 0.0002678561223411392,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22513029128313064,
      "num_tokens": 114573988.0,
      "step": 66050
    },
    {
      "entropy": 5.872666835784912,
      "epoch": 5.139272515074889,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002678284207813163,
      "loss": 4.8198,
      "mean_token_accuracy": 0.22817371189594268,
      "num_tokens": 114583155.0,
      "step": 66055
    },
    {
      "entropy": 5.853002738952637,
      "epoch": 5.139661544446605,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002678007193303107,
      "loss": 4.889,
      "mean_token_accuracy": 0.227026005089283,
      "num_tokens": 114591491.0,
      "step": 66060
    },
    {
      "entropy": 5.823988723754883,
      "epoch": 5.140050573818323,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002677730179885427,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2195958226919174,
      "num_tokens": 114600275.0,
      "step": 66065
    },
    {
      "entropy": 5.885872840881348,
      "epoch": 5.140439603190041,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026774531675643267,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21493610441684724,
      "num_tokens": 114608999.0,
      "step": 66070
    },
    {
      "entropy": 5.844340896606445,
      "epoch": 5.1408286325617585,
      "grad_norm": 1.375,
      "learning_rate": 0.0002677176156344009,
      "loss": 4.7524,
      "mean_token_accuracy": 0.23454027771949768,
      "num_tokens": 114617750.0,
      "step": 66075
    },
    {
      "entropy": 5.875127601623535,
      "epoch": 5.141217661933476,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002676899146228678,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2248332306742668,
      "num_tokens": 114626610.0,
      "step": 66080
    },
    {
      "entropy": 5.906566095352173,
      "epoch": 5.141606691305194,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026766221372225354,
      "loss": 4.9563,
      "mean_token_accuracy": 0.215479576587677,
      "num_tokens": 114635199.0,
      "step": 66085
    },
    {
      "entropy": 5.884940814971924,
      "epoch": 5.141995720676911,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002676345129329785,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21726604104042052,
      "num_tokens": 114644104.0,
      "step": 66090
    },
    {
      "entropy": 5.850084447860718,
      "epoch": 5.142384750048628,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026760681225546296,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2195482909679413,
      "num_tokens": 114653562.0,
      "step": 66095
    },
    {
      "entropy": 5.85883731842041,
      "epoch": 5.142773779420346,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002675791116901273,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22434998601675032,
      "num_tokens": 114661888.0,
      "step": 66100
    },
    {
      "entropy": 5.8832379341125485,
      "epoch": 5.143162808792064,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002675514112373918,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21074043065309525,
      "num_tokens": 114670942.0,
      "step": 66105
    },
    {
      "entropy": 5.874914073944092,
      "epoch": 5.1435518381637815,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026752371089767686,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22310996055603027,
      "num_tokens": 114679570.0,
      "step": 66110
    },
    {
      "entropy": 5.921423292160034,
      "epoch": 5.143940867535499,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002674960106714026,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2123573899269104,
      "num_tokens": 114687833.0,
      "step": 66115
    },
    {
      "entropy": 5.862590599060058,
      "epoch": 5.144329896907217,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026746831055898953,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22230245769023896,
      "num_tokens": 114695534.0,
      "step": 66120
    },
    {
      "entropy": 5.891458988189697,
      "epoch": 5.144718926278934,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002674406105608578,
      "loss": 4.8409,
      "mean_token_accuracy": 0.21625454127788543,
      "num_tokens": 114704443.0,
      "step": 66125
    },
    {
      "entropy": 5.828302574157715,
      "epoch": 5.145107955650651,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002674129106774278,
      "loss": 4.8642,
      "mean_token_accuracy": 0.21996913999319076,
      "num_tokens": 114713103.0,
      "step": 66130
    },
    {
      "entropy": 5.884413623809815,
      "epoch": 5.145496985022369,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026738521090911984,
      "loss": 4.914,
      "mean_token_accuracy": 0.21563147604465485,
      "num_tokens": 114721222.0,
      "step": 66135
    },
    {
      "entropy": 5.901731824874878,
      "epoch": 5.145886014394087,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002673575112563541,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20807105749845506,
      "num_tokens": 114730499.0,
      "step": 66140
    },
    {
      "entropy": 5.885125255584716,
      "epoch": 5.1462750437658045,
      "grad_norm": 1.453125,
      "learning_rate": 0.000267329811719551,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22027895897626876,
      "num_tokens": 114739536.0,
      "step": 66145
    },
    {
      "entropy": 5.92445936203003,
      "epoch": 5.146664073137522,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002673021122991308,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21122844368219376,
      "num_tokens": 114748062.0,
      "step": 66150
    },
    {
      "entropy": 5.966323804855347,
      "epoch": 5.14705310250924,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002672744129955138,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22188374102115632,
      "num_tokens": 114756315.0,
      "step": 66155
    },
    {
      "entropy": 5.84259443283081,
      "epoch": 5.147442131880957,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002672467138091204,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2332136243581772,
      "num_tokens": 114764714.0,
      "step": 66160
    },
    {
      "entropy": 5.805589199066162,
      "epoch": 5.147831161252674,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002672190147403705,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2252497136592865,
      "num_tokens": 114773357.0,
      "step": 66165
    },
    {
      "entropy": 5.87189040184021,
      "epoch": 5.148220190624392,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002671913157896849,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21112050116062164,
      "num_tokens": 114782100.0,
      "step": 66170
    },
    {
      "entropy": 5.950364875793457,
      "epoch": 5.14860921999611,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026716361695748353,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20136291533708572,
      "num_tokens": 114790827.0,
      "step": 66175
    },
    {
      "entropy": 5.870655298233032,
      "epoch": 5.1489982493678275,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002671359182441868,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22124303579330445,
      "num_tokens": 114799365.0,
      "step": 66180
    },
    {
      "entropy": 5.86721396446228,
      "epoch": 5.149387278739545,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026710821965021507,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2175467163324356,
      "num_tokens": 114807891.0,
      "step": 66185
    },
    {
      "entropy": 5.884500408172608,
      "epoch": 5.149776308111262,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002670805211759886,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22707600444555281,
      "num_tokens": 114816537.0,
      "step": 66190
    },
    {
      "entropy": 5.886652660369873,
      "epoch": 5.15016533748298,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002670528228219275,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22133893817663192,
      "num_tokens": 114824886.0,
      "step": 66195
    },
    {
      "entropy": 5.883154344558716,
      "epoch": 5.150554366854697,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026702512458845224,
      "loss": 4.875,
      "mean_token_accuracy": 0.22279927730560303,
      "num_tokens": 114833626.0,
      "step": 66200
    },
    {
      "entropy": 5.921316385269165,
      "epoch": 5.150943396226415,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026699742647598294,
      "loss": 4.9853,
      "mean_token_accuracy": 0.22070241272449492,
      "num_tokens": 114841678.0,
      "step": 66205
    },
    {
      "entropy": 5.980746746063232,
      "epoch": 5.151332425598133,
      "grad_norm": 1.5,
      "learning_rate": 0.00026696972848494005,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20526908189058304,
      "num_tokens": 114850948.0,
      "step": 66210
    },
    {
      "entropy": 5.925118637084961,
      "epoch": 5.1517214549698505,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026694203061574366,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21908355504274368,
      "num_tokens": 114859643.0,
      "step": 66215
    },
    {
      "entropy": 5.915058422088623,
      "epoch": 5.152110484341568,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026691433286881415,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21707459688186645,
      "num_tokens": 114868276.0,
      "step": 66220
    },
    {
      "entropy": 5.949935054779052,
      "epoch": 5.152499513713285,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002668866352445718,
      "loss": 4.915,
      "mean_token_accuracy": 0.22504731863737107,
      "num_tokens": 114876836.0,
      "step": 66225
    },
    {
      "entropy": 5.801215934753418,
      "epoch": 5.152888543085003,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002668589377434368,
      "loss": 4.7947,
      "mean_token_accuracy": 0.22922674864530562,
      "num_tokens": 114885736.0,
      "step": 66230
    },
    {
      "entropy": 5.893365859985352,
      "epoch": 5.15327757245672,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002668312403658295,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22239357382059097,
      "num_tokens": 114894040.0,
      "step": 66235
    },
    {
      "entropy": 5.842454433441162,
      "epoch": 5.153666601828438,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026680354311217006,
      "loss": 4.8665,
      "mean_token_accuracy": 0.226260307431221,
      "num_tokens": 114903338.0,
      "step": 66240
    },
    {
      "entropy": 5.78943362236023,
      "epoch": 5.154055631200156,
      "grad_norm": 1.53125,
      "learning_rate": 0.00026677584598287893,
      "loss": 4.8022,
      "mean_token_accuracy": 0.2276190534234047,
      "num_tokens": 114912195.0,
      "step": 66245
    },
    {
      "entropy": 5.866451263427734,
      "epoch": 5.1544446605718734,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026674814897837616,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2283419266343117,
      "num_tokens": 114920409.0,
      "step": 66250
    },
    {
      "entropy": 5.883718204498291,
      "epoch": 5.154833689943591,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026672045209908215,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2126406878232956,
      "num_tokens": 114929927.0,
      "step": 66255
    },
    {
      "entropy": 5.926264381408691,
      "epoch": 5.155222719315308,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026669275534541703,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22446409314870835,
      "num_tokens": 114938581.0,
      "step": 66260
    },
    {
      "entropy": 5.866660070419312,
      "epoch": 5.155611748687026,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00026666505871780115,
      "loss": 4.8768,
      "mean_token_accuracy": 0.2292337104678154,
      "num_tokens": 114946905.0,
      "step": 66265
    },
    {
      "entropy": 5.8527459621429445,
      "epoch": 5.156000778058743,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002666373622166548,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22785453498363495,
      "num_tokens": 114955846.0,
      "step": 66270
    },
    {
      "entropy": 5.921048259735107,
      "epoch": 5.156389807430461,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026660966584239815,
      "loss": 5.0243,
      "mean_token_accuracy": 0.20934616029262543,
      "num_tokens": 114965736.0,
      "step": 66275
    },
    {
      "entropy": 5.964421272277832,
      "epoch": 5.156778836802179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026658196959545145,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21834589242935182,
      "num_tokens": 114974026.0,
      "step": 66280
    },
    {
      "entropy": 5.93393874168396,
      "epoch": 5.157167866173896,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026655427347623494,
      "loss": 4.8853,
      "mean_token_accuracy": 0.22582802325487136,
      "num_tokens": 114982961.0,
      "step": 66285
    },
    {
      "entropy": 5.895879077911377,
      "epoch": 5.157556895545614,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002665265774851689,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21263739913702012,
      "num_tokens": 114991634.0,
      "step": 66290
    },
    {
      "entropy": 5.893559885025025,
      "epoch": 5.157945924917331,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026649888162267357,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2133638694882393,
      "num_tokens": 115000730.0,
      "step": 66295
    },
    {
      "entropy": 5.892694139480591,
      "epoch": 5.158334954289049,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002664711858891691,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21855069547891617,
      "num_tokens": 115009782.0,
      "step": 66300
    },
    {
      "entropy": 5.8914374828338625,
      "epoch": 5.158723983660766,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002664434902850759,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2207781195640564,
      "num_tokens": 115018650.0,
      "step": 66305
    },
    {
      "entropy": 5.858518028259278,
      "epoch": 5.159113013032484,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026641579481081405,
      "loss": 4.8212,
      "mean_token_accuracy": 0.23053707778453827,
      "num_tokens": 115026971.0,
      "step": 66310
    },
    {
      "entropy": 5.841846656799317,
      "epoch": 5.159502042404202,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026638809946680393,
      "loss": 4.883,
      "mean_token_accuracy": 0.22177408039569854,
      "num_tokens": 115034833.0,
      "step": 66315
    },
    {
      "entropy": 5.877322673797607,
      "epoch": 5.159891071775919,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026636040425346565,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21605104207992554,
      "num_tokens": 115043574.0,
      "step": 66320
    },
    {
      "entropy": 5.8099061965942385,
      "epoch": 5.160280101147636,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002663327091712195,
      "loss": 4.767,
      "mean_token_accuracy": 0.2256719022989273,
      "num_tokens": 115051970.0,
      "step": 66325
    },
    {
      "entropy": 5.876535367965698,
      "epoch": 5.160669130519354,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002663050142204856,
      "loss": 4.9642,
      "mean_token_accuracy": 0.21116408258676528,
      "num_tokens": 115060758.0,
      "step": 66330
    },
    {
      "entropy": 5.999392652511597,
      "epoch": 5.161058159891072,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00026627731940168445,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20914485305547714,
      "num_tokens": 115068632.0,
      "step": 66335
    },
    {
      "entropy": 6.010627126693725,
      "epoch": 5.161447189262789,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026624962471523595,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2086430236697197,
      "num_tokens": 115077608.0,
      "step": 66340
    },
    {
      "entropy": 5.917259645462036,
      "epoch": 5.161836218634507,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026622193016156054,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2151756078004837,
      "num_tokens": 115085900.0,
      "step": 66345
    },
    {
      "entropy": 5.941770267486572,
      "epoch": 5.162225248006225,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002661942357410784,
      "loss": 4.9682,
      "mean_token_accuracy": 0.21971530318260193,
      "num_tokens": 115094903.0,
      "step": 66350
    },
    {
      "entropy": 5.878421783447266,
      "epoch": 5.162614277377942,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002661665414542097,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2301975205540657,
      "num_tokens": 115103915.0,
      "step": 66355
    },
    {
      "entropy": 5.883051109313965,
      "epoch": 5.163003306749659,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026613884730137454,
      "loss": 4.8956,
      "mean_token_accuracy": 0.21756810694932938,
      "num_tokens": 115112322.0,
      "step": 66360
    },
    {
      "entropy": 5.949410486221313,
      "epoch": 5.163392336121377,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002661111532829933,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21123673766851425,
      "num_tokens": 115121620.0,
      "step": 66365
    },
    {
      "entropy": 5.911853551864624,
      "epoch": 5.163781365493095,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026608345939948625,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22496387958526612,
      "num_tokens": 115130116.0,
      "step": 66370
    },
    {
      "entropy": 5.887026071548462,
      "epoch": 5.164170394864812,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002660557656512734,
      "loss": 4.8918,
      "mean_token_accuracy": 0.21781501024961472,
      "num_tokens": 115138833.0,
      "step": 66375
    },
    {
      "entropy": 5.884005880355835,
      "epoch": 5.16455942423653,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026602807203877525,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22562648057937623,
      "num_tokens": 115147512.0,
      "step": 66380
    },
    {
      "entropy": 5.886314630508423,
      "epoch": 5.164948453608248,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026600037856241174,
      "loss": 4.874,
      "mean_token_accuracy": 0.22554809898138045,
      "num_tokens": 115155878.0,
      "step": 66385
    },
    {
      "entropy": 5.831168746948242,
      "epoch": 5.1653374829799645,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026597268522260317,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21856729239225386,
      "num_tokens": 115163879.0,
      "step": 66390
    },
    {
      "entropy": 5.9472376823425295,
      "epoch": 5.165726512351682,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026594499201976966,
      "loss": 5.0334,
      "mean_token_accuracy": 0.2057293325662613,
      "num_tokens": 115172665.0,
      "step": 66395
    },
    {
      "entropy": 5.89753155708313,
      "epoch": 5.1661155417234,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026591729895433157,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22132029831409455,
      "num_tokens": 115181449.0,
      "step": 66400
    },
    {
      "entropy": 5.922393655776977,
      "epoch": 5.166504571095118,
      "grad_norm": 1.375,
      "learning_rate": 0.00026588960602670896,
      "loss": 4.9135,
      "mean_token_accuracy": 0.21673181802034377,
      "num_tokens": 115190327.0,
      "step": 66405
    },
    {
      "entropy": 5.857445621490479,
      "epoch": 5.166893600466835,
      "grad_norm": 1.53125,
      "learning_rate": 0.00026586191323732213,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22496640980243682,
      "num_tokens": 115199336.0,
      "step": 66410
    },
    {
      "entropy": 5.926786088943482,
      "epoch": 5.167282629838553,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002658342205865912,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21693017482757568,
      "num_tokens": 115208768.0,
      "step": 66415
    },
    {
      "entropy": 5.8828271389007565,
      "epoch": 5.167671659210271,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002658065280749363,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21831827610731125,
      "num_tokens": 115217681.0,
      "step": 66420
    },
    {
      "entropy": 5.876974105834961,
      "epoch": 5.1680606885819875,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026577883570277786,
      "loss": 4.8674,
      "mean_token_accuracy": 0.21840519905090333,
      "num_tokens": 115227090.0,
      "step": 66425
    },
    {
      "entropy": 5.870783710479737,
      "epoch": 5.168449717953705,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026575114347053574,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21677094250917434,
      "num_tokens": 115236984.0,
      "step": 66430
    },
    {
      "entropy": 5.932261610031128,
      "epoch": 5.168838747325423,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002657234513786304,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2178018346428871,
      "num_tokens": 115245259.0,
      "step": 66435
    },
    {
      "entropy": 5.896965265274048,
      "epoch": 5.169227776697141,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002656957594274819,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21606735289096832,
      "num_tokens": 115254450.0,
      "step": 66440
    },
    {
      "entropy": 5.968612194061279,
      "epoch": 5.169616806068858,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00026566806761751043,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22056339234113692,
      "num_tokens": 115263312.0,
      "step": 66445
    },
    {
      "entropy": 5.918842506408692,
      "epoch": 5.170005835440576,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002656403759491362,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21694356799125672,
      "num_tokens": 115271546.0,
      "step": 66450
    },
    {
      "entropy": 5.946949291229248,
      "epoch": 5.170394864812294,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002656126844227794,
      "loss": 4.9589,
      "mean_token_accuracy": 0.21896385103464128,
      "num_tokens": 115280298.0,
      "step": 66455
    },
    {
      "entropy": 5.911055231094361,
      "epoch": 5.1707838941840105,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026558499303886015,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21614856719970704,
      "num_tokens": 115289295.0,
      "step": 66460
    },
    {
      "entropy": 5.905653572082519,
      "epoch": 5.171172923555728,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026555730179779864,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2210756003856659,
      "num_tokens": 115298374.0,
      "step": 66465
    },
    {
      "entropy": 5.937005519866943,
      "epoch": 5.171561952927446,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000265529610700015,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2228921353816986,
      "num_tokens": 115307152.0,
      "step": 66470
    },
    {
      "entropy": 5.937046718597412,
      "epoch": 5.171950982299164,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002655019197459295,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2136289045214653,
      "num_tokens": 115316226.0,
      "step": 66475
    },
    {
      "entropy": 5.9344195365905765,
      "epoch": 5.172340011670881,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026547422893596223,
      "loss": 4.9655,
      "mean_token_accuracy": 0.20836678445339202,
      "num_tokens": 115324848.0,
      "step": 66480
    },
    {
      "entropy": 5.880370235443115,
      "epoch": 5.172729041042599,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026544653827053336,
      "loss": 4.944,
      "mean_token_accuracy": 0.22749677151441575,
      "num_tokens": 115333719.0,
      "step": 66485
    },
    {
      "entropy": 5.929104137420654,
      "epoch": 5.173118070414317,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002654188477500631,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21282721310853958,
      "num_tokens": 115341676.0,
      "step": 66490
    },
    {
      "entropy": 5.8462766170501705,
      "epoch": 5.1735070997860335,
      "grad_norm": 1.375,
      "learning_rate": 0.0002653911573749715,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22088954746723174,
      "num_tokens": 115350487.0,
      "step": 66495
    },
    {
      "entropy": 5.926163864135742,
      "epoch": 5.173896129157751,
      "grad_norm": 1.578125,
      "learning_rate": 0.00026536346714567887,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21988250315189362,
      "num_tokens": 115359607.0,
      "step": 66500
    },
    {
      "entropy": 5.883273696899414,
      "epoch": 5.174285158529469,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026533577706260524,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21950936019420625,
      "num_tokens": 115368326.0,
      "step": 66505
    },
    {
      "entropy": 5.857379484176636,
      "epoch": 5.174674187901187,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002653080871261709,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22179691195487977,
      "num_tokens": 115376693.0,
      "step": 66510
    },
    {
      "entropy": 5.826855802536011,
      "epoch": 5.175063217272904,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002652803973367958,
      "loss": 4.8873,
      "mean_token_accuracy": 0.21160331666469573,
      "num_tokens": 115384708.0,
      "step": 66515
    },
    {
      "entropy": 5.903528499603271,
      "epoch": 5.175452246644622,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026525270769490024,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22562988847494125,
      "num_tokens": 115393190.0,
      "step": 66520
    },
    {
      "entropy": 5.912107229232788,
      "epoch": 5.175841276016339,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026522501820090433,
      "loss": 4.938,
      "mean_token_accuracy": 0.22097258269786835,
      "num_tokens": 115401768.0,
      "step": 66525
    },
    {
      "entropy": 5.824725246429443,
      "epoch": 5.1762303053880565,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026519732885522826,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22406095266342163,
      "num_tokens": 115410919.0,
      "step": 66530
    },
    {
      "entropy": 5.847902059555054,
      "epoch": 5.176619334759774,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026516963965829204,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2188459575176239,
      "num_tokens": 115419053.0,
      "step": 66535
    },
    {
      "entropy": 5.8656830310821535,
      "epoch": 5.177008364131492,
      "grad_norm": 1.515625,
      "learning_rate": 0.000265141950610516,
      "loss": 4.9362,
      "mean_token_accuracy": 0.221359221637249,
      "num_tokens": 115427456.0,
      "step": 66540
    },
    {
      "entropy": 5.921991968154908,
      "epoch": 5.17739739350321,
      "grad_norm": 1.5,
      "learning_rate": 0.0002651142617123201,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21659674644470214,
      "num_tokens": 115435619.0,
      "step": 66545
    },
    {
      "entropy": 5.927452802658081,
      "epoch": 5.177786422874927,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002650865729641246,
      "loss": 4.98,
      "mean_token_accuracy": 0.22005217671394348,
      "num_tokens": 115443958.0,
      "step": 66550
    },
    {
      "entropy": 5.869166469573974,
      "epoch": 5.178175452246645,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002650588843663495,
      "loss": 4.902,
      "mean_token_accuracy": 0.22233424484729766,
      "num_tokens": 115452676.0,
      "step": 66555
    },
    {
      "entropy": 5.883228015899658,
      "epoch": 5.178564481618362,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000265031195919415,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21622713506221772,
      "num_tokens": 115461292.0,
      "step": 66560
    },
    {
      "entropy": 5.858842658996582,
      "epoch": 5.1789535109900795,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026500350762374115,
      "loss": 4.8482,
      "mean_token_accuracy": 0.2284151315689087,
      "num_tokens": 115469739.0,
      "step": 66565
    },
    {
      "entropy": 5.766324996948242,
      "epoch": 5.179342540361797,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026497581947974836,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23931365758180617,
      "num_tokens": 115478569.0,
      "step": 66570
    },
    {
      "entropy": 5.870415735244751,
      "epoch": 5.179731569733515,
      "grad_norm": 1.578125,
      "learning_rate": 0.00026494813148785645,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2259065717458725,
      "num_tokens": 115486949.0,
      "step": 66575
    },
    {
      "entropy": 5.856617164611817,
      "epoch": 5.180120599105233,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00026492044364848566,
      "loss": 4.835,
      "mean_token_accuracy": 0.22525329738855362,
      "num_tokens": 115495007.0,
      "step": 66580
    },
    {
      "entropy": 5.874849081039429,
      "epoch": 5.18050962847695,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002648927559620561,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2272611603140831,
      "num_tokens": 115503053.0,
      "step": 66585
    },
    {
      "entropy": 5.872139406204224,
      "epoch": 5.180898657848668,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002648650684289879,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2183742806315422,
      "num_tokens": 115511950.0,
      "step": 66590
    },
    {
      "entropy": 5.864149522781372,
      "epoch": 5.181287687220385,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002648373810497011,
      "loss": 4.8544,
      "mean_token_accuracy": 0.22028339952230452,
      "num_tokens": 115520272.0,
      "step": 66595
    },
    {
      "entropy": 5.8790405750274655,
      "epoch": 5.1816767165921025,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002648096938246159,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22629420906305314,
      "num_tokens": 115529000.0,
      "step": 66600
    },
    {
      "entropy": 5.94224967956543,
      "epoch": 5.18206574596382,
      "grad_norm": 1.5,
      "learning_rate": 0.0002647820067541524,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21430711299180985,
      "num_tokens": 115537783.0,
      "step": 66605
    },
    {
      "entropy": 5.911783218383789,
      "epoch": 5.182454775335538,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026475431983873066,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21028370410203934,
      "num_tokens": 115546689.0,
      "step": 66610
    },
    {
      "entropy": 5.968821048736572,
      "epoch": 5.182843804707256,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002647266330787708,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21254965513944626,
      "num_tokens": 115554965.0,
      "step": 66615
    },
    {
      "entropy": 5.898732757568359,
      "epoch": 5.183232834078973,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026469894647469295,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21675666570663452,
      "num_tokens": 115563402.0,
      "step": 66620
    },
    {
      "entropy": 5.855556392669678,
      "epoch": 5.183621863450691,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002646712600269172,
      "loss": 4.8972,
      "mean_token_accuracy": 0.21870721131563187,
      "num_tokens": 115572950.0,
      "step": 66625
    },
    {
      "entropy": 5.860622787475586,
      "epoch": 5.184010892822408,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002646435737358635,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22866910845041274,
      "num_tokens": 115581927.0,
      "step": 66630
    },
    {
      "entropy": 5.914406490325928,
      "epoch": 5.1843999221941255,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002646158876019522,
      "loss": 4.904,
      "mean_token_accuracy": 0.21869576275348662,
      "num_tokens": 115590334.0,
      "step": 66635
    },
    {
      "entropy": 5.837580966949463,
      "epoch": 5.184788951565843,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002645882016256033,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2292750209569931,
      "num_tokens": 115599863.0,
      "step": 66640
    },
    {
      "entropy": 5.869169425964356,
      "epoch": 5.185177980937561,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002645605158072368,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2125752553343773,
      "num_tokens": 115608303.0,
      "step": 66645
    },
    {
      "entropy": 5.84347128868103,
      "epoch": 5.185567010309279,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000264532830147273,
      "loss": 4.8538,
      "mean_token_accuracy": 0.2244350492954254,
      "num_tokens": 115616677.0,
      "step": 66650
    },
    {
      "entropy": 5.896841287612915,
      "epoch": 5.185956039680996,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026450514464613177,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21330585777759553,
      "num_tokens": 115625522.0,
      "step": 66655
    },
    {
      "entropy": 5.900094318389892,
      "epoch": 5.186345069052713,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002644774593042332,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21754818111658097,
      "num_tokens": 115633826.0,
      "step": 66660
    },
    {
      "entropy": 5.926422595977783,
      "epoch": 5.186734098424431,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002644497741219975,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20958569496870041,
      "num_tokens": 115642862.0,
      "step": 66665
    },
    {
      "entropy": 5.897913551330566,
      "epoch": 5.1871231277961485,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002644220890998447,
      "loss": 4.953,
      "mean_token_accuracy": 0.2159480094909668,
      "num_tokens": 115651364.0,
      "step": 66670
    },
    {
      "entropy": 5.953419208526611,
      "epoch": 5.187512157167866,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026439440423819483,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21189212501049043,
      "num_tokens": 115659975.0,
      "step": 66675
    },
    {
      "entropy": 5.867889785766602,
      "epoch": 5.187901186539584,
      "grad_norm": 1.484375,
      "learning_rate": 0.000264366719537468,
      "loss": 4.965,
      "mean_token_accuracy": 0.21710327118635178,
      "num_tokens": 115669058.0,
      "step": 66680
    },
    {
      "entropy": 5.8232215404510494,
      "epoch": 5.188290215911302,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002643390349980843,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22096076905727385,
      "num_tokens": 115677443.0,
      "step": 66685
    },
    {
      "entropy": 5.881569862365723,
      "epoch": 5.188679245283019,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002643113506204637,
      "loss": 4.8512,
      "mean_token_accuracy": 0.22627180367708205,
      "num_tokens": 115685441.0,
      "step": 66690
    },
    {
      "entropy": 5.914648389816284,
      "epoch": 5.189068274654736,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002642836664050263,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2170115143060684,
      "num_tokens": 115695037.0,
      "step": 66695
    },
    {
      "entropy": 5.855237007141113,
      "epoch": 5.189457304026454,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026425598235219235,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22932253181934356,
      "num_tokens": 115704237.0,
      "step": 66700
    },
    {
      "entropy": 5.93607530593872,
      "epoch": 5.1898463333981715,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026422829846238174,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21257587522268295,
      "num_tokens": 115712608.0,
      "step": 66705
    },
    {
      "entropy": 5.840132331848144,
      "epoch": 5.190235362769889,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026420061473601445,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22687792927026748,
      "num_tokens": 115721542.0,
      "step": 66710
    },
    {
      "entropy": 5.844367837905883,
      "epoch": 5.190624392141607,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026417293117351077,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21926034688949586,
      "num_tokens": 115729982.0,
      "step": 66715
    },
    {
      "entropy": 5.8385780334472654,
      "epoch": 5.191013421513325,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026414524777529055,
      "loss": 4.8374,
      "mean_token_accuracy": 0.23138161152601242,
      "num_tokens": 115739126.0,
      "step": 66720
    },
    {
      "entropy": 5.851501893997193,
      "epoch": 5.191402450885041,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002641175645417739,
      "loss": 4.8825,
      "mean_token_accuracy": 0.21858133673667907,
      "num_tokens": 115747588.0,
      "step": 66725
    },
    {
      "entropy": 5.938333511352539,
      "epoch": 5.191791480256759,
      "grad_norm": 1.578125,
      "learning_rate": 0.000264089881473381,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21610334813594817,
      "num_tokens": 115756040.0,
      "step": 66730
    },
    {
      "entropy": 5.903454923629761,
      "epoch": 5.192180509628477,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002640621985705318,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21261221766471863,
      "num_tokens": 115764888.0,
      "step": 66735
    },
    {
      "entropy": 5.953672075271607,
      "epoch": 5.1925695390001945,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026403451583364616,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21494019478559495,
      "num_tokens": 115774036.0,
      "step": 66740
    },
    {
      "entropy": 5.917304849624633,
      "epoch": 5.192958568371912,
      "grad_norm": 1.5,
      "learning_rate": 0.0002640068332631444,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22461487650871276,
      "num_tokens": 115782341.0,
      "step": 66745
    },
    {
      "entropy": 5.881608963012695,
      "epoch": 5.19334759774363,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002639791508594464,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22036605030298234,
      "num_tokens": 115791908.0,
      "step": 66750
    },
    {
      "entropy": 5.9237569808959964,
      "epoch": 5.193736627115348,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026395146862297225,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21254721134901047,
      "num_tokens": 115801182.0,
      "step": 66755
    },
    {
      "entropy": 5.892680025100708,
      "epoch": 5.194125656487064,
      "grad_norm": 1.40625,
      "learning_rate": 0.000263923786554142,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22193896025419235,
      "num_tokens": 115810210.0,
      "step": 66760
    },
    {
      "entropy": 5.89268684387207,
      "epoch": 5.194514685858782,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002638961046533756,
      "loss": 4.8888,
      "mean_token_accuracy": 0.21785345524549485,
      "num_tokens": 115818987.0,
      "step": 66765
    },
    {
      "entropy": 5.905697059631348,
      "epoch": 5.1949037152305,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002638684229210932,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22455980777740478,
      "num_tokens": 115827164.0,
      "step": 66770
    },
    {
      "entropy": 5.913054800033569,
      "epoch": 5.1952927446022175,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026384074135771474,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2152501329779625,
      "num_tokens": 115837601.0,
      "step": 66775
    },
    {
      "entropy": 5.830989599227905,
      "epoch": 5.195681773973935,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002638130599636603,
      "loss": 4.8615,
      "mean_token_accuracy": 0.22214706838130951,
      "num_tokens": 115845636.0,
      "step": 66780
    },
    {
      "entropy": 5.888315629959107,
      "epoch": 5.196070803345653,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026378537873934983,
      "loss": 4.8773,
      "mean_token_accuracy": 0.21925183683633803,
      "num_tokens": 115854380.0,
      "step": 66785
    },
    {
      "entropy": 5.918946838378906,
      "epoch": 5.1964598327173706,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026375769768520344,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22070348113775254,
      "num_tokens": 115863045.0,
      "step": 66790
    },
    {
      "entropy": 5.896366977691651,
      "epoch": 5.196848862089087,
      "grad_norm": 1.46875,
      "learning_rate": 0.000263730016801641,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21918170601129533,
      "num_tokens": 115871066.0,
      "step": 66795
    },
    {
      "entropy": 5.845680236816406,
      "epoch": 5.197237891460805,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002637023360890827,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2230381414294243,
      "num_tokens": 115880401.0,
      "step": 66800
    },
    {
      "entropy": 5.7885599613189695,
      "epoch": 5.197626920832523,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026367465554794843,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23228487968444825,
      "num_tokens": 115888630.0,
      "step": 66805
    },
    {
      "entropy": 5.911736726760864,
      "epoch": 5.1980159502042405,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026364697517865834,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22549426406621934,
      "num_tokens": 115897145.0,
      "step": 66810
    },
    {
      "entropy": 5.900034189224243,
      "epoch": 5.198404979575958,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002636192949816322,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22327499389648436,
      "num_tokens": 115905895.0,
      "step": 66815
    },
    {
      "entropy": 5.79327392578125,
      "epoch": 5.198794008947676,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002635916149572902,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22235584110021592,
      "num_tokens": 115914977.0,
      "step": 66820
    },
    {
      "entropy": 5.855564928054809,
      "epoch": 5.1991830383193935,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026356393510605224,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21617656648159028,
      "num_tokens": 115923630.0,
      "step": 66825
    },
    {
      "entropy": 5.853539037704468,
      "epoch": 5.19957206769111,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002635362554283383,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22370741069316863,
      "num_tokens": 115932828.0,
      "step": 66830
    },
    {
      "entropy": 5.947035884857177,
      "epoch": 5.199961097062828,
      "grad_norm": 1.5,
      "learning_rate": 0.0002635085759245686,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21474708914756774,
      "num_tokens": 115942328.0,
      "step": 66835
    },
    {
      "entropy": 5.937947607040405,
      "epoch": 5.200350126434546,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002634808965951629,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21636297255754472,
      "num_tokens": 115951062.0,
      "step": 66840
    },
    {
      "entropy": 5.854681825637817,
      "epoch": 5.2007391558062634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002634532174405413,
      "loss": 4.8403,
      "mean_token_accuracy": 0.22606565654277802,
      "num_tokens": 115959716.0,
      "step": 66845
    },
    {
      "entropy": 5.820920038223266,
      "epoch": 5.201128185177981,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026342553846112364,
      "loss": 4.8354,
      "mean_token_accuracy": 0.22617035806179048,
      "num_tokens": 115967623.0,
      "step": 66850
    },
    {
      "entropy": 5.865072631835938,
      "epoch": 5.201517214549699,
      "grad_norm": 1.5,
      "learning_rate": 0.0002633978596573301,
      "loss": 4.8741,
      "mean_token_accuracy": 0.2224920228123665,
      "num_tokens": 115976944.0,
      "step": 66855
    },
    {
      "entropy": 5.826416540145874,
      "epoch": 5.201906243921416,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026337018102958056,
      "loss": 4.8718,
      "mean_token_accuracy": 0.21866938918828965,
      "num_tokens": 115986432.0,
      "step": 66860
    },
    {
      "entropy": 5.862338829040527,
      "epoch": 5.202295273293133,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000263342502578295,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2172921985387802,
      "num_tokens": 115995386.0,
      "step": 66865
    },
    {
      "entropy": 5.8694549083709715,
      "epoch": 5.202684302664851,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026331482430389343,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22362824231386186,
      "num_tokens": 116004048.0,
      "step": 66870
    },
    {
      "entropy": 5.880117607116699,
      "epoch": 5.203073332036569,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026328714620679576,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2106212168931961,
      "num_tokens": 116013948.0,
      "step": 66875
    },
    {
      "entropy": 5.914364242553711,
      "epoch": 5.203462361408286,
      "grad_norm": 1.515625,
      "learning_rate": 0.00026325946828742203,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22134124338626862,
      "num_tokens": 116023017.0,
      "step": 66880
    },
    {
      "entropy": 5.911861562728882,
      "epoch": 5.203851390780004,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00026323179054619215,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21681919693946838,
      "num_tokens": 116031912.0,
      "step": 66885
    },
    {
      "entropy": 5.897346067428589,
      "epoch": 5.204240420151722,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002632041129835262,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22524741888046265,
      "num_tokens": 116040294.0,
      "step": 66890
    },
    {
      "entropy": 5.951008272171021,
      "epoch": 5.204629449523439,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026317643559984396,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2123206987977028,
      "num_tokens": 116048222.0,
      "step": 66895
    },
    {
      "entropy": 5.852381324768066,
      "epoch": 5.205018478895156,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002631487583955655,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21047160923480987,
      "num_tokens": 116057595.0,
      "step": 66900
    },
    {
      "entropy": 5.794004631042481,
      "epoch": 5.205407508266874,
      "grad_norm": 1.5,
      "learning_rate": 0.00026312108137111086,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23198727071285247,
      "num_tokens": 116065229.0,
      "step": 66905
    },
    {
      "entropy": 5.836598348617554,
      "epoch": 5.205796537638592,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002630934045268998,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22179910987615586,
      "num_tokens": 116073620.0,
      "step": 66910
    },
    {
      "entropy": 5.8712241649627686,
      "epoch": 5.206185567010309,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002630657278633525,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21909717619419097,
      "num_tokens": 116082019.0,
      "step": 66915
    },
    {
      "entropy": 5.870947408676147,
      "epoch": 5.206574596382027,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002630380513808887,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2148544117808342,
      "num_tokens": 116090628.0,
      "step": 66920
    },
    {
      "entropy": 5.907679414749145,
      "epoch": 5.206963625753745,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002630103750799284,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21176630407571792,
      "num_tokens": 116098954.0,
      "step": 66925
    },
    {
      "entropy": 5.903951358795166,
      "epoch": 5.207352655125462,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00026298269896089164,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21052053570747375,
      "num_tokens": 116107844.0,
      "step": 66930
    },
    {
      "entropy": 5.889689302444458,
      "epoch": 5.207741684497179,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026295502302419826,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2199919953942299,
      "num_tokens": 116117130.0,
      "step": 66935
    },
    {
      "entropy": 5.890400075912476,
      "epoch": 5.208130713868897,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002629273472702683,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2140430450439453,
      "num_tokens": 116125760.0,
      "step": 66940
    },
    {
      "entropy": 5.875790405273437,
      "epoch": 5.208519743240615,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002628996716995215,
      "loss": 4.8625,
      "mean_token_accuracy": 0.21772147715091705,
      "num_tokens": 116134324.0,
      "step": 66945
    },
    {
      "entropy": 5.875073194503784,
      "epoch": 5.208908772612332,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000262871996312378,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2169422388076782,
      "num_tokens": 116142570.0,
      "step": 66950
    },
    {
      "entropy": 5.954105949401855,
      "epoch": 5.20929780198405,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026284432110925765,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21435638517141342,
      "num_tokens": 116150831.0,
      "step": 66955
    },
    {
      "entropy": 5.932734060287475,
      "epoch": 5.209686831355767,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026281664609058037,
      "loss": 4.9773,
      "mean_token_accuracy": 0.20973667055368422,
      "num_tokens": 116159626.0,
      "step": 66960
    },
    {
      "entropy": 5.897079610824585,
      "epoch": 5.210075860727485,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002627889712567661,
      "loss": 4.8331,
      "mean_token_accuracy": 0.22387586534023285,
      "num_tokens": 116168780.0,
      "step": 66965
    },
    {
      "entropy": 5.89501919746399,
      "epoch": 5.210464890099202,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026276129660823475,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23094754070043563,
      "num_tokens": 116177294.0,
      "step": 66970
    },
    {
      "entropy": 5.8401710987091064,
      "epoch": 5.21085391947092,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026273362214540623,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2284397602081299,
      "num_tokens": 116185844.0,
      "step": 66975
    },
    {
      "entropy": 5.840371894836426,
      "epoch": 5.211242948842638,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002627059478687005,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21286423057317733,
      "num_tokens": 116194489.0,
      "step": 66980
    },
    {
      "entropy": 5.828110218048096,
      "epoch": 5.211631978214355,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002626782737785374,
      "loss": 4.917,
      "mean_token_accuracy": 0.2179950162768364,
      "num_tokens": 116203022.0,
      "step": 66985
    },
    {
      "entropy": 5.84049620628357,
      "epoch": 5.212021007586073,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026265059987533697,
      "loss": 4.8701,
      "mean_token_accuracy": 0.21906978785991668,
      "num_tokens": 116210838.0,
      "step": 66990
    },
    {
      "entropy": 5.8754777908325195,
      "epoch": 5.21241003695779,
      "grad_norm": 1.515625,
      "learning_rate": 0.00026262292615951893,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2227970004081726,
      "num_tokens": 116219356.0,
      "step": 66995
    },
    {
      "entropy": 5.809766006469727,
      "epoch": 5.212799066329508,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002625952526315034,
      "loss": 4.911,
      "mean_token_accuracy": 0.2198785051703453,
      "num_tokens": 116228355.0,
      "step": 67000
    },
    {
      "entropy": 5.892364406585694,
      "epoch": 5.213188095701225,
      "grad_norm": 1.5,
      "learning_rate": 0.0002625675792917101,
      "loss": 4.9021,
      "mean_token_accuracy": 0.21953105032444,
      "num_tokens": 116237070.0,
      "step": 67005
    },
    {
      "entropy": 5.896638584136963,
      "epoch": 5.213577125072943,
      "grad_norm": 1.421875,
      "learning_rate": 0.000262539906140559,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22197256237268448,
      "num_tokens": 116245912.0,
      "step": 67010
    },
    {
      "entropy": 5.912347841262817,
      "epoch": 5.213966154444661,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026251223317847003,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21724906712770461,
      "num_tokens": 116254315.0,
      "step": 67015
    },
    {
      "entropy": 5.931737995147705,
      "epoch": 5.214355183816378,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000262484560405863,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21388865262269974,
      "num_tokens": 116263053.0,
      "step": 67020
    },
    {
      "entropy": 5.906467437744141,
      "epoch": 5.214744213188096,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026245688782315787,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21803633421659468,
      "num_tokens": 116271482.0,
      "step": 67025
    },
    {
      "entropy": 5.948765277862549,
      "epoch": 5.215133242559813,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00026242921543077435,
      "loss": 4.927,
      "mean_token_accuracy": 0.2193041056394577,
      "num_tokens": 116280297.0,
      "step": 67030
    },
    {
      "entropy": 5.842594623565674,
      "epoch": 5.215522271931531,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002624015432291326,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22340085804462434,
      "num_tokens": 116287860.0,
      "step": 67035
    },
    {
      "entropy": 5.8514588356018065,
      "epoch": 5.215911301303248,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002623738712186525,
      "loss": 4.8747,
      "mean_token_accuracy": 0.21723172068595886,
      "num_tokens": 116295844.0,
      "step": 67040
    },
    {
      "entropy": 5.872314548492431,
      "epoch": 5.216300330674966,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026234619939975365,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22092931866645812,
      "num_tokens": 116304746.0,
      "step": 67045
    },
    {
      "entropy": 5.893154239654541,
      "epoch": 5.216689360046684,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00026231852777285606,
      "loss": 4.8154,
      "mean_token_accuracy": 0.22430939227342606,
      "num_tokens": 116313428.0,
      "step": 67050
    },
    {
      "entropy": 5.9300566673278805,
      "epoch": 5.217078389418401,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026229085633837975,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22198998928070068,
      "num_tokens": 116321776.0,
      "step": 67055
    },
    {
      "entropy": 5.915703773498535,
      "epoch": 5.217467418790118,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002622631850967444,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2206052213907242,
      "num_tokens": 116330804.0,
      "step": 67060
    },
    {
      "entropy": 5.876027011871338,
      "epoch": 5.217856448161836,
      "grad_norm": 1.53125,
      "learning_rate": 0.00026223551404837,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2091422438621521,
      "num_tokens": 116339948.0,
      "step": 67065
    },
    {
      "entropy": 5.891392421722412,
      "epoch": 5.218245477533554,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026220784319367626,
      "loss": 4.8976,
      "mean_token_accuracy": 0.21987392157316207,
      "num_tokens": 116348996.0,
      "step": 67070
    },
    {
      "entropy": 5.910856103897094,
      "epoch": 5.218634506905271,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002621801725330832,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22370724231004716,
      "num_tokens": 116357909.0,
      "step": 67075
    },
    {
      "entropy": 5.837147283554077,
      "epoch": 5.219023536276989,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002621525020670106,
      "loss": 4.8702,
      "mean_token_accuracy": 0.21973688006401063,
      "num_tokens": 116367054.0,
      "step": 67080
    },
    {
      "entropy": 5.933539295196534,
      "epoch": 5.219412565648707,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002621248317958784,
      "loss": 5.0017,
      "mean_token_accuracy": 0.20808265656232833,
      "num_tokens": 116375597.0,
      "step": 67085
    },
    {
      "entropy": 5.851144790649414,
      "epoch": 5.219801595020424,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026209716172010636,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22649372816085817,
      "num_tokens": 116383800.0,
      "step": 67090
    },
    {
      "entropy": 5.852349948883057,
      "epoch": 5.220190624392141,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002620694918401142,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2199783056974411,
      "num_tokens": 116394190.0,
      "step": 67095
    },
    {
      "entropy": 5.849413204193115,
      "epoch": 5.220579653763859,
      "grad_norm": 1.421875,
      "learning_rate": 0.000262041822156322,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22887897193431855,
      "num_tokens": 116402733.0,
      "step": 67100
    },
    {
      "entropy": 5.877736568450928,
      "epoch": 5.220968683135577,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002620141526691496,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21220042407512665,
      "num_tokens": 116411788.0,
      "step": 67105
    },
    {
      "entropy": 5.844420862197876,
      "epoch": 5.221357712507294,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026198648337901674,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23267367333173752,
      "num_tokens": 116420040.0,
      "step": 67110
    },
    {
      "entropy": 5.82170615196228,
      "epoch": 5.221746741879012,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026195881428634334,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22638074159622193,
      "num_tokens": 116428181.0,
      "step": 67115
    },
    {
      "entropy": 5.8831531524658205,
      "epoch": 5.22213577125073,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026193114539154903,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21132851839065553,
      "num_tokens": 116436911.0,
      "step": 67120
    },
    {
      "entropy": 5.933945751190185,
      "epoch": 5.222524800622447,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002619034766950539,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2095859929919243,
      "num_tokens": 116446234.0,
      "step": 67125
    },
    {
      "entropy": 5.917168045043946,
      "epoch": 5.222913829994164,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002618758081972776,
      "loss": 4.9301,
      "mean_token_accuracy": 0.2152769759297371,
      "num_tokens": 116454848.0,
      "step": 67130
    },
    {
      "entropy": 5.969388437271118,
      "epoch": 5.223302859365882,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002618481398986401,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22084282487630844,
      "num_tokens": 116464852.0,
      "step": 67135
    },
    {
      "entropy": 5.883950805664062,
      "epoch": 5.2236918887376,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026182047179956107,
      "loss": 4.8815,
      "mean_token_accuracy": 0.223748841881752,
      "num_tokens": 116473043.0,
      "step": 67140
    },
    {
      "entropy": 5.874982643127441,
      "epoch": 5.224080918109317,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026179280390046047,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22069993615150452,
      "num_tokens": 116481245.0,
      "step": 67145
    },
    {
      "entropy": 5.900635242462158,
      "epoch": 5.224469947481035,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026176513620175795,
      "loss": 4.953,
      "mean_token_accuracy": 0.21784802228212358,
      "num_tokens": 116489977.0,
      "step": 67150
    },
    {
      "entropy": 5.832939863204956,
      "epoch": 5.224858976852753,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002617374687038735,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22005811035633088,
      "num_tokens": 116499289.0,
      "step": 67155
    },
    {
      "entropy": 5.851674652099609,
      "epoch": 5.22524800622447,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026170980140722677,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2202461764216423,
      "num_tokens": 116508331.0,
      "step": 67160
    },
    {
      "entropy": 5.9518561363220215,
      "epoch": 5.225637035596187,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026168213431223777,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22675934582948684,
      "num_tokens": 116516305.0,
      "step": 67165
    },
    {
      "entropy": 5.905051279067993,
      "epoch": 5.226026064967905,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026165446741932606,
      "loss": 4.944,
      "mean_token_accuracy": 0.2239120990037918,
      "num_tokens": 116524642.0,
      "step": 67170
    },
    {
      "entropy": 5.850819635391235,
      "epoch": 5.226415094339623,
      "grad_norm": 1.53125,
      "learning_rate": 0.00026162680072891166,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22330676168203353,
      "num_tokens": 116533713.0,
      "step": 67175
    },
    {
      "entropy": 5.872627782821655,
      "epoch": 5.22680412371134,
      "grad_norm": 1.515625,
      "learning_rate": 0.00026159913424141426,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22323849499225618,
      "num_tokens": 116542350.0,
      "step": 67180
    },
    {
      "entropy": 5.8786962032318115,
      "epoch": 5.227193153083058,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002615714679572536,
      "loss": 4.923,
      "mean_token_accuracy": 0.2190012440085411,
      "num_tokens": 116550788.0,
      "step": 67185
    },
    {
      "entropy": 5.890914106369019,
      "epoch": 5.227582182454776,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026154380187684963,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21663843244314193,
      "num_tokens": 116559984.0,
      "step": 67190
    },
    {
      "entropy": 5.8632618427276615,
      "epoch": 5.2279712118264925,
      "grad_norm": 1.4375,
      "learning_rate": 0.000261516136000622,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21169033795595169,
      "num_tokens": 116568738.0,
      "step": 67195
    },
    {
      "entropy": 5.802139520645142,
      "epoch": 5.22836024119821,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002614884703289906,
      "loss": 4.8191,
      "mean_token_accuracy": 0.2344991758465767,
      "num_tokens": 116576722.0,
      "step": 67200
    },
    {
      "entropy": 5.842132329940796,
      "epoch": 5.228749270569928,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026146080486237513,
      "loss": 4.933,
      "mean_token_accuracy": 0.22037400156259537,
      "num_tokens": 116585323.0,
      "step": 67205
    },
    {
      "entropy": 5.861917924880982,
      "epoch": 5.229138299941646,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026143313960119534,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2211926355957985,
      "num_tokens": 116593215.0,
      "step": 67210
    },
    {
      "entropy": 5.867314624786377,
      "epoch": 5.229527329313363,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002614054745458711,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2223244935274124,
      "num_tokens": 116601599.0,
      "step": 67215
    },
    {
      "entropy": 5.897385931015014,
      "epoch": 5.229916358685081,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00026137780969682215,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2315788447856903,
      "num_tokens": 116610309.0,
      "step": 67220
    },
    {
      "entropy": 5.9587305068969725,
      "epoch": 5.230305388056799,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002613501450544683,
      "loss": 5.0265,
      "mean_token_accuracy": 0.213980470597744,
      "num_tokens": 116618555.0,
      "step": 67225
    },
    {
      "entropy": 5.886282539367675,
      "epoch": 5.2306944174285155,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002613224806192291,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22403783202171326,
      "num_tokens": 116627639.0,
      "step": 67230
    },
    {
      "entropy": 5.894217538833618,
      "epoch": 5.231083446800233,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002612948163915247,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2140920177102089,
      "num_tokens": 116637077.0,
      "step": 67235
    },
    {
      "entropy": 5.874657821655274,
      "epoch": 5.231472476171951,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002612671523717745,
      "loss": 4.884,
      "mean_token_accuracy": 0.22267655283212662,
      "num_tokens": 116645570.0,
      "step": 67240
    },
    {
      "entropy": 5.832493877410888,
      "epoch": 5.231861505543669,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002612394885603985,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2233529731631279,
      "num_tokens": 116654140.0,
      "step": 67245
    },
    {
      "entropy": 5.831371307373047,
      "epoch": 5.232250534915386,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002612118249578163,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22032055258750916,
      "num_tokens": 116662740.0,
      "step": 67250
    },
    {
      "entropy": 5.897946834564209,
      "epoch": 5.232639564287104,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002611841615644477,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21394146531820296,
      "num_tokens": 116670813.0,
      "step": 67255
    },
    {
      "entropy": 5.904433727264404,
      "epoch": 5.233028593658822,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026115649838071253,
      "loss": 4.9487,
      "mean_token_accuracy": 0.20981094986200333,
      "num_tokens": 116679897.0,
      "step": 67260
    },
    {
      "entropy": 5.891775703430175,
      "epoch": 5.2334176230305385,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002611288354070304,
      "loss": 4.9535,
      "mean_token_accuracy": 0.2096189945936203,
      "num_tokens": 116688910.0,
      "step": 67265
    },
    {
      "entropy": 5.944534492492676,
      "epoch": 5.233806652402256,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002611011726438212,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21259674280881882,
      "num_tokens": 116697114.0,
      "step": 67270
    },
    {
      "entropy": 5.80322847366333,
      "epoch": 5.234195681773974,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026107351009150453,
      "loss": 4.8221,
      "mean_token_accuracy": 0.22969842255115508,
      "num_tokens": 116705623.0,
      "step": 67275
    },
    {
      "entropy": 5.898817157745361,
      "epoch": 5.234584711145692,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002610458477505002,
      "loss": 4.8443,
      "mean_token_accuracy": 0.22299365252256392,
      "num_tokens": 116713954.0,
      "step": 67280
    },
    {
      "entropy": 5.884688758850098,
      "epoch": 5.234973740517409,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002610181856212279,
      "loss": 4.908,
      "mean_token_accuracy": 0.21802364140748978,
      "num_tokens": 116722378.0,
      "step": 67285
    },
    {
      "entropy": 5.862059593200684,
      "epoch": 5.235362769889127,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002609905237041074,
      "loss": 4.935,
      "mean_token_accuracy": 0.22016392946243285,
      "num_tokens": 116731339.0,
      "step": 67290
    },
    {
      "entropy": 5.791627693176269,
      "epoch": 5.235751799260844,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002609628619995582,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22428351193666457,
      "num_tokens": 116739949.0,
      "step": 67295
    },
    {
      "entropy": 5.8831977367401125,
      "epoch": 5.2361408286325615,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026093520050800046,
      "loss": 4.913,
      "mean_token_accuracy": 0.2229623943567276,
      "num_tokens": 116747810.0,
      "step": 67300
    },
    {
      "entropy": 5.87790150642395,
      "epoch": 5.236529858004279,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002609075392298536,
      "loss": 4.8343,
      "mean_token_accuracy": 0.21886638998985292,
      "num_tokens": 116756523.0,
      "step": 67305
    },
    {
      "entropy": 5.8684978008270265,
      "epoch": 5.236918887375997,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002608798781655374,
      "loss": 4.9062,
      "mean_token_accuracy": 0.21671493202447892,
      "num_tokens": 116765328.0,
      "step": 67310
    },
    {
      "entropy": 5.884585237503051,
      "epoch": 5.237307916747715,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002608522173154716,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2154143288731575,
      "num_tokens": 116773752.0,
      "step": 67315
    },
    {
      "entropy": 5.888968658447266,
      "epoch": 5.237696946119432,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002608245566800759,
      "loss": 4.9347,
      "mean_token_accuracy": 0.2158031314611435,
      "num_tokens": 116782130.0,
      "step": 67320
    },
    {
      "entropy": 5.897051048278809,
      "epoch": 5.23808597549115,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026079689625976995,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22446153312921524,
      "num_tokens": 116790965.0,
      "step": 67325
    },
    {
      "entropy": 5.84403920173645,
      "epoch": 5.238475004862867,
      "grad_norm": 1.484375,
      "learning_rate": 0.00026076923605497347,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22390540689229965,
      "num_tokens": 116800567.0,
      "step": 67330
    },
    {
      "entropy": 5.921960258483887,
      "epoch": 5.2388640342345845,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026074157606610626,
      "loss": 4.917,
      "mean_token_accuracy": 0.2217146024107933,
      "num_tokens": 116809819.0,
      "step": 67335
    },
    {
      "entropy": 5.940748167037964,
      "epoch": 5.239253063606302,
      "grad_norm": 1.5,
      "learning_rate": 0.00026071391629358786,
      "loss": 4.9265,
      "mean_token_accuracy": 0.21651136577129365,
      "num_tokens": 116818498.0,
      "step": 67340
    },
    {
      "entropy": 5.884336471557617,
      "epoch": 5.23964209297802,
      "grad_norm": 1.4375,
      "learning_rate": 0.00026068625673783807,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21122610419988633,
      "num_tokens": 116827013.0,
      "step": 67345
    },
    {
      "entropy": 5.896107769012451,
      "epoch": 5.240031122349738,
      "grad_norm": 1.46875,
      "learning_rate": 0.00026065859739927657,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20991375297307968,
      "num_tokens": 116835935.0,
      "step": 67350
    },
    {
      "entropy": 5.864320850372314,
      "epoch": 5.240420151721455,
      "grad_norm": 1.390625,
      "learning_rate": 0.000260630938278323,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2287273585796356,
      "num_tokens": 116844040.0,
      "step": 67355
    },
    {
      "entropy": 5.939917373657226,
      "epoch": 5.240809181093173,
      "grad_norm": 1.40625,
      "learning_rate": 0.00026060327937539704,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21367578506469725,
      "num_tokens": 116852506.0,
      "step": 67360
    },
    {
      "entropy": 5.896703481674194,
      "epoch": 5.24119821046489,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00026057562069091845,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2179089218378067,
      "num_tokens": 116861528.0,
      "step": 67365
    },
    {
      "entropy": 5.940378904342651,
      "epoch": 5.2415872398366075,
      "grad_norm": 1.609375,
      "learning_rate": 0.00026054796222530684,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21729314029216767,
      "num_tokens": 116869758.0,
      "step": 67370
    },
    {
      "entropy": 5.94890284538269,
      "epoch": 5.241976269208325,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002605203039789819,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22087464183568956,
      "num_tokens": 116878239.0,
      "step": 67375
    },
    {
      "entropy": 5.885954809188843,
      "epoch": 5.242365298580043,
      "grad_norm": 1.5,
      "learning_rate": 0.0002604926459523633,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2185305029153824,
      "num_tokens": 116886998.0,
      "step": 67380
    },
    {
      "entropy": 5.797305774688721,
      "epoch": 5.2427543279517606,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026046498814587076,
      "loss": 4.8224,
      "mean_token_accuracy": 0.22675151079893113,
      "num_tokens": 116895535.0,
      "step": 67385
    },
    {
      "entropy": 5.923039674758911,
      "epoch": 5.243143357323478,
      "grad_norm": 1.5,
      "learning_rate": 0.00026043733055992386,
      "loss": 5.0094,
      "mean_token_accuracy": 0.20485444068908693,
      "num_tokens": 116904096.0,
      "step": 67390
    },
    {
      "entropy": 5.9373468399047855,
      "epoch": 5.243532386695195,
      "grad_norm": 1.421875,
      "learning_rate": 0.00026040967319494225,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22203624099493027,
      "num_tokens": 116911955.0,
      "step": 67395
    },
    {
      "entropy": 5.83544397354126,
      "epoch": 5.243921416066913,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026038201605134567,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22670974135398864,
      "num_tokens": 116921179.0,
      "step": 67400
    },
    {
      "entropy": 5.848672151565552,
      "epoch": 5.2443104454386305,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026035435912955375,
      "loss": 4.9174,
      "mean_token_accuracy": 0.21601036489009856,
      "num_tokens": 116929347.0,
      "step": 67405
    },
    {
      "entropy": 5.836112833023071,
      "epoch": 5.244699474810348,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002603267024299861,
      "loss": 4.917,
      "mean_token_accuracy": 0.21424152404069902,
      "num_tokens": 116937858.0,
      "step": 67410
    },
    {
      "entropy": 5.929918479919434,
      "epoch": 5.245088504182066,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002602990459530623,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21778079569339753,
      "num_tokens": 116946645.0,
      "step": 67415
    },
    {
      "entropy": 5.944310426712036,
      "epoch": 5.2454775335537835,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002602713896992022,
      "loss": 4.821,
      "mean_token_accuracy": 0.23370299041271209,
      "num_tokens": 116954601.0,
      "step": 67420
    },
    {
      "entropy": 5.956086921691894,
      "epoch": 5.245866562925501,
      "grad_norm": 1.375,
      "learning_rate": 0.0002602437336688252,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21491467356681823,
      "num_tokens": 116963173.0,
      "step": 67425
    },
    {
      "entropy": 5.834650897979737,
      "epoch": 5.246255592297218,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002602160778623511,
      "loss": 4.7381,
      "mean_token_accuracy": 0.2353695347905159,
      "num_tokens": 116971583.0,
      "step": 67430
    },
    {
      "entropy": 5.850420904159546,
      "epoch": 5.246644621668936,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026018842228019953,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2281804695725441,
      "num_tokens": 116979746.0,
      "step": 67435
    },
    {
      "entropy": 5.95275912284851,
      "epoch": 5.2470336510406534,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00026016076692279003,
      "loss": 5.029,
      "mean_token_accuracy": 0.2150035709142685,
      "num_tokens": 116988513.0,
      "step": 67440
    },
    {
      "entropy": 5.886801910400391,
      "epoch": 5.247422680412371,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002601331117905423,
      "loss": 4.8894,
      "mean_token_accuracy": 0.21818059235811232,
      "num_tokens": 116997213.0,
      "step": 67445
    },
    {
      "entropy": 5.948919582366943,
      "epoch": 5.247811709784089,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026010545688387595,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2081000417470932,
      "num_tokens": 117005571.0,
      "step": 67450
    },
    {
      "entropy": 5.961368417739868,
      "epoch": 5.2482007391558065,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00026007780220321054,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2115231767296791,
      "num_tokens": 117014393.0,
      "step": 67455
    },
    {
      "entropy": 5.846679735183716,
      "epoch": 5.248589768527524,
      "grad_norm": 1.546875,
      "learning_rate": 0.00026005014774896574,
      "loss": 4.8307,
      "mean_token_accuracy": 0.23029744178056716,
      "num_tokens": 117023482.0,
      "step": 67460
    },
    {
      "entropy": 5.933190727233887,
      "epoch": 5.248978797899241,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026002249352156124,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22121233493089676,
      "num_tokens": 117031784.0,
      "step": 67465
    },
    {
      "entropy": 5.918664932250977,
      "epoch": 5.249367827270959,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002599948395214165,
      "loss": 4.8906,
      "mean_token_accuracy": 0.2299582302570343,
      "num_tokens": 117040455.0,
      "step": 67470
    },
    {
      "entropy": 5.854545211791992,
      "epoch": 5.249756856642676,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00025996718574895116,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22336635440587999,
      "num_tokens": 117048452.0,
      "step": 67475
    },
    {
      "entropy": 5.877184629440308,
      "epoch": 5.250145886014394,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025993953220458487,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22021687626838685,
      "num_tokens": 117057263.0,
      "step": 67480
    },
    {
      "entropy": 5.856005477905273,
      "epoch": 5.250534915386112,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002599118788887372,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21525786370038985,
      "num_tokens": 117066214.0,
      "step": 67485
    },
    {
      "entropy": 5.863619613647461,
      "epoch": 5.2509239447578295,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025988422580182775,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2247301459312439,
      "num_tokens": 117075139.0,
      "step": 67490
    },
    {
      "entropy": 5.9252355098724365,
      "epoch": 5.251312974129547,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002598565729442761,
      "loss": 4.9403,
      "mean_token_accuracy": 0.219345261156559,
      "num_tokens": 117083855.0,
      "step": 67495
    },
    {
      "entropy": 5.879280471801758,
      "epoch": 5.251702003501264,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025982892031650186,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22372221648693086,
      "num_tokens": 117092339.0,
      "step": 67500
    },
    {
      "entropy": 5.834828186035156,
      "epoch": 5.252091032872982,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002598012679189246,
      "loss": 4.8959,
      "mean_token_accuracy": 0.21889061182737352,
      "num_tokens": 117100996.0,
      "step": 67505
    },
    {
      "entropy": 5.794737529754639,
      "epoch": 5.252480062244699,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00025977361575196394,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22175722271203996,
      "num_tokens": 117109124.0,
      "step": 67510
    },
    {
      "entropy": 5.923291158676148,
      "epoch": 5.252869091616417,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00025974596381603943,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21626615673303604,
      "num_tokens": 117117710.0,
      "step": 67515
    },
    {
      "entropy": 5.9296900749206545,
      "epoch": 5.253258120988135,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002597183121115707,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2250870496034622,
      "num_tokens": 117126543.0,
      "step": 67520
    },
    {
      "entropy": 5.968899202346802,
      "epoch": 5.2536471503598525,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002596906606389771,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21744668781757354,
      "num_tokens": 117134992.0,
      "step": 67525
    },
    {
      "entropy": 5.9787322044372555,
      "epoch": 5.254036179731569,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025966300939867843,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2072967529296875,
      "num_tokens": 117144641.0,
      "step": 67530
    },
    {
      "entropy": 5.8928173065185545,
      "epoch": 5.254425209103287,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002596353583910942,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21453773677349092,
      "num_tokens": 117153834.0,
      "step": 67535
    },
    {
      "entropy": 5.942238426208496,
      "epoch": 5.254814238475005,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00025960770761664396,
      "loss": 4.911,
      "mean_token_accuracy": 0.21444918364286422,
      "num_tokens": 117162003.0,
      "step": 67540
    },
    {
      "entropy": 5.893832445144653,
      "epoch": 5.255203267846722,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025958005707574726,
      "loss": 4.8631,
      "mean_token_accuracy": 0.223610782623291,
      "num_tokens": 117169765.0,
      "step": 67545
    },
    {
      "entropy": 5.847612380981445,
      "epoch": 5.25559229721844,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002595524067688237,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21991652399301528,
      "num_tokens": 117178736.0,
      "step": 67550
    },
    {
      "entropy": 5.8742828369140625,
      "epoch": 5.255981326590158,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00025952475669629267,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22097132503986358,
      "num_tokens": 117187209.0,
      "step": 67555
    },
    {
      "entropy": 5.857695484161377,
      "epoch": 5.2563703559618755,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00025949710685857383,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23106579631567,
      "num_tokens": 117195581.0,
      "step": 67560
    },
    {
      "entropy": 5.976848173141479,
      "epoch": 5.256759385333592,
      "grad_norm": 1.59375,
      "learning_rate": 0.00025946945725608675,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21694038957357406,
      "num_tokens": 117204384.0,
      "step": 67565
    },
    {
      "entropy": 5.906661939620972,
      "epoch": 5.25714841470531,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002594418078892509,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21820637434720994,
      "num_tokens": 117213243.0,
      "step": 67570
    },
    {
      "entropy": 5.858357572555542,
      "epoch": 5.257537444077028,
      "grad_norm": 1.390625,
      "learning_rate": 0.000259414158758486,
      "loss": 4.8584,
      "mean_token_accuracy": 0.21784919798374175,
      "num_tokens": 117222509.0,
      "step": 67575
    },
    {
      "entropy": 5.93952202796936,
      "epoch": 5.257926473448745,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025938650986421125,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2162969321012497,
      "num_tokens": 117231146.0,
      "step": 67580
    },
    {
      "entropy": 5.930944871902466,
      "epoch": 5.258315502820463,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002593588612068465,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21613504588603974,
      "num_tokens": 117239798.0,
      "step": 67585
    },
    {
      "entropy": 5.9070967674255375,
      "epoch": 5.258704532192181,
      "grad_norm": 1.40625,
      "learning_rate": 0.000259331212786811,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21774909347295762,
      "num_tokens": 117248574.0,
      "step": 67590
    },
    {
      "entropy": 5.876617527008056,
      "epoch": 5.259093561563898,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025930356460452454,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2210534244775772,
      "num_tokens": 117258213.0,
      "step": 67595
    },
    {
      "entropy": 5.8371346950531,
      "epoch": 5.259482590935615,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002592759166604064,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23086679875850677,
      "num_tokens": 117266371.0,
      "step": 67600
    },
    {
      "entropy": 5.969144201278686,
      "epoch": 5.259871620307333,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025924826895487623,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2118205040693283,
      "num_tokens": 117274895.0,
      "step": 67605
    },
    {
      "entropy": 5.892315196990967,
      "epoch": 5.260260649679051,
      "grad_norm": 1.625,
      "learning_rate": 0.00025922062148835357,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21808929443359376,
      "num_tokens": 117283771.0,
      "step": 67610
    },
    {
      "entropy": 5.8611406803131105,
      "epoch": 5.260649679050768,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002591929742612578,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21932263821363449,
      "num_tokens": 117293968.0,
      "step": 67615
    },
    {
      "entropy": 5.938511562347412,
      "epoch": 5.261038708422486,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002591653272740085,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22923264801502227,
      "num_tokens": 117302468.0,
      "step": 67620
    },
    {
      "entropy": 5.890681362152099,
      "epoch": 5.261427737794204,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025913768052702514,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2165292426943779,
      "num_tokens": 117311363.0,
      "step": 67625
    },
    {
      "entropy": 5.947332096099854,
      "epoch": 5.261816767165921,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002591100340207272,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21536273658275604,
      "num_tokens": 117320940.0,
      "step": 67630
    },
    {
      "entropy": 5.8825990676879885,
      "epoch": 5.262205796537638,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002590823877555343,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22942212224006653,
      "num_tokens": 117329629.0,
      "step": 67635
    },
    {
      "entropy": 5.833219242095947,
      "epoch": 5.262594825909356,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002590547417318657,
      "loss": 4.8596,
      "mean_token_accuracy": 0.22257792055606843,
      "num_tokens": 117338713.0,
      "step": 67640
    },
    {
      "entropy": 5.915221166610718,
      "epoch": 5.262983855281074,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002590270959501411,
      "loss": 4.919,
      "mean_token_accuracy": 0.2254884198307991,
      "num_tokens": 117347114.0,
      "step": 67645
    },
    {
      "entropy": 5.863712358474731,
      "epoch": 5.263372884652791,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00025899945041077995,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22042381912469863,
      "num_tokens": 117354953.0,
      "step": 67650
    },
    {
      "entropy": 5.870214986801147,
      "epoch": 5.263761914024509,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025897180511420164,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21727113723754882,
      "num_tokens": 117363459.0,
      "step": 67655
    },
    {
      "entropy": 5.915839385986328,
      "epoch": 5.264150943396227,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025894416006082564,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21095409244298935,
      "num_tokens": 117371463.0,
      "step": 67660
    },
    {
      "entropy": 5.914535188674927,
      "epoch": 5.264539972767944,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025891651525107145,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21387271583080292,
      "num_tokens": 117380386.0,
      "step": 67665
    },
    {
      "entropy": 5.904464387893677,
      "epoch": 5.264929002139661,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00025888887068535853,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21889828741550446,
      "num_tokens": 117389533.0,
      "step": 67670
    },
    {
      "entropy": 5.97153868675232,
      "epoch": 5.265318031511379,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025886122636410643,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21230579167604446,
      "num_tokens": 117397806.0,
      "step": 67675
    },
    {
      "entropy": 5.871523427963257,
      "epoch": 5.265707060883097,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025883358228773445,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2221473142504692,
      "num_tokens": 117407391.0,
      "step": 67680
    },
    {
      "entropy": 5.8514550685882565,
      "epoch": 5.266096090254814,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002588059384566622,
      "loss": 4.819,
      "mean_token_accuracy": 0.2305518001317978,
      "num_tokens": 117416228.0,
      "step": 67685
    },
    {
      "entropy": 5.810494089126587,
      "epoch": 5.266485119626532,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025877829487130894,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2205313354730606,
      "num_tokens": 117424839.0,
      "step": 67690
    },
    {
      "entropy": 5.844971799850464,
      "epoch": 5.26687414899825,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002587506515320943,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22161851823329926,
      "num_tokens": 117433463.0,
      "step": 67695
    },
    {
      "entropy": 5.950062656402588,
      "epoch": 5.267263178369967,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002587230084394377,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21276509910821914,
      "num_tokens": 117441785.0,
      "step": 67700
    },
    {
      "entropy": 5.931679344177246,
      "epoch": 5.267652207741684,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002586953655937585,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21448352485895156,
      "num_tokens": 117450426.0,
      "step": 67705
    },
    {
      "entropy": 5.89712119102478,
      "epoch": 5.268041237113402,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002586677229954761,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22230238169431688,
      "num_tokens": 117458817.0,
      "step": 67710
    },
    {
      "entropy": 5.904269886016846,
      "epoch": 5.26843026648512,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002586400806450102,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21600616872310638,
      "num_tokens": 117467241.0,
      "step": 67715
    },
    {
      "entropy": 5.858381080627441,
      "epoch": 5.268819295856837,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00025861243854277986,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22042580544948578,
      "num_tokens": 117475859.0,
      "step": 67720
    },
    {
      "entropy": 5.85347352027893,
      "epoch": 5.269208325228555,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002585847966892048,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2111980438232422,
      "num_tokens": 117485202.0,
      "step": 67725
    },
    {
      "entropy": 5.915950441360474,
      "epoch": 5.269597354600272,
      "grad_norm": 1.625,
      "learning_rate": 0.0002585571550847042,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2161031737923622,
      "num_tokens": 117494156.0,
      "step": 67730
    },
    {
      "entropy": 5.915547037124634,
      "epoch": 5.26998638397199,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002585295137296978,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21331336945295334,
      "num_tokens": 117502850.0,
      "step": 67735
    },
    {
      "entropy": 5.970816135406494,
      "epoch": 5.270375413343707,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002585018726246047,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21634238660335542,
      "num_tokens": 117511400.0,
      "step": 67740
    },
    {
      "entropy": 5.938521528244019,
      "epoch": 5.270764442715425,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002584742317698444,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22227482497692108,
      "num_tokens": 117519326.0,
      "step": 67745
    },
    {
      "entropy": 5.862645435333252,
      "epoch": 5.271153472087143,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025844659116583635,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21507801711559296,
      "num_tokens": 117527945.0,
      "step": 67750
    },
    {
      "entropy": 5.829906415939331,
      "epoch": 5.27154250145886,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002584189508129999,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21933382451534272,
      "num_tokens": 117535892.0,
      "step": 67755
    },
    {
      "entropy": 5.982791519165039,
      "epoch": 5.271931530830578,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002583913107117544,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22133725583553315,
      "num_tokens": 117544387.0,
      "step": 67760
    },
    {
      "entropy": 5.861956644058227,
      "epoch": 5.272320560202295,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002583636708625195,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21146388500928878,
      "num_tokens": 117552774.0,
      "step": 67765
    },
    {
      "entropy": 5.864036417007446,
      "epoch": 5.272709589574013,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002583360312657143,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21191056817770004,
      "num_tokens": 117561741.0,
      "step": 67770
    },
    {
      "entropy": 5.97829647064209,
      "epoch": 5.27309861894573,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025830839192175835,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22493544965982437,
      "num_tokens": 117570185.0,
      "step": 67775
    },
    {
      "entropy": 5.8773880958557125,
      "epoch": 5.273487648317448,
      "grad_norm": 1.5625,
      "learning_rate": 0.000258280752831071,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2094309523701668,
      "num_tokens": 117578547.0,
      "step": 67780
    },
    {
      "entropy": 5.908707904815674,
      "epoch": 5.273876677689166,
      "grad_norm": 1.5,
      "learning_rate": 0.00025825311399407157,
      "loss": 4.9911,
      "mean_token_accuracy": 0.20932831317186357,
      "num_tokens": 117586795.0,
      "step": 67785
    },
    {
      "entropy": 5.903944873809815,
      "epoch": 5.274265707060883,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002582254754111795,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22859781384468078,
      "num_tokens": 117595219.0,
      "step": 67790
    },
    {
      "entropy": 5.84618353843689,
      "epoch": 5.274654736432601,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002581978370828142,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23000494986772538,
      "num_tokens": 117603947.0,
      "step": 67795
    },
    {
      "entropy": 5.8910047054290775,
      "epoch": 5.275043765804318,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002581701990093949,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21675776988267897,
      "num_tokens": 117612529.0,
      "step": 67800
    },
    {
      "entropy": 5.923612451553344,
      "epoch": 5.275432795176036,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002581425611913411,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21027622669935225,
      "num_tokens": 117620929.0,
      "step": 67805
    },
    {
      "entropy": 5.915188360214233,
      "epoch": 5.275821824547753,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002581149236290721,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22170062959194184,
      "num_tokens": 117629354.0,
      "step": 67810
    },
    {
      "entropy": 5.841807174682617,
      "epoch": 5.276210853919471,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002580872863230072,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22112427055835723,
      "num_tokens": 117637481.0,
      "step": 67815
    },
    {
      "entropy": 5.89835000038147,
      "epoch": 5.276599883291189,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025805964927356584,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21275329291820527,
      "num_tokens": 117645661.0,
      "step": 67820
    },
    {
      "entropy": 5.88832950592041,
      "epoch": 5.276988912662906,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025803201248116725,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2148388609290123,
      "num_tokens": 117654405.0,
      "step": 67825
    },
    {
      "entropy": 5.900545120239258,
      "epoch": 5.277377942034624,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025800437594623095,
      "loss": 4.9121,
      "mean_token_accuracy": 0.214832104742527,
      "num_tokens": 117663585.0,
      "step": 67830
    },
    {
      "entropy": 5.807180690765381,
      "epoch": 5.277766971406341,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002579767396691762,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22937777489423752,
      "num_tokens": 117672141.0,
      "step": 67835
    },
    {
      "entropy": 5.91658878326416,
      "epoch": 5.278156000778059,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00025794910365042233,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21431930512189865,
      "num_tokens": 117680755.0,
      "step": 67840
    },
    {
      "entropy": 5.928325128555298,
      "epoch": 5.278545030149776,
      "grad_norm": 1.484375,
      "learning_rate": 0.00025792146789038875,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2170793816447258,
      "num_tokens": 117689430.0,
      "step": 67845
    },
    {
      "entropy": 5.883506202697754,
      "epoch": 5.278934059521494,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002578938323894946,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22267838716506957,
      "num_tokens": 117698134.0,
      "step": 67850
    },
    {
      "entropy": 5.789335012435913,
      "epoch": 5.279323088893212,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025786619714815933,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22640684694051744,
      "num_tokens": 117707475.0,
      "step": 67855
    },
    {
      "entropy": 5.895144033432007,
      "epoch": 5.279712118264929,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002578385621668023,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22923634499311446,
      "num_tokens": 117716353.0,
      "step": 67860
    },
    {
      "entropy": 5.859406852722168,
      "epoch": 5.280101147636646,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002578109274458427,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22423188537359237,
      "num_tokens": 117725244.0,
      "step": 67865
    },
    {
      "entropy": 5.926948690414429,
      "epoch": 5.280490177008364,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025778329298569997,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22163361608982085,
      "num_tokens": 117733419.0,
      "step": 67870
    },
    {
      "entropy": 5.920617008209229,
      "epoch": 5.280879206380082,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002577556587867934,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2092956393957138,
      "num_tokens": 117742256.0,
      "step": 67875
    },
    {
      "entropy": 5.939651346206665,
      "epoch": 5.281268235751799,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025772802484954217,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21982627660036086,
      "num_tokens": 117750982.0,
      "step": 67880
    },
    {
      "entropy": 5.858913612365723,
      "epoch": 5.281657265123517,
      "grad_norm": 1.65625,
      "learning_rate": 0.00025770039117436566,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2208586648106575,
      "num_tokens": 117758629.0,
      "step": 67885
    },
    {
      "entropy": 5.847533226013184,
      "epoch": 5.282046294495235,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002576727577616832,
      "loss": 4.9336,
      "mean_token_accuracy": 0.21533840745687485,
      "num_tokens": 117766620.0,
      "step": 67890
    },
    {
      "entropy": 5.824357891082764,
      "epoch": 5.282435323866952,
      "grad_norm": 1.546875,
      "learning_rate": 0.00025764512461191413,
      "loss": 4.831,
      "mean_token_accuracy": 0.2270228698849678,
      "num_tokens": 117774906.0,
      "step": 67895
    },
    {
      "entropy": 5.8857419967651365,
      "epoch": 5.282824353238669,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002576174917254776,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21486251652240754,
      "num_tokens": 117784140.0,
      "step": 67900
    },
    {
      "entropy": 5.875032424926758,
      "epoch": 5.283213382610387,
      "grad_norm": 1.359375,
      "learning_rate": 0.00025758985910279304,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22593943476676942,
      "num_tokens": 117792927.0,
      "step": 67905
    },
    {
      "entropy": 5.9111357688903805,
      "epoch": 5.283602411982105,
      "grad_norm": 1.5,
      "learning_rate": 0.0002575622267442796,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22253002673387529,
      "num_tokens": 117801433.0,
      "step": 67910
    },
    {
      "entropy": 5.964916658401489,
      "epoch": 5.283991441353822,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002575345946503566,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21213325411081313,
      "num_tokens": 117810027.0,
      "step": 67915
    },
    {
      "entropy": 5.828066635131836,
      "epoch": 5.28438047072554,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025750696282144335,
      "loss": 4.8367,
      "mean_token_accuracy": 0.22768062651157378,
      "num_tokens": 117818244.0,
      "step": 67920
    },
    {
      "entropy": 5.94966835975647,
      "epoch": 5.284769500097258,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002574793312579592,
      "loss": 5.0725,
      "mean_token_accuracy": 0.2102821335196495,
      "num_tokens": 117827908.0,
      "step": 67925
    },
    {
      "entropy": 5.927512741088867,
      "epoch": 5.2851585294689745,
      "grad_norm": 1.375,
      "learning_rate": 0.00025745169996032317,
      "loss": 4.9983,
      "mean_token_accuracy": 0.20632197707891464,
      "num_tokens": 117836678.0,
      "step": 67930
    },
    {
      "entropy": 5.925325393676758,
      "epoch": 5.285547558840692,
      "grad_norm": 1.5,
      "learning_rate": 0.0002574240689289547,
      "loss": 4.9022,
      "mean_token_accuracy": 0.21965862065553665,
      "num_tokens": 117845132.0,
      "step": 67935
    },
    {
      "entropy": 5.925509595870972,
      "epoch": 5.28593658821241,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025739643816427305,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2124182403087616,
      "num_tokens": 117854450.0,
      "step": 67940
    },
    {
      "entropy": 5.87840781211853,
      "epoch": 5.286325617584128,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002573688076666974,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22372539937496186,
      "num_tokens": 117863781.0,
      "step": 67945
    },
    {
      "entropy": 5.908760213851929,
      "epoch": 5.286714646955845,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025734117743664704,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22636251002550126,
      "num_tokens": 117872054.0,
      "step": 67950
    },
    {
      "entropy": 5.850053310394287,
      "epoch": 5.287103676327563,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002573135474745412,
      "loss": 4.8353,
      "mean_token_accuracy": 0.22148197144269943,
      "num_tokens": 117880095.0,
      "step": 67955
    },
    {
      "entropy": 5.875826692581176,
      "epoch": 5.287492705699281,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025728591778079906,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21551012098789216,
      "num_tokens": 117889105.0,
      "step": 67960
    },
    {
      "entropy": 5.911121225357055,
      "epoch": 5.2878817350709975,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00025725828835584,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22085015922784806,
      "num_tokens": 117897421.0,
      "step": 67965
    },
    {
      "entropy": 5.9332763671875,
      "epoch": 5.288270764442715,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025723065920008314,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2259612798690796,
      "num_tokens": 117906688.0,
      "step": 67970
    },
    {
      "entropy": 5.928823614120484,
      "epoch": 5.288659793814433,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002572030303139477,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21617251634597778,
      "num_tokens": 117914355.0,
      "step": 67975
    },
    {
      "entropy": 5.8620137691497805,
      "epoch": 5.2890488231861505,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025717540169785304,
      "loss": 4.9063,
      "mean_token_accuracy": 0.21718096882104873,
      "num_tokens": 117923358.0,
      "step": 67980
    },
    {
      "entropy": 5.840695238113403,
      "epoch": 5.289437852557868,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002571477733522183,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21440148949623108,
      "num_tokens": 117932258.0,
      "step": 67985
    },
    {
      "entropy": 5.8731626033782955,
      "epoch": 5.289826881929586,
      "grad_norm": 1.53125,
      "learning_rate": 0.00025712014527746264,
      "loss": 4.874,
      "mean_token_accuracy": 0.22133521735668182,
      "num_tokens": 117941357.0,
      "step": 67990
    },
    {
      "entropy": 5.942467069625854,
      "epoch": 5.290215911301304,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002570925174740053,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21716598570346832,
      "num_tokens": 117950737.0,
      "step": 67995
    },
    {
      "entropy": 5.887917327880859,
      "epoch": 5.2906049406730205,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002570648899422655,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21766560077667235,
      "num_tokens": 117959531.0,
      "step": 68000
    },
    {
      "entropy": 5.893392324447632,
      "epoch": 5.290993970044738,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002570372626826625,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21929726749658585,
      "num_tokens": 117967679.0,
      "step": 68005
    },
    {
      "entropy": 5.798775005340576,
      "epoch": 5.291382999416456,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002570096356956155,
      "loss": 4.8473,
      "mean_token_accuracy": 0.22092106491327285,
      "num_tokens": 117976076.0,
      "step": 68010
    },
    {
      "entropy": 5.8942817687988285,
      "epoch": 5.2917720287881735,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025698200898154354,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2162758782505989,
      "num_tokens": 117984629.0,
      "step": 68015
    },
    {
      "entropy": 5.920121765136718,
      "epoch": 5.292161058159891,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025695438254086595,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22065748572349547,
      "num_tokens": 117993955.0,
      "step": 68020
    },
    {
      "entropy": 5.8249279975891115,
      "epoch": 5.292550087531609,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002569267563740018,
      "loss": 4.8436,
      "mean_token_accuracy": 0.22861330211162567,
      "num_tokens": 118002239.0,
      "step": 68025
    },
    {
      "entropy": 5.85436520576477,
      "epoch": 5.292939116903327,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002568991304813704,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22957465648651124,
      "num_tokens": 118011622.0,
      "step": 68030
    },
    {
      "entropy": 5.891383504867553,
      "epoch": 5.2933281462750434,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025687150486339094,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2137890487909317,
      "num_tokens": 118021084.0,
      "step": 68035
    },
    {
      "entropy": 5.9022111892700195,
      "epoch": 5.293717175646761,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025684387952048247,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21629287153482438,
      "num_tokens": 118029911.0,
      "step": 68040
    },
    {
      "entropy": 5.986558532714843,
      "epoch": 5.294106205018479,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025681625445306435,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21188804656267166,
      "num_tokens": 118037787.0,
      "step": 68045
    },
    {
      "entropy": 5.847724342346192,
      "epoch": 5.2944952343901965,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002567886296615555,
      "loss": 4.9035,
      "mean_token_accuracy": 0.21422518342733382,
      "num_tokens": 118046753.0,
      "step": 68050
    },
    {
      "entropy": 5.856315183639526,
      "epoch": 5.294884263761914,
      "grad_norm": 1.5,
      "learning_rate": 0.00025676100514637523,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22548332065343857,
      "num_tokens": 118055425.0,
      "step": 68055
    },
    {
      "entropy": 5.86829137802124,
      "epoch": 5.295273293133632,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025673338090794276,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22401642054319382,
      "num_tokens": 118063857.0,
      "step": 68060
    },
    {
      "entropy": 5.826662063598633,
      "epoch": 5.295662322505349,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002567057569466771,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21844796538352967,
      "num_tokens": 118072243.0,
      "step": 68065
    },
    {
      "entropy": 5.899838495254516,
      "epoch": 5.296051351877066,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025667813326299746,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21064009368419648,
      "num_tokens": 118081019.0,
      "step": 68070
    },
    {
      "entropy": 5.877289295196533,
      "epoch": 5.296440381248784,
      "grad_norm": 1.515625,
      "learning_rate": 0.00025665050985732295,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22061519026756288,
      "num_tokens": 118089662.0,
      "step": 68075
    },
    {
      "entropy": 5.882540559768676,
      "epoch": 5.296829410620502,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00025662288673007286,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2262355461716652,
      "num_tokens": 118098229.0,
      "step": 68080
    },
    {
      "entropy": 5.84581446647644,
      "epoch": 5.2972184399922195,
      "grad_norm": 1.359375,
      "learning_rate": 0.00025659526388166617,
      "loss": 4.7805,
      "mean_token_accuracy": 0.23151797652244568,
      "num_tokens": 118107016.0,
      "step": 68085
    },
    {
      "entropy": 5.870253801345825,
      "epoch": 5.297607469363937,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025656764131252194,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2163043275475502,
      "num_tokens": 118115090.0,
      "step": 68090
    },
    {
      "entropy": 5.880966472625732,
      "epoch": 5.297996498735655,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002565400190230595,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2227540761232376,
      "num_tokens": 118125044.0,
      "step": 68095
    },
    {
      "entropy": 5.865283393859864,
      "epoch": 5.298385528107372,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002565123970136978,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2226044565439224,
      "num_tokens": 118133609.0,
      "step": 68100
    },
    {
      "entropy": 5.864976501464843,
      "epoch": 5.298774557479089,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002564847752848562,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21915333718061447,
      "num_tokens": 118142217.0,
      "step": 68105
    },
    {
      "entropy": 5.872900867462159,
      "epoch": 5.299163586850807,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025645715383695357,
      "loss": 4.9501,
      "mean_token_accuracy": 0.221360881626606,
      "num_tokens": 118150215.0,
      "step": 68110
    },
    {
      "entropy": 5.887674570083618,
      "epoch": 5.299552616222525,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002564295326704092,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22114074230194092,
      "num_tokens": 118158909.0,
      "step": 68115
    },
    {
      "entropy": 5.890580415725708,
      "epoch": 5.2999416455942425,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002564019117856421,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22198819667100905,
      "num_tokens": 118168445.0,
      "step": 68120
    },
    {
      "entropy": 5.915107250213623,
      "epoch": 5.30033067496596,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025637429118307133,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21482883542776107,
      "num_tokens": 118177521.0,
      "step": 68125
    },
    {
      "entropy": 5.895406246185303,
      "epoch": 5.300719704337677,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002563466708631161,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21379590332508086,
      "num_tokens": 118185929.0,
      "step": 68130
    },
    {
      "entropy": 5.895614576339722,
      "epoch": 5.301108733709395,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002563190508261954,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2244717448949814,
      "num_tokens": 118194758.0,
      "step": 68135
    },
    {
      "entropy": 5.802139854431152,
      "epoch": 5.301497763081112,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025629143107272837,
      "loss": 4.83,
      "mean_token_accuracy": 0.2341532438993454,
      "num_tokens": 118203496.0,
      "step": 68140
    },
    {
      "entropy": 5.873927354812622,
      "epoch": 5.30188679245283,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002562638116031341,
      "loss": 4.8851,
      "mean_token_accuracy": 0.21559060513973236,
      "num_tokens": 118212348.0,
      "step": 68145
    },
    {
      "entropy": 5.823506212234497,
      "epoch": 5.302275821824548,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002562361924178317,
      "loss": 4.7847,
      "mean_token_accuracy": 0.2399173393845558,
      "num_tokens": 118220589.0,
      "step": 68150
    },
    {
      "entropy": 5.881532955169678,
      "epoch": 5.3026648511962655,
      "grad_norm": 1.53125,
      "learning_rate": 0.00025620857351724017,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21270554959774018,
      "num_tokens": 118229348.0,
      "step": 68155
    },
    {
      "entropy": 5.905920076370239,
      "epoch": 5.303053880567983,
      "grad_norm": 1.53125,
      "learning_rate": 0.00025618095490177864,
      "loss": 4.9722,
      "mean_token_accuracy": 0.22141170799732207,
      "num_tokens": 118238097.0,
      "step": 68160
    },
    {
      "entropy": 5.8781332015991214,
      "epoch": 5.303442909939701,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025615333657186615,
      "loss": 4.8421,
      "mean_token_accuracy": 0.22938079833984376,
      "num_tokens": 118246724.0,
      "step": 68165
    },
    {
      "entropy": 5.985379552841186,
      "epoch": 5.303831939311418,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002561257185279218,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2127864733338356,
      "num_tokens": 118255832.0,
      "step": 68170
    },
    {
      "entropy": 5.846642923355103,
      "epoch": 5.304220968683135,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025609810077036467,
      "loss": 4.8639,
      "mean_token_accuracy": 0.23114868104457856,
      "num_tokens": 118265848.0,
      "step": 68175
    },
    {
      "entropy": 5.860721445083618,
      "epoch": 5.304609998054853,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002560704832996137,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2232258141040802,
      "num_tokens": 118273966.0,
      "step": 68180
    },
    {
      "entropy": 5.939294290542603,
      "epoch": 5.304999027426571,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002560428661160881,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21015969812870025,
      "num_tokens": 118282796.0,
      "step": 68185
    },
    {
      "entropy": 5.911509418487549,
      "epoch": 5.3053880567982885,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025601524922020676,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21690237522125244,
      "num_tokens": 118291023.0,
      "step": 68190
    },
    {
      "entropy": 5.8414980411529545,
      "epoch": 5.305777086170006,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002559876326123889,
      "loss": 4.911,
      "mean_token_accuracy": 0.2187328338623047,
      "num_tokens": 118299183.0,
      "step": 68195
    },
    {
      "entropy": 5.911609220504761,
      "epoch": 5.306166115541723,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002559600162930534,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21019780039787292,
      "num_tokens": 118308001.0,
      "step": 68200
    },
    {
      "entropy": 5.89142165184021,
      "epoch": 5.306555144913441,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002559324002626193,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22005068510770798,
      "num_tokens": 118316619.0,
      "step": 68205
    },
    {
      "entropy": 5.921667146682739,
      "epoch": 5.306944174285158,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002559047845215057,
      "loss": 4.9656,
      "mean_token_accuracy": 0.20941032469272614,
      "num_tokens": 118325014.0,
      "step": 68210
    },
    {
      "entropy": 5.881661653518677,
      "epoch": 5.307333203656876,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002558771690701316,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2178819641470909,
      "num_tokens": 118334987.0,
      "step": 68215
    },
    {
      "entropy": 5.82679591178894,
      "epoch": 5.307722233028594,
      "grad_norm": 1.515625,
      "learning_rate": 0.000255849553908916,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22475523203611375,
      "num_tokens": 118343053.0,
      "step": 68220
    },
    {
      "entropy": 5.862268447875977,
      "epoch": 5.3081112624003115,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025582193903827795,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22428847551345826,
      "num_tokens": 118351217.0,
      "step": 68225
    },
    {
      "entropy": 5.908346509933471,
      "epoch": 5.308500291772029,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002557943244586365,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21032453924417496,
      "num_tokens": 118359829.0,
      "step": 68230
    },
    {
      "entropy": 5.93742036819458,
      "epoch": 5.308889321143746,
      "grad_norm": 1.375,
      "learning_rate": 0.00025576671017041053,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21180319041013718,
      "num_tokens": 118369114.0,
      "step": 68235
    },
    {
      "entropy": 5.928803300857544,
      "epoch": 5.309278350515464,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002557390961740192,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21763966828584672,
      "num_tokens": 118377684.0,
      "step": 68240
    },
    {
      "entropy": 5.833543729782105,
      "epoch": 5.309667379887181,
      "grad_norm": 1.59375,
      "learning_rate": 0.00025571148246988137,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2290648803114891,
      "num_tokens": 118385799.0,
      "step": 68245
    },
    {
      "entropy": 5.895347642898559,
      "epoch": 5.310056409258899,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025568386905841614,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2111326426267624,
      "num_tokens": 118394131.0,
      "step": 68250
    },
    {
      "entropy": 5.84665813446045,
      "epoch": 5.310445438630617,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002556562559400424,
      "loss": 4.8971,
      "mean_token_accuracy": 0.21838960349559783,
      "num_tokens": 118403795.0,
      "step": 68255
    },
    {
      "entropy": 5.899920320510864,
      "epoch": 5.3108344680023345,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002556286431151792,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22230299562215805,
      "num_tokens": 118412818.0,
      "step": 68260
    },
    {
      "entropy": 5.961867809295654,
      "epoch": 5.311223497374051,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002556010305842456,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21594969928264618,
      "num_tokens": 118421365.0,
      "step": 68265
    },
    {
      "entropy": 5.881282567977905,
      "epoch": 5.311612526745769,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002555734183476603,
      "loss": 4.8733,
      "mean_token_accuracy": 0.21725194454193114,
      "num_tokens": 118429993.0,
      "step": 68270
    },
    {
      "entropy": 5.914928388595581,
      "epoch": 5.312001556117487,
      "grad_norm": 1.5,
      "learning_rate": 0.0002555458064058427,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21357848942279817,
      "num_tokens": 118438701.0,
      "step": 68275
    },
    {
      "entropy": 5.875366687774658,
      "epoch": 5.312390585489204,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025551819475921135,
      "loss": 4.9391,
      "mean_token_accuracy": 0.21513456404209136,
      "num_tokens": 118447358.0,
      "step": 68280
    },
    {
      "entropy": 5.818239212036133,
      "epoch": 5.312779614860922,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002554905834081854,
      "loss": 4.8353,
      "mean_token_accuracy": 0.22087589800357818,
      "num_tokens": 118455736.0,
      "step": 68285
    },
    {
      "entropy": 5.940585279464722,
      "epoch": 5.31316864423264,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002554629723531838,
      "loss": 4.983,
      "mean_token_accuracy": 0.21524973213672638,
      "num_tokens": 118464765.0,
      "step": 68290
    },
    {
      "entropy": 5.8969581604003904,
      "epoch": 5.3135576736043575,
      "grad_norm": 1.5,
      "learning_rate": 0.0002554353615946255,
      "loss": 4.9079,
      "mean_token_accuracy": 0.21740695238113403,
      "num_tokens": 118473181.0,
      "step": 68295
    },
    {
      "entropy": 5.921140384674072,
      "epoch": 5.313946702976074,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025540775113292943,
      "loss": 4.959,
      "mean_token_accuracy": 0.22132176756858826,
      "num_tokens": 118481849.0,
      "step": 68300
    },
    {
      "entropy": 5.835493087768555,
      "epoch": 5.314335732347792,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002553801409685146,
      "loss": 4.8883,
      "mean_token_accuracy": 0.227797931432724,
      "num_tokens": 118491381.0,
      "step": 68305
    },
    {
      "entropy": 5.894591665267944,
      "epoch": 5.31472476171951,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025535253110179994,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21446572989225388,
      "num_tokens": 118500575.0,
      "step": 68310
    },
    {
      "entropy": 5.813097333908081,
      "epoch": 5.315113791091227,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002553249215332043,
      "loss": 4.8068,
      "mean_token_accuracy": 0.23089618235826492,
      "num_tokens": 118509037.0,
      "step": 68315
    },
    {
      "entropy": 5.817120027542114,
      "epoch": 5.315502820462945,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025529731226314663,
      "loss": 4.8055,
      "mean_token_accuracy": 0.2249295011162758,
      "num_tokens": 118517397.0,
      "step": 68320
    },
    {
      "entropy": 5.882214975357056,
      "epoch": 5.315891849834663,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002552697032920459,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21396974325180054,
      "num_tokens": 118525878.0,
      "step": 68325
    },
    {
      "entropy": 5.922155523300171,
      "epoch": 5.3162808792063805,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025524209462032105,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22554733753204345,
      "num_tokens": 118533615.0,
      "step": 68330
    },
    {
      "entropy": 5.940156650543213,
      "epoch": 5.316669908578097,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002552144862483909,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20806398391723632,
      "num_tokens": 118543149.0,
      "step": 68335
    },
    {
      "entropy": 5.901089668273926,
      "epoch": 5.317058937949815,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002551868781766745,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22267400175333024,
      "num_tokens": 118551888.0,
      "step": 68340
    },
    {
      "entropy": 5.906593084335327,
      "epoch": 5.317447967321533,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002551592704055907,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2153277188539505,
      "num_tokens": 118560426.0,
      "step": 68345
    },
    {
      "entropy": 5.864302253723144,
      "epoch": 5.31783699669325,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025513166293555836,
      "loss": 4.9353,
      "mean_token_accuracy": 0.2263246327638626,
      "num_tokens": 118569081.0,
      "step": 68350
    },
    {
      "entropy": 5.981005144119263,
      "epoch": 5.318226026064968,
      "grad_norm": 1.546875,
      "learning_rate": 0.00025510405576699636,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2149076282978058,
      "num_tokens": 118577276.0,
      "step": 68355
    },
    {
      "entropy": 5.875511646270752,
      "epoch": 5.318615055436686,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002550764489003237,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21489230990409852,
      "num_tokens": 118586746.0,
      "step": 68360
    },
    {
      "entropy": 5.970089483261108,
      "epoch": 5.3190040848084035,
      "grad_norm": 1.625,
      "learning_rate": 0.0002550488423359592,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21742721796035766,
      "num_tokens": 118594954.0,
      "step": 68365
    },
    {
      "entropy": 5.946256923675537,
      "epoch": 5.31939311418012,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00025502123607432174,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21895347386598588,
      "num_tokens": 118603282.0,
      "step": 68370
    },
    {
      "entropy": 5.901910877227783,
      "epoch": 5.319782143551838,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025499363011583024,
      "loss": 4.8861,
      "mean_token_accuracy": 0.21976556181907653,
      "num_tokens": 118611571.0,
      "step": 68375
    },
    {
      "entropy": 5.887943935394287,
      "epoch": 5.320171172923556,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002549660244609036,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21880963742733,
      "num_tokens": 118620594.0,
      "step": 68380
    },
    {
      "entropy": 5.972258234024048,
      "epoch": 5.320560202295273,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002549384191099607,
      "loss": 4.999,
      "mean_token_accuracy": 0.21272812634706498,
      "num_tokens": 118629113.0,
      "step": 68385
    },
    {
      "entropy": 5.840529155731201,
      "epoch": 5.320949231666991,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025491081406342025,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22505107372999192,
      "num_tokens": 118638111.0,
      "step": 68390
    },
    {
      "entropy": 5.872589826583862,
      "epoch": 5.321338261038709,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00025488320932170127,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2110362783074379,
      "num_tokens": 118646169.0,
      "step": 68395
    },
    {
      "entropy": 5.838713598251343,
      "epoch": 5.321727290410426,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002548556048852226,
      "loss": 4.8148,
      "mean_token_accuracy": 0.22953622788190842,
      "num_tokens": 118654321.0,
      "step": 68400
    },
    {
      "entropy": 5.885142135620117,
      "epoch": 5.322116319782143,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025482800075440296,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21816242039203643,
      "num_tokens": 118663177.0,
      "step": 68405
    },
    {
      "entropy": 5.860215568542481,
      "epoch": 5.322505349153861,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002548003969296614,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22706662565469743,
      "num_tokens": 118671159.0,
      "step": 68410
    },
    {
      "entropy": 5.844998216629028,
      "epoch": 5.322894378525579,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025477279341141665,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21758589893579483,
      "num_tokens": 118679955.0,
      "step": 68415
    },
    {
      "entropy": 5.847661781311035,
      "epoch": 5.323283407897296,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002547451902000875,
      "loss": 4.8896,
      "mean_token_accuracy": 0.21909790635108947,
      "num_tokens": 118689222.0,
      "step": 68420
    },
    {
      "entropy": 5.8374275207519535,
      "epoch": 5.323672437269014,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00025471758729609284,
      "loss": 4.8746,
      "mean_token_accuracy": 0.2160216525197029,
      "num_tokens": 118698177.0,
      "step": 68425
    },
    {
      "entropy": 5.846771669387818,
      "epoch": 5.324061466640732,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025468998469985155,
      "loss": 4.8764,
      "mean_token_accuracy": 0.21855004578828813,
      "num_tokens": 118706699.0,
      "step": 68430
    },
    {
      "entropy": 5.809262466430664,
      "epoch": 5.324450496012449,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002546623824117824,
      "loss": 4.7656,
      "mean_token_accuracy": 0.2333296298980713,
      "num_tokens": 118715469.0,
      "step": 68435
    },
    {
      "entropy": 5.886678838729859,
      "epoch": 5.324839525384166,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002546347804323043,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2199527993798256,
      "num_tokens": 118723765.0,
      "step": 68440
    },
    {
      "entropy": 5.865315914154053,
      "epoch": 5.325228554755884,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002546071787618359,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2208289161324501,
      "num_tokens": 118733011.0,
      "step": 68445
    },
    {
      "entropy": 5.835000896453858,
      "epoch": 5.325617584127602,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002545795774007962,
      "loss": 4.7844,
      "mean_token_accuracy": 0.2351594462990761,
      "num_tokens": 118741437.0,
      "step": 68450
    },
    {
      "entropy": 5.864903402328491,
      "epoch": 5.326006613499319,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002545519763496038,
      "loss": 4.8308,
      "mean_token_accuracy": 0.22937202751636504,
      "num_tokens": 118750423.0,
      "step": 68455
    },
    {
      "entropy": 5.905500173568726,
      "epoch": 5.326395642871037,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00025452437560867774,
      "loss": 4.9578,
      "mean_token_accuracy": 0.21856200098991393,
      "num_tokens": 118759005.0,
      "step": 68460
    },
    {
      "entropy": 5.9192205429077145,
      "epoch": 5.326784672242754,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002544967751784367,
      "loss": 4.942,
      "mean_token_accuracy": 0.21852443218231202,
      "num_tokens": 118767273.0,
      "step": 68465
    },
    {
      "entropy": 5.992331314086914,
      "epoch": 5.327173701614472,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025446917505929937,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21180498450994492,
      "num_tokens": 118777153.0,
      "step": 68470
    },
    {
      "entropy": 5.910095882415772,
      "epoch": 5.327562730986189,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002544415752516846,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21108947694301605,
      "num_tokens": 118786819.0,
      "step": 68475
    },
    {
      "entropy": 5.851808357238769,
      "epoch": 5.327951760357907,
      "grad_norm": 1.53125,
      "learning_rate": 0.00025441397575601125,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22210266888141633,
      "num_tokens": 118795444.0,
      "step": 68480
    },
    {
      "entropy": 5.925428152084351,
      "epoch": 5.328340789729625,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025438637657269805,
      "loss": 5.037,
      "mean_token_accuracy": 0.20812795013189317,
      "num_tokens": 118804539.0,
      "step": 68485
    },
    {
      "entropy": 5.834197664260865,
      "epoch": 5.328729819101342,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002543587777021636,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22053764015436172,
      "num_tokens": 118812799.0,
      "step": 68490
    },
    {
      "entropy": 5.921358346939087,
      "epoch": 5.32911884847306,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002543311791448271,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2165755197405815,
      "num_tokens": 118821415.0,
      "step": 68495
    },
    {
      "entropy": 5.926783800125122,
      "epoch": 5.329507877844778,
      "grad_norm": 1.625,
      "learning_rate": 0.00025430358090110705,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2162941262125969,
      "num_tokens": 118829354.0,
      "step": 68500
    },
    {
      "entropy": 5.8595507621765135,
      "epoch": 5.329896907216495,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025427598297142213,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22090301811695098,
      "num_tokens": 118837924.0,
      "step": 68505
    },
    {
      "entropy": 5.9099733352661135,
      "epoch": 5.330285936588212,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002542483853561912,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21547487676143645,
      "num_tokens": 118847342.0,
      "step": 68510
    },
    {
      "entropy": 5.8886724472045895,
      "epoch": 5.33067496595993,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000254220788055833,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22894565910100936,
      "num_tokens": 118856135.0,
      "step": 68515
    },
    {
      "entropy": 5.9507856369018555,
      "epoch": 5.331063995331648,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002541931910707662,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21925931870937349,
      "num_tokens": 118864898.0,
      "step": 68520
    },
    {
      "entropy": 5.87237663269043,
      "epoch": 5.331453024703365,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002541655944014097,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2275308445096016,
      "num_tokens": 118874276.0,
      "step": 68525
    },
    {
      "entropy": 5.882525968551636,
      "epoch": 5.331842054075083,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002541379980481821,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21678184866905212,
      "num_tokens": 118883315.0,
      "step": 68530
    },
    {
      "entropy": 5.94589409828186,
      "epoch": 5.3322310834468,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002541104020115022,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21140984296798707,
      "num_tokens": 118891228.0,
      "step": 68535
    },
    {
      "entropy": 5.88700385093689,
      "epoch": 5.3326201128185176,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025408280629178864,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22441271245479583,
      "num_tokens": 118899757.0,
      "step": 68540
    },
    {
      "entropy": 5.873735952377319,
      "epoch": 5.333009142190235,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002540552108894603,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2203931987285614,
      "num_tokens": 118908723.0,
      "step": 68545
    },
    {
      "entropy": 5.77337965965271,
      "epoch": 5.333398171561953,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002540276158049358,
      "loss": 4.7652,
      "mean_token_accuracy": 0.23447226881980895,
      "num_tokens": 118917711.0,
      "step": 68550
    },
    {
      "entropy": 5.821710443496704,
      "epoch": 5.333787200933671,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00025400002103863367,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22639009803533555,
      "num_tokens": 118925970.0,
      "step": 68555
    },
    {
      "entropy": 5.892201709747314,
      "epoch": 5.334176230305388,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025397242659097293,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21569897085428238,
      "num_tokens": 118934747.0,
      "step": 68560
    },
    {
      "entropy": 5.930914402008057,
      "epoch": 5.334565259677106,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002539448324623722,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2148421123623848,
      "num_tokens": 118943206.0,
      "step": 68565
    },
    {
      "entropy": 5.965063571929932,
      "epoch": 5.334954289048823,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025391723865325,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22332533150911332,
      "num_tokens": 118951751.0,
      "step": 68570
    },
    {
      "entropy": 5.876570701599121,
      "epoch": 5.3353433184205405,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025388964516402527,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22254080772399903,
      "num_tokens": 118960297.0,
      "step": 68575
    },
    {
      "entropy": 5.871348237991333,
      "epoch": 5.335732347792258,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002538620519951165,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21320126652717591,
      "num_tokens": 118969283.0,
      "step": 68580
    },
    {
      "entropy": 5.824328947067261,
      "epoch": 5.336121377163976,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025383445914694245,
      "loss": 4.811,
      "mean_token_accuracy": 0.22077952474355697,
      "num_tokens": 118977946.0,
      "step": 68585
    },
    {
      "entropy": 5.859393692016601,
      "epoch": 5.336510406535694,
      "grad_norm": 1.5,
      "learning_rate": 0.0002538068666199218,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22745316177606584,
      "num_tokens": 118986303.0,
      "step": 68590
    },
    {
      "entropy": 5.924205732345581,
      "epoch": 5.336899435907411,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002537792744144732,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20896050184965134,
      "num_tokens": 118995174.0,
      "step": 68595
    },
    {
      "entropy": 5.887504005432129,
      "epoch": 5.337288465279128,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002537516825310154,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21333954632282257,
      "num_tokens": 119004849.0,
      "step": 68600
    },
    {
      "entropy": 5.8827667236328125,
      "epoch": 5.337677494650846,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002537240909699669,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2223087877035141,
      "num_tokens": 119013423.0,
      "step": 68605
    },
    {
      "entropy": 5.875261735916138,
      "epoch": 5.3380665240225635,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002536964997317465,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2201731786131859,
      "num_tokens": 119021422.0,
      "step": 68610
    },
    {
      "entropy": 5.83810715675354,
      "epoch": 5.338455553394281,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025366890881677284,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22475724220275878,
      "num_tokens": 119029306.0,
      "step": 68615
    },
    {
      "entropy": 5.934020948410034,
      "epoch": 5.338844582765999,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002536413182254645,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2176676169037819,
      "num_tokens": 119037768.0,
      "step": 68620
    },
    {
      "entropy": 5.880868148803711,
      "epoch": 5.339233612137717,
      "grad_norm": 1.53125,
      "learning_rate": 0.00025361372795824015,
      "loss": 4.907,
      "mean_token_accuracy": 0.21996604800224304,
      "num_tokens": 119045902.0,
      "step": 68625
    },
    {
      "entropy": 5.858265686035156,
      "epoch": 5.339622641509434,
      "grad_norm": 1.375,
      "learning_rate": 0.0002535861380155185,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22123787999153138,
      "num_tokens": 119054245.0,
      "step": 68630
    },
    {
      "entropy": 5.828171586990356,
      "epoch": 5.340011670881151,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025355854839771805,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22752593755722045,
      "num_tokens": 119062957.0,
      "step": 68635
    },
    {
      "entropy": 5.898113441467285,
      "epoch": 5.340400700252869,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025353095910525753,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2094024732708931,
      "num_tokens": 119073209.0,
      "step": 68640
    },
    {
      "entropy": 5.860943031311035,
      "epoch": 5.3407897296245865,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002535033701385555,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22649920284748076,
      "num_tokens": 119081493.0,
      "step": 68645
    },
    {
      "entropy": 5.901688575744629,
      "epoch": 5.341178758996304,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002534757814980307,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21407009959220885,
      "num_tokens": 119090703.0,
      "step": 68650
    },
    {
      "entropy": 5.957004499435425,
      "epoch": 5.341567788368022,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025344819318410167,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21590329557657242,
      "num_tokens": 119099599.0,
      "step": 68655
    },
    {
      "entropy": 5.952355003356933,
      "epoch": 5.34195681773974,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025342060519718693,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21975032538175582,
      "num_tokens": 119108329.0,
      "step": 68660
    },
    {
      "entropy": 5.900430727005005,
      "epoch": 5.342345847111457,
      "grad_norm": 1.5625,
      "learning_rate": 0.00025339301753770515,
      "loss": 4.933,
      "mean_token_accuracy": 0.2170292168855667,
      "num_tokens": 119117034.0,
      "step": 68665
    },
    {
      "entropy": 5.82502589225769,
      "epoch": 5.342734876483174,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000253365430206075,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2162778303027153,
      "num_tokens": 119126316.0,
      "step": 68670
    },
    {
      "entropy": 5.829318189620972,
      "epoch": 5.343123905854892,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002533378432027149,
      "loss": 4.8528,
      "mean_token_accuracy": 0.22669486105442047,
      "num_tokens": 119134740.0,
      "step": 68675
    },
    {
      "entropy": 5.874153518676758,
      "epoch": 5.3435129352266095,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002533102565280437,
      "loss": 4.898,
      "mean_token_accuracy": 0.22926861494779588,
      "num_tokens": 119143841.0,
      "step": 68680
    },
    {
      "entropy": 5.871552324295044,
      "epoch": 5.343901964598327,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002532826701824797,
      "loss": 4.9057,
      "mean_token_accuracy": 0.21307575851678848,
      "num_tokens": 119151697.0,
      "step": 68685
    },
    {
      "entropy": 5.952570962905884,
      "epoch": 5.344290993970045,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002532550841664416,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2164118230342865,
      "num_tokens": 119160705.0,
      "step": 68690
    },
    {
      "entropy": 5.847532033920288,
      "epoch": 5.344680023341763,
      "grad_norm": 1.5,
      "learning_rate": 0.0002532274984803481,
      "loss": 4.9046,
      "mean_token_accuracy": 0.21675250679254532,
      "num_tokens": 119168969.0,
      "step": 68695
    },
    {
      "entropy": 5.965555620193482,
      "epoch": 5.34506905271348,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002531999131246176,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21198669224977493,
      "num_tokens": 119176948.0,
      "step": 68700
    },
    {
      "entropy": 5.872636938095093,
      "epoch": 5.345458082085197,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002531723280996687,
      "loss": 4.9221,
      "mean_token_accuracy": 0.21458568722009658,
      "num_tokens": 119184978.0,
      "step": 68705
    },
    {
      "entropy": 5.84907546043396,
      "epoch": 5.345847111456915,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025314474340592,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2155911609530449,
      "num_tokens": 119193646.0,
      "step": 68710
    },
    {
      "entropy": 5.88135871887207,
      "epoch": 5.3462361408286325,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025311715904379004,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21183569580316544,
      "num_tokens": 119202733.0,
      "step": 68715
    },
    {
      "entropy": 5.863332223892212,
      "epoch": 5.34662517020035,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025308957501369737,
      "loss": 4.8614,
      "mean_token_accuracy": 0.23205315172672272,
      "num_tokens": 119211312.0,
      "step": 68720
    },
    {
      "entropy": 5.856861639022827,
      "epoch": 5.347014199572068,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025306199131606055,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21553645730018617,
      "num_tokens": 119219874.0,
      "step": 68725
    },
    {
      "entropy": 5.869893026351929,
      "epoch": 5.347403228943786,
      "grad_norm": 1.375,
      "learning_rate": 0.0002530344079512981,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2119730904698372,
      "num_tokens": 119229589.0,
      "step": 68730
    },
    {
      "entropy": 5.84339656829834,
      "epoch": 5.347792258315502,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025300682491982853,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22014718353748322,
      "num_tokens": 119237811.0,
      "step": 68735
    },
    {
      "entropy": 5.881813907623291,
      "epoch": 5.34818128768722,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002529792422220704,
      "loss": 4.91,
      "mean_token_accuracy": 0.22206678986549377,
      "num_tokens": 119245980.0,
      "step": 68740
    },
    {
      "entropy": 5.914845418930054,
      "epoch": 5.348570317058938,
      "grad_norm": 1.375,
      "learning_rate": 0.00025295165985844215,
      "loss": 5.012,
      "mean_token_accuracy": 0.21366143822669983,
      "num_tokens": 119255757.0,
      "step": 68745
    },
    {
      "entropy": 5.812852764129639,
      "epoch": 5.3489593464306555,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002529240778293624,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22038663178682327,
      "num_tokens": 119264446.0,
      "step": 68750
    },
    {
      "entropy": 5.8950036525726315,
      "epoch": 5.349348375802373,
      "grad_norm": 1.484375,
      "learning_rate": 0.00025289649613524954,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2230082258582115,
      "num_tokens": 119272904.0,
      "step": 68755
    },
    {
      "entropy": 5.83568754196167,
      "epoch": 5.349737405174091,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002528689147765223,
      "loss": 4.8585,
      "mean_token_accuracy": 0.22607823461294174,
      "num_tokens": 119281127.0,
      "step": 68760
    },
    {
      "entropy": 5.877138757705689,
      "epoch": 5.350126434545809,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00025284133375359904,
      "loss": 4.8968,
      "mean_token_accuracy": 0.21732288748025894,
      "num_tokens": 119289362.0,
      "step": 68765
    },
    {
      "entropy": 5.819321250915527,
      "epoch": 5.350515463917525,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025281375306689824,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22518203258514405,
      "num_tokens": 119298551.0,
      "step": 68770
    },
    {
      "entropy": 5.889164018630981,
      "epoch": 5.350904493289243,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025278617271683846,
      "loss": 4.8312,
      "mean_token_accuracy": 0.22748586982488633,
      "num_tokens": 119307291.0,
      "step": 68775
    },
    {
      "entropy": 5.916347074508667,
      "epoch": 5.351293522660961,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002527585927038381,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2251664087176323,
      "num_tokens": 119315811.0,
      "step": 68780
    },
    {
      "entropy": 5.893291473388672,
      "epoch": 5.3516825520326785,
      "grad_norm": 1.59375,
      "learning_rate": 0.00025273101302831575,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21012038439512254,
      "num_tokens": 119323936.0,
      "step": 68785
    },
    {
      "entropy": 5.8305823802948,
      "epoch": 5.352071581404396,
      "grad_norm": 1.484375,
      "learning_rate": 0.00025270343369068976,
      "loss": 4.8835,
      "mean_token_accuracy": 0.21994901746511458,
      "num_tokens": 119332752.0,
      "step": 68790
    },
    {
      "entropy": 5.809795188903808,
      "epoch": 5.352460610776114,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002526758546913787,
      "loss": 4.8147,
      "mean_token_accuracy": 0.22716590911149978,
      "num_tokens": 119341591.0,
      "step": 68795
    },
    {
      "entropy": 5.893950748443603,
      "epoch": 5.352849640147831,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002526482760308011,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21745316237211226,
      "num_tokens": 119351190.0,
      "step": 68800
    },
    {
      "entropy": 5.900557565689087,
      "epoch": 5.353238669519548,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025262069770937525,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22687359154224396,
      "num_tokens": 119360306.0,
      "step": 68805
    },
    {
      "entropy": 5.824986314773559,
      "epoch": 5.353627698891266,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002525931197275197,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22688221037387848,
      "num_tokens": 119369196.0,
      "step": 68810
    },
    {
      "entropy": 5.880152940750122,
      "epoch": 5.354016728262984,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025256554208565296,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2100459471344948,
      "num_tokens": 119377671.0,
      "step": 68815
    },
    {
      "entropy": 5.86359133720398,
      "epoch": 5.3544057576347015,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025253796478419323,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2237601950764656,
      "num_tokens": 119386853.0,
      "step": 68820
    },
    {
      "entropy": 5.929996013641357,
      "epoch": 5.354794787006419,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002525103878235593,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21045673936605452,
      "num_tokens": 119395894.0,
      "step": 68825
    },
    {
      "entropy": 5.9023144245147705,
      "epoch": 5.355183816378137,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002524828112041694,
      "loss": 4.9224,
      "mean_token_accuracy": 0.21858870834112168,
      "num_tokens": 119404035.0,
      "step": 68830
    },
    {
      "entropy": 5.919142389297486,
      "epoch": 5.355572845749854,
      "grad_norm": 1.546875,
      "learning_rate": 0.00025245523492644197,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2167753279209137,
      "num_tokens": 119413547.0,
      "step": 68835
    },
    {
      "entropy": 5.932954549789429,
      "epoch": 5.355961875121571,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002524276589907955,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2163742184638977,
      "num_tokens": 119422204.0,
      "step": 68840
    },
    {
      "entropy": 5.919292640686035,
      "epoch": 5.356350904493289,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00025240008339764844,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21403691321611404,
      "num_tokens": 119430858.0,
      "step": 68845
    },
    {
      "entropy": 5.8192863941192625,
      "epoch": 5.356739933865007,
      "grad_norm": 1.359375,
      "learning_rate": 0.000252372508147419,
      "loss": 4.8342,
      "mean_token_accuracy": 0.22788862884044647,
      "num_tokens": 119439525.0,
      "step": 68850
    },
    {
      "entropy": 5.934349822998047,
      "epoch": 5.3571289632367245,
      "grad_norm": 1.5,
      "learning_rate": 0.0002523449332405258,
      "loss": 4.9853,
      "mean_token_accuracy": 0.20827164500951767,
      "num_tokens": 119447131.0,
      "step": 68855
    },
    {
      "entropy": 5.874726390838623,
      "epoch": 5.357517992608442,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002523173586773872,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21438563019037246,
      "num_tokens": 119455599.0,
      "step": 68860
    },
    {
      "entropy": 5.879022264480591,
      "epoch": 5.35790702198016,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002522897844584216,
      "loss": 4.9566,
      "mean_token_accuracy": 0.21701999604701996,
      "num_tokens": 119463786.0,
      "step": 68865
    },
    {
      "entropy": 5.908893871307373,
      "epoch": 5.358296051351877,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025226221058404733,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22020077854394912,
      "num_tokens": 119472432.0,
      "step": 68870
    },
    {
      "entropy": 5.922171783447266,
      "epoch": 5.358685080723594,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002522346370546828,
      "loss": 4.982,
      "mean_token_accuracy": 0.216733418405056,
      "num_tokens": 119482020.0,
      "step": 68875
    },
    {
      "entropy": 5.880100440979004,
      "epoch": 5.359074110095312,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002522070638707465,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21981246620416642,
      "num_tokens": 119490150.0,
      "step": 68880
    },
    {
      "entropy": 5.888078212738037,
      "epoch": 5.35946313946703,
      "grad_norm": 1.390625,
      "learning_rate": 0.00025217949103265665,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21875652223825454,
      "num_tokens": 119498463.0,
      "step": 68885
    },
    {
      "entropy": 5.892721891403198,
      "epoch": 5.3598521688387475,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025215191854083166,
      "loss": 5.027,
      "mean_token_accuracy": 0.2122641071677208,
      "num_tokens": 119507429.0,
      "step": 68890
    },
    {
      "entropy": 5.858903551101685,
      "epoch": 5.360241198210465,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025212434639568996,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2265573501586914,
      "num_tokens": 119516666.0,
      "step": 68895
    },
    {
      "entropy": 5.89514741897583,
      "epoch": 5.360630227582183,
      "grad_norm": 1.40625,
      "learning_rate": 0.00025209677459765,
      "loss": 4.8884,
      "mean_token_accuracy": 0.2195172354578972,
      "num_tokens": 119525842.0,
      "step": 68900
    },
    {
      "entropy": 5.8139424324035645,
      "epoch": 5.3610192569539,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002520692031471299,
      "loss": 4.758,
      "mean_token_accuracy": 0.23216582387685775,
      "num_tokens": 119534225.0,
      "step": 68905
    },
    {
      "entropy": 5.904566049575806,
      "epoch": 5.361408286325617,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025204163204454826,
      "loss": 4.935,
      "mean_token_accuracy": 0.21795859038829804,
      "num_tokens": 119542146.0,
      "step": 68910
    },
    {
      "entropy": 5.832619714736938,
      "epoch": 5.361797315697335,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002520140612903233,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2214763343334198,
      "num_tokens": 119550289.0,
      "step": 68915
    },
    {
      "entropy": 5.840014123916626,
      "epoch": 5.362186345069053,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002519864908848733,
      "loss": 4.8411,
      "mean_token_accuracy": 0.22868757843971252,
      "num_tokens": 119558710.0,
      "step": 68920
    },
    {
      "entropy": 5.854184150695801,
      "epoch": 5.3625753744407705,
      "grad_norm": 1.609375,
      "learning_rate": 0.00025195892082861674,
      "loss": 4.8872,
      "mean_token_accuracy": 0.2257773369550705,
      "num_tokens": 119567702.0,
      "step": 68925
    },
    {
      "entropy": 5.846275711059571,
      "epoch": 5.362964403812488,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002519313511219719,
      "loss": 4.8692,
      "mean_token_accuracy": 0.21893164217472078,
      "num_tokens": 119576197.0,
      "step": 68930
    },
    {
      "entropy": 5.853884601593018,
      "epoch": 5.363353433184205,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025190378176535703,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22229641526937485,
      "num_tokens": 119584624.0,
      "step": 68935
    },
    {
      "entropy": 5.91581654548645,
      "epoch": 5.363742462555923,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002518762127591905,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21773041486740113,
      "num_tokens": 119593741.0,
      "step": 68940
    },
    {
      "entropy": 5.842016410827637,
      "epoch": 5.36413149192764,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002518486441038907,
      "loss": 4.8922,
      "mean_token_accuracy": 0.21947795897722244,
      "num_tokens": 119602288.0,
      "step": 68945
    },
    {
      "entropy": 5.898153352737427,
      "epoch": 5.364520521299358,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002518210757998758,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21633385270833969,
      "num_tokens": 119610787.0,
      "step": 68950
    },
    {
      "entropy": 5.890827369689942,
      "epoch": 5.364909550671076,
      "grad_norm": 1.375,
      "learning_rate": 0.0002517935078475642,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22389369308948517,
      "num_tokens": 119619175.0,
      "step": 68955
    },
    {
      "entropy": 5.81711163520813,
      "epoch": 5.3652985800427935,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002517659402473742,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22528464198112488,
      "num_tokens": 119627828.0,
      "step": 68960
    },
    {
      "entropy": 5.800579404830932,
      "epoch": 5.365687609414511,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002517383729997241,
      "loss": 4.7721,
      "mean_token_accuracy": 0.23346325159072875,
      "num_tokens": 119635639.0,
      "step": 68965
    },
    {
      "entropy": 5.845604419708252,
      "epoch": 5.366076638786228,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002517108061050321,
      "loss": 4.923,
      "mean_token_accuracy": 0.21171995401382446,
      "num_tokens": 119644807.0,
      "step": 68970
    },
    {
      "entropy": 5.866562175750732,
      "epoch": 5.366465668157946,
      "grad_norm": 1.375,
      "learning_rate": 0.00025168323956371657,
      "loss": 4.8495,
      "mean_token_accuracy": 0.22503067553043365,
      "num_tokens": 119653409.0,
      "step": 68975
    },
    {
      "entropy": 5.82502121925354,
      "epoch": 5.366854697529663,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025165567337619583,
      "loss": 4.8233,
      "mean_token_accuracy": 0.2237764850258827,
      "num_tokens": 119661703.0,
      "step": 68980
    },
    {
      "entropy": 5.994072723388672,
      "epoch": 5.367243726901381,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000251628107542888,
      "loss": 5.0267,
      "mean_token_accuracy": 0.213287852704525,
      "num_tokens": 119669853.0,
      "step": 68985
    },
    {
      "entropy": 5.9194553852081295,
      "epoch": 5.367632756273099,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00025160054206421143,
      "loss": 5.0075,
      "mean_token_accuracy": 0.207281956076622,
      "num_tokens": 119679035.0,
      "step": 68990
    },
    {
      "entropy": 5.8558997631073,
      "epoch": 5.3680217856448165,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002515729769405845,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21935223042964935,
      "num_tokens": 119687213.0,
      "step": 68995
    },
    {
      "entropy": 5.877433490753174,
      "epoch": 5.368410815016533,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002515454121724253,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21387919783592224,
      "num_tokens": 119695751.0,
      "step": 69000
    },
    {
      "epoch": 5.368410815016533,
      "eval_entropy": 5.566619871853635,
      "eval_loss": 5.029155731201172,
      "eval_mean_token_accuracy": 0.22246949281740694,
      "eval_num_tokens": 119695751.0,
      "eval_runtime": 22.0426,
      "eval_samples_per_second": 1885.349,
      "eval_steps_per_second": 235.68,
      "step": 69000
    },
    {
      "entropy": 5.877942895889282,
      "epoch": 5.368799844388251,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025151784776015216,
      "loss": 4.8869,
      "mean_token_accuracy": 0.21971699148416518,
      "num_tokens": 119704242.0,
      "step": 69005
    },
    {
      "entropy": 5.812563419342041,
      "epoch": 5.369188873759969,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002514902837041833,
      "loss": 4.7782,
      "mean_token_accuracy": 0.22920262962579727,
      "num_tokens": 119712677.0,
      "step": 69010
    },
    {
      "entropy": 5.849329137802124,
      "epoch": 5.369577903131686,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025146272000493695,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2273086354136467,
      "num_tokens": 119721594.0,
      "step": 69015
    },
    {
      "entropy": 5.90934190750122,
      "epoch": 5.369966932503404,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002514351566628313,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22464912086725236,
      "num_tokens": 119729917.0,
      "step": 69020
    },
    {
      "entropy": 5.832656192779541,
      "epoch": 5.370355961875122,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002514075936782847,
      "loss": 4.875,
      "mean_token_accuracy": 0.22217466980218886,
      "num_tokens": 119737940.0,
      "step": 69025
    },
    {
      "entropy": 5.865193557739258,
      "epoch": 5.3707449912468395,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002513800310517153,
      "loss": 4.9694,
      "mean_token_accuracy": 0.21317492425441742,
      "num_tokens": 119746843.0,
      "step": 69030
    },
    {
      "entropy": 5.817890930175781,
      "epoch": 5.371134020618557,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002513524687835414,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2206287369132042,
      "num_tokens": 119754697.0,
      "step": 69035
    },
    {
      "entropy": 5.955285739898682,
      "epoch": 5.371523049990274,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002513249068741811,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20849117785692214,
      "num_tokens": 119763005.0,
      "step": 69040
    },
    {
      "entropy": 5.8645713329315186,
      "epoch": 5.371912079361992,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025129734532405265,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22328405678272248,
      "num_tokens": 119771319.0,
      "step": 69045
    },
    {
      "entropy": 5.826034164428711,
      "epoch": 5.372301108733709,
      "grad_norm": 1.484375,
      "learning_rate": 0.00025126978413357424,
      "loss": 4.8743,
      "mean_token_accuracy": 0.21972937136888504,
      "num_tokens": 119780555.0,
      "step": 69050
    },
    {
      "entropy": 5.853900051116943,
      "epoch": 5.372690138105427,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002512422233031641,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2285275712609291,
      "num_tokens": 119789116.0,
      "step": 69055
    },
    {
      "entropy": 5.8706118106842045,
      "epoch": 5.373079167477145,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002512146628332404,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2122945874929428,
      "num_tokens": 119797737.0,
      "step": 69060
    },
    {
      "entropy": 5.879394006729126,
      "epoch": 5.3734681968488625,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002511871027242214,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2241018831729889,
      "num_tokens": 119805716.0,
      "step": 69065
    },
    {
      "entropy": 5.896802520751953,
      "epoch": 5.373857226220579,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002511595429765251,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2214080035686493,
      "num_tokens": 119815538.0,
      "step": 69070
    },
    {
      "entropy": 5.905560493469238,
      "epoch": 5.374246255592297,
      "grad_norm": 1.515625,
      "learning_rate": 0.00025113198359056986,
      "loss": 4.954,
      "mean_token_accuracy": 0.21926838159561157,
      "num_tokens": 119824775.0,
      "step": 69075
    },
    {
      "entropy": 5.912995862960815,
      "epoch": 5.374635284964015,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002511044245667738,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22192323058843613,
      "num_tokens": 119833454.0,
      "step": 69080
    },
    {
      "entropy": 5.930603647232056,
      "epoch": 5.375024314335732,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000251076865905555,
      "loss": 5.0155,
      "mean_token_accuracy": 0.20840980112552643,
      "num_tokens": 119841606.0,
      "step": 69085
    },
    {
      "entropy": 5.911882781982422,
      "epoch": 5.37541334370745,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025104930760733163,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21499580293893814,
      "num_tokens": 119849618.0,
      "step": 69090
    },
    {
      "entropy": 5.89327187538147,
      "epoch": 5.375802373079168,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000251021749672522,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2206882655620575,
      "num_tokens": 119858453.0,
      "step": 69095
    },
    {
      "entropy": 5.973142385482788,
      "epoch": 5.3761914024508854,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025099419210154413,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21770086288452148,
      "num_tokens": 119867607.0,
      "step": 69100
    },
    {
      "entropy": 5.844433307647705,
      "epoch": 5.376580431822602,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002509666348948162,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21360281258821487,
      "num_tokens": 119876498.0,
      "step": 69105
    },
    {
      "entropy": 5.921897268295288,
      "epoch": 5.37696946119432,
      "grad_norm": 1.5625,
      "learning_rate": 0.00025093907805275636,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2185054525732994,
      "num_tokens": 119884229.0,
      "step": 69110
    },
    {
      "entropy": 5.900941944122314,
      "epoch": 5.377358490566038,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025091152157578266,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21984490901231765,
      "num_tokens": 119893076.0,
      "step": 69115
    },
    {
      "entropy": 5.889987802505493,
      "epoch": 5.377747519937755,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002508839654643133,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22053925544023514,
      "num_tokens": 119901847.0,
      "step": 69120
    },
    {
      "entropy": 5.899180316925049,
      "epoch": 5.378136549309473,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002508564097187664,
      "loss": 4.9194,
      "mean_token_accuracy": 0.22105727046728135,
      "num_tokens": 119910745.0,
      "step": 69125
    },
    {
      "entropy": 5.873424100875854,
      "epoch": 5.378525578681191,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00025082885433956007,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22316674143075943,
      "num_tokens": 119920100.0,
      "step": 69130
    },
    {
      "entropy": 5.863775253295898,
      "epoch": 5.3789146080529076,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025080129932711234,
      "loss": 4.7944,
      "mean_token_accuracy": 0.2347757861018181,
      "num_tokens": 119928792.0,
      "step": 69135
    },
    {
      "entropy": 5.861487150192261,
      "epoch": 5.379303637424625,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002507737446818414,
      "loss": 4.883,
      "mean_token_accuracy": 0.2249917671084404,
      "num_tokens": 119937536.0,
      "step": 69140
    },
    {
      "entropy": 5.909967279434204,
      "epoch": 5.379692666796343,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002507461904041653,
      "loss": 5.0112,
      "mean_token_accuracy": 0.20842734575271607,
      "num_tokens": 119946441.0,
      "step": 69145
    },
    {
      "entropy": 5.88842887878418,
      "epoch": 5.380081696168061,
      "grad_norm": 1.421875,
      "learning_rate": 0.00025071863649450216,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22013828307390212,
      "num_tokens": 119954796.0,
      "step": 69150
    },
    {
      "entropy": 5.924608755111694,
      "epoch": 5.380470725539778,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025069108295327006,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2157398819923401,
      "num_tokens": 119964262.0,
      "step": 69155
    },
    {
      "entropy": 5.900449895858765,
      "epoch": 5.380859754911496,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025066352978088706,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2225585788488388,
      "num_tokens": 119973386.0,
      "step": 69160
    },
    {
      "entropy": 5.9590412139892575,
      "epoch": 5.381248784283214,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002506359769777713,
      "loss": 4.959,
      "mean_token_accuracy": 0.22015082985162734,
      "num_tokens": 119981816.0,
      "step": 69165
    },
    {
      "entropy": 5.863586521148681,
      "epoch": 5.3816378136549305,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002506084245443408,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21485460698604583,
      "num_tokens": 119990352.0,
      "step": 69170
    },
    {
      "entropy": 5.907462167739868,
      "epoch": 5.382026843026648,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002505808724810137,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21606376767158508,
      "num_tokens": 119998648.0,
      "step": 69175
    },
    {
      "entropy": 5.829870986938476,
      "epoch": 5.382415872398366,
      "grad_norm": 1.6875,
      "learning_rate": 0.00025055332078820794,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2289383053779602,
      "num_tokens": 120006395.0,
      "step": 69180
    },
    {
      "entropy": 5.8599165916442875,
      "epoch": 5.382804901770084,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002505257694663416,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2160799279808998,
      "num_tokens": 120015425.0,
      "step": 69185
    },
    {
      "entropy": 5.9029261589050295,
      "epoch": 5.383193931141801,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002504982185158328,
      "loss": 4.976,
      "mean_token_accuracy": 0.21001326888799668,
      "num_tokens": 120023864.0,
      "step": 69190
    },
    {
      "entropy": 5.93074312210083,
      "epoch": 5.383582960513519,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002504706679370995,
      "loss": 4.9827,
      "mean_token_accuracy": 0.212491175532341,
      "num_tokens": 120032569.0,
      "step": 69195
    },
    {
      "entropy": 5.880759620666504,
      "epoch": 5.383971989885237,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002504431177305598,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21399974226951599,
      "num_tokens": 120041024.0,
      "step": 69200
    },
    {
      "entropy": 5.879503345489502,
      "epoch": 5.3843610192569535,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002504155678966317,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21851942092180252,
      "num_tokens": 120050469.0,
      "step": 69205
    },
    {
      "entropy": 5.877401828765869,
      "epoch": 5.384750048628671,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002503880184357333,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2229369178414345,
      "num_tokens": 120058854.0,
      "step": 69210
    },
    {
      "entropy": 5.845195341110229,
      "epoch": 5.385139078000389,
      "grad_norm": 1.515625,
      "learning_rate": 0.00025036046934828244,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2297510549426079,
      "num_tokens": 120066961.0,
      "step": 69215
    },
    {
      "entropy": 5.879139375686646,
      "epoch": 5.385528107372107,
      "grad_norm": 1.5625,
      "learning_rate": 0.00025033292063469714,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21203892230987548,
      "num_tokens": 120075967.0,
      "step": 69220
    },
    {
      "entropy": 5.9994566440582275,
      "epoch": 5.385917136743824,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00025030537229539567,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21765843480825425,
      "num_tokens": 120084190.0,
      "step": 69225
    },
    {
      "entropy": 5.9163719654083256,
      "epoch": 5.386306166115542,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002502778243307958,
      "loss": 4.9893,
      "mean_token_accuracy": 0.20981699377298355,
      "num_tokens": 120092533.0,
      "step": 69230
    },
    {
      "entropy": 5.879611015319824,
      "epoch": 5.38669519548726,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025025027674131567,
      "loss": 4.957,
      "mean_token_accuracy": 0.2219676449894905,
      "num_tokens": 120100569.0,
      "step": 69235
    },
    {
      "entropy": 5.893147039413452,
      "epoch": 5.3870842248589765,
      "grad_norm": 1.5,
      "learning_rate": 0.00025022272952737317,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21738879680633544,
      "num_tokens": 120109531.0,
      "step": 69240
    },
    {
      "entropy": 5.988116073608398,
      "epoch": 5.387473254230694,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002501951826893863,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2205149605870247,
      "num_tokens": 120118419.0,
      "step": 69245
    },
    {
      "entropy": 5.859396934509277,
      "epoch": 5.387862283602412,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002501676362277731,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22374484986066817,
      "num_tokens": 120127554.0,
      "step": 69250
    },
    {
      "entropy": 5.8123798847198485,
      "epoch": 5.38825131297413,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025014009014295136,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22404297292232514,
      "num_tokens": 120136076.0,
      "step": 69255
    },
    {
      "entropy": 5.807517862319946,
      "epoch": 5.388640342345847,
      "grad_norm": 1.5,
      "learning_rate": 0.0002501125444353393,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23198066502809525,
      "num_tokens": 120144295.0,
      "step": 69260
    },
    {
      "entropy": 5.8630022525787355,
      "epoch": 5.389029371717565,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002500849991053547,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22258503139019012,
      "num_tokens": 120153190.0,
      "step": 69265
    },
    {
      "entropy": 5.823506641387939,
      "epoch": 5.389418401089282,
      "grad_norm": 1.359375,
      "learning_rate": 0.00025005745415341565,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22435268014669418,
      "num_tokens": 120161683.0,
      "step": 69270
    },
    {
      "entropy": 5.877452182769775,
      "epoch": 5.3898074304609995,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00025002990957994,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2179741308093071,
      "num_tokens": 120170133.0,
      "step": 69275
    },
    {
      "entropy": 5.862102317810058,
      "epoch": 5.390196459832717,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00025000236538534577,
      "loss": 4.9321,
      "mean_token_accuracy": 0.217137710750103,
      "num_tokens": 120179151.0,
      "step": 69280
    },
    {
      "entropy": 5.894723415374756,
      "epoch": 5.390585489204435,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024997482157005074,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21530358195304872,
      "num_tokens": 120187425.0,
      "step": 69285
    },
    {
      "entropy": 5.861607074737549,
      "epoch": 5.390974518576153,
      "grad_norm": 1.484375,
      "learning_rate": 0.000249947278134473,
      "loss": 4.876,
      "mean_token_accuracy": 0.2241993859410286,
      "num_tokens": 120196166.0,
      "step": 69290
    },
    {
      "entropy": 5.866091632843018,
      "epoch": 5.39136354794787,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002499197350790305,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22650136798620224,
      "num_tokens": 120204889.0,
      "step": 69295
    },
    {
      "entropy": 5.849116563796997,
      "epoch": 5.391752577319588,
      "grad_norm": 1.625,
      "learning_rate": 0.00024989219240414105,
      "loss": 4.834,
      "mean_token_accuracy": 0.2311936065554619,
      "num_tokens": 120212984.0,
      "step": 69300
    },
    {
      "entropy": 5.87714467048645,
      "epoch": 5.392141606691305,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00024986465011022263,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21252454221248626,
      "num_tokens": 120221147.0,
      "step": 69305
    },
    {
      "entropy": 5.8738504409790036,
      "epoch": 5.3925306360630225,
      "grad_norm": 1.375,
      "learning_rate": 0.00024983710819769325,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21855110675096512,
      "num_tokens": 120230105.0,
      "step": 69310
    },
    {
      "entropy": 5.86072268486023,
      "epoch": 5.39291966543474,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002498095666669706,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22696622163057328,
      "num_tokens": 120239361.0,
      "step": 69315
    },
    {
      "entropy": 5.823500490188598,
      "epoch": 5.393308694806458,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002497820255184727,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22716987878084183,
      "num_tokens": 120248154.0,
      "step": 69320
    },
    {
      "entropy": 5.845702695846557,
      "epoch": 5.393697724178176,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024975448475261737,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22400232255458832,
      "num_tokens": 120256033.0,
      "step": 69325
    },
    {
      "entropy": 5.857286882400513,
      "epoch": 5.394086753549893,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024972694436982255,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21649876087903977,
      "num_tokens": 120265269.0,
      "step": 69330
    },
    {
      "entropy": 5.882098913192749,
      "epoch": 5.39447578292161,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002496994043705062,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2217616245150566,
      "num_tokens": 120274472.0,
      "step": 69335
    },
    {
      "entropy": 5.929987859725952,
      "epoch": 5.394864812293328,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002496718647550861,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22107165157794953,
      "num_tokens": 120283833.0,
      "step": 69340
    },
    {
      "entropy": 5.938629007339477,
      "epoch": 5.3952538416650455,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002496443255239801,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22974551767110823,
      "num_tokens": 120292340.0,
      "step": 69345
    },
    {
      "entropy": 5.8549083232879635,
      "epoch": 5.395642871036763,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002496167866776061,
      "loss": 4.8469,
      "mean_token_accuracy": 0.22552416026592254,
      "num_tokens": 120300505.0,
      "step": 69350
    },
    {
      "entropy": 5.903234958648682,
      "epoch": 5.396031900408481,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024958924821638195,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22066121995449067,
      "num_tokens": 120308938.0,
      "step": 69355
    },
    {
      "entropy": 5.851533985137939,
      "epoch": 5.396420929780199,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024956171014072555,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2212996855378151,
      "num_tokens": 120318282.0,
      "step": 69360
    },
    {
      "entropy": 5.926537799835205,
      "epoch": 5.396809959151916,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024953417245105467,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2087308332324028,
      "num_tokens": 120327414.0,
      "step": 69365
    },
    {
      "entropy": 5.926514339447022,
      "epoch": 5.397198988523634,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00024950663514778726,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22006155848503112,
      "num_tokens": 120334794.0,
      "step": 69370
    },
    {
      "entropy": 5.9539604663848875,
      "epoch": 5.397588017895351,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00024947909823134107,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2159801244735718,
      "num_tokens": 120343405.0,
      "step": 69375
    },
    {
      "entropy": 5.773233413696289,
      "epoch": 5.3979770472670685,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002494515617021339,
      "loss": 4.8559,
      "mean_token_accuracy": 0.22582780420780182,
      "num_tokens": 120351720.0,
      "step": 69380
    },
    {
      "entropy": 5.892366504669189,
      "epoch": 5.398366076638786,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00024942402556058364,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2203717902302742,
      "num_tokens": 120360291.0,
      "step": 69385
    },
    {
      "entropy": 5.923715209960937,
      "epoch": 5.398755106010504,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002493964898071081,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21336133033037186,
      "num_tokens": 120369089.0,
      "step": 69390
    },
    {
      "entropy": 5.854427766799927,
      "epoch": 5.399144135382222,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002493689544421251,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22352437525987626,
      "num_tokens": 120378182.0,
      "step": 69395
    },
    {
      "entropy": 5.832686471939087,
      "epoch": 5.399533164753939,
      "grad_norm": 1.40625,
      "learning_rate": 0.00024934141946605237,
      "loss": 4.8753,
      "mean_token_accuracy": 0.21651578992605208,
      "num_tokens": 120386724.0,
      "step": 69400
    },
    {
      "entropy": 5.904237747192383,
      "epoch": 5.399922194125656,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00024931388487930777,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21638444662094117,
      "num_tokens": 120394889.0,
      "step": 69405
    },
    {
      "entropy": 5.872105503082276,
      "epoch": 5.400311223497374,
      "grad_norm": 1.5,
      "learning_rate": 0.000249286350682309,
      "loss": 4.93,
      "mean_token_accuracy": 0.2185324802994728,
      "num_tokens": 120403211.0,
      "step": 69410
    },
    {
      "entropy": 5.851849985122681,
      "epoch": 5.4007002528690915,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002492588168754741,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22184375822544097,
      "num_tokens": 120412539.0,
      "step": 69415
    },
    {
      "entropy": 5.807697868347168,
      "epoch": 5.401089282240809,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024923128345922055,
      "loss": 4.8246,
      "mean_token_accuracy": 0.2274965077638626,
      "num_tokens": 120421316.0,
      "step": 69420
    },
    {
      "entropy": 5.825650215148926,
      "epoch": 5.401478311612527,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002492037504339663,
      "loss": 4.8664,
      "mean_token_accuracy": 0.22584414035081862,
      "num_tokens": 120430146.0,
      "step": 69425
    },
    {
      "entropy": 5.921231985092163,
      "epoch": 5.401867340984245,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024917621780012905,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2142064243555069,
      "num_tokens": 120438881.0,
      "step": 69430
    },
    {
      "entropy": 5.941400194168091,
      "epoch": 5.402256370355962,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024914868555812663,
      "loss": 4.9601,
      "mean_token_accuracy": 0.21080868244171141,
      "num_tokens": 120447060.0,
      "step": 69435
    },
    {
      "entropy": 5.85520191192627,
      "epoch": 5.402645399727679,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002491211537083768,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22535262405872344,
      "num_tokens": 120455053.0,
      "step": 69440
    },
    {
      "entropy": 5.806003904342651,
      "epoch": 5.403034429099397,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002490936222512972,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21925103813409805,
      "num_tokens": 120463161.0,
      "step": 69445
    },
    {
      "entropy": 5.8722187042236325,
      "epoch": 5.4034234584711145,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002490660911873057,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2107679381966591,
      "num_tokens": 120472242.0,
      "step": 69450
    },
    {
      "entropy": 5.916609477996826,
      "epoch": 5.403812487842832,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024903856051681995,
      "loss": 4.9097,
      "mean_token_accuracy": 0.21595146507024765,
      "num_tokens": 120480711.0,
      "step": 69455
    },
    {
      "entropy": 5.9163306713104244,
      "epoch": 5.40420151721455,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002490110302402577,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21871328800916673,
      "num_tokens": 120489798.0,
      "step": 69460
    },
    {
      "entropy": 5.868837213516235,
      "epoch": 5.404590546586268,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002489835003580367,
      "loss": 4.9493,
      "mean_token_accuracy": 0.20643870830535888,
      "num_tokens": 120498294.0,
      "step": 69465
    },
    {
      "entropy": 5.8625524044036865,
      "epoch": 5.404979575957984,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024895597087057483,
      "loss": 4.9526,
      "mean_token_accuracy": 0.2208401158452034,
      "num_tokens": 120506711.0,
      "step": 69470
    },
    {
      "entropy": 5.798846864700318,
      "epoch": 5.405368605329702,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002489284417782895,
      "loss": 4.8696,
      "mean_token_accuracy": 0.21850617229938507,
      "num_tokens": 120515241.0,
      "step": 69475
    },
    {
      "entropy": 5.87580509185791,
      "epoch": 5.40575763470142,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002489009130815986,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22165576070547105,
      "num_tokens": 120524523.0,
      "step": 69480
    },
    {
      "entropy": 5.883918380737304,
      "epoch": 5.4061466640731375,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002488733847809197,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22820912748575212,
      "num_tokens": 120533384.0,
      "step": 69485
    },
    {
      "entropy": 5.856443548202515,
      "epoch": 5.406535693444855,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002488458568766707,
      "loss": 4.8631,
      "mean_token_accuracy": 0.21724938601255417,
      "num_tokens": 120540965.0,
      "step": 69490
    },
    {
      "entropy": 5.8871289730072025,
      "epoch": 5.406924722816573,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002488183293692692,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2185433253645897,
      "num_tokens": 120549905.0,
      "step": 69495
    },
    {
      "entropy": 5.843818235397339,
      "epoch": 5.407313752188291,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002487908022591329,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22395817041397095,
      "num_tokens": 120558547.0,
      "step": 69500
    },
    {
      "entropy": 5.865789127349854,
      "epoch": 5.407702781560007,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024876327554667943,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2228495016694069,
      "num_tokens": 120567086.0,
      "step": 69505
    },
    {
      "entropy": 5.842902612686157,
      "epoch": 5.408091810931725,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002487357492323265,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2165142998099327,
      "num_tokens": 120575767.0,
      "step": 69510
    },
    {
      "entropy": 5.810247564315796,
      "epoch": 5.408480840303443,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002487082233164918,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22351898550987243,
      "num_tokens": 120584952.0,
      "step": 69515
    },
    {
      "entropy": 5.896504354476929,
      "epoch": 5.4088698696751605,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002486806977995929,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21054230630397797,
      "num_tokens": 120594221.0,
      "step": 69520
    },
    {
      "entropy": 5.849171209335327,
      "epoch": 5.409258899046878,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002486531726820476,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23185864835977554,
      "num_tokens": 120602533.0,
      "step": 69525
    },
    {
      "entropy": 5.832099199295044,
      "epoch": 5.409647928418596,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024862564796427337,
      "loss": 4.844,
      "mean_token_accuracy": 0.22502754777669906,
      "num_tokens": 120611534.0,
      "step": 69530
    },
    {
      "entropy": 5.824452447891235,
      "epoch": 5.410036957790314,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000248598123646688,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22650086283683776,
      "num_tokens": 120620357.0,
      "step": 69535
    },
    {
      "entropy": 5.858800315856934,
      "epoch": 5.41042598716203,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002485705997297091,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22339128255844115,
      "num_tokens": 120628940.0,
      "step": 69540
    },
    {
      "entropy": 5.883877992630005,
      "epoch": 5.410815016533748,
      "grad_norm": 1.453125,
      "learning_rate": 0.00024854307621375427,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2207343652844429,
      "num_tokens": 120637872.0,
      "step": 69545
    },
    {
      "entropy": 5.878601503372193,
      "epoch": 5.411204045905466,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002485155530992411,
      "loss": 4.9077,
      "mean_token_accuracy": 0.21310213655233384,
      "num_tokens": 120646326.0,
      "step": 69550
    },
    {
      "entropy": 5.948483180999756,
      "epoch": 5.4115930752771835,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002484880303865873,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21148795038461685,
      "num_tokens": 120654689.0,
      "step": 69555
    },
    {
      "entropy": 5.874451923370361,
      "epoch": 5.411982104648901,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002484605080762104,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23045978993177413,
      "num_tokens": 120662656.0,
      "step": 69560
    },
    {
      "entropy": 5.812439966201782,
      "epoch": 5.412371134020619,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002484329861685281,
      "loss": 4.8476,
      "mean_token_accuracy": 0.2201933056116104,
      "num_tokens": 120671493.0,
      "step": 69565
    },
    {
      "entropy": 5.839184188842774,
      "epoch": 5.412760163392337,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024840546466395796,
      "loss": 4.9324,
      "mean_token_accuracy": 0.21320036947727203,
      "num_tokens": 120680276.0,
      "step": 69570
    },
    {
      "entropy": 5.912174797058105,
      "epoch": 5.413149192764053,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024837794356291756,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21608465760946274,
      "num_tokens": 120689115.0,
      "step": 69575
    },
    {
      "entropy": 5.9935235500335695,
      "epoch": 5.413538222135771,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002483504228658245,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2139122426509857,
      "num_tokens": 120697550.0,
      "step": 69580
    },
    {
      "entropy": 5.9713341236114506,
      "epoch": 5.413927251507489,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002483229025730963,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20711686909198762,
      "num_tokens": 120706629.0,
      "step": 69585
    },
    {
      "entropy": 5.903969049453735,
      "epoch": 5.4143162808792065,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002482953826851506,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2100489154458046,
      "num_tokens": 120715538.0,
      "step": 69590
    },
    {
      "entropy": 5.976001787185669,
      "epoch": 5.414705310250924,
      "grad_norm": 1.5625,
      "learning_rate": 0.000248267863202405,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21625385880470277,
      "num_tokens": 120723861.0,
      "step": 69595
    },
    {
      "entropy": 5.883838701248169,
      "epoch": 5.415094339622642,
      "grad_norm": 1.75,
      "learning_rate": 0.00024824034412527704,
      "loss": 4.903,
      "mean_token_accuracy": 0.22541527450084686,
      "num_tokens": 120732499.0,
      "step": 69600
    },
    {
      "entropy": 5.8660331726074215,
      "epoch": 5.415483368994359,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002482128254541842,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21767235547304153,
      "num_tokens": 120741436.0,
      "step": 69605
    },
    {
      "entropy": 5.969306325912475,
      "epoch": 5.415872398366076,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002481853071895442,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21886663585901261,
      "num_tokens": 120750079.0,
      "step": 69610
    },
    {
      "entropy": 5.896447467803955,
      "epoch": 5.416261427737794,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002481577893317744,
      "loss": 4.901,
      "mean_token_accuracy": 0.21875064671039582,
      "num_tokens": 120757843.0,
      "step": 69615
    },
    {
      "entropy": 5.847191429138183,
      "epoch": 5.416650457109512,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002481302718812924,
      "loss": 4.8612,
      "mean_token_accuracy": 0.22933880239725113,
      "num_tokens": 120766308.0,
      "step": 69620
    },
    {
      "entropy": 5.831346893310547,
      "epoch": 5.4170394864812295,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002481027548385158,
      "loss": 4.877,
      "mean_token_accuracy": 0.2182373061776161,
      "num_tokens": 120775717.0,
      "step": 69625
    },
    {
      "entropy": 5.837592267990113,
      "epoch": 5.417428515852947,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024807523820386207,
      "loss": 4.8999,
      "mean_token_accuracy": 0.21881067305803298,
      "num_tokens": 120784112.0,
      "step": 69630
    },
    {
      "entropy": 5.91919584274292,
      "epoch": 5.417817545224665,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002480477219777488,
      "loss": 5.0216,
      "mean_token_accuracy": 0.217192941904068,
      "num_tokens": 120793812.0,
      "step": 69635
    },
    {
      "entropy": 5.905784225463867,
      "epoch": 5.418206574596382,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002480202061605933,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2232805609703064,
      "num_tokens": 120802564.0,
      "step": 69640
    },
    {
      "entropy": 5.844201612472534,
      "epoch": 5.418595603968099,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002479926907528134,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2213527664542198,
      "num_tokens": 120811113.0,
      "step": 69645
    },
    {
      "entropy": 5.807879161834717,
      "epoch": 5.418984633339817,
      "grad_norm": 1.421875,
      "learning_rate": 0.00024796517575482633,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22408051639795304,
      "num_tokens": 120820498.0,
      "step": 69650
    },
    {
      "entropy": 5.89239912033081,
      "epoch": 5.419373662711535,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002479376611670498,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2239508405327797,
      "num_tokens": 120829510.0,
      "step": 69655
    },
    {
      "entropy": 5.878476572036743,
      "epoch": 5.4197626920832525,
      "grad_norm": 1.421875,
      "learning_rate": 0.00024791014698990116,
      "loss": 4.9467,
      "mean_token_accuracy": 0.21319446414709092,
      "num_tokens": 120838314.0,
      "step": 69660
    },
    {
      "entropy": 5.891367673873901,
      "epoch": 5.42015172145497,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00024788263322379783,
      "loss": 5.0096,
      "mean_token_accuracy": 0.20777628421783448,
      "num_tokens": 120847373.0,
      "step": 69665
    },
    {
      "entropy": 5.900431537628174,
      "epoch": 5.420540750826687,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024785511986915746,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22088544070720673,
      "num_tokens": 120856237.0,
      "step": 69670
    },
    {
      "entropy": 5.857255458831787,
      "epoch": 5.420929780198405,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002478276069263974,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2104274421930313,
      "num_tokens": 120864865.0,
      "step": 69675
    },
    {
      "entropy": 5.7898133277893065,
      "epoch": 5.421318809570122,
      "grad_norm": 1.40625,
      "learning_rate": 0.00024780009439593514,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2245606392621994,
      "num_tokens": 120874040.0,
      "step": 69680
    },
    {
      "entropy": 5.9104047298431395,
      "epoch": 5.42170783894184,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002477725822781881,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2222105845808983,
      "num_tokens": 120882695.0,
      "step": 69685
    },
    {
      "entropy": 5.917834281921387,
      "epoch": 5.422096868313558,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002477450705735739,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2137506902217865,
      "num_tokens": 120891256.0,
      "step": 69690
    },
    {
      "entropy": 5.905517482757569,
      "epoch": 5.4224858976852754,
      "grad_norm": 1.328125,
      "learning_rate": 0.00024771755928250984,
      "loss": 4.904,
      "mean_token_accuracy": 0.22329799085855484,
      "num_tokens": 120900122.0,
      "step": 69695
    },
    {
      "entropy": 5.987349128723144,
      "epoch": 5.422874927056993,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024769004840541337,
      "loss": 5.0624,
      "mean_token_accuracy": 0.20581938326358795,
      "num_tokens": 120909121.0,
      "step": 69700
    },
    {
      "entropy": 5.869535303115844,
      "epoch": 5.42326395642871,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000247662537942702,
      "loss": 4.9175,
      "mean_token_accuracy": 0.21817582696676255,
      "num_tokens": 120918575.0,
      "step": 69705
    },
    {
      "entropy": 5.886776924133301,
      "epoch": 5.423652985800428,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002476350278947931,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21841296553611755,
      "num_tokens": 120927874.0,
      "step": 69710
    },
    {
      "entropy": 5.905745553970337,
      "epoch": 5.424042015172145,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000247607518262104,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2170095130801201,
      "num_tokens": 120936625.0,
      "step": 69715
    },
    {
      "entropy": 5.896506977081299,
      "epoch": 5.424431044543863,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002475800090450523,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21640065163373948,
      "num_tokens": 120945105.0,
      "step": 69720
    },
    {
      "entropy": 5.83589015007019,
      "epoch": 5.424820073915581,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002475525002440553,
      "loss": 4.8179,
      "mean_token_accuracy": 0.2301688477396965,
      "num_tokens": 120954840.0,
      "step": 69725
    },
    {
      "entropy": 5.836387825012207,
      "epoch": 5.425209103287298,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024752499185953043,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22304360270500184,
      "num_tokens": 120963638.0,
      "step": 69730
    },
    {
      "entropy": 5.911069345474243,
      "epoch": 5.425598132659016,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002474974838918951,
      "loss": 4.9731,
      "mean_token_accuracy": 0.2195759207010269,
      "num_tokens": 120972621.0,
      "step": 69735
    },
    {
      "entropy": 5.911742496490478,
      "epoch": 5.425987162030733,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002474699763415667,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2123311161994934,
      "num_tokens": 120980788.0,
      "step": 69740
    },
    {
      "entropy": 5.8309423446655275,
      "epoch": 5.426376191402451,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00024744246920896243,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22210921198129654,
      "num_tokens": 120989624.0,
      "step": 69745
    },
    {
      "entropy": 5.866896677017212,
      "epoch": 5.426765220774168,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002474149624944999,
      "loss": 4.955,
      "mean_token_accuracy": 0.2195325493812561,
      "num_tokens": 120998036.0,
      "step": 69750
    },
    {
      "entropy": 5.861516427993775,
      "epoch": 5.427154250145886,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024738745619859643,
      "loss": 4.821,
      "mean_token_accuracy": 0.23034973740577697,
      "num_tokens": 121006115.0,
      "step": 69755
    },
    {
      "entropy": 5.874679899215698,
      "epoch": 5.427543279517604,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024735995032166934,
      "loss": 4.9498,
      "mean_token_accuracy": 0.21132907569408416,
      "num_tokens": 121015569.0,
      "step": 69760
    },
    {
      "entropy": 5.901677370071411,
      "epoch": 5.427932308889321,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002473324448641361,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2151007503271103,
      "num_tokens": 121023891.0,
      "step": 69765
    },
    {
      "entropy": 5.857754230499268,
      "epoch": 5.428321338261039,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024730493982641396,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22547693848609923,
      "num_tokens": 121032733.0,
      "step": 69770
    },
    {
      "entropy": 5.868499422073365,
      "epoch": 5.428710367632756,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002472774352089202,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2275269642472267,
      "num_tokens": 121041668.0,
      "step": 69775
    },
    {
      "entropy": 5.8739098072052,
      "epoch": 5.429099397004474,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002472499310120723,
      "loss": 4.967,
      "mean_token_accuracy": 0.2221992313861847,
      "num_tokens": 121051173.0,
      "step": 69780
    },
    {
      "entropy": 5.836136436462402,
      "epoch": 5.429488426376191,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002472224272362875,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22239420115947722,
      "num_tokens": 121059576.0,
      "step": 69785
    },
    {
      "entropy": 5.921668004989624,
      "epoch": 5.429877455747909,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024719492388198313,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21343456208705902,
      "num_tokens": 121068089.0,
      "step": 69790
    },
    {
      "entropy": 5.8960349559783936,
      "epoch": 5.430266485119627,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002471674209495765,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2157305434346199,
      "num_tokens": 121077146.0,
      "step": 69795
    },
    {
      "entropy": 5.8640378475189205,
      "epoch": 5.430655514491344,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024713991843948506,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2152519091963768,
      "num_tokens": 121085575.0,
      "step": 69800
    },
    {
      "entropy": 5.867755937576294,
      "epoch": 5.431044543863061,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024711241635212597,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21450530290603637,
      "num_tokens": 121093784.0,
      "step": 69805
    },
    {
      "entropy": 5.895231866836548,
      "epoch": 5.431433573234779,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024708491468791656,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21802643984556197,
      "num_tokens": 121102962.0,
      "step": 69810
    },
    {
      "entropy": 5.823903799057007,
      "epoch": 5.431822602606497,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002470574134472741,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23221516013145446,
      "num_tokens": 121111782.0,
      "step": 69815
    },
    {
      "entropy": 5.906509876251221,
      "epoch": 5.432211631978214,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00024702991263061595,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22134843617677688,
      "num_tokens": 121120075.0,
      "step": 69820
    },
    {
      "entropy": 5.868521213531494,
      "epoch": 5.432600661349932,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024700241223835935,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2157566085457802,
      "num_tokens": 121129106.0,
      "step": 69825
    },
    {
      "entropy": 5.919453859329224,
      "epoch": 5.43298969072165,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024697491227092163,
      "loss": 5.0832,
      "mean_token_accuracy": 0.20318902730941774,
      "num_tokens": 121137273.0,
      "step": 69830
    },
    {
      "entropy": 5.94287223815918,
      "epoch": 5.433378720093367,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024694741272872,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2222316801548004,
      "num_tokens": 121145632.0,
      "step": 69835
    },
    {
      "entropy": 5.9062751770019535,
      "epoch": 5.433767749465084,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024691991361217166,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22614223062992095,
      "num_tokens": 121154469.0,
      "step": 69840
    },
    {
      "entropy": 5.812202787399292,
      "epoch": 5.434156778836802,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024689241492169403,
      "loss": 4.8112,
      "mean_token_accuracy": 0.22566297352313996,
      "num_tokens": 121162826.0,
      "step": 69845
    },
    {
      "entropy": 5.88845739364624,
      "epoch": 5.43454580820852,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00024686491665770424,
      "loss": 5.0288,
      "mean_token_accuracy": 0.20887815505266188,
      "num_tokens": 121170562.0,
      "step": 69850
    },
    {
      "entropy": 5.858035135269165,
      "epoch": 5.434934837580237,
      "grad_norm": 1.578125,
      "learning_rate": 0.00024683741882061954,
      "loss": 4.947,
      "mean_token_accuracy": 0.21516374349594117,
      "num_tokens": 121179582.0,
      "step": 69855
    },
    {
      "entropy": 5.981293439865112,
      "epoch": 5.435323866951955,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002468099214108572,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20693484097719192,
      "num_tokens": 121188599.0,
      "step": 69860
    },
    {
      "entropy": 5.930628252029419,
      "epoch": 5.435712896323673,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00024678242442883444,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22614535093307495,
      "num_tokens": 121196762.0,
      "step": 69865
    },
    {
      "entropy": 5.885363912582397,
      "epoch": 5.43610192569539,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024675492787496844,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2217929929494858,
      "num_tokens": 121205736.0,
      "step": 69870
    },
    {
      "entropy": 5.8096733570098875,
      "epoch": 5.436490955067107,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024672743174967645,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22104507982730864,
      "num_tokens": 121214186.0,
      "step": 69875
    },
    {
      "entropy": 5.852441930770874,
      "epoch": 5.436879984438825,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002466999360533757,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2231011673808098,
      "num_tokens": 121222656.0,
      "step": 69880
    },
    {
      "entropy": 5.84800157546997,
      "epoch": 5.437269013810543,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002466724407864833,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2223466530442238,
      "num_tokens": 121231287.0,
      "step": 69885
    },
    {
      "entropy": 5.894041061401367,
      "epoch": 5.43765804318226,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024664494594941664,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22935065925121306,
      "num_tokens": 121239783.0,
      "step": 69890
    },
    {
      "entropy": 5.836501169204712,
      "epoch": 5.438047072553978,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00024661745154259273,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22122283577919005,
      "num_tokens": 121248131.0,
      "step": 69895
    },
    {
      "entropy": 5.8036092758178714,
      "epoch": 5.438436101925696,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002465899575664289,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21683369427919388,
      "num_tokens": 121256719.0,
      "step": 69900
    },
    {
      "entropy": 5.825444889068604,
      "epoch": 5.438825131297413,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024656246402134216,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22483257949352264,
      "num_tokens": 121265199.0,
      "step": 69905
    },
    {
      "entropy": 5.821436738967895,
      "epoch": 5.43921416066913,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002465349709077498,
      "loss": 4.816,
      "mean_token_accuracy": 0.2257923513650894,
      "num_tokens": 121274648.0,
      "step": 69910
    },
    {
      "entropy": 5.886503267288208,
      "epoch": 5.439603190040848,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002465074782260689,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21682190597057344,
      "num_tokens": 121284159.0,
      "step": 69915
    },
    {
      "entropy": 5.925439167022705,
      "epoch": 5.439992219412566,
      "grad_norm": 1.359375,
      "learning_rate": 0.00024647998597671674,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22496234476566315,
      "num_tokens": 121292635.0,
      "step": 69920
    },
    {
      "entropy": 5.9155323028564455,
      "epoch": 5.440381248784283,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002464524941601103,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22101916819810868,
      "num_tokens": 121301688.0,
      "step": 69925
    },
    {
      "entropy": 5.870996379852295,
      "epoch": 5.440770278156001,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024642500277666696,
      "loss": 4.9593,
      "mean_token_accuracy": 0.20758458375930786,
      "num_tokens": 121310234.0,
      "step": 69930
    },
    {
      "entropy": 5.850041055679322,
      "epoch": 5.441159307527719,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00024639751182680363,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22240880280733108,
      "num_tokens": 121318757.0,
      "step": 69935
    },
    {
      "entropy": 5.947680997848511,
      "epoch": 5.4415483368994355,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002463700213109376,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21812837123870848,
      "num_tokens": 121327881.0,
      "step": 69940
    },
    {
      "entropy": 5.945798254013061,
      "epoch": 5.441937366271153,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024634253122948584,
      "loss": 4.9955,
      "mean_token_accuracy": 0.20732611566781997,
      "num_tokens": 121336444.0,
      "step": 69945
    },
    {
      "entropy": 5.961141204833984,
      "epoch": 5.442326395642871,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024631504158286556,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2212778463959694,
      "num_tokens": 121344571.0,
      "step": 69950
    },
    {
      "entropy": 5.902258443832397,
      "epoch": 5.442715425014589,
      "grad_norm": 1.5,
      "learning_rate": 0.00024628755237149385,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22164823412895202,
      "num_tokens": 121352806.0,
      "step": 69955
    },
    {
      "entropy": 5.881486988067627,
      "epoch": 5.443104454386306,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024626006359578797,
      "loss": 4.8905,
      "mean_token_accuracy": 0.21954032629728318,
      "num_tokens": 121361189.0,
      "step": 69960
    },
    {
      "entropy": 5.91632490158081,
      "epoch": 5.443493483758024,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024623257525616484,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21740350276231765,
      "num_tokens": 121369845.0,
      "step": 69965
    },
    {
      "entropy": 5.846297311782837,
      "epoch": 5.443882513129742,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024620508735304146,
      "loss": 4.912,
      "mean_token_accuracy": 0.21787936687469484,
      "num_tokens": 121377923.0,
      "step": 69970
    },
    {
      "entropy": 5.8934637069702145,
      "epoch": 5.4442715425014585,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002461775998868352,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22255182713270188,
      "num_tokens": 121386442.0,
      "step": 69975
    },
    {
      "entropy": 5.929491949081421,
      "epoch": 5.444660571873176,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024615011285796293,
      "loss": 4.9325,
      "mean_token_accuracy": 0.21644456386566163,
      "num_tokens": 121394918.0,
      "step": 69980
    },
    {
      "entropy": 5.939820909500122,
      "epoch": 5.445049601244894,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024612262626684176,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2131143882870674,
      "num_tokens": 121403509.0,
      "step": 69985
    },
    {
      "entropy": 5.9830351829528805,
      "epoch": 5.445438630616612,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002460951401138888,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21672654151916504,
      "num_tokens": 121411673.0,
      "step": 69990
    },
    {
      "entropy": 5.968668651580811,
      "epoch": 5.445827659988329,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024606765439952105,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21530264914035796,
      "num_tokens": 121420843.0,
      "step": 69995
    },
    {
      "entropy": 5.8963369846344,
      "epoch": 5.446216689360047,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002460401691241556,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21451497972011566,
      "num_tokens": 121429828.0,
      "step": 70000
    },
    {
      "entropy": 5.88644585609436,
      "epoch": 5.446605718731764,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024601268428820947,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22093952745199202,
      "num_tokens": 121438058.0,
      "step": 70005
    },
    {
      "entropy": 5.841746664047241,
      "epoch": 5.4469947481034815,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024598519989209983,
      "loss": 4.8186,
      "mean_token_accuracy": 0.229946006834507,
      "num_tokens": 121447113.0,
      "step": 70010
    },
    {
      "entropy": 5.924835252761841,
      "epoch": 5.447383777475199,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002459577159362434,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21294353753328324,
      "num_tokens": 121455217.0,
      "step": 70015
    },
    {
      "entropy": 5.934188175201416,
      "epoch": 5.447772806846917,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002459302324210575,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22458408921957015,
      "num_tokens": 121463208.0,
      "step": 70020
    },
    {
      "entropy": 5.92068076133728,
      "epoch": 5.448161836218635,
      "grad_norm": 1.5,
      "learning_rate": 0.0002459027493469591,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2200424328446388,
      "num_tokens": 121472400.0,
      "step": 70025
    },
    {
      "entropy": 5.919198274612427,
      "epoch": 5.448550865590352,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00024587526671436517,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21414900571107864,
      "num_tokens": 121481400.0,
      "step": 70030
    },
    {
      "entropy": 5.888467454910279,
      "epoch": 5.44893989496207,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024584778452369264,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21520513743162156,
      "num_tokens": 121490004.0,
      "step": 70035
    },
    {
      "entropy": 5.911335039138794,
      "epoch": 5.449328924333787,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024582030277535866,
      "loss": 5.0176,
      "mean_token_accuracy": 0.2107431024312973,
      "num_tokens": 121500136.0,
      "step": 70040
    },
    {
      "entropy": 5.8785295486450195,
      "epoch": 5.4497179537055045,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024579282146978016,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22354887425899506,
      "num_tokens": 121508711.0,
      "step": 70045
    },
    {
      "entropy": 5.886579608917236,
      "epoch": 5.450106983077222,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000245765340607374,
      "loss": 4.9089,
      "mean_token_accuracy": 0.21476557850837708,
      "num_tokens": 121516971.0,
      "step": 70050
    },
    {
      "entropy": 5.843554162979126,
      "epoch": 5.45049601244894,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002457378601885573,
      "loss": 4.958,
      "mean_token_accuracy": 0.2163891687989235,
      "num_tokens": 121526359.0,
      "step": 70055
    },
    {
      "entropy": 5.95477032661438,
      "epoch": 5.450885041820658,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024571038021374705,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21252834796905518,
      "num_tokens": 121535001.0,
      "step": 70060
    },
    {
      "entropy": 5.885092401504517,
      "epoch": 5.451274071192375,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024568290068336014,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2156697019934654,
      "num_tokens": 121543938.0,
      "step": 70065
    },
    {
      "entropy": 5.8728879451751705,
      "epoch": 5.451663100564093,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002456554215978135,
      "loss": 4.9091,
      "mean_token_accuracy": 0.21648670732975006,
      "num_tokens": 121552009.0,
      "step": 70070
    },
    {
      "entropy": 5.889917659759521,
      "epoch": 5.45205212993581,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002456279429575242,
      "loss": 4.947,
      "mean_token_accuracy": 0.22220084071159363,
      "num_tokens": 121560117.0,
      "step": 70075
    },
    {
      "entropy": 5.909624242782593,
      "epoch": 5.4524411593075275,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024560046476290905,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2108195036649704,
      "num_tokens": 121568796.0,
      "step": 70080
    },
    {
      "entropy": 5.845491886138916,
      "epoch": 5.452830188679245,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00024557298701438504,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21641364246606826,
      "num_tokens": 121577650.0,
      "step": 70085
    },
    {
      "entropy": 5.9159543991088865,
      "epoch": 5.453219218050963,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002455455097123691,
      "loss": 4.946,
      "mean_token_accuracy": 0.2233739748597145,
      "num_tokens": 121585997.0,
      "step": 70090
    },
    {
      "entropy": 5.869453573226929,
      "epoch": 5.453608247422681,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002455180328572783,
      "loss": 4.9526,
      "mean_token_accuracy": 0.20883786976337432,
      "num_tokens": 121594753.0,
      "step": 70095
    },
    {
      "entropy": 5.838490629196167,
      "epoch": 5.453997276794398,
      "grad_norm": 1.625,
      "learning_rate": 0.00024549055644952926,
      "loss": 4.8379,
      "mean_token_accuracy": 0.22815800309181214,
      "num_tokens": 121602593.0,
      "step": 70100
    },
    {
      "entropy": 5.931815338134766,
      "epoch": 5.454386306166116,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002454630804895392,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21055189967155458,
      "num_tokens": 121612073.0,
      "step": 70105
    },
    {
      "entropy": 5.915041971206665,
      "epoch": 5.454775335537833,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002454356049777248,
      "loss": 4.9168,
      "mean_token_accuracy": 0.21592997312545775,
      "num_tokens": 121620847.0,
      "step": 70110
    },
    {
      "entropy": 5.8787009716033936,
      "epoch": 5.4551643649095505,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00024540812991450306,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2222980961203575,
      "num_tokens": 121629440.0,
      "step": 70115
    },
    {
      "entropy": 5.791359186172485,
      "epoch": 5.455553394281268,
      "grad_norm": 1.375,
      "learning_rate": 0.00024538065530029093,
      "loss": 4.8319,
      "mean_token_accuracy": 0.22363897114992143,
      "num_tokens": 121637539.0,
      "step": 70120
    },
    {
      "entropy": 5.840430736541748,
      "epoch": 5.455942423652986,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002453531811355051,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21694067269563674,
      "num_tokens": 121646081.0,
      "step": 70125
    },
    {
      "entropy": 5.880767583847046,
      "epoch": 5.456331453024704,
      "grad_norm": 1.453125,
      "learning_rate": 0.00024532570742056257,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21445721685886382,
      "num_tokens": 121655138.0,
      "step": 70130
    },
    {
      "entropy": 5.9068340301513675,
      "epoch": 5.456720482396421,
      "grad_norm": 1.59375,
      "learning_rate": 0.00024529823415588025,
      "loss": 4.971,
      "mean_token_accuracy": 0.2148249551653862,
      "num_tokens": 121663528.0,
      "step": 70135
    },
    {
      "entropy": 5.848746919631958,
      "epoch": 5.457109511768138,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002452707613418749,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23226836770772935,
      "num_tokens": 121671818.0,
      "step": 70140
    },
    {
      "entropy": 5.837978076934815,
      "epoch": 5.457498541139856,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00024524328897896345,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22009336799383164,
      "num_tokens": 121680803.0,
      "step": 70145
    },
    {
      "entropy": 5.873604011535645,
      "epoch": 5.4578875705115735,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024521581706756266,
      "loss": 4.9368,
      "mean_token_accuracy": 0.21695126891136168,
      "num_tokens": 121689917.0,
      "step": 70150
    },
    {
      "entropy": 5.839484977722168,
      "epoch": 5.458276599883291,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024518834560808954,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22134851515293122,
      "num_tokens": 121698281.0,
      "step": 70155
    },
    {
      "entropy": 5.9200358390808105,
      "epoch": 5.458665629255009,
      "grad_norm": 1.40625,
      "learning_rate": 0.00024516087460096074,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21614780873060227,
      "num_tokens": 121707311.0,
      "step": 70160
    },
    {
      "entropy": 5.8499139785766605,
      "epoch": 5.459054658626727,
      "grad_norm": 1.5,
      "learning_rate": 0.0002451334040465933,
      "loss": 4.8317,
      "mean_token_accuracy": 0.2256646141409874,
      "num_tokens": 121717187.0,
      "step": 70165
    },
    {
      "entropy": 5.860132265090942,
      "epoch": 5.459443687998444,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024510593394540383,
      "loss": 4.9191,
      "mean_token_accuracy": 0.20968862622976303,
      "num_tokens": 121725713.0,
      "step": 70170
    },
    {
      "entropy": 5.931229114532471,
      "epoch": 5.459832717370161,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002450784642978092,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2226481944322586,
      "num_tokens": 121734089.0,
      "step": 70175
    },
    {
      "entropy": 5.880706071853638,
      "epoch": 5.460221746741879,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002450509951042263,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22196077257394792,
      "num_tokens": 121742615.0,
      "step": 70180
    },
    {
      "entropy": 5.883300733566284,
      "epoch": 5.4606107761135965,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00024502352636507184,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2176334649324417,
      "num_tokens": 121751042.0,
      "step": 70185
    },
    {
      "entropy": 5.854633426666259,
      "epoch": 5.460999805485314,
      "grad_norm": 1.375,
      "learning_rate": 0.00024499605808076273,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22110584378242493,
      "num_tokens": 121759716.0,
      "step": 70190
    },
    {
      "entropy": 5.7913196086883545,
      "epoch": 5.461388834857032,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024496859025171564,
      "loss": 4.8257,
      "mean_token_accuracy": 0.22525720596313475,
      "num_tokens": 121768151.0,
      "step": 70195
    },
    {
      "entropy": 5.831320667266846,
      "epoch": 5.46177786422875,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002449411228783474,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22174605429172517,
      "num_tokens": 121776734.0,
      "step": 70200
    },
    {
      "entropy": 5.871917772293091,
      "epoch": 5.462166893600466,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024491365596107477,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22642588317394258,
      "num_tokens": 121784598.0,
      "step": 70205
    },
    {
      "entropy": 5.863680648803711,
      "epoch": 5.462555922972184,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024488618950031457,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22186751961708068,
      "num_tokens": 121793953.0,
      "step": 70210
    },
    {
      "entropy": 5.813996982574463,
      "epoch": 5.462944952343902,
      "grad_norm": 1.5625,
      "learning_rate": 0.00024485872349648335,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2226432040333748,
      "num_tokens": 121802909.0,
      "step": 70215
    },
    {
      "entropy": 5.879582834243775,
      "epoch": 5.4633339817156195,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024483125794999815,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21645007729530336,
      "num_tokens": 121812752.0,
      "step": 70220
    },
    {
      "entropy": 5.918142175674438,
      "epoch": 5.463723011087337,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002448037928612756,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21603080928325652,
      "num_tokens": 121821293.0,
      "step": 70225
    },
    {
      "entropy": 5.949479675292968,
      "epoch": 5.464112040459055,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002447763282307324,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2240629956126213,
      "num_tokens": 121829980.0,
      "step": 70230
    },
    {
      "entropy": 5.909882974624634,
      "epoch": 5.4645010698307726,
      "grad_norm": 1.375,
      "learning_rate": 0.0002447488640587854,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2191912516951561,
      "num_tokens": 121838496.0,
      "step": 70235
    },
    {
      "entropy": 5.90141806602478,
      "epoch": 5.46489009920249,
      "grad_norm": 1.5,
      "learning_rate": 0.00024472140034585113,
      "loss": 4.9055,
      "mean_token_accuracy": 0.21905417740345,
      "num_tokens": 121846744.0,
      "step": 70240
    },
    {
      "entropy": 5.789193820953369,
      "epoch": 5.465279128574207,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024469393709234647,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22844281792640686,
      "num_tokens": 121855059.0,
      "step": 70245
    },
    {
      "entropy": 5.858499765396118,
      "epoch": 5.465668157945925,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024466647429868804,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22334830462932587,
      "num_tokens": 121863654.0,
      "step": 70250
    },
    {
      "entropy": 5.833910608291626,
      "epoch": 5.4660571873176425,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002446390119652926,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22110218703746795,
      "num_tokens": 121872034.0,
      "step": 70255
    },
    {
      "entropy": 5.822856044769287,
      "epoch": 5.46644621668936,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024461155009257684,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23349737823009492,
      "num_tokens": 121881173.0,
      "step": 70260
    },
    {
      "entropy": 5.890830707550049,
      "epoch": 5.466835246061078,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024458408868095744,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22202355712652205,
      "num_tokens": 121889996.0,
      "step": 70265
    },
    {
      "entropy": 5.93950629234314,
      "epoch": 5.4672242754327955,
      "grad_norm": 1.375,
      "learning_rate": 0.00024455662773085107,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2144548401236534,
      "num_tokens": 121899032.0,
      "step": 70270
    },
    {
      "entropy": 5.957671451568603,
      "epoch": 5.467613304804512,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002445291672426744,
      "loss": 4.927,
      "mean_token_accuracy": 0.21856434643268585,
      "num_tokens": 121907953.0,
      "step": 70275
    },
    {
      "entropy": 5.89250111579895,
      "epoch": 5.46800233417623,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024450170721684416,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21921935826539993,
      "num_tokens": 121916492.0,
      "step": 70280
    },
    {
      "entropy": 5.960846662521362,
      "epoch": 5.468391363547948,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002444742476537769,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2174324333667755,
      "num_tokens": 121925762.0,
      "step": 70285
    },
    {
      "entropy": 5.808321475982666,
      "epoch": 5.4687803929196654,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00024444678855388934,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23028063923120498,
      "num_tokens": 121935085.0,
      "step": 70290
    },
    {
      "entropy": 5.865771484375,
      "epoch": 5.469169422291383,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024441932991759824,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21919080764055252,
      "num_tokens": 121943887.0,
      "step": 70295
    },
    {
      "entropy": 5.879450130462646,
      "epoch": 5.469558451663101,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002443918717453201,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21358540803194045,
      "num_tokens": 121952211.0,
      "step": 70300
    },
    {
      "entropy": 5.861865091323852,
      "epoch": 5.4699474810348185,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002443644140374715,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2119327887892723,
      "num_tokens": 121960413.0,
      "step": 70305
    },
    {
      "entropy": 5.895777654647827,
      "epoch": 5.470336510406535,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024433695679446934,
      "loss": 4.8796,
      "mean_token_accuracy": 0.2266196131706238,
      "num_tokens": 121968314.0,
      "step": 70310
    },
    {
      "entropy": 5.8208451747894285,
      "epoch": 5.470725539778253,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002443095000167299,
      "loss": 4.8453,
      "mean_token_accuracy": 0.22463129013776778,
      "num_tokens": 121977432.0,
      "step": 70315
    },
    {
      "entropy": 5.85915732383728,
      "epoch": 5.471114569149971,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024428204370467,
      "loss": 4.944,
      "mean_token_accuracy": 0.21576273441314697,
      "num_tokens": 121986649.0,
      "step": 70320
    },
    {
      "entropy": 5.863361167907715,
      "epoch": 5.471503598521688,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024425458785870617,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2120680943131447,
      "num_tokens": 121995408.0,
      "step": 70325
    },
    {
      "entropy": 5.867340612411499,
      "epoch": 5.471892627893406,
      "grad_norm": 1.5,
      "learning_rate": 0.00024422713247925507,
      "loss": 4.8461,
      "mean_token_accuracy": 0.21574461162090303,
      "num_tokens": 122003775.0,
      "step": 70330
    },
    {
      "entropy": 5.895418882369995,
      "epoch": 5.472281657265124,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002441996775667332,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22133471667766572,
      "num_tokens": 122011249.0,
      "step": 70335
    },
    {
      "entropy": 5.87779393196106,
      "epoch": 5.472670686636841,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002441722231215572,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21311632096767424,
      "num_tokens": 122019895.0,
      "step": 70340
    },
    {
      "entropy": 5.852601814270019,
      "epoch": 5.473059716008558,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024414476914414373,
      "loss": 4.966,
      "mean_token_accuracy": 0.21349520981311798,
      "num_tokens": 122028452.0,
      "step": 70345
    },
    {
      "entropy": 5.7951802730560305,
      "epoch": 5.473448745380276,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024411731563490918,
      "loss": 4.845,
      "mean_token_accuracy": 0.22670210599899293,
      "num_tokens": 122036648.0,
      "step": 70350
    },
    {
      "entropy": 5.859739303588867,
      "epoch": 5.473837774751994,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00024408986259427025,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22684519588947297,
      "num_tokens": 122044621.0,
      "step": 70355
    },
    {
      "entropy": 5.895005941390991,
      "epoch": 5.474226804123711,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024406241002264346,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21889501363039016,
      "num_tokens": 122053525.0,
      "step": 70360
    },
    {
      "entropy": 5.90136866569519,
      "epoch": 5.474615833495429,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024403495792044532,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21378229111433028,
      "num_tokens": 122062247.0,
      "step": 70365
    },
    {
      "entropy": 5.911420202255249,
      "epoch": 5.475004862867147,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024400750628809247,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22436732798814774,
      "num_tokens": 122071683.0,
      "step": 70370
    },
    {
      "entropy": 5.870447587966919,
      "epoch": 5.475393892238864,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024398005512600134,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22033074498176575,
      "num_tokens": 122079902.0,
      "step": 70375
    },
    {
      "entropy": 5.835946273803711,
      "epoch": 5.475782921610581,
      "grad_norm": 1.5,
      "learning_rate": 0.00024395260443458856,
      "loss": 4.8184,
      "mean_token_accuracy": 0.22671363204717637,
      "num_tokens": 122088200.0,
      "step": 70380
    },
    {
      "entropy": 5.895721817016602,
      "epoch": 5.476171950982299,
      "grad_norm": 1.421875,
      "learning_rate": 0.00024392515421427054,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22096757143735885,
      "num_tokens": 122096338.0,
      "step": 70385
    },
    {
      "entropy": 5.817822027206421,
      "epoch": 5.476560980354017,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024389770446546388,
      "loss": 4.7914,
      "mean_token_accuracy": 0.2322947785258293,
      "num_tokens": 122104811.0,
      "step": 70390
    },
    {
      "entropy": 5.873475551605225,
      "epoch": 5.476950009725734,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024387025518858503,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2219479575753212,
      "num_tokens": 122113422.0,
      "step": 70395
    },
    {
      "entropy": 5.857664489746094,
      "epoch": 5.477339039097452,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002438428063840505,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21425848007202147,
      "num_tokens": 122121951.0,
      "step": 70400
    },
    {
      "entropy": 5.810535907745361,
      "epoch": 5.47772806846917,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00024381535805227677,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23977371156215668,
      "num_tokens": 122130648.0,
      "step": 70405
    },
    {
      "entropy": 5.892986726760864,
      "epoch": 5.478117097840887,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024378791019368036,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2219313934445381,
      "num_tokens": 122139143.0,
      "step": 70410
    },
    {
      "entropy": 5.860080337524414,
      "epoch": 5.478506127212604,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024376046280867765,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2216213583946228,
      "num_tokens": 122147116.0,
      "step": 70415
    },
    {
      "entropy": 5.883833694458008,
      "epoch": 5.478895156584322,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024373301589768531,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22029977440834045,
      "num_tokens": 122155336.0,
      "step": 70420
    },
    {
      "entropy": 5.892518424987793,
      "epoch": 5.47928418595604,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002437055694611197,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21999952346086502,
      "num_tokens": 122164318.0,
      "step": 70425
    },
    {
      "entropy": 5.981170177459717,
      "epoch": 5.479673215327757,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00024367812349939723,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21360650956630706,
      "num_tokens": 122173744.0,
      "step": 70430
    },
    {
      "entropy": 5.9328819751739506,
      "epoch": 5.480062244699475,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024365067801293438,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21956901699304582,
      "num_tokens": 122181976.0,
      "step": 70435
    },
    {
      "entropy": 5.865108585357666,
      "epoch": 5.480451274071193,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024362323300214755,
      "loss": 4.8431,
      "mean_token_accuracy": 0.22158541083335875,
      "num_tokens": 122190548.0,
      "step": 70440
    },
    {
      "entropy": 5.850777912139892,
      "epoch": 5.48084030344291,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002435957884674533,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21562442183494568,
      "num_tokens": 122199276.0,
      "step": 70445
    },
    {
      "entropy": 5.877582550048828,
      "epoch": 5.481229332814627,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002435683444092679,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2194230630993843,
      "num_tokens": 122208252.0,
      "step": 70450
    },
    {
      "entropy": 5.8184239864349365,
      "epoch": 5.481618362186345,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002435409008280079,
      "loss": 4.7988,
      "mean_token_accuracy": 0.2301386460661888,
      "num_tokens": 122217336.0,
      "step": 70455
    },
    {
      "entropy": 5.918790292739868,
      "epoch": 5.482007391558063,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002435134577240896,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2185482755303383,
      "num_tokens": 122226158.0,
      "step": 70460
    },
    {
      "entropy": 5.7809507846832275,
      "epoch": 5.48239642092978,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024348601509792956,
      "loss": 4.7286,
      "mean_token_accuracy": 0.2381079152226448,
      "num_tokens": 122234737.0,
      "step": 70465
    },
    {
      "entropy": 5.883880090713501,
      "epoch": 5.482785450301498,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024345857294994405,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21722276359796525,
      "num_tokens": 122243079.0,
      "step": 70470
    },
    {
      "entropy": 5.904862689971924,
      "epoch": 5.483174479673215,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024343113128054945,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2185622438788414,
      "num_tokens": 122252467.0,
      "step": 70475
    },
    {
      "entropy": 5.910847234725952,
      "epoch": 5.483563509044933,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002434036900901621,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2124548599123955,
      "num_tokens": 122261930.0,
      "step": 70480
    },
    {
      "entropy": 5.987967443466187,
      "epoch": 5.48395253841665,
      "grad_norm": 1.453125,
      "learning_rate": 0.00024337624937919856,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20829004496335984,
      "num_tokens": 122270707.0,
      "step": 70485
    },
    {
      "entropy": 5.853066253662109,
      "epoch": 5.484341567788368,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024334880914807512,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22321339696645737,
      "num_tokens": 122279711.0,
      "step": 70490
    },
    {
      "entropy": 5.858671617507935,
      "epoch": 5.484730597160086,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024332136939720806,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22278978377580644,
      "num_tokens": 122287518.0,
      "step": 70495
    },
    {
      "entropy": 5.819047975540161,
      "epoch": 5.485119626531803,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024329393012701396,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2194607138633728,
      "num_tokens": 122295816.0,
      "step": 70500
    },
    {
      "entropy": 5.873182916641236,
      "epoch": 5.485508655903521,
      "grad_norm": 1.671875,
      "learning_rate": 0.00024326649133790885,
      "loss": 4.8437,
      "mean_token_accuracy": 0.2168092757463455,
      "num_tokens": 122304231.0,
      "step": 70505
    },
    {
      "entropy": 5.851479244232178,
      "epoch": 5.485897685275238,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024323905303030925,
      "loss": 4.8012,
      "mean_token_accuracy": 0.22867920994758606,
      "num_tokens": 122313051.0,
      "step": 70510
    },
    {
      "entropy": 5.879076051712036,
      "epoch": 5.486286714646956,
      "grad_norm": 1.5,
      "learning_rate": 0.00024321161520463147,
      "loss": 4.8027,
      "mean_token_accuracy": 0.23242637664079666,
      "num_tokens": 122321467.0,
      "step": 70515
    },
    {
      "entropy": 5.927892208099365,
      "epoch": 5.486675744018673,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024318417786129188,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21385224610567094,
      "num_tokens": 122329761.0,
      "step": 70520
    },
    {
      "entropy": 5.944902658462524,
      "epoch": 5.487064773390391,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024315674100070672,
      "loss": 4.984,
      "mean_token_accuracy": 0.21199873089790344,
      "num_tokens": 122338714.0,
      "step": 70525
    },
    {
      "entropy": 5.87919430732727,
      "epoch": 5.487453802762109,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002431293046232923,
      "loss": 4.8994,
      "mean_token_accuracy": 0.21970379948616028,
      "num_tokens": 122347378.0,
      "step": 70530
    },
    {
      "entropy": 5.8648277759552006,
      "epoch": 5.487842832133826,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00024310186872946498,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2287174642086029,
      "num_tokens": 122355092.0,
      "step": 70535
    },
    {
      "entropy": 5.930278635025024,
      "epoch": 5.488231861505543,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024307443331964107,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21015854477882384,
      "num_tokens": 122363855.0,
      "step": 70540
    },
    {
      "entropy": 5.899869966506958,
      "epoch": 5.488620890877261,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024304699839423677,
      "loss": 4.9132,
      "mean_token_accuracy": 0.2202589064836502,
      "num_tokens": 122372298.0,
      "step": 70545
    },
    {
      "entropy": 5.865393733978271,
      "epoch": 5.489009920248979,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00024301956395366838,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22467932999134063,
      "num_tokens": 122380916.0,
      "step": 70550
    },
    {
      "entropy": 5.795569133758545,
      "epoch": 5.489398949620696,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024299212999835223,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21357515156269075,
      "num_tokens": 122389638.0,
      "step": 70555
    },
    {
      "entropy": 5.9335315227508545,
      "epoch": 5.489787978992414,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024296469652870456,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21384841203689575,
      "num_tokens": 122397983.0,
      "step": 70560
    },
    {
      "entropy": 5.926889944076538,
      "epoch": 5.490177008364132,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024293726354514166,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22419352382421492,
      "num_tokens": 122406395.0,
      "step": 70565
    },
    {
      "entropy": 5.888062286376953,
      "epoch": 5.490566037735849,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002429098310480798,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22472501546144485,
      "num_tokens": 122414551.0,
      "step": 70570
    },
    {
      "entropy": 5.898818016052246,
      "epoch": 5.490955067107567,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024288239903793508,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2245040312409401,
      "num_tokens": 122422757.0,
      "step": 70575
    },
    {
      "entropy": 5.879982900619507,
      "epoch": 5.491344096479284,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024285496751512393,
      "loss": 4.962,
      "mean_token_accuracy": 0.21537859588861466,
      "num_tokens": 122432085.0,
      "step": 70580
    },
    {
      "entropy": 5.899964141845703,
      "epoch": 5.491733125851002,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002428275364800624,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23134653270244598,
      "num_tokens": 122441838.0,
      "step": 70585
    },
    {
      "entropy": 5.877882957458496,
      "epoch": 5.492122155222719,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024280010593316683,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2212147071957588,
      "num_tokens": 122450114.0,
      "step": 70590
    },
    {
      "entropy": 5.932202863693237,
      "epoch": 5.492511184594437,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002427726758748534,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21399688124656677,
      "num_tokens": 122458806.0,
      "step": 70595
    },
    {
      "entropy": 5.898769569396973,
      "epoch": 5.492900213966155,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002427452463055383,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2062707394361496,
      "num_tokens": 122467252.0,
      "step": 70600
    },
    {
      "entropy": 5.946411514282227,
      "epoch": 5.493289243337872,
      "grad_norm": 1.453125,
      "learning_rate": 0.00024271781722563773,
      "loss": 4.928,
      "mean_token_accuracy": 0.22092086523771287,
      "num_tokens": 122475751.0,
      "step": 70605
    },
    {
      "entropy": 5.943218803405761,
      "epoch": 5.493678272709589,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024269038863556798,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21298309713602065,
      "num_tokens": 122484126.0,
      "step": 70610
    },
    {
      "entropy": 5.883031797409058,
      "epoch": 5.494067302081307,
      "grad_norm": 1.5,
      "learning_rate": 0.00024266296053574506,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22110921889543533,
      "num_tokens": 122492714.0,
      "step": 70615
    },
    {
      "entropy": 5.842199516296387,
      "epoch": 5.494456331453025,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002426355329265853,
      "loss": 4.8268,
      "mean_token_accuracy": 0.22903233617544175,
      "num_tokens": 122501466.0,
      "step": 70620
    },
    {
      "entropy": 5.831239938735962,
      "epoch": 5.494845360824742,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024260810580850479,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21231242567300795,
      "num_tokens": 122510098.0,
      "step": 70625
    },
    {
      "entropy": 5.903623867034912,
      "epoch": 5.49523439019646,
      "grad_norm": 1.59375,
      "learning_rate": 0.00024258067918191971,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21722180694341658,
      "num_tokens": 122518883.0,
      "step": 70630
    },
    {
      "entropy": 5.971438360214234,
      "epoch": 5.495623419568178,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002425532530472463,
      "loss": 4.931,
      "mean_token_accuracy": 0.21792835593223572,
      "num_tokens": 122527408.0,
      "step": 70635
    },
    {
      "entropy": 5.86884388923645,
      "epoch": 5.496012448939895,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024252582740490053,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22278157472610474,
      "num_tokens": 122535859.0,
      "step": 70640
    },
    {
      "entropy": 5.872522211074829,
      "epoch": 5.496401478311612,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024249840225529874,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2259932279586792,
      "num_tokens": 122544360.0,
      "step": 70645
    },
    {
      "entropy": 5.853538513183594,
      "epoch": 5.49679050768333,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002424709775988569,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2178159922361374,
      "num_tokens": 122552457.0,
      "step": 70650
    },
    {
      "entropy": 5.8580464839935305,
      "epoch": 5.497179537055048,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00024244355343599116,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21358455419540406,
      "num_tokens": 122560153.0,
      "step": 70655
    },
    {
      "entropy": 5.805581521987915,
      "epoch": 5.497568566426765,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024241612976711786,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22413382679224014,
      "num_tokens": 122568602.0,
      "step": 70660
    },
    {
      "entropy": 5.907294797897339,
      "epoch": 5.497957595798483,
      "grad_norm": 1.5625,
      "learning_rate": 0.00024238870659265278,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22667179107666016,
      "num_tokens": 122576663.0,
      "step": 70665
    },
    {
      "entropy": 5.965759897232056,
      "epoch": 5.498346625170201,
      "grad_norm": 1.5625,
      "learning_rate": 0.00024236128391301215,
      "loss": 4.93,
      "mean_token_accuracy": 0.21304364800453185,
      "num_tokens": 122585835.0,
      "step": 70670
    },
    {
      "entropy": 5.979984474182129,
      "epoch": 5.4987356545419175,
      "grad_norm": 1.453125,
      "learning_rate": 0.00024233386172861217,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21105094850063325,
      "num_tokens": 122595374.0,
      "step": 70675
    },
    {
      "entropy": 5.913734102249146,
      "epoch": 5.499124683913635,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024230644003986869,
      "loss": 5.0632,
      "mean_token_accuracy": 0.2016451060771942,
      "num_tokens": 122604836.0,
      "step": 70680
    },
    {
      "entropy": 5.901591014862061,
      "epoch": 5.499513713285353,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024227901884719806,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21935382336378098,
      "num_tokens": 122612817.0,
      "step": 70685
    },
    {
      "entropy": 5.895628166198731,
      "epoch": 5.499902742657071,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024225159815101626,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2140609934926033,
      "num_tokens": 122621624.0,
      "step": 70690
    },
    {
      "entropy": 5.89120078086853,
      "epoch": 5.500291772028788,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024222417795173929,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21693207025527955,
      "num_tokens": 122630122.0,
      "step": 70695
    },
    {
      "entropy": 5.913450384140015,
      "epoch": 5.500680801400506,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00024219675824978332,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21708621978759765,
      "num_tokens": 122639185.0,
      "step": 70700
    },
    {
      "entropy": 5.899397563934326,
      "epoch": 5.501069830772224,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024216933904556425,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21679517328739167,
      "num_tokens": 122648285.0,
      "step": 70705
    },
    {
      "entropy": 5.897579145431519,
      "epoch": 5.501458860143941,
      "grad_norm": 1.375,
      "learning_rate": 0.00024214192033949827,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2199350044131279,
      "num_tokens": 122657287.0,
      "step": 70710
    },
    {
      "entropy": 5.914361906051636,
      "epoch": 5.501847889515658,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00024211450213200132,
      "loss": 4.8184,
      "mean_token_accuracy": 0.22597637325525283,
      "num_tokens": 122665666.0,
      "step": 70715
    },
    {
      "entropy": 5.868224382400513,
      "epoch": 5.502236918887376,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00024208708442348943,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21766099482774734,
      "num_tokens": 122674098.0,
      "step": 70720
    },
    {
      "entropy": 5.814623880386352,
      "epoch": 5.502625948259094,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024205966721437867,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2253967821598053,
      "num_tokens": 122682201.0,
      "step": 70725
    },
    {
      "entropy": 5.794929838180542,
      "epoch": 5.503014977630811,
      "grad_norm": 1.5,
      "learning_rate": 0.00024203225050508503,
      "loss": 4.8997,
      "mean_token_accuracy": 0.21132864207029342,
      "num_tokens": 122690453.0,
      "step": 70730
    },
    {
      "entropy": 5.83031964302063,
      "epoch": 5.503404007002529,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002420048342960245,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22359938323497772,
      "num_tokens": 122699690.0,
      "step": 70735
    },
    {
      "entropy": 5.886711978912354,
      "epoch": 5.503793036374246,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024197741858761318,
      "loss": 4.821,
      "mean_token_accuracy": 0.2310731902718544,
      "num_tokens": 122708090.0,
      "step": 70740
    },
    {
      "entropy": 5.906991577148437,
      "epoch": 5.5041820657459635,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002419500033802669,
      "loss": 4.8488,
      "mean_token_accuracy": 0.21641291975975036,
      "num_tokens": 122716570.0,
      "step": 70745
    },
    {
      "entropy": 5.862079429626465,
      "epoch": 5.504571095117681,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002419225886744016,
      "loss": 4.8783,
      "mean_token_accuracy": 0.21815048903226852,
      "num_tokens": 122725680.0,
      "step": 70750
    },
    {
      "entropy": 5.938792943954468,
      "epoch": 5.504960124489399,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024189517447043352,
      "loss": 5.0127,
      "mean_token_accuracy": 0.20855545550584792,
      "num_tokens": 122734187.0,
      "step": 70755
    },
    {
      "entropy": 5.940947484970093,
      "epoch": 5.505349153861117,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024186776076877843,
      "loss": 5.0211,
      "mean_token_accuracy": 0.22191838026046753,
      "num_tokens": 122742465.0,
      "step": 70760
    },
    {
      "entropy": 5.900040912628174,
      "epoch": 5.505738183232834,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024184034756985235,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21057606786489486,
      "num_tokens": 122751132.0,
      "step": 70765
    },
    {
      "entropy": 5.91504545211792,
      "epoch": 5.506127212604552,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024181293487407125,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21571872234344483,
      "num_tokens": 122761015.0,
      "step": 70770
    },
    {
      "entropy": 5.875730371475219,
      "epoch": 5.50651624197627,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000241785522681851,
      "loss": 4.9109,
      "mean_token_accuracy": 0.2188305139541626,
      "num_tokens": 122769357.0,
      "step": 70775
    },
    {
      "entropy": 5.871375370025635,
      "epoch": 5.5069052713479865,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024175811099360757,
      "loss": 4.8887,
      "mean_token_accuracy": 0.21324459016323088,
      "num_tokens": 122777446.0,
      "step": 70780
    },
    {
      "entropy": 5.812668943405152,
      "epoch": 5.507294300719704,
      "grad_norm": 1.59375,
      "learning_rate": 0.00024173069980975691,
      "loss": 4.8209,
      "mean_token_accuracy": 0.22518251687288285,
      "num_tokens": 122785650.0,
      "step": 70785
    },
    {
      "entropy": 5.845976972579956,
      "epoch": 5.507683330091422,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024170328913071493,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22716088891029357,
      "num_tokens": 122794212.0,
      "step": 70790
    },
    {
      "entropy": 5.897595977783203,
      "epoch": 5.5080723594631396,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024167587895689753,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21925583928823472,
      "num_tokens": 122803210.0,
      "step": 70795
    },
    {
      "entropy": 5.845195770263672,
      "epoch": 5.508461388834857,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024164846928872064,
      "loss": 4.8518,
      "mean_token_accuracy": 0.21931227445602416,
      "num_tokens": 122811351.0,
      "step": 70800
    },
    {
      "entropy": 5.908354043960571,
      "epoch": 5.508850418206575,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024162106012660012,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22653854191303252,
      "num_tokens": 122819809.0,
      "step": 70805
    },
    {
      "entropy": 5.829726839065552,
      "epoch": 5.509239447578292,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002415936514709519,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22574039846658706,
      "num_tokens": 122828677.0,
      "step": 70810
    },
    {
      "entropy": 5.935666751861572,
      "epoch": 5.5096284769500095,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002415662433221918,
      "loss": 5.003,
      "mean_token_accuracy": 0.21530951857566832,
      "num_tokens": 122836991.0,
      "step": 70815
    },
    {
      "entropy": 5.8208420753479,
      "epoch": 5.510017506321727,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00024153883568073576,
      "loss": 4.8385,
      "mean_token_accuracy": 0.21939749717712403,
      "num_tokens": 122845059.0,
      "step": 70820
    },
    {
      "entropy": 5.8439061641693115,
      "epoch": 5.510406535693445,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002415114285469997,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21516021937131882,
      "num_tokens": 122853590.0,
      "step": 70825
    },
    {
      "entropy": 5.8516028881072994,
      "epoch": 5.5107955650651625,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024148402192139935,
      "loss": 4.916,
      "mean_token_accuracy": 0.2161286488175392,
      "num_tokens": 122861824.0,
      "step": 70830
    },
    {
      "entropy": 5.931172895431518,
      "epoch": 5.51118459443688,
      "grad_norm": 1.5,
      "learning_rate": 0.00024145661580435063,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22111879289150238,
      "num_tokens": 122870583.0,
      "step": 70835
    },
    {
      "entropy": 5.858726453781128,
      "epoch": 5.511573623808598,
      "grad_norm": 1.5,
      "learning_rate": 0.00024142921019626933,
      "loss": 4.8475,
      "mean_token_accuracy": 0.22723608911037446,
      "num_tokens": 122878920.0,
      "step": 70840
    },
    {
      "entropy": 5.868453073501587,
      "epoch": 5.511962653180315,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024140180509757143,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21629225760698317,
      "num_tokens": 122887686.0,
      "step": 70845
    },
    {
      "entropy": 5.921877241134643,
      "epoch": 5.5123516825520325,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024137440050867265,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22163681983947753,
      "num_tokens": 122897007.0,
      "step": 70850
    },
    {
      "entropy": 5.973113489151001,
      "epoch": 5.51274071192375,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024134699642998882,
      "loss": 5.0276,
      "mean_token_accuracy": 0.207808618247509,
      "num_tokens": 122905710.0,
      "step": 70855
    },
    {
      "entropy": 5.976579189300537,
      "epoch": 5.513129741295468,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024131959286193572,
      "loss": 5.006,
      "mean_token_accuracy": 0.2133365899324417,
      "num_tokens": 122915538.0,
      "step": 70860
    },
    {
      "entropy": 5.942613887786865,
      "epoch": 5.5135187706671855,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024129218980492917,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21609801352024077,
      "num_tokens": 122924027.0,
      "step": 70865
    },
    {
      "entropy": 5.838892126083374,
      "epoch": 5.513907800038903,
      "grad_norm": 1.421875,
      "learning_rate": 0.000241264787259385,
      "loss": 4.873,
      "mean_token_accuracy": 0.2217567205429077,
      "num_tokens": 122933406.0,
      "step": 70870
    },
    {
      "entropy": 5.893123579025269,
      "epoch": 5.51429682941062,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024123738522571902,
      "loss": 4.9767,
      "mean_token_accuracy": 0.20833575427532197,
      "num_tokens": 122941777.0,
      "step": 70875
    },
    {
      "entropy": 5.911886215209961,
      "epoch": 5.514685858782338,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002412099837043469,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2112295925617218,
      "num_tokens": 122950249.0,
      "step": 70880
    },
    {
      "entropy": 5.852481079101563,
      "epoch": 5.515074888154055,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024118258269568456,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2267287403345108,
      "num_tokens": 122959440.0,
      "step": 70885
    },
    {
      "entropy": 5.918970918655395,
      "epoch": 5.515463917525773,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00024115518220014771,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22516084611415862,
      "num_tokens": 122967703.0,
      "step": 70890
    },
    {
      "entropy": 5.913910865783691,
      "epoch": 5.515852946897491,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024112778221815206,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21686048209667205,
      "num_tokens": 122976058.0,
      "step": 70895
    },
    {
      "entropy": 5.834472608566284,
      "epoch": 5.5162419762692085,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024110038275011343,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22217390686273575,
      "num_tokens": 122985522.0,
      "step": 70900
    },
    {
      "entropy": 5.893016242980957,
      "epoch": 5.516631005640926,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024107298379644754,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21776696890592576,
      "num_tokens": 122994110.0,
      "step": 70905
    },
    {
      "entropy": 5.948649978637695,
      "epoch": 5.517020035012644,
      "grad_norm": 1.5,
      "learning_rate": 0.0002410455853575701,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21296276450157164,
      "num_tokens": 123003532.0,
      "step": 70910
    },
    {
      "entropy": 5.915206813812256,
      "epoch": 5.517409064384361,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024101818743389687,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22741225212812424,
      "num_tokens": 123012054.0,
      "step": 70915
    },
    {
      "entropy": 5.851957511901856,
      "epoch": 5.517798093756078,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024099079002584355,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22578116953372956,
      "num_tokens": 123021264.0,
      "step": 70920
    },
    {
      "entropy": 5.849913740158081,
      "epoch": 5.518187123127796,
      "grad_norm": 1.5625,
      "learning_rate": 0.00024096339313382586,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2147666797041893,
      "num_tokens": 123029546.0,
      "step": 70925
    },
    {
      "entropy": 5.7046126365661625,
      "epoch": 5.518576152499514,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00024093599675825954,
      "loss": 4.839,
      "mean_token_accuracy": 0.2178676173090935,
      "num_tokens": 123037959.0,
      "step": 70930
    },
    {
      "entropy": 5.847097969055175,
      "epoch": 5.5189651818712315,
      "grad_norm": 1.5,
      "learning_rate": 0.00024090860089956024,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2248156487941742,
      "num_tokens": 123046714.0,
      "step": 70935
    },
    {
      "entropy": 5.851279544830322,
      "epoch": 5.519354211242949,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002408812055581437,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22261950224637986,
      "num_tokens": 123055248.0,
      "step": 70940
    },
    {
      "entropy": 5.869126844406128,
      "epoch": 5.519743240614666,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024085381073442536,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23422433137893678,
      "num_tokens": 123063924.0,
      "step": 70945
    },
    {
      "entropy": 5.880530500411988,
      "epoch": 5.520132269986384,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024082641642882125,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21926818192005157,
      "num_tokens": 123072969.0,
      "step": 70950
    },
    {
      "entropy": 5.874475622177124,
      "epoch": 5.520521299358101,
      "grad_norm": 1.40625,
      "learning_rate": 0.00024079902264174686,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21264958530664443,
      "num_tokens": 123081893.0,
      "step": 70955
    },
    {
      "entropy": 5.939900732040405,
      "epoch": 5.520910328729819,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002407716293736179,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2108771562576294,
      "num_tokens": 123090011.0,
      "step": 70960
    },
    {
      "entropy": 5.756151580810547,
      "epoch": 5.521299358101537,
      "grad_norm": 2.3125,
      "learning_rate": 0.00024074423662484996,
      "loss": 4.7069,
      "mean_token_accuracy": 0.2429073706269264,
      "num_tokens": 123098451.0,
      "step": 70965
    },
    {
      "entropy": 5.930081939697265,
      "epoch": 5.5216883874732545,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002407168443958587,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21592566967010499,
      "num_tokens": 123106417.0,
      "step": 70970
    },
    {
      "entropy": 5.854647636413574,
      "epoch": 5.522077416844972,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024068945268705978,
      "loss": 4.88,
      "mean_token_accuracy": 0.22433900088071823,
      "num_tokens": 123115173.0,
      "step": 70975
    },
    {
      "entropy": 5.799871492385864,
      "epoch": 5.522466446216689,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00024066206149886876,
      "loss": 4.7568,
      "mean_token_accuracy": 0.23893425911664962,
      "num_tokens": 123123579.0,
      "step": 70980
    },
    {
      "entropy": 5.8202595710754395,
      "epoch": 5.522855475588407,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024063467083170132,
      "loss": 4.782,
      "mean_token_accuracy": 0.2245902732014656,
      "num_tokens": 123132035.0,
      "step": 70985
    },
    {
      "entropy": 5.936633348464966,
      "epoch": 5.523244504960124,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002406072806859731,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21390410810708999,
      "num_tokens": 123140955.0,
      "step": 70990
    },
    {
      "entropy": 5.85298433303833,
      "epoch": 5.523633534331842,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002405798910620996,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22086906731128692,
      "num_tokens": 123149411.0,
      "step": 70995
    },
    {
      "entropy": 5.902180242538452,
      "epoch": 5.52402256370356,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002405525019604965,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22430809885263442,
      "num_tokens": 123157777.0,
      "step": 71000
    },
    {
      "entropy": 5.900916624069214,
      "epoch": 5.5244115930752775,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00024052511338157928,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2209058955311775,
      "num_tokens": 123166164.0,
      "step": 71005
    },
    {
      "entropy": 5.856133508682251,
      "epoch": 5.524800622446994,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024049772532576364,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21820068806409837,
      "num_tokens": 123174085.0,
      "step": 71010
    },
    {
      "entropy": 5.916501092910766,
      "epoch": 5.525189651818712,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024047033779346505,
      "loss": 4.9532,
      "mean_token_accuracy": 0.21858211755752563,
      "num_tokens": 123183315.0,
      "step": 71015
    },
    {
      "entropy": 5.888610696792602,
      "epoch": 5.52557868119043,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002404429507850992,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2172047734260559,
      "num_tokens": 123192521.0,
      "step": 71020
    },
    {
      "entropy": 5.97922830581665,
      "epoch": 5.525967710562147,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002404155643010815,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21639478206634521,
      "num_tokens": 123201385.0,
      "step": 71025
    },
    {
      "entropy": 5.933700609207153,
      "epoch": 5.526356739933865,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024038817834182762,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22150099724531175,
      "num_tokens": 123209929.0,
      "step": 71030
    },
    {
      "entropy": 5.904707241058349,
      "epoch": 5.526745769305583,
      "grad_norm": 1.53125,
      "learning_rate": 0.00024036079290775308,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22024631053209304,
      "num_tokens": 123217707.0,
      "step": 71035
    },
    {
      "entropy": 5.881718635559082,
      "epoch": 5.5271347986773005,
      "grad_norm": 1.46875,
      "learning_rate": 0.00024033340799927338,
      "loss": 4.847,
      "mean_token_accuracy": 0.22537967562675476,
      "num_tokens": 123226409.0,
      "step": 71040
    },
    {
      "entropy": 5.8941045761108395,
      "epoch": 5.527523828049017,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000240306023616804,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2212715521454811,
      "num_tokens": 123234512.0,
      "step": 71045
    },
    {
      "entropy": 5.86834397315979,
      "epoch": 5.527912857420735,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00024027863976076047,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21857585608959199,
      "num_tokens": 123243553.0,
      "step": 71050
    },
    {
      "entropy": 5.812940168380737,
      "epoch": 5.528301886792453,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002402512564315583,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23126472681760787,
      "num_tokens": 123252337.0,
      "step": 71055
    },
    {
      "entropy": 5.842457389831543,
      "epoch": 5.52869091616417,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00024022387362961306,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22423473596572877,
      "num_tokens": 123261818.0,
      "step": 71060
    },
    {
      "entropy": 5.91920166015625,
      "epoch": 5.529079945535888,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024019649135534015,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22065941095352173,
      "num_tokens": 123271109.0,
      "step": 71065
    },
    {
      "entropy": 5.927752733230591,
      "epoch": 5.529468974907606,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00024016910960915512,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22036345601081847,
      "num_tokens": 123279854.0,
      "step": 71070
    },
    {
      "entropy": 5.792593908309937,
      "epoch": 5.529858004279323,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024014172839147336,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22208085358142854,
      "num_tokens": 123288594.0,
      "step": 71075
    },
    {
      "entropy": 5.815707731246948,
      "epoch": 5.53024703365104,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024011434770271033,
      "loss": 4.8905,
      "mean_token_accuracy": 0.2224251702427864,
      "num_tokens": 123297129.0,
      "step": 71080
    },
    {
      "entropy": 5.907900333404541,
      "epoch": 5.530636063022758,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0002400869675432817,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21807628870010376,
      "num_tokens": 123306404.0,
      "step": 71085
    },
    {
      "entropy": 5.908203554153443,
      "epoch": 5.531025092394476,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00024005958791360266,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2182743325829506,
      "num_tokens": 123315362.0,
      "step": 71090
    },
    {
      "entropy": 5.959443473815918,
      "epoch": 5.531414121766193,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024003220881408882,
      "loss": 4.9431,
      "mean_token_accuracy": 0.20926050543785096,
      "num_tokens": 123323661.0,
      "step": 71095
    },
    {
      "entropy": 5.908237075805664,
      "epoch": 5.531803151137911,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024000483024515552,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21583239436149598,
      "num_tokens": 123331508.0,
      "step": 71100
    },
    {
      "entropy": 5.8730137825012205,
      "epoch": 5.532192180509629,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023997745220721823,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22094352543354034,
      "num_tokens": 123340573.0,
      "step": 71105
    },
    {
      "entropy": 5.886836099624634,
      "epoch": 5.5325812098813465,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023995007470069236,
      "loss": 4.9681,
      "mean_token_accuracy": 0.21462431848049163,
      "num_tokens": 123348966.0,
      "step": 71110
    },
    {
      "entropy": 5.880202054977417,
      "epoch": 5.532970239253063,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023992269772599335,
      "loss": 4.8393,
      "mean_token_accuracy": 0.23065102100372314,
      "num_tokens": 123357136.0,
      "step": 71115
    },
    {
      "entropy": 5.884918642044068,
      "epoch": 5.533359268624781,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023989532128353654,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21662943065166473,
      "num_tokens": 123366282.0,
      "step": 71120
    },
    {
      "entropy": 5.917414379119873,
      "epoch": 5.533748297996499,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002398679453737374,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21515192091464996,
      "num_tokens": 123376235.0,
      "step": 71125
    },
    {
      "entropy": 5.829476356506348,
      "epoch": 5.534137327368216,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023984056999701121,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21856586635112762,
      "num_tokens": 123385184.0,
      "step": 71130
    },
    {
      "entropy": 5.896203708648682,
      "epoch": 5.534526356739934,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00023981319515377347,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21635591089725495,
      "num_tokens": 123393507.0,
      "step": 71135
    },
    {
      "entropy": 5.874002981185913,
      "epoch": 5.534915386111652,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00023978582084443939,
      "loss": 4.9587,
      "mean_token_accuracy": 0.20964184105396272,
      "num_tokens": 123402347.0,
      "step": 71140
    },
    {
      "entropy": 5.8775359153747555,
      "epoch": 5.535304415483369,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002397584470694244,
      "loss": 4.8438,
      "mean_token_accuracy": 0.2240394324064255,
      "num_tokens": 123410417.0,
      "step": 71145
    },
    {
      "entropy": 5.845563888549805,
      "epoch": 5.535693444855086,
      "grad_norm": 1.515625,
      "learning_rate": 0.000239731073829144,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22936154454946517,
      "num_tokens": 123418851.0,
      "step": 71150
    },
    {
      "entropy": 5.858168363571167,
      "epoch": 5.536082474226804,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002397037011240134,
      "loss": 4.8862,
      "mean_token_accuracy": 0.21338058412075042,
      "num_tokens": 123427188.0,
      "step": 71155
    },
    {
      "entropy": 5.827371311187744,
      "epoch": 5.536471503598522,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00023967632895444796,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2228875294327736,
      "num_tokens": 123435412.0,
      "step": 71160
    },
    {
      "entropy": 5.901863479614258,
      "epoch": 5.536860532970239,
      "grad_norm": 1.453125,
      "learning_rate": 0.000239648957320863,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22370035499334334,
      "num_tokens": 123443548.0,
      "step": 71165
    },
    {
      "entropy": 5.823812627792359,
      "epoch": 5.537249562341957,
      "grad_norm": 1.5,
      "learning_rate": 0.0002396215862236738,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2151074767112732,
      "num_tokens": 123452803.0,
      "step": 71170
    },
    {
      "entropy": 5.831398677825928,
      "epoch": 5.537638591713675,
      "grad_norm": 1.5625,
      "learning_rate": 0.00023959421566329574,
      "loss": 4.8128,
      "mean_token_accuracy": 0.22561281323432922,
      "num_tokens": 123461532.0,
      "step": 71175
    },
    {
      "entropy": 5.8820453643798825,
      "epoch": 5.538027621085392,
      "grad_norm": 1.40625,
      "learning_rate": 0.00023956684564014415,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22349590957164764,
      "num_tokens": 123471282.0,
      "step": 71180
    },
    {
      "entropy": 5.888096809387207,
      "epoch": 5.538416650457109,
      "grad_norm": 1.515625,
      "learning_rate": 0.00023953947615463423,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21931317150592805,
      "num_tokens": 123480142.0,
      "step": 71185
    },
    {
      "entropy": 5.808627367019653,
      "epoch": 5.538805679828827,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002395121072071813,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2217669129371643,
      "num_tokens": 123488963.0,
      "step": 71190
    },
    {
      "entropy": 5.917604160308838,
      "epoch": 5.539194709200545,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023948473879820072,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2157071828842163,
      "num_tokens": 123498681.0,
      "step": 71195
    },
    {
      "entropy": 5.907694482803345,
      "epoch": 5.539583738572262,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023945737092810764,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21854503750801085,
      "num_tokens": 123507935.0,
      "step": 71200
    },
    {
      "entropy": 5.85070013999939,
      "epoch": 5.53997276794398,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002394300035973175,
      "loss": 4.8997,
      "mean_token_accuracy": 0.22331675589084626,
      "num_tokens": 123516674.0,
      "step": 71205
    },
    {
      "entropy": 5.836267280578613,
      "epoch": 5.540361797315697,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023940263680624535,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21735917627811432,
      "num_tokens": 123525078.0,
      "step": 71210
    },
    {
      "entropy": 5.8304518699646,
      "epoch": 5.540750826687415,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023937527055530643,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22267885655164718,
      "num_tokens": 123533000.0,
      "step": 71215
    },
    {
      "entropy": 5.871550798416138,
      "epoch": 5.541139856059132,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002393479048449162,
      "loss": 4.9743,
      "mean_token_accuracy": 0.21853255927562715,
      "num_tokens": 123541410.0,
      "step": 71220
    },
    {
      "entropy": 5.933336162567139,
      "epoch": 5.54152888543085,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00023932053967548971,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22237518131732942,
      "num_tokens": 123550016.0,
      "step": 71225
    },
    {
      "entropy": 5.914285612106323,
      "epoch": 5.541917914802568,
      "grad_norm": 1.59375,
      "learning_rate": 0.00023929317504744235,
      "loss": 4.8213,
      "mean_token_accuracy": 0.23212036043405532,
      "num_tokens": 123557961.0,
      "step": 71230
    },
    {
      "entropy": 5.890051746368409,
      "epoch": 5.542306944174285,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00023926581096118916,
      "loss": 4.9335,
      "mean_token_accuracy": 0.21679433286190034,
      "num_tokens": 123565541.0,
      "step": 71235
    },
    {
      "entropy": 5.883606243133545,
      "epoch": 5.542695973546003,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002392384474171454,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2160152465105057,
      "num_tokens": 123574902.0,
      "step": 71240
    },
    {
      "entropy": 5.88239221572876,
      "epoch": 5.543085002917721,
      "grad_norm": 1.578125,
      "learning_rate": 0.00023921108441572625,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22684556990861893,
      "num_tokens": 123583278.0,
      "step": 71245
    },
    {
      "entropy": 5.961444425582886,
      "epoch": 5.543474032289438,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000239183721957347,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2188818171620369,
      "num_tokens": 123591754.0,
      "step": 71250
    },
    {
      "entropy": 5.913015365600586,
      "epoch": 5.543863061661155,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00023915636004242274,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21619240194559097,
      "num_tokens": 123600432.0,
      "step": 71255
    },
    {
      "entropy": 5.889128351211548,
      "epoch": 5.544252091032873,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002391289986713686,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2184408888220787,
      "num_tokens": 123608832.0,
      "step": 71260
    },
    {
      "entropy": 5.8438990116119385,
      "epoch": 5.544641120404591,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023910163784459986,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22291169911623002,
      "num_tokens": 123617615.0,
      "step": 71265
    },
    {
      "entropy": 5.872749662399292,
      "epoch": 5.545030149776308,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002390742775625317,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22159286439418793,
      "num_tokens": 123625536.0,
      "step": 71270
    },
    {
      "entropy": 5.860332775115967,
      "epoch": 5.545419179148025,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023904691782557908,
      "loss": 4.9099,
      "mean_token_accuracy": 0.21916413307189941,
      "num_tokens": 123635300.0,
      "step": 71275
    },
    {
      "entropy": 5.868891954421997,
      "epoch": 5.545808208519743,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00023901955863415725,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22071472704410552,
      "num_tokens": 123643180.0,
      "step": 71280
    },
    {
      "entropy": 5.796931743621826,
      "epoch": 5.546197237891461,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00023899219998868142,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22456008195877075,
      "num_tokens": 123651281.0,
      "step": 71285
    },
    {
      "entropy": 5.840266370773316,
      "epoch": 5.546586267263178,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023896484188956664,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23620522767305374,
      "num_tokens": 123659353.0,
      "step": 71290
    },
    {
      "entropy": 5.883699655532837,
      "epoch": 5.546975296634896,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000238937484337228,
      "loss": 4.929,
      "mean_token_accuracy": 0.219449745118618,
      "num_tokens": 123667561.0,
      "step": 71295
    },
    {
      "entropy": 5.8165771484375,
      "epoch": 5.547364326006614,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00023891012733208062,
      "loss": 4.8074,
      "mean_token_accuracy": 0.2314310625195503,
      "num_tokens": 123675847.0,
      "step": 71300
    },
    {
      "entropy": 5.911809253692627,
      "epoch": 5.547753355378331,
      "grad_norm": 1.578125,
      "learning_rate": 0.00023888277087453962,
      "loss": 4.9327,
      "mean_token_accuracy": 0.21171299815177919,
      "num_tokens": 123684055.0,
      "step": 71305
    },
    {
      "entropy": 5.923636817932129,
      "epoch": 5.548142384750049,
      "grad_norm": 1.578125,
      "learning_rate": 0.00023885541496502017,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20587049424648285,
      "num_tokens": 123692104.0,
      "step": 71310
    },
    {
      "entropy": 5.8724524021148685,
      "epoch": 5.548531414121766,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00023882805960393722,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22192499041557312,
      "num_tokens": 123700815.0,
      "step": 71315
    },
    {
      "entropy": 5.969444942474365,
      "epoch": 5.548920443493484,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002388007047917059,
      "loss": 5.011,
      "mean_token_accuracy": 0.21902494877576828,
      "num_tokens": 123708931.0,
      "step": 71320
    },
    {
      "entropy": 5.888339471817017,
      "epoch": 5.549309472865201,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023877335052874122,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21822427362203597,
      "num_tokens": 123717691.0,
      "step": 71325
    },
    {
      "entropy": 5.886244821548462,
      "epoch": 5.549698502236919,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023874599681545833,
      "loss": 4.9743,
      "mean_token_accuracy": 0.21292585283517837,
      "num_tokens": 123726798.0,
      "step": 71330
    },
    {
      "entropy": 5.84439697265625,
      "epoch": 5.550087531608637,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002387186436522722,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21453494131565093,
      "num_tokens": 123735806.0,
      "step": 71335
    },
    {
      "entropy": 5.89442138671875,
      "epoch": 5.550476560980354,
      "grad_norm": 1.5,
      "learning_rate": 0.0002386912910395979,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21643429845571518,
      "num_tokens": 123744484.0,
      "step": 71340
    },
    {
      "entropy": 5.880376958847046,
      "epoch": 5.550865590352071,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023866393897785049,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22609055638313294,
      "num_tokens": 123753431.0,
      "step": 71345
    },
    {
      "entropy": 5.914479827880859,
      "epoch": 5.551254619723789,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023863658746744494,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21664968878030777,
      "num_tokens": 123762179.0,
      "step": 71350
    },
    {
      "entropy": 5.913131284713745,
      "epoch": 5.551643649095507,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023860923650879636,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22453399002552032,
      "num_tokens": 123771205.0,
      "step": 71355
    },
    {
      "entropy": 5.919298267364502,
      "epoch": 5.552032678467224,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002385818861023197,
      "loss": 4.8674,
      "mean_token_accuracy": 0.2230959415435791,
      "num_tokens": 123779365.0,
      "step": 71360
    },
    {
      "entropy": 5.865307903289795,
      "epoch": 5.552421707838942,
      "grad_norm": 1.421875,
      "learning_rate": 0.00023855453624842986,
      "loss": 4.922,
      "mean_token_accuracy": 0.21516790837049485,
      "num_tokens": 123787684.0,
      "step": 71365
    },
    {
      "entropy": 5.837750673294067,
      "epoch": 5.55281073721066,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000238527186947542,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22070905119180678,
      "num_tokens": 123797116.0,
      "step": 71370
    },
    {
      "entropy": 5.914835262298584,
      "epoch": 5.553199766582377,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000238499838200071,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21112614572048188,
      "num_tokens": 123806033.0,
      "step": 71375
    },
    {
      "entropy": 5.9647355556488035,
      "epoch": 5.553588795954094,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023847249000643185,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21138681918382646,
      "num_tokens": 123815181.0,
      "step": 71380
    },
    {
      "entropy": 5.875647449493409,
      "epoch": 5.553977825325812,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00023844514236703952,
      "loss": 4.9437,
      "mean_token_accuracy": 0.22360238581895828,
      "num_tokens": 123824450.0,
      "step": 71385
    },
    {
      "entropy": 5.86277027130127,
      "epoch": 5.5543668546975296,
      "grad_norm": 1.515625,
      "learning_rate": 0.000238417795282309,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2161511018872261,
      "num_tokens": 123833159.0,
      "step": 71390
    },
    {
      "entropy": 5.9063958644866945,
      "epoch": 5.554755884069247,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00023839044875265526,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2193934887647629,
      "num_tokens": 123841305.0,
      "step": 71395
    },
    {
      "entropy": 5.858890533447266,
      "epoch": 5.555144913440965,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002383631027784931,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2237251415848732,
      "num_tokens": 123850523.0,
      "step": 71400
    },
    {
      "entropy": 5.87549204826355,
      "epoch": 5.555533942812683,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023833575736023754,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2199530377984047,
      "num_tokens": 123859047.0,
      "step": 71405
    },
    {
      "entropy": 5.863650798797607,
      "epoch": 5.5559229721843995,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023830841249830348,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22800556272268296,
      "num_tokens": 123867708.0,
      "step": 71410
    },
    {
      "entropy": 5.863533544540405,
      "epoch": 5.556312001556117,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002382810681931059,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2248632863163948,
      "num_tokens": 123876653.0,
      "step": 71415
    },
    {
      "entropy": 5.84511170387268,
      "epoch": 5.556701030927835,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00023825372444505967,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22102579176425935,
      "num_tokens": 123885626.0,
      "step": 71420
    },
    {
      "entropy": 5.861737585067749,
      "epoch": 5.5570900602995525,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023822638125457974,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22003293633461,
      "num_tokens": 123894409.0,
      "step": 71425
    },
    {
      "entropy": 5.898362827301026,
      "epoch": 5.55747908967127,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023819903862208093,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22036586999893187,
      "num_tokens": 123903174.0,
      "step": 71430
    },
    {
      "entropy": 5.897608661651612,
      "epoch": 5.557868119042988,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002381716965479781,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22472616881132126,
      "num_tokens": 123911367.0,
      "step": 71435
    },
    {
      "entropy": 5.809305477142334,
      "epoch": 5.558257148414706,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023814435503268617,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23224235475063323,
      "num_tokens": 123919719.0,
      "step": 71440
    },
    {
      "entropy": 5.90183310508728,
      "epoch": 5.558646177786423,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023811701407661995,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2156620055437088,
      "num_tokens": 123928109.0,
      "step": 71445
    },
    {
      "entropy": 5.84024133682251,
      "epoch": 5.55903520715814,
      "grad_norm": 1.515625,
      "learning_rate": 0.00023808967368019434,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22007009834051133,
      "num_tokens": 123936701.0,
      "step": 71450
    },
    {
      "entropy": 5.874726963043213,
      "epoch": 5.559424236529858,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002380623338438242,
      "loss": 4.82,
      "mean_token_accuracy": 0.2251657336950302,
      "num_tokens": 123945466.0,
      "step": 71455
    },
    {
      "entropy": 5.897744274139404,
      "epoch": 5.5598132659015755,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002380349945679244,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22062162756919862,
      "num_tokens": 123954016.0,
      "step": 71460
    },
    {
      "entropy": 5.842208385467529,
      "epoch": 5.560202295273293,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023800765585290967,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2231471136212349,
      "num_tokens": 123962608.0,
      "step": 71465
    },
    {
      "entropy": 5.8327594757080075,
      "epoch": 5.560591324645011,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023798031769919488,
      "loss": 4.8545,
      "mean_token_accuracy": 0.21783513575792313,
      "num_tokens": 123970985.0,
      "step": 71470
    },
    {
      "entropy": 5.859819841384888,
      "epoch": 5.560980354016729,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002379529801071949,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22451525628566743,
      "num_tokens": 123978977.0,
      "step": 71475
    },
    {
      "entropy": 5.842158412933349,
      "epoch": 5.561369383388445,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002379256430773244,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22186188995838166,
      "num_tokens": 123987497.0,
      "step": 71480
    },
    {
      "entropy": 5.917280435562134,
      "epoch": 5.561758412760163,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00023789830660999822,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20742530524730682,
      "num_tokens": 123995890.0,
      "step": 71485
    },
    {
      "entropy": 5.892652606964111,
      "epoch": 5.562147442131881,
      "grad_norm": 1.5,
      "learning_rate": 0.0002378709707056313,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21924246847629547,
      "num_tokens": 124004108.0,
      "step": 71490
    },
    {
      "entropy": 5.8609894752502445,
      "epoch": 5.5625364715035985,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00023784363536463827,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22600919008255005,
      "num_tokens": 124012736.0,
      "step": 71495
    },
    {
      "entropy": 5.85231237411499,
      "epoch": 5.562925500875316,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00023781630058743392,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22204794287681578,
      "num_tokens": 124021315.0,
      "step": 71500
    },
    {
      "entropy": 5.830487871170044,
      "epoch": 5.563314530247034,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002377889663744331,
      "loss": 4.8588,
      "mean_token_accuracy": 0.2235457956790924,
      "num_tokens": 124029502.0,
      "step": 71505
    },
    {
      "entropy": 5.852705240249634,
      "epoch": 5.563703559618752,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002377616327260504,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2234371080994606,
      "num_tokens": 124039009.0,
      "step": 71510
    },
    {
      "entropy": 5.916640949249268,
      "epoch": 5.564092588990468,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023773429964270065,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21472072452306748,
      "num_tokens": 124048142.0,
      "step": 71515
    },
    {
      "entropy": 5.8968117237091064,
      "epoch": 5.564481618362186,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002377069671247986,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21522412747144698,
      "num_tokens": 124057005.0,
      "step": 71520
    },
    {
      "entropy": 5.875715065002441,
      "epoch": 5.564870647733904,
      "grad_norm": 1.375,
      "learning_rate": 0.00023767963517275903,
      "loss": 4.9094,
      "mean_token_accuracy": 0.21324016153812408,
      "num_tokens": 124066234.0,
      "step": 71525
    },
    {
      "entropy": 5.84584321975708,
      "epoch": 5.5652596771056215,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002376523037869965,
      "loss": 4.7759,
      "mean_token_accuracy": 0.22949904948472977,
      "num_tokens": 124074838.0,
      "step": 71530
    },
    {
      "entropy": 5.8684920310974125,
      "epoch": 5.565648706477339,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023762497296792585,
      "loss": 4.8341,
      "mean_token_accuracy": 0.2179991438984871,
      "num_tokens": 124083638.0,
      "step": 71535
    },
    {
      "entropy": 5.879748725891114,
      "epoch": 5.566037735849057,
      "grad_norm": 1.5,
      "learning_rate": 0.00023759764271596173,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23106204718351364,
      "num_tokens": 124092056.0,
      "step": 71540
    },
    {
      "entropy": 5.91393723487854,
      "epoch": 5.566426765220774,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023757031303151888,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2164360076189041,
      "num_tokens": 124100682.0,
      "step": 71545
    },
    {
      "entropy": 5.86744704246521,
      "epoch": 5.566815794592491,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023754298391501195,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21989387273788452,
      "num_tokens": 124109007.0,
      "step": 71550
    },
    {
      "entropy": 5.90716962814331,
      "epoch": 5.567204823964209,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002375156553668556,
      "loss": 4.954,
      "mean_token_accuracy": 0.21844569742679595,
      "num_tokens": 124117891.0,
      "step": 71555
    },
    {
      "entropy": 5.872766590118408,
      "epoch": 5.567593853335927,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002374883273874645,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2165599435567856,
      "num_tokens": 124126016.0,
      "step": 71560
    },
    {
      "entropy": 5.841198110580445,
      "epoch": 5.5679828827076445,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002374609999772534,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21689030528068542,
      "num_tokens": 124134550.0,
      "step": 71565
    },
    {
      "entropy": 5.854798316955566,
      "epoch": 5.568371912079362,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00023743367313663682,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2210424557328224,
      "num_tokens": 124142890.0,
      "step": 71570
    },
    {
      "entropy": 5.929487085342407,
      "epoch": 5.56876094145108,
      "grad_norm": 1.5,
      "learning_rate": 0.00023740634686602945,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21308027505874633,
      "num_tokens": 124151922.0,
      "step": 71575
    },
    {
      "entropy": 5.916885614395142,
      "epoch": 5.569149970822798,
      "grad_norm": 1.375,
      "learning_rate": 0.00023737902116584598,
      "loss": 4.985,
      "mean_token_accuracy": 0.20874694734811783,
      "num_tokens": 124160233.0,
      "step": 71580
    },
    {
      "entropy": 5.835113954544068,
      "epoch": 5.569539000194514,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023735169603650104,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22175232768058778,
      "num_tokens": 124167735.0,
      "step": 71585
    },
    {
      "entropy": 5.880920124053955,
      "epoch": 5.569928029566232,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023732437147840906,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21433311253786086,
      "num_tokens": 124176773.0,
      "step": 71590
    },
    {
      "entropy": 5.93673038482666,
      "epoch": 5.57031705893795,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00023729704749198482,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21146591305732726,
      "num_tokens": 124185573.0,
      "step": 71595
    },
    {
      "entropy": 5.892629766464234,
      "epoch": 5.5707060883096675,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023726972407764286,
      "loss": 4.7779,
      "mean_token_accuracy": 0.23135612308979034,
      "num_tokens": 124193750.0,
      "step": 71600
    },
    {
      "entropy": 5.964342403411865,
      "epoch": 5.571095117681385,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023724240123579782,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20909804552793504,
      "num_tokens": 124203082.0,
      "step": 71605
    },
    {
      "entropy": 5.873932647705078,
      "epoch": 5.571484147053102,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023721507896686412,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22478119283914566,
      "num_tokens": 124211748.0,
      "step": 71610
    },
    {
      "entropy": 5.925532817840576,
      "epoch": 5.57187317642482,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00023718775727125658,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21337562054395676,
      "num_tokens": 124220626.0,
      "step": 71615
    },
    {
      "entropy": 5.927702569961548,
      "epoch": 5.572262205796537,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00023716043614938965,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2146678552031517,
      "num_tokens": 124229213.0,
      "step": 71620
    },
    {
      "entropy": 5.950141763687133,
      "epoch": 5.572651235168255,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023713311560167783,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2136816844344139,
      "num_tokens": 124238142.0,
      "step": 71625
    },
    {
      "entropy": 5.922205591201783,
      "epoch": 5.573040264539973,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023710579562853573,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21990497261285782,
      "num_tokens": 124246828.0,
      "step": 71630
    },
    {
      "entropy": 5.8808447360992435,
      "epoch": 5.5734292939116905,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023707847623037783,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22464536279439926,
      "num_tokens": 124255950.0,
      "step": 71635
    },
    {
      "entropy": 5.864926528930664,
      "epoch": 5.573818323283408,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00023705115740761874,
      "loss": 4.8268,
      "mean_token_accuracy": 0.2276832193136215,
      "num_tokens": 124264328.0,
      "step": 71640
    },
    {
      "entropy": 5.898495435714722,
      "epoch": 5.574207352655126,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023702383916067294,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22460647374391557,
      "num_tokens": 124272392.0,
      "step": 71645
    },
    {
      "entropy": 5.879801893234253,
      "epoch": 5.574596382026843,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00023699652148995488,
      "loss": 4.8391,
      "mean_token_accuracy": 0.226133069396019,
      "num_tokens": 124280635.0,
      "step": 71650
    },
    {
      "entropy": 5.82718505859375,
      "epoch": 5.57498541139856,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00023696920439587917,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2218429833650589,
      "num_tokens": 124289549.0,
      "step": 71655
    },
    {
      "entropy": 5.843733310699463,
      "epoch": 5.575374440770278,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023694188787886023,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22525503784418105,
      "num_tokens": 124298498.0,
      "step": 71660
    },
    {
      "entropy": 5.9289967060089115,
      "epoch": 5.575763470141996,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002369145719393126,
      "loss": 4.9582,
      "mean_token_accuracy": 0.21312253922224045,
      "num_tokens": 124306840.0,
      "step": 71665
    },
    {
      "entropy": 5.880760526657104,
      "epoch": 5.5761524995137135,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002368872565776507,
      "loss": 4.939,
      "mean_token_accuracy": 0.21875345706939697,
      "num_tokens": 124315600.0,
      "step": 71670
    },
    {
      "entropy": 5.878965425491333,
      "epoch": 5.576541528885431,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023685994179428887,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20883547514677048,
      "num_tokens": 124325042.0,
      "step": 71675
    },
    {
      "entropy": 5.938892936706543,
      "epoch": 5.576930558257148,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00023683262758964185,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22147416770458223,
      "num_tokens": 124333882.0,
      "step": 71680
    },
    {
      "entropy": 5.938407945632934,
      "epoch": 5.577319587628866,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023680531396412397,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2097380205988884,
      "num_tokens": 124342771.0,
      "step": 71685
    },
    {
      "entropy": 5.8764979362487795,
      "epoch": 5.577708617000583,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023677800091814966,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2193641409277916,
      "num_tokens": 124350832.0,
      "step": 71690
    },
    {
      "entropy": 5.835262680053711,
      "epoch": 5.578097646372301,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002367506884521334,
      "loss": 4.8517,
      "mean_token_accuracy": 0.21996341198682784,
      "num_tokens": 124359506.0,
      "step": 71695
    },
    {
      "entropy": 5.835062694549561,
      "epoch": 5.578486675744019,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023672337656648951,
      "loss": 4.956,
      "mean_token_accuracy": 0.21816186755895614,
      "num_tokens": 124368062.0,
      "step": 71700
    },
    {
      "entropy": 5.883343267440796,
      "epoch": 5.5788757051157365,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002366960652616324,
      "loss": 4.919,
      "mean_token_accuracy": 0.21482980251312256,
      "num_tokens": 124376619.0,
      "step": 71705
    },
    {
      "entropy": 5.852959299087525,
      "epoch": 5.579264734487454,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002366687545379766,
      "loss": 4.8663,
      "mean_token_accuracy": 0.2251165419816971,
      "num_tokens": 124385696.0,
      "step": 71710
    },
    {
      "entropy": 5.824714088439942,
      "epoch": 5.579653763859171,
      "grad_norm": 1.515625,
      "learning_rate": 0.00023664144439593644,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22657761871814727,
      "num_tokens": 124393981.0,
      "step": 71715
    },
    {
      "entropy": 5.906927013397217,
      "epoch": 5.580042793230889,
      "grad_norm": 1.640625,
      "learning_rate": 0.00023661413483592626,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21867086887359619,
      "num_tokens": 124402629.0,
      "step": 71720
    },
    {
      "entropy": 5.8916889190673825,
      "epoch": 5.580431822602606,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023658682585836049,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22116289138793946,
      "num_tokens": 124411704.0,
      "step": 71725
    },
    {
      "entropy": 5.91919903755188,
      "epoch": 5.580820851974324,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00023655951746365356,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21445553600788117,
      "num_tokens": 124420858.0,
      "step": 71730
    },
    {
      "entropy": 5.933084869384766,
      "epoch": 5.581209881346042,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023653220965221973,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2231084555387497,
      "num_tokens": 124430164.0,
      "step": 71735
    },
    {
      "entropy": 5.925378227233887,
      "epoch": 5.5815989107177595,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00023650490242447338,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22054435163736344,
      "num_tokens": 124438818.0,
      "step": 71740
    },
    {
      "entropy": 5.905390691757202,
      "epoch": 5.581987940089476,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023647759578082883,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21851639598608016,
      "num_tokens": 124447520.0,
      "step": 71745
    },
    {
      "entropy": 5.86979489326477,
      "epoch": 5.582376969461194,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00023645028972170046,
      "loss": 4.839,
      "mean_token_accuracy": 0.22222554683685303,
      "num_tokens": 124456340.0,
      "step": 71750
    },
    {
      "entropy": 5.848875665664673,
      "epoch": 5.582765998832912,
      "grad_norm": 1.5,
      "learning_rate": 0.00023642298424750264,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2159243181347847,
      "num_tokens": 124464897.0,
      "step": 71755
    },
    {
      "entropy": 5.870050096511841,
      "epoch": 5.583155028204629,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023639567935864958,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2175465151667595,
      "num_tokens": 124473912.0,
      "step": 71760
    },
    {
      "entropy": 5.843994522094727,
      "epoch": 5.583544057576347,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002363683750555557,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2141558364033699,
      "num_tokens": 124482086.0,
      "step": 71765
    },
    {
      "entropy": 5.849083137512207,
      "epoch": 5.583933086948065,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002363410713386352,
      "loss": 4.894,
      "mean_token_accuracy": 0.22688110917806625,
      "num_tokens": 124491134.0,
      "step": 71770
    },
    {
      "entropy": 5.890615606307984,
      "epoch": 5.5843221163197825,
      "grad_norm": 1.375,
      "learning_rate": 0.00023631376820830248,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22537148147821426,
      "num_tokens": 124499266.0,
      "step": 71775
    },
    {
      "entropy": 5.918025970458984,
      "epoch": 5.5847111456915,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002362864656649717,
      "loss": 4.889,
      "mean_token_accuracy": 0.2198265850543976,
      "num_tokens": 124508183.0,
      "step": 71780
    },
    {
      "entropy": 5.927316474914551,
      "epoch": 5.585100175063217,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023625916370905714,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22319251596927642,
      "num_tokens": 124516700.0,
      "step": 71785
    },
    {
      "entropy": 5.916544198989868,
      "epoch": 5.585489204434935,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023623186234097315,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21733037680387496,
      "num_tokens": 124525529.0,
      "step": 71790
    },
    {
      "entropy": 5.899367904663086,
      "epoch": 5.585878233806652,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00023620456156113397,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22569195032119752,
      "num_tokens": 124533770.0,
      "step": 71795
    },
    {
      "entropy": 5.826066303253174,
      "epoch": 5.58626726317837,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023617726136995377,
      "loss": 4.8696,
      "mean_token_accuracy": 0.21504239737987518,
      "num_tokens": 124543605.0,
      "step": 71800
    },
    {
      "entropy": 5.94782338142395,
      "epoch": 5.586656292550088,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023614996176784687,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2155761420726776,
      "num_tokens": 124552377.0,
      "step": 71805
    },
    {
      "entropy": 5.857422494888306,
      "epoch": 5.5870453219218055,
      "grad_norm": 1.421875,
      "learning_rate": 0.00023612266275522737,
      "loss": 4.8677,
      "mean_token_accuracy": 0.229761104285717,
      "num_tokens": 124560557.0,
      "step": 71810
    },
    {
      "entropy": 5.914006662368775,
      "epoch": 5.587434351293522,
      "grad_norm": 1.734375,
      "learning_rate": 0.00023609536433250966,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22052199989557267,
      "num_tokens": 124568405.0,
      "step": 71815
    },
    {
      "entropy": 5.8679533958435055,
      "epoch": 5.58782338066524,
      "grad_norm": 1.421875,
      "learning_rate": 0.00023606806650010786,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21538047045469283,
      "num_tokens": 124576658.0,
      "step": 71820
    },
    {
      "entropy": 5.844242525100708,
      "epoch": 5.588212410036958,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023604076925843622,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2277559369802475,
      "num_tokens": 124584945.0,
      "step": 71825
    },
    {
      "entropy": 5.877267265319825,
      "epoch": 5.588601439408675,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023601347260790884,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21626728177070617,
      "num_tokens": 124593920.0,
      "step": 71830
    },
    {
      "entropy": 5.96894941329956,
      "epoch": 5.588990468780393,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00023598617654894006,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2106095477938652,
      "num_tokens": 124603076.0,
      "step": 71835
    },
    {
      "entropy": 5.850470876693725,
      "epoch": 5.589379498152111,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023595888108194385,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2315011590719223,
      "num_tokens": 124611561.0,
      "step": 71840
    },
    {
      "entropy": 5.871891164779663,
      "epoch": 5.5897685275238285,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002359315862073345,
      "loss": 4.97,
      "mean_token_accuracy": 0.20848801583051682,
      "num_tokens": 124619751.0,
      "step": 71845
    },
    {
      "entropy": 5.923875570297241,
      "epoch": 5.590157556895545,
      "grad_norm": 1.375,
      "learning_rate": 0.0002359042919255262,
      "loss": 4.9408,
      "mean_token_accuracy": 0.219897823035717,
      "num_tokens": 124628886.0,
      "step": 71850
    },
    {
      "entropy": 5.934640407562256,
      "epoch": 5.590546586267263,
      "grad_norm": 1.40625,
      "learning_rate": 0.00023587699823693304,
      "loss": 4.9869,
      "mean_token_accuracy": 0.20579180419445037,
      "num_tokens": 124637632.0,
      "step": 71855
    },
    {
      "entropy": 5.866040134429932,
      "epoch": 5.590935615638981,
      "grad_norm": 1.5625,
      "learning_rate": 0.00023584970514196912,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2247046172618866,
      "num_tokens": 124646597.0,
      "step": 71860
    },
    {
      "entropy": 5.828413629531861,
      "epoch": 5.591324645010698,
      "grad_norm": 1.515625,
      "learning_rate": 0.00023582241264104864,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2165502443909645,
      "num_tokens": 124655203.0,
      "step": 71865
    },
    {
      "entropy": 5.861328649520874,
      "epoch": 5.591713674382416,
      "grad_norm": 1.40625,
      "learning_rate": 0.00023579512073458565,
      "loss": 4.9027,
      "mean_token_accuracy": 0.21877923756837844,
      "num_tokens": 124663451.0,
      "step": 71870
    },
    {
      "entropy": 5.932674598693848,
      "epoch": 5.592102703754134,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00023576782942299424,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2149149551987648,
      "num_tokens": 124671593.0,
      "step": 71875
    },
    {
      "entropy": 5.88585467338562,
      "epoch": 5.592491733125851,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002357405387066886,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21367121934890748,
      "num_tokens": 124680901.0,
      "step": 71880
    },
    {
      "entropy": 5.852211904525757,
      "epoch": 5.592880762497568,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002357132485860829,
      "loss": 4.8388,
      "mean_token_accuracy": 0.2252211168408394,
      "num_tokens": 124689404.0,
      "step": 71885
    },
    {
      "entropy": 5.897069597244263,
      "epoch": 5.593269791869286,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023568595906159107,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21659686118364335,
      "num_tokens": 124697813.0,
      "step": 71890
    },
    {
      "entropy": 5.89212703704834,
      "epoch": 5.593658821241004,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00023565867013362725,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21245740801095964,
      "num_tokens": 124706975.0,
      "step": 71895
    },
    {
      "entropy": 5.9368408203125,
      "epoch": 5.594047850612721,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00023563138180260552,
      "loss": 4.9144,
      "mean_token_accuracy": 0.21478965878486633,
      "num_tokens": 124715648.0,
      "step": 71900
    },
    {
      "entropy": 5.832048130035401,
      "epoch": 5.594436879984439,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023560409406893986,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21714415550231933,
      "num_tokens": 124725661.0,
      "step": 71905
    },
    {
      "entropy": 5.8279478549957275,
      "epoch": 5.594825909356157,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023557680693304435,
      "loss": 4.867,
      "mean_token_accuracy": 0.2206057369709015,
      "num_tokens": 124734529.0,
      "step": 71910
    },
    {
      "entropy": 5.904497337341309,
      "epoch": 5.595214938727874,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00023554952039533307,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21208282560110092,
      "num_tokens": 124743034.0,
      "step": 71915
    },
    {
      "entropy": 5.847658157348633,
      "epoch": 5.595603968099591,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023552223445622006,
      "loss": 4.9282,
      "mean_token_accuracy": 0.21852352023124694,
      "num_tokens": 124752138.0,
      "step": 71920
    },
    {
      "entropy": 5.896258211135864,
      "epoch": 5.595992997471309,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00023549494911611924,
      "loss": 4.954,
      "mean_token_accuracy": 0.21962095499038697,
      "num_tokens": 124760757.0,
      "step": 71925
    },
    {
      "entropy": 5.881213521957397,
      "epoch": 5.596382026843027,
      "grad_norm": 1.484375,
      "learning_rate": 0.00023546766437544477,
      "loss": 4.9005,
      "mean_token_accuracy": 0.21265427768230438,
      "num_tokens": 124769558.0,
      "step": 71930
    },
    {
      "entropy": 5.880981349945069,
      "epoch": 5.596771056214744,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00023544038023461057,
      "loss": 4.8123,
      "mean_token_accuracy": 0.22515979260206223,
      "num_tokens": 124777682.0,
      "step": 71935
    },
    {
      "entropy": 5.842745208740235,
      "epoch": 5.597160085586462,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002354130966940306,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21385494321584703,
      "num_tokens": 124786624.0,
      "step": 71940
    },
    {
      "entropy": 5.829004716873169,
      "epoch": 5.597549114958179,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00023538581375411876,
      "loss": 4.8525,
      "mean_token_accuracy": 0.22054546624422072,
      "num_tokens": 124794948.0,
      "step": 71945
    },
    {
      "entropy": 5.844397687911988,
      "epoch": 5.597938144329897,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00023535853141528923,
      "loss": 4.9184,
      "mean_token_accuracy": 0.21097736954689025,
      "num_tokens": 124803628.0,
      "step": 71950
    },
    {
      "entropy": 5.877474689483643,
      "epoch": 5.598327173701614,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00023533124967795594,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2235736832022667,
      "num_tokens": 124812594.0,
      "step": 71955
    },
    {
      "entropy": 5.7657074451446535,
      "epoch": 5.598716203073332,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023530396854253276,
      "loss": 4.8764,
      "mean_token_accuracy": 0.2240551680326462,
      "num_tokens": 124821823.0,
      "step": 71960
    },
    {
      "entropy": 5.8217638492584225,
      "epoch": 5.59910523244505,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00023527668800943374,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2220994785428047,
      "num_tokens": 124830068.0,
      "step": 71965
    },
    {
      "entropy": 5.953456783294678,
      "epoch": 5.599494261816767,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002352494080790727,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20699155628681182,
      "num_tokens": 124837846.0,
      "step": 71970
    },
    {
      "entropy": 5.911177444458008,
      "epoch": 5.599883291188485,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002352221287518636,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21448818147182463,
      "num_tokens": 124846820.0,
      "step": 71975
    },
    {
      "entropy": 5.824293613433838,
      "epoch": 5.600272320560203,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00023519485002822038,
      "loss": 4.855,
      "mean_token_accuracy": 0.22541943788528443,
      "num_tokens": 124855785.0,
      "step": 71980
    },
    {
      "entropy": 5.919091129302979,
      "epoch": 5.6006613499319196,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023516757190855705,
      "loss": 5.017,
      "mean_token_accuracy": 0.21589028388261794,
      "num_tokens": 124863550.0,
      "step": 71985
    },
    {
      "entropy": 5.910682535171508,
      "epoch": 5.601050379303637,
      "grad_norm": 1.390625,
      "learning_rate": 0.00023514029439328734,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22280191034078597,
      "num_tokens": 124872546.0,
      "step": 71990
    },
    {
      "entropy": 5.922741317749024,
      "epoch": 5.601439408675355,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002351130174828252,
      "loss": 4.9323,
      "mean_token_accuracy": 0.21954950094223022,
      "num_tokens": 124881343.0,
      "step": 71995
    },
    {
      "entropy": 5.903824377059936,
      "epoch": 5.601828438047073,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00023508574117758453,
      "loss": 4.9763,
      "mean_token_accuracy": 0.2081823006272316,
      "num_tokens": 124889650.0,
      "step": 72000
    },
    {
      "epoch": 5.601828438047073,
      "eval_entropy": 5.545164722703304,
      "eval_loss": 5.028749942779541,
      "eval_mean_token_accuracy": 0.22333821078638713,
      "eval_num_tokens": 124889650.0,
      "eval_runtime": 21.6767,
      "eval_samples_per_second": 1917.177,
      "eval_steps_per_second": 239.659,
      "step": 72000
    }
  ],
  "logging_steps": 5,
  "max_steps": 128520,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.514688695493632e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}