{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4668352460610776,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.9352446556091305,
      "epoch": 0.00038902937171756465,
      "grad_norm": 13.375,
      "learning_rate": 2e-06,
      "loss": 14.3638,
      "mean_token_accuracy": 0.0,
      "num_tokens": 9243.0,
      "step": 5
    },
    {
      "entropy": 4.933586740493775,
      "epoch": 0.0007780587434351293,
      "grad_norm": 13.5625,
      "learning_rate": 4.5e-06,
      "loss": 14.3351,
      "mean_token_accuracy": 0.0,
      "num_tokens": 18424.0,
      "step": 10
    },
    {
      "entropy": 4.954997825622558,
      "epoch": 0.001167088115152694,
      "grad_norm": 14.8125,
      "learning_rate": 7e-06,
      "loss": 14.2409,
      "mean_token_accuracy": 0.0,
      "num_tokens": 27006.0,
      "step": 15
    },
    {
      "entropy": 4.98590784072876,
      "epoch": 0.0015561174868702586,
      "grad_norm": 17.625,
      "learning_rate": 9.5e-06,
      "loss": 14.0569,
      "mean_token_accuracy": 0.0,
      "num_tokens": 35577.0,
      "step": 20
    },
    {
      "entropy": 5.102123737335205,
      "epoch": 0.0019451468585878235,
      "grad_norm": 21.875,
      "learning_rate": 1.2e-05,
      "loss": 13.6317,
      "mean_token_accuracy": 0.0,
      "num_tokens": 43706.0,
      "step": 25
    },
    {
      "entropy": 5.41047158241272,
      "epoch": 0.002334176230305388,
      "grad_norm": 28.25,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.0807,
      "mean_token_accuracy": 0.0,
      "num_tokens": 52571.0,
      "step": 30
    },
    {
      "entropy": 6.994736385345459,
      "epoch": 0.0027232056020229526,
      "grad_norm": 23.125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9622,
      "mean_token_accuracy": 0.0,
      "num_tokens": 61900.0,
      "step": 35
    },
    {
      "entropy": 10.197742652893066,
      "epoch": 0.0031122349737405172,
      "grad_norm": 3.296875,
      "learning_rate": 1.95e-05,
      "loss": 10.8537,
      "mean_token_accuracy": 0.0009407172445207834,
      "num_tokens": 71234.0,
      "step": 40
    },
    {
      "entropy": 10.716931438446045,
      "epoch": 0.003501264345458082,
      "grad_norm": 3.46875,
      "learning_rate": 2.2e-05,
      "loss": 10.6567,
      "mean_token_accuracy": 0.014731284510344268,
      "num_tokens": 79610.0,
      "step": 45
    },
    {
      "entropy": 10.737910270690918,
      "epoch": 0.003890293717175647,
      "grad_norm": 3.0625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5096,
      "mean_token_accuracy": 0.021222466975450514,
      "num_tokens": 87527.0,
      "step": 50
    },
    {
      "entropy": 10.68083963394165,
      "epoch": 0.004279323088893211,
      "grad_norm": 2.234375,
      "learning_rate": 2.7e-05,
      "loss": 10.3385,
      "mean_token_accuracy": 0.021166695561259986,
      "num_tokens": 95512.0,
      "step": 55
    },
    {
      "entropy": 10.617268085479736,
      "epoch": 0.004668352460610776,
      "grad_norm": 2.171875,
      "learning_rate": 2.95e-05,
      "loss": 10.2086,
      "mean_token_accuracy": 0.018889906629920006,
      "num_tokens": 104395.0,
      "step": 60
    },
    {
      "entropy": 10.707407093048095,
      "epoch": 0.0050573818323283405,
      "grad_norm": 2.03125,
      "learning_rate": 3.2e-05,
      "loss": 10.0297,
      "mean_token_accuracy": 0.02151527740061283,
      "num_tokens": 112562.0,
      "step": 65
    },
    {
      "entropy": 10.726566791534424,
      "epoch": 0.005446411204045905,
      "grad_norm": 1.984375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9483,
      "mean_token_accuracy": 0.03697771243751049,
      "num_tokens": 121962.0,
      "step": 70
    },
    {
      "entropy": 10.715249347686768,
      "epoch": 0.00583544057576347,
      "grad_norm": 1.75,
      "learning_rate": 3.7e-05,
      "loss": 9.9014,
      "mean_token_accuracy": 0.05064300559461117,
      "num_tokens": 130714.0,
      "step": 75
    },
    {
      "entropy": 10.702756023406982,
      "epoch": 0.0062244699474810344,
      "grad_norm": 1.71875,
      "learning_rate": 3.95e-05,
      "loss": 9.8032,
      "mean_token_accuracy": 0.03924348317086697,
      "num_tokens": 139674.0,
      "step": 80
    },
    {
      "entropy": 10.711318397521973,
      "epoch": 0.006613499319198599,
      "grad_norm": 1.8515625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6945,
      "mean_token_accuracy": 0.05025743767619133,
      "num_tokens": 148654.0,
      "step": 85
    },
    {
      "entropy": 10.706228637695313,
      "epoch": 0.007002528690916164,
      "grad_norm": 1.8984375,
      "learning_rate": 4.45e-05,
      "loss": 9.5288,
      "mean_token_accuracy": 0.05727620646357536,
      "num_tokens": 156180.0,
      "step": 90
    },
    {
      "entropy": 10.672399044036865,
      "epoch": 0.007391558062633729,
      "grad_norm": 1.8359375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5166,
      "mean_token_accuracy": 0.04150119200348854,
      "num_tokens": 165608.0,
      "step": 95
    },
    {
      "entropy": 10.661020946502685,
      "epoch": 0.007780587434351294,
      "grad_norm": 1.8125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3671,
      "mean_token_accuracy": 0.052654581516981123,
      "num_tokens": 173871.0,
      "step": 100
    },
    {
      "entropy": 10.639804267883301,
      "epoch": 0.008169616806068859,
      "grad_norm": 1.765625,
      "learning_rate": 5.2e-05,
      "loss": 9.278,
      "mean_token_accuracy": 0.053280239924788475,
      "num_tokens": 182356.0,
      "step": 105
    },
    {
      "entropy": 10.586706256866455,
      "epoch": 0.008558646177786422,
      "grad_norm": 1.6328125,
      "learning_rate": 5.45e-05,
      "loss": 9.1642,
      "mean_token_accuracy": 0.05218262672424316,
      "num_tokens": 190361.0,
      "step": 110
    },
    {
      "entropy": 10.545366954803466,
      "epoch": 0.008947675549503988,
      "grad_norm": 1.671875,
      "learning_rate": 5.7e-05,
      "loss": 9.038,
      "mean_token_accuracy": 0.06131550520658493,
      "num_tokens": 199253.0,
      "step": 115
    },
    {
      "entropy": 10.458304595947265,
      "epoch": 0.009336704921221552,
      "grad_norm": 1.7890625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.941,
      "mean_token_accuracy": 0.05900274179875851,
      "num_tokens": 208181.0,
      "step": 120
    },
    {
      "entropy": 10.377948665618897,
      "epoch": 0.009725734292939117,
      "grad_norm": 1.703125,
      "learning_rate": 6.2e-05,
      "loss": 8.8182,
      "mean_token_accuracy": 0.06022760644555092,
      "num_tokens": 216898.0,
      "step": 125
    },
    {
      "entropy": 10.340231990814209,
      "epoch": 0.010114763664656681,
      "grad_norm": 1.5234375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.7216,
      "mean_token_accuracy": 0.059017321839928624,
      "num_tokens": 226051.0,
      "step": 130
    },
    {
      "entropy": 10.248167514801025,
      "epoch": 0.010503793036374246,
      "grad_norm": 1.328125,
      "learning_rate": 6.7e-05,
      "loss": 8.6303,
      "mean_token_accuracy": 0.05895141884684563,
      "num_tokens": 235018.0,
      "step": 135
    },
    {
      "entropy": 10.186167526245118,
      "epoch": 0.01089282240809181,
      "grad_norm": 1.4921875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4974,
      "mean_token_accuracy": 0.056320498883724215,
      "num_tokens": 244268.0,
      "step": 140
    },
    {
      "entropy": 10.056542587280273,
      "epoch": 0.011281851779809376,
      "grad_norm": 1.3046875,
      "learning_rate": 7.2e-05,
      "loss": 8.401,
      "mean_token_accuracy": 0.05752755627036095,
      "num_tokens": 253545.0,
      "step": 145
    },
    {
      "entropy": 9.921157741546631,
      "epoch": 0.01167088115152694,
      "grad_norm": 1.3359375,
      "learning_rate": 7.45e-05,
      "loss": 8.201,
      "mean_token_accuracy": 0.06480571478605271,
      "num_tokens": 261767.0,
      "step": 150
    },
    {
      "entropy": 9.763555145263672,
      "epoch": 0.012059910523244505,
      "grad_norm": 1.125,
      "learning_rate": 7.7e-05,
      "loss": 8.1694,
      "mean_token_accuracy": 0.060174553841352466,
      "num_tokens": 270372.0,
      "step": 155
    },
    {
      "entropy": 9.612780094146729,
      "epoch": 0.012448939894962069,
      "grad_norm": 1.140625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.13,
      "mean_token_accuracy": 0.05902123637497425,
      "num_tokens": 279726.0,
      "step": 160
    },
    {
      "entropy": 9.454668998718262,
      "epoch": 0.012837969266679634,
      "grad_norm": 0.99609375,
      "learning_rate": 8.2e-05,
      "loss": 8.0495,
      "mean_token_accuracy": 0.05770065151154995,
      "num_tokens": 288955.0,
      "step": 165
    },
    {
      "entropy": 9.253509902954102,
      "epoch": 0.013226998638397198,
      "grad_norm": 0.890625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.9255,
      "mean_token_accuracy": 0.05795506499707699,
      "num_tokens": 297898.0,
      "step": 170
    },
    {
      "entropy": 9.027568244934082,
      "epoch": 0.013616028010114764,
      "grad_norm": 0.9921875,
      "learning_rate": 8.7e-05,
      "loss": 7.8475,
      "mean_token_accuracy": 0.06472490318119525,
      "num_tokens": 306189.0,
      "step": 175
    },
    {
      "entropy": 8.855317306518554,
      "epoch": 0.014005057381832327,
      "grad_norm": 0.75390625,
      "learning_rate": 8.95e-05,
      "loss": 7.7903,
      "mean_token_accuracy": 0.05992711745202541,
      "num_tokens": 314881.0,
      "step": 180
    },
    {
      "entropy": 8.634145164489746,
      "epoch": 0.014394086753549893,
      "grad_norm": 0.78515625,
      "learning_rate": 9.2e-05,
      "loss": 7.7628,
      "mean_token_accuracy": 0.05949850864708424,
      "num_tokens": 323357.0,
      "step": 185
    },
    {
      "entropy": 8.52868824005127,
      "epoch": 0.014783116125267459,
      "grad_norm": 0.828125,
      "learning_rate": 9.45e-05,
      "loss": 7.7934,
      "mean_token_accuracy": 0.06202235743403435,
      "num_tokens": 331799.0,
      "step": 190
    },
    {
      "entropy": 8.411410522460937,
      "epoch": 0.015172145496985022,
      "grad_norm": 0.75390625,
      "learning_rate": 9.7e-05,
      "loss": 7.6607,
      "mean_token_accuracy": 0.06408823989331722,
      "num_tokens": 340320.0,
      "step": 195
    },
    {
      "entropy": 8.311204147338866,
      "epoch": 0.015561174868702588,
      "grad_norm": 0.74609375,
      "learning_rate": 9.95e-05,
      "loss": 7.6695,
      "mean_token_accuracy": 0.06163757778704167,
      "num_tokens": 348990.0,
      "step": 200
    },
    {
      "entropy": 8.226022624969483,
      "epoch": 0.01595020424042015,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000102,
      "loss": 7.6319,
      "mean_token_accuracy": 0.06188234388828277,
      "num_tokens": 357661.0,
      "step": 205
    },
    {
      "entropy": 8.203551959991454,
      "epoch": 0.016339233612137717,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.5957,
      "mean_token_accuracy": 0.06355586163699627,
      "num_tokens": 366099.0,
      "step": 210
    },
    {
      "entropy": 8.117009258270263,
      "epoch": 0.016728262983855283,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000107,
      "loss": 7.7405,
      "mean_token_accuracy": 0.05980301797389984,
      "num_tokens": 374926.0,
      "step": 215
    },
    {
      "entropy": 8.17076997756958,
      "epoch": 0.017117292355572845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001095,
      "loss": 7.6617,
      "mean_token_accuracy": 0.06195650584995747,
      "num_tokens": 383846.0,
      "step": 220
    },
    {
      "entropy": 8.05511932373047,
      "epoch": 0.01750632172729041,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000112,
      "loss": 7.6259,
      "mean_token_accuracy": 0.06483461409807205,
      "num_tokens": 393678.0,
      "step": 225
    },
    {
      "entropy": 8.157504177093506,
      "epoch": 0.017895351099007976,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0001145,
      "loss": 7.661,
      "mean_token_accuracy": 0.0653709027916193,
      "num_tokens": 401866.0,
      "step": 230
    },
    {
      "entropy": 8.015466785430908,
      "epoch": 0.01828438047072554,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.6346,
      "mean_token_accuracy": 0.06497575715184212,
      "num_tokens": 411057.0,
      "step": 235
    },
    {
      "entropy": 8.082311630249023,
      "epoch": 0.018673409842443103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.6129,
      "mean_token_accuracy": 0.0683998104184866,
      "num_tokens": 419083.0,
      "step": 240
    },
    {
      "entropy": 8.020170497894288,
      "epoch": 0.01906243921416067,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000122,
      "loss": 7.5134,
      "mean_token_accuracy": 0.06742828860878944,
      "num_tokens": 427503.0,
      "step": 245
    },
    {
      "entropy": 8.029362392425536,
      "epoch": 0.019451468585878234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001245,
      "loss": 7.6228,
      "mean_token_accuracy": 0.06421508118510247,
      "num_tokens": 435820.0,
      "step": 250
    },
    {
      "entropy": 7.950299835205078,
      "epoch": 0.0198404979575958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000127,
      "loss": 7.4872,
      "mean_token_accuracy": 0.07130805253982545,
      "num_tokens": 444692.0,
      "step": 255
    },
    {
      "entropy": 8.0076434135437,
      "epoch": 0.020229527329313362,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0001295,
      "loss": 7.5393,
      "mean_token_accuracy": 0.06418065093457699,
      "num_tokens": 453536.0,
      "step": 260
    },
    {
      "entropy": 7.979480600357055,
      "epoch": 0.020618556701030927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000132,
      "loss": 7.5602,
      "mean_token_accuracy": 0.06866644658148288,
      "num_tokens": 462264.0,
      "step": 265
    },
    {
      "entropy": 7.999968433380127,
      "epoch": 0.021007586072748493,
      "grad_norm": 1.09375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.6853,
      "mean_token_accuracy": 0.06377483941614628,
      "num_tokens": 471819.0,
      "step": 270
    },
    {
      "entropy": 7.984496641159057,
      "epoch": 0.02139661544446606,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.5893,
      "mean_token_accuracy": 0.06685386300086975,
      "num_tokens": 480878.0,
      "step": 275
    },
    {
      "entropy": 8.042268657684327,
      "epoch": 0.02178564481618362,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0001395,
      "loss": 7.5368,
      "mean_token_accuracy": 0.0704406276345253,
      "num_tokens": 489454.0,
      "step": 280
    },
    {
      "entropy": 7.937283277511597,
      "epoch": 0.022174674187901186,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.4434,
      "mean_token_accuracy": 0.06957585588097573,
      "num_tokens": 498378.0,
      "step": 285
    },
    {
      "entropy": 7.92456955909729,
      "epoch": 0.02256370355961875,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001445,
      "loss": 7.6546,
      "mean_token_accuracy": 0.06681204475462436,
      "num_tokens": 507638.0,
      "step": 290
    },
    {
      "entropy": 8.000158309936523,
      "epoch": 0.022952732931336317,
      "grad_norm": 0.875,
      "learning_rate": 0.000147,
      "loss": 7.584,
      "mean_token_accuracy": 0.06461342237889767,
      "num_tokens": 517047.0,
      "step": 295
    },
    {
      "entropy": 7.9436469078063965,
      "epoch": 0.02334176230305388,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0001495,
      "loss": 7.5004,
      "mean_token_accuracy": 0.07071815803647041,
      "num_tokens": 525287.0,
      "step": 300
    },
    {
      "entropy": 7.946035575866699,
      "epoch": 0.023730791674771445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000152,
      "loss": 7.447,
      "mean_token_accuracy": 0.06887780912220479,
      "num_tokens": 533266.0,
      "step": 305
    },
    {
      "entropy": 7.880048036575317,
      "epoch": 0.02411982104648901,
      "grad_norm": 0.859375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.5603,
      "mean_token_accuracy": 0.07069731503725052,
      "num_tokens": 542448.0,
      "step": 310
    },
    {
      "entropy": 7.958608961105346,
      "epoch": 0.024508850418206576,
      "grad_norm": 1.0,
      "learning_rate": 0.000157,
      "loss": 7.4477,
      "mean_token_accuracy": 0.06876055970788002,
      "num_tokens": 550872.0,
      "step": 315
    },
    {
      "entropy": 7.892773580551148,
      "epoch": 0.024897879789924138,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0001595,
      "loss": 7.5034,
      "mean_token_accuracy": 0.07373907268047333,
      "num_tokens": 559616.0,
      "step": 320
    },
    {
      "entropy": 7.829233169555664,
      "epoch": 0.025286909161641703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000162,
      "loss": 7.4818,
      "mean_token_accuracy": 0.07325205206871033,
      "num_tokens": 568379.0,
      "step": 325
    },
    {
      "entropy": 7.9177818775177,
      "epoch": 0.02567593853335927,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.4785,
      "mean_token_accuracy": 0.07421944811940193,
      "num_tokens": 577288.0,
      "step": 330
    },
    {
      "entropy": 7.906116104125976,
      "epoch": 0.026064967905076834,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.4467,
      "mean_token_accuracy": 0.06836865991353988,
      "num_tokens": 587054.0,
      "step": 335
    },
    {
      "entropy": 7.869630527496338,
      "epoch": 0.026453997276794396,
      "grad_norm": 1.515625,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.5577,
      "mean_token_accuracy": 0.06964600048959255,
      "num_tokens": 595405.0,
      "step": 340
    },
    {
      "entropy": 7.920015335083008,
      "epoch": 0.026843026648511962,
      "grad_norm": 1.28125,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.4283,
      "mean_token_accuracy": 0.07099934928119182,
      "num_tokens": 604095.0,
      "step": 345
    },
    {
      "entropy": 7.7977094650268555,
      "epoch": 0.027232056020229527,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.4641,
      "mean_token_accuracy": 0.06749536134302617,
      "num_tokens": 613518.0,
      "step": 350
    },
    {
      "entropy": 7.825807666778564,
      "epoch": 0.027621085391947093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000177,
      "loss": 7.329,
      "mean_token_accuracy": 0.07268133610486985,
      "num_tokens": 621212.0,
      "step": 355
    },
    {
      "entropy": 7.8151813507080075,
      "epoch": 0.028010114763664655,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001795,
      "loss": 7.3993,
      "mean_token_accuracy": 0.07544071115553379,
      "num_tokens": 629326.0,
      "step": 360
    },
    {
      "entropy": 7.786165714263916,
      "epoch": 0.02839914413538222,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000182,
      "loss": 7.4069,
      "mean_token_accuracy": 0.07520471625030041,
      "num_tokens": 637034.0,
      "step": 365
    },
    {
      "entropy": 7.750944328308106,
      "epoch": 0.028788173507099786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001845,
      "loss": 7.315,
      "mean_token_accuracy": 0.07539180293679237,
      "num_tokens": 645924.0,
      "step": 370
    },
    {
      "entropy": 7.882585048675537,
      "epoch": 0.02917720287881735,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000187,
      "loss": 7.5043,
      "mean_token_accuracy": 0.07769677974283695,
      "num_tokens": 655020.0,
      "step": 375
    },
    {
      "entropy": 7.858515930175781,
      "epoch": 0.029566232250534917,
      "grad_norm": 1.328125,
      "learning_rate": 0.0001895,
      "loss": 7.3506,
      "mean_token_accuracy": 0.0738551527261734,
      "num_tokens": 663451.0,
      "step": 380
    },
    {
      "entropy": 7.7064581394195555,
      "epoch": 0.02995526162225248,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000192,
      "loss": 7.4189,
      "mean_token_accuracy": 0.07526476085186004,
      "num_tokens": 671973.0,
      "step": 385
    },
    {
      "entropy": 7.7602927684783936,
      "epoch": 0.030344290993970045,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001945,
      "loss": 7.3259,
      "mean_token_accuracy": 0.07666941098868847,
      "num_tokens": 680583.0,
      "step": 390
    },
    {
      "entropy": 7.750089979171753,
      "epoch": 0.03073332036568761,
      "grad_norm": 1.453125,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.3341,
      "mean_token_accuracy": 0.0714889034628868,
      "num_tokens": 689000.0,
      "step": 395
    },
    {
      "entropy": 7.787451219558716,
      "epoch": 0.031122349737405176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.3268,
      "mean_token_accuracy": 0.0762193225324154,
      "num_tokens": 698047.0,
      "step": 400
    },
    {
      "entropy": 7.7701332569122314,
      "epoch": 0.03151137910912274,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000202,
      "loss": 7.4381,
      "mean_token_accuracy": 0.07519823797047138,
      "num_tokens": 707192.0,
      "step": 405
    },
    {
      "entropy": 7.660664463043213,
      "epoch": 0.0319004084808403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.2728,
      "mean_token_accuracy": 0.07804172001779079,
      "num_tokens": 715895.0,
      "step": 410
    },
    {
      "entropy": 7.707487344741821,
      "epoch": 0.03228943785255787,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000207,
      "loss": 7.3356,
      "mean_token_accuracy": 0.07376265488564968,
      "num_tokens": 724891.0,
      "step": 415
    },
    {
      "entropy": 7.820835208892822,
      "epoch": 0.032678467224275434,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002095,
      "loss": 7.3764,
      "mean_token_accuracy": 0.07474377714097499,
      "num_tokens": 733378.0,
      "step": 420
    },
    {
      "entropy": 7.726008749008178,
      "epoch": 0.033067496595993,
      "grad_norm": 1.203125,
      "learning_rate": 0.000212,
      "loss": 7.3418,
      "mean_token_accuracy": 0.07988023310899735,
      "num_tokens": 741098.0,
      "step": 425
    },
    {
      "entropy": 7.787885046005249,
      "epoch": 0.033456525967710565,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002145,
      "loss": 7.3086,
      "mean_token_accuracy": 0.07603151835501194,
      "num_tokens": 749902.0,
      "step": 430
    },
    {
      "entropy": 7.594846105575561,
      "epoch": 0.033845555339428124,
      "grad_norm": 1.125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.2836,
      "mean_token_accuracy": 0.08008698970079423,
      "num_tokens": 758638.0,
      "step": 435
    },
    {
      "entropy": 7.749899053573609,
      "epoch": 0.03423458471114569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002195,
      "loss": 7.3592,
      "mean_token_accuracy": 0.07718191593885422,
      "num_tokens": 768170.0,
      "step": 440
    },
    {
      "entropy": 7.675568771362305,
      "epoch": 0.034623614082863255,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000222,
      "loss": 7.2449,
      "mean_token_accuracy": 0.07738250121474266,
      "num_tokens": 776745.0,
      "step": 445
    },
    {
      "entropy": 7.658681774139405,
      "epoch": 0.03501264345458082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002245,
      "loss": 7.3473,
      "mean_token_accuracy": 0.077309750020504,
      "num_tokens": 785030.0,
      "step": 450
    },
    {
      "entropy": 7.6965330123901365,
      "epoch": 0.035401672826298386,
      "grad_norm": 1.609375,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.3159,
      "mean_token_accuracy": 0.07549342140555382,
      "num_tokens": 793543.0,
      "step": 455
    },
    {
      "entropy": 7.600979042053223,
      "epoch": 0.03579070219801595,
      "grad_norm": 0.9375,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.1834,
      "mean_token_accuracy": 0.0792169462889433,
      "num_tokens": 802401.0,
      "step": 460
    },
    {
      "entropy": 7.680768060684204,
      "epoch": 0.03617973156973352,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.3003,
      "mean_token_accuracy": 0.07450498901307583,
      "num_tokens": 811625.0,
      "step": 465
    },
    {
      "entropy": 7.673035144805908,
      "epoch": 0.03656876094145108,
      "grad_norm": 1.40625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.2218,
      "mean_token_accuracy": 0.07750762216746807,
      "num_tokens": 820742.0,
      "step": 470
    },
    {
      "entropy": 7.5899158954620365,
      "epoch": 0.03695779031316864,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000237,
      "loss": 7.3166,
      "mean_token_accuracy": 0.07351075634360313,
      "num_tokens": 829580.0,
      "step": 475
    },
    {
      "entropy": 7.651634788513183,
      "epoch": 0.03734681968488621,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002395,
      "loss": 7.257,
      "mean_token_accuracy": 0.07667978703975678,
      "num_tokens": 838818.0,
      "step": 480
    },
    {
      "entropy": 7.68389048576355,
      "epoch": 0.03773584905660377,
      "grad_norm": 1.375,
      "learning_rate": 0.000242,
      "loss": 7.2846,
      "mean_token_accuracy": 0.06978295557200909,
      "num_tokens": 847804.0,
      "step": 485
    },
    {
      "entropy": 7.529942798614502,
      "epoch": 0.03812487842832134,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002445,
      "loss": 7.0454,
      "mean_token_accuracy": 0.08401452973484994,
      "num_tokens": 856395.0,
      "step": 490
    },
    {
      "entropy": 7.63438572883606,
      "epoch": 0.0385139078000389,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000247,
      "loss": 7.2677,
      "mean_token_accuracy": 0.08493128195405006,
      "num_tokens": 865219.0,
      "step": 495
    },
    {
      "entropy": 7.5453051090240475,
      "epoch": 0.03890293717175647,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002495,
      "loss": 7.0764,
      "mean_token_accuracy": 0.08479876518249511,
      "num_tokens": 873868.0,
      "step": 500
    },
    {
      "entropy": 7.546178579330444,
      "epoch": 0.039291966543474034,
      "grad_norm": 1.09375,
      "learning_rate": 0.000252,
      "loss": 7.1791,
      "mean_token_accuracy": 0.07757280953228474,
      "num_tokens": 882698.0,
      "step": 505
    },
    {
      "entropy": 7.52639946937561,
      "epoch": 0.0396809959151916,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002545,
      "loss": 7.1175,
      "mean_token_accuracy": 0.08353231772780419,
      "num_tokens": 891280.0,
      "step": 510
    },
    {
      "entropy": 7.560566186904907,
      "epoch": 0.04007002528690916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000257,
      "loss": 7.1573,
      "mean_token_accuracy": 0.08208137303590775,
      "num_tokens": 900044.0,
      "step": 515
    },
    {
      "entropy": 7.605652093887329,
      "epoch": 0.040459054658626724,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002595,
      "loss": 7.1293,
      "mean_token_accuracy": 0.08118686527013778,
      "num_tokens": 909297.0,
      "step": 520
    },
    {
      "entropy": 7.494569969177246,
      "epoch": 0.04084808403034429,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000262,
      "loss": 7.1054,
      "mean_token_accuracy": 0.08253288716077804,
      "num_tokens": 917514.0,
      "step": 525
    },
    {
      "entropy": 7.536244630813599,
      "epoch": 0.041237113402061855,
      "grad_norm": 1.65625,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.0974,
      "mean_token_accuracy": 0.08544723689556122,
      "num_tokens": 925842.0,
      "step": 530
    },
    {
      "entropy": 7.4475737571716305,
      "epoch": 0.04162614277377942,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.1702,
      "mean_token_accuracy": 0.07806997671723366,
      "num_tokens": 934676.0,
      "step": 535
    },
    {
      "entropy": 7.52036681175232,
      "epoch": 0.042015172145496986,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.0967,
      "mean_token_accuracy": 0.08425900638103485,
      "num_tokens": 943763.0,
      "step": 540
    },
    {
      "entropy": 7.469168663024902,
      "epoch": 0.04240420151721455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.0821,
      "mean_token_accuracy": 0.0887335106730461,
      "num_tokens": 952721.0,
      "step": 545
    },
    {
      "entropy": 7.503802490234375,
      "epoch": 0.04279323088893212,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002745,
      "loss": 7.0916,
      "mean_token_accuracy": 0.086705182492733,
      "num_tokens": 960899.0,
      "step": 550
    },
    {
      "entropy": 7.405792284011841,
      "epoch": 0.04318226026064968,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000277,
      "loss": 7.0739,
      "mean_token_accuracy": 0.08868521377444268,
      "num_tokens": 969362.0,
      "step": 555
    },
    {
      "entropy": 7.486908721923828,
      "epoch": 0.04357128963236724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002795,
      "loss": 7.0449,
      "mean_token_accuracy": 0.0890669658780098,
      "num_tokens": 978028.0,
      "step": 560
    },
    {
      "entropy": 7.465371990203858,
      "epoch": 0.04396031900408481,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.0362,
      "mean_token_accuracy": 0.08907110318541527,
      "num_tokens": 985709.0,
      "step": 565
    },
    {
      "entropy": 7.462168502807617,
      "epoch": 0.04434934837580237,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002845,
      "loss": 7.1305,
      "mean_token_accuracy": 0.0881653033196926,
      "num_tokens": 994769.0,
      "step": 570
    },
    {
      "entropy": 7.491409683227539,
      "epoch": 0.04473837774751994,
      "grad_norm": 1.125,
      "learning_rate": 0.000287,
      "loss": 7.1044,
      "mean_token_accuracy": 0.08935707733035088,
      "num_tokens": 1003997.0,
      "step": 575
    },
    {
      "entropy": 7.457375764846802,
      "epoch": 0.0451274071192375,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002895,
      "loss": 7.0944,
      "mean_token_accuracy": 0.08739758059382438,
      "num_tokens": 1012931.0,
      "step": 580
    },
    {
      "entropy": 7.459478521347046,
      "epoch": 0.04551643649095507,
      "grad_norm": 1.46875,
      "learning_rate": 0.000292,
      "loss": 7.0883,
      "mean_token_accuracy": 0.08481152728199959,
      "num_tokens": 1021149.0,
      "step": 585
    },
    {
      "entropy": 7.538705492019654,
      "epoch": 0.045905465862672634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002945,
      "loss": 7.0841,
      "mean_token_accuracy": 0.08799395263195038,
      "num_tokens": 1029421.0,
      "step": 590
    },
    {
      "entropy": 7.462785053253174,
      "epoch": 0.0462944952343902,
      "grad_norm": 1.203125,
      "learning_rate": 0.000297,
      "loss": 7.1743,
      "mean_token_accuracy": 0.08678501397371292,
      "num_tokens": 1037222.0,
      "step": 595
    },
    {
      "entropy": 7.437726593017578,
      "epoch": 0.04668352460610776,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002995,
      "loss": 7.0995,
      "mean_token_accuracy": 0.08787578642368317,
      "num_tokens": 1046903.0,
      "step": 600
    },
    {
      "entropy": 7.512172842025757,
      "epoch": 0.047072553977825324,
      "grad_norm": 1.34375,
      "learning_rate": 0.000302,
      "loss": 7.1473,
      "mean_token_accuracy": 0.0882769763469696,
      "num_tokens": 1055516.0,
      "step": 605
    },
    {
      "entropy": 7.3588001251220705,
      "epoch": 0.04746158334954289,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003045,
      "loss": 7.062,
      "mean_token_accuracy": 0.08846223801374435,
      "num_tokens": 1063846.0,
      "step": 610
    },
    {
      "entropy": 7.445447111129761,
      "epoch": 0.047850612721260455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000307,
      "loss": 7.0266,
      "mean_token_accuracy": 0.08862499296665191,
      "num_tokens": 1072500.0,
      "step": 615
    },
    {
      "entropy": 7.362279796600342,
      "epoch": 0.04823964209297802,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003095,
      "loss": 7.0448,
      "mean_token_accuracy": 0.08868532851338387,
      "num_tokens": 1081399.0,
      "step": 620
    },
    {
      "entropy": 7.348788785934448,
      "epoch": 0.048628671464695586,
      "grad_norm": 1.15625,
      "learning_rate": 0.000312,
      "loss": 7.0033,
      "mean_token_accuracy": 0.09508873224258423,
      "num_tokens": 1089666.0,
      "step": 625
    },
    {
      "entropy": 7.375779914855957,
      "epoch": 0.04901770083641315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003145,
      "loss": 6.8973,
      "mean_token_accuracy": 0.09559790790081024,
      "num_tokens": 1097496.0,
      "step": 630
    },
    {
      "entropy": 7.29237060546875,
      "epoch": 0.04940673020813072,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000317,
      "loss": 7.0035,
      "mean_token_accuracy": 0.09299874976277352,
      "num_tokens": 1105840.0,
      "step": 635
    },
    {
      "entropy": 7.402494239807129,
      "epoch": 0.049795759579848276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003195,
      "loss": 7.0719,
      "mean_token_accuracy": 0.08872134163975716,
      "num_tokens": 1115507.0,
      "step": 640
    },
    {
      "entropy": 7.369076442718506,
      "epoch": 0.05018478895156584,
      "grad_norm": 1.09375,
      "learning_rate": 0.000322,
      "loss": 7.0236,
      "mean_token_accuracy": 0.0886545293033123,
      "num_tokens": 1124820.0,
      "step": 645
    },
    {
      "entropy": 7.28340630531311,
      "epoch": 0.05057381832328341,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.874,
      "mean_token_accuracy": 0.09596052095293998,
      "num_tokens": 1133641.0,
      "step": 650
    },
    {
      "entropy": 7.213095664978027,
      "epoch": 0.05096284769500097,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.9189,
      "mean_token_accuracy": 0.09583729729056359,
      "num_tokens": 1142182.0,
      "step": 655
    },
    {
      "entropy": 7.313646078109741,
      "epoch": 0.05135187706671854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.0063,
      "mean_token_accuracy": 0.09615088105201722,
      "num_tokens": 1150737.0,
      "step": 660
    },
    {
      "entropy": 7.298776626586914,
      "epoch": 0.0517409064384361,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.8665,
      "mean_token_accuracy": 0.09271117597818375,
      "num_tokens": 1158899.0,
      "step": 665
    },
    {
      "entropy": 7.385696363449097,
      "epoch": 0.05212993581015367,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.949,
      "mean_token_accuracy": 0.10016907081007957,
      "num_tokens": 1167397.0,
      "step": 670
    },
    {
      "entropy": 7.342027235031128,
      "epoch": 0.052518965181871234,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000337,
      "loss": 6.9648,
      "mean_token_accuracy": 0.09081067815423012,
      "num_tokens": 1177044.0,
      "step": 675
    },
    {
      "entropy": 7.156559610366822,
      "epoch": 0.05290799455358879,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003395,
      "loss": 6.8077,
      "mean_token_accuracy": 0.09764151573181153,
      "num_tokens": 1185740.0,
      "step": 680
    },
    {
      "entropy": 7.273555040359497,
      "epoch": 0.05329702392530636,
      "grad_norm": 1.40625,
      "learning_rate": 0.000342,
      "loss": 6.959,
      "mean_token_accuracy": 0.09533077627420425,
      "num_tokens": 1194970.0,
      "step": 685
    },
    {
      "entropy": 7.26585021018982,
      "epoch": 0.053686053297023924,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.9193,
      "mean_token_accuracy": 0.09548866376280785,
      "num_tokens": 1203565.0,
      "step": 690
    },
    {
      "entropy": 7.345769262313842,
      "epoch": 0.05407508266874149,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000347,
      "loss": 6.9552,
      "mean_token_accuracy": 0.09232278615236282,
      "num_tokens": 1213226.0,
      "step": 695
    },
    {
      "entropy": 7.133775281906128,
      "epoch": 0.054464112040459055,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003495,
      "loss": 6.8723,
      "mean_token_accuracy": 0.0922174483537674,
      "num_tokens": 1221710.0,
      "step": 700
    },
    {
      "entropy": 7.19491286277771,
      "epoch": 0.05485314141217662,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000352,
      "loss": 6.931,
      "mean_token_accuracy": 0.088956368714571,
      "num_tokens": 1231125.0,
      "step": 705
    },
    {
      "entropy": 7.321548891067505,
      "epoch": 0.055242170783894186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003545,
      "loss": 6.8845,
      "mean_token_accuracy": 0.09706432223320008,
      "num_tokens": 1239521.0,
      "step": 710
    },
    {
      "entropy": 7.171043634414673,
      "epoch": 0.05563120015561175,
      "grad_norm": 1.125,
      "learning_rate": 0.000357,
      "loss": 6.7669,
      "mean_token_accuracy": 0.10296523198485374,
      "num_tokens": 1247363.0,
      "step": 715
    },
    {
      "entropy": 7.1869182109832765,
      "epoch": 0.05602022952732931,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003595,
      "loss": 6.8672,
      "mean_token_accuracy": 0.09592388719320297,
      "num_tokens": 1255395.0,
      "step": 720
    },
    {
      "entropy": 7.241849136352539,
      "epoch": 0.056409258899046875,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000362,
      "loss": 6.9361,
      "mean_token_accuracy": 0.09305860325694085,
      "num_tokens": 1264135.0,
      "step": 725
    },
    {
      "entropy": 7.2033219814300535,
      "epoch": 0.05679828827076444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003645,
      "loss": 6.858,
      "mean_token_accuracy": 0.09347002804279328,
      "num_tokens": 1272375.0,
      "step": 730
    },
    {
      "entropy": 7.255517768859863,
      "epoch": 0.057187317642482006,
      "grad_norm": 1.109375,
      "learning_rate": 0.000367,
      "loss": 6.9055,
      "mean_token_accuracy": 0.08993236273527146,
      "num_tokens": 1280960.0,
      "step": 735
    },
    {
      "entropy": 7.214277744293213,
      "epoch": 0.05757634701419957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003695,
      "loss": 6.9302,
      "mean_token_accuracy": 0.09800290465354919,
      "num_tokens": 1289273.0,
      "step": 740
    },
    {
      "entropy": 7.18172402381897,
      "epoch": 0.05796537638591714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000372,
      "loss": 6.8479,
      "mean_token_accuracy": 0.09940816089510918,
      "num_tokens": 1297571.0,
      "step": 745
    },
    {
      "entropy": 7.171477317810059,
      "epoch": 0.0583544057576347,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003745,
      "loss": 6.9001,
      "mean_token_accuracy": 0.0956696018576622,
      "num_tokens": 1305686.0,
      "step": 750
    },
    {
      "entropy": 7.218287420272827,
      "epoch": 0.05874343512935227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000377,
      "loss": 6.9046,
      "mean_token_accuracy": 0.09595917537808418,
      "num_tokens": 1314483.0,
      "step": 755
    },
    {
      "entropy": 7.148498582839966,
      "epoch": 0.059132464501069834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003795,
      "loss": 6.7823,
      "mean_token_accuracy": 0.09254824668169022,
      "num_tokens": 1323183.0,
      "step": 760
    },
    {
      "entropy": 7.194712829589844,
      "epoch": 0.05952149387278739,
      "grad_norm": 1.390625,
      "learning_rate": 0.000382,
      "loss": 6.8996,
      "mean_token_accuracy": 0.09819645509123802,
      "num_tokens": 1331964.0,
      "step": 765
    },
    {
      "entropy": 7.1077686786651615,
      "epoch": 0.05991052324450496,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003845,
      "loss": 6.8611,
      "mean_token_accuracy": 0.09804551675915718,
      "num_tokens": 1340185.0,
      "step": 770
    },
    {
      "entropy": 7.263053369522095,
      "epoch": 0.060299552616222524,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.9342,
      "mean_token_accuracy": 0.09536710008978844,
      "num_tokens": 1349522.0,
      "step": 775
    },
    {
      "entropy": 7.275899028778076,
      "epoch": 0.06068858198794009,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.896,
      "mean_token_accuracy": 0.09667107537388801,
      "num_tokens": 1357893.0,
      "step": 780
    },
    {
      "entropy": 7.106265592575073,
      "epoch": 0.061077611359657655,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.7809,
      "mean_token_accuracy": 0.10025013461709023,
      "num_tokens": 1366685.0,
      "step": 785
    },
    {
      "entropy": 7.080127429962158,
      "epoch": 0.06146664073137522,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.7265,
      "mean_token_accuracy": 0.09952032342553138,
      "num_tokens": 1375245.0,
      "step": 790
    },
    {
      "entropy": 7.122674369812012,
      "epoch": 0.061855670103092786,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.8964,
      "mean_token_accuracy": 0.09939667508006096,
      "num_tokens": 1383724.0,
      "step": 795
    },
    {
      "entropy": 7.244520807266236,
      "epoch": 0.06224469947481035,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003995,
      "loss": 6.888,
      "mean_token_accuracy": 0.09881971329450608,
      "num_tokens": 1392455.0,
      "step": 800
    },
    {
      "entropy": 7.216006183624268,
      "epoch": 0.06263372884652792,
      "grad_norm": 1.03125,
      "learning_rate": 0.000402,
      "loss": 6.891,
      "mean_token_accuracy": 0.10223213583230972,
      "num_tokens": 1401262.0,
      "step": 805
    },
    {
      "entropy": 7.012789487838745,
      "epoch": 0.06302275821824548,
      "grad_norm": 1.125,
      "learning_rate": 0.0004045,
      "loss": 6.77,
      "mean_token_accuracy": 0.10305121541023254,
      "num_tokens": 1409991.0,
      "step": 810
    },
    {
      "entropy": 7.0912278175354,
      "epoch": 0.06341178758996305,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.7739,
      "mean_token_accuracy": 0.10144544690847397,
      "num_tokens": 1418208.0,
      "step": 815
    },
    {
      "entropy": 7.041250562667846,
      "epoch": 0.0638008169616806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095,
      "loss": 6.7687,
      "mean_token_accuracy": 0.10150073319673539,
      "num_tokens": 1426665.0,
      "step": 820
    },
    {
      "entropy": 7.040017032623291,
      "epoch": 0.06418984633339817,
      "grad_norm": 1.34375,
      "learning_rate": 0.000412,
      "loss": 6.7003,
      "mean_token_accuracy": 0.10085978209972382,
      "num_tokens": 1435247.0,
      "step": 825
    },
    {
      "entropy": 7.063506841659546,
      "epoch": 0.06457887570511574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004145,
      "loss": 6.7405,
      "mean_token_accuracy": 0.10099377259612083,
      "num_tokens": 1443678.0,
      "step": 830
    },
    {
      "entropy": 7.011990594863891,
      "epoch": 0.0649679050768333,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000417,
      "loss": 6.6843,
      "mean_token_accuracy": 0.10724351927638054,
      "num_tokens": 1451567.0,
      "step": 835
    },
    {
      "entropy": 7.009270477294922,
      "epoch": 0.06535693444855087,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004195,
      "loss": 6.704,
      "mean_token_accuracy": 0.10393429771065713,
      "num_tokens": 1460618.0,
      "step": 840
    },
    {
      "entropy": 7.021320343017578,
      "epoch": 0.06574596382026843,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000422,
      "loss": 6.7753,
      "mean_token_accuracy": 0.10534258186817169,
      "num_tokens": 1469821.0,
      "step": 845
    },
    {
      "entropy": 6.965624380111694,
      "epoch": 0.066134993191986,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245,
      "loss": 6.8413,
      "mean_token_accuracy": 0.0993105836212635,
      "num_tokens": 1477989.0,
      "step": 850
    },
    {
      "entropy": 7.1604921340942385,
      "epoch": 0.06652402256370356,
      "grad_norm": 1.25,
      "learning_rate": 0.000427,
      "loss": 6.8444,
      "mean_token_accuracy": 0.10425978153944016,
      "num_tokens": 1486822.0,
      "step": 855
    },
    {
      "entropy": 7.088742446899414,
      "epoch": 0.06691305193542113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004295,
      "loss": 6.7305,
      "mean_token_accuracy": 0.09952922314405441,
      "num_tokens": 1496604.0,
      "step": 860
    },
    {
      "entropy": 7.047960662841797,
      "epoch": 0.06730208130713869,
      "grad_norm": 1.34375,
      "learning_rate": 0.000432,
      "loss": 6.7496,
      "mean_token_accuracy": 0.10054385960102082,
      "num_tokens": 1504725.0,
      "step": 865
    },
    {
      "entropy": 6.983655691146851,
      "epoch": 0.06769111067885625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004345,
      "loss": 6.7616,
      "mean_token_accuracy": 0.10596725270152092,
      "num_tokens": 1513222.0,
      "step": 870
    },
    {
      "entropy": 7.112746620178223,
      "epoch": 0.06808014005057382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437,
      "loss": 6.7105,
      "mean_token_accuracy": 0.10374617874622345,
      "num_tokens": 1522183.0,
      "step": 875
    },
    {
      "entropy": 7.01748366355896,
      "epoch": 0.06846916942229138,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004395,
      "loss": 6.7634,
      "mean_token_accuracy": 0.10354626253247261,
      "num_tokens": 1531316.0,
      "step": 880
    },
    {
      "entropy": 7.0218301773071286,
      "epoch": 0.06885819879400895,
      "grad_norm": 1.234375,
      "learning_rate": 0.000442,
      "loss": 6.8291,
      "mean_token_accuracy": 0.10073515474796295,
      "num_tokens": 1540053.0,
      "step": 885
    },
    {
      "entropy": 7.033105850219727,
      "epoch": 0.06924722816572651,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004445,
      "loss": 6.6775,
      "mean_token_accuracy": 0.11033806353807449,
      "num_tokens": 1549391.0,
      "step": 890
    },
    {
      "entropy": 7.096326446533203,
      "epoch": 0.06963625753744408,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000447,
      "loss": 6.7919,
      "mean_token_accuracy": 0.10401200950145721,
      "num_tokens": 1557364.0,
      "step": 895
    },
    {
      "entropy": 6.939074897766114,
      "epoch": 0.07002528690916164,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.6375,
      "mean_token_accuracy": 0.0982839398086071,
      "num_tokens": 1567011.0,
      "step": 900
    },
    {
      "entropy": 6.94973726272583,
      "epoch": 0.0704143162808792,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.7271,
      "mean_token_accuracy": 0.10176745280623437,
      "num_tokens": 1575441.0,
      "step": 905
    },
    {
      "entropy": 6.9638408660888675,
      "epoch": 0.07080334565259677,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.6876,
      "mean_token_accuracy": 0.10327148288488389,
      "num_tokens": 1583757.0,
      "step": 910
    },
    {
      "entropy": 7.055232191085816,
      "epoch": 0.07119237502431433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.7335,
      "mean_token_accuracy": 0.10632129982113839,
      "num_tokens": 1592406.0,
      "step": 915
    },
    {
      "entropy": 6.91970043182373,
      "epoch": 0.0715814043960319,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.7293,
      "mean_token_accuracy": 0.10444770827889442,
      "num_tokens": 1601665.0,
      "step": 920
    },
    {
      "entropy": 7.091557741165161,
      "epoch": 0.07197043376774946,
      "grad_norm": 1.234375,
      "learning_rate": 0.000462,
      "loss": 6.7245,
      "mean_token_accuracy": 0.10248482674360275,
      "num_tokens": 1609788.0,
      "step": 925
    },
    {
      "entropy": 6.889610815048218,
      "epoch": 0.07235946313946703,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004645,
      "loss": 6.7202,
      "mean_token_accuracy": 0.10887994691729545,
      "num_tokens": 1618030.0,
      "step": 930
    },
    {
      "entropy": 6.963936805725098,
      "epoch": 0.07274849251118459,
      "grad_norm": 1.1875,
      "learning_rate": 0.000467,
      "loss": 6.7281,
      "mean_token_accuracy": 0.10066362768411637,
      "num_tokens": 1626901.0,
      "step": 935
    },
    {
      "entropy": 6.932220506668091,
      "epoch": 0.07313752188290217,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004695,
      "loss": 6.6348,
      "mean_token_accuracy": 0.1041472889482975,
      "num_tokens": 1635372.0,
      "step": 940
    },
    {
      "entropy": 6.882161235809326,
      "epoch": 0.07352655125461972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472,
      "loss": 6.6485,
      "mean_token_accuracy": 0.10415024533867837,
      "num_tokens": 1643791.0,
      "step": 945
    },
    {
      "entropy": 6.876175880432129,
      "epoch": 0.07391558062633728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004745,
      "loss": 6.6031,
      "mean_token_accuracy": 0.11224255934357644,
      "num_tokens": 1652161.0,
      "step": 950
    },
    {
      "entropy": 6.945423316955567,
      "epoch": 0.07430460999805485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000477,
      "loss": 6.6567,
      "mean_token_accuracy": 0.10526313111186028,
      "num_tokens": 1661077.0,
      "step": 955
    },
    {
      "entropy": 6.903894758224487,
      "epoch": 0.07469363936977241,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004795,
      "loss": 6.6253,
      "mean_token_accuracy": 0.10890860855579376,
      "num_tokens": 1669010.0,
      "step": 960
    },
    {
      "entropy": 6.9699595928192135,
      "epoch": 0.07508266874148999,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000482,
      "loss": 6.8042,
      "mean_token_accuracy": 0.10613740161061287,
      "num_tokens": 1678002.0,
      "step": 965
    },
    {
      "entropy": 6.982291316986084,
      "epoch": 0.07547169811320754,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004845,
      "loss": 6.7491,
      "mean_token_accuracy": 0.10167021304368973,
      "num_tokens": 1687403.0,
      "step": 970
    },
    {
      "entropy": 6.9213052749633786,
      "epoch": 0.07586072748492512,
      "grad_norm": 1.15625,
      "learning_rate": 0.000487,
      "loss": 6.6526,
      "mean_token_accuracy": 0.10744511783123016,
      "num_tokens": 1695512.0,
      "step": 975
    },
    {
      "entropy": 6.9510541439056395,
      "epoch": 0.07624975685664268,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004895,
      "loss": 6.6685,
      "mean_token_accuracy": 0.10527175143361092,
      "num_tokens": 1705116.0,
      "step": 980
    },
    {
      "entropy": 6.837292337417603,
      "epoch": 0.07663878622836025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492,
      "loss": 6.5608,
      "mean_token_accuracy": 0.11234913319349289,
      "num_tokens": 1713715.0,
      "step": 985
    },
    {
      "entropy": 6.849586534500122,
      "epoch": 0.0770278156000778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945,
      "loss": 6.6274,
      "mean_token_accuracy": 0.11056514084339142,
      "num_tokens": 1721881.0,
      "step": 990
    },
    {
      "entropy": 6.907543706893921,
      "epoch": 0.07741684497179536,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497,
      "loss": 6.6461,
      "mean_token_accuracy": 0.11055096238851547,
      "num_tokens": 1730716.0,
      "step": 995
    },
    {
      "entropy": 7.061227083206177,
      "epoch": 0.07780587434351294,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995,
      "loss": 6.8026,
      "mean_token_accuracy": 0.10393945574760437,
      "num_tokens": 1740257.0,
      "step": 1000
    },
    {
      "entropy": 6.850629186630249,
      "epoch": 0.0781949037152305,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999999989075146,
      "loss": 6.7019,
      "mean_token_accuracy": 0.10508826747536659,
      "num_tokens": 1749071.0,
      "step": 1005
    },
    {
      "entropy": 6.91280608177185,
      "epoch": 0.07858393308694807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999944692927,
      "loss": 6.7081,
      "mean_token_accuracy": 0.09934931620955467,
      "num_tokens": 1757952.0,
      "step": 1010
    },
    {
      "entropy": 6.924489593505859,
      "epoch": 0.07897296245866563,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499999986617054,
      "loss": 6.6445,
      "mean_token_accuracy": 0.1096059374511242,
      "num_tokens": 1766027.0,
      "step": 1015
    },
    {
      "entropy": 6.844860982894898,
      "epoch": 0.0793619918303832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999753507986,
      "loss": 6.6349,
      "mean_token_accuracy": 0.10513899624347686,
      "num_tokens": 1775039.0,
      "step": 1020
    },
    {
      "entropy": 6.822795629501343,
      "epoch": 0.07975102120210076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999606705267,
      "loss": 6.6415,
      "mean_token_accuracy": 0.10872110053896904,
      "num_tokens": 1783971.0,
      "step": 1025
    },
    {
      "entropy": 6.864479064941406,
      "epoch": 0.08014005057381832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999425762385,
      "loss": 6.5852,
      "mean_token_accuracy": 0.10642140656709671,
      "num_tokens": 1792743.0,
      "step": 1030
    },
    {
      "entropy": 6.888079357147217,
      "epoch": 0.08052907994553589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999999210679344,
      "loss": 6.7187,
      "mean_token_accuracy": 0.10097150802612305,
      "num_tokens": 1801435.0,
      "step": 1035
    },
    {
      "entropy": 6.943144989013672,
      "epoch": 0.08091810931725345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999998961456145,
      "loss": 6.6372,
      "mean_token_accuracy": 0.10501615256071091,
      "num_tokens": 1810147.0,
      "step": 1040
    },
    {
      "entropy": 6.91335916519165,
      "epoch": 0.08130713868897102,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999998678092794,
      "loss": 6.5481,
      "mean_token_accuracy": 0.11189196705818176,
      "num_tokens": 1818672.0,
      "step": 1045
    },
    {
      "entropy": 6.825189781188965,
      "epoch": 0.08169616806068858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999998360589293,
      "loss": 6.7051,
      "mean_token_accuracy": 0.10792158544063568,
      "num_tokens": 1827590.0,
      "step": 1050
    },
    {
      "entropy": 6.842350673675537,
      "epoch": 0.08208519743240615,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499999800894565,
      "loss": 6.6064,
      "mean_token_accuracy": 0.10936003550887108,
      "num_tokens": 1836467.0,
      "step": 1055
    },
    {
      "entropy": 6.7909856796264645,
      "epoch": 0.08247422680412371,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999997623161866,
      "loss": 6.5501,
      "mean_token_accuracy": 0.1102816753089428,
      "num_tokens": 1845537.0,
      "step": 1060
    },
    {
      "entropy": 6.798137998580932,
      "epoch": 0.08286325617584128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999997203237951,
      "loss": 6.5568,
      "mean_token_accuracy": 0.11228980720043183,
      "num_tokens": 1854433.0,
      "step": 1065
    },
    {
      "entropy": 6.846316385269165,
      "epoch": 0.08325228554755884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999674917391,
      "loss": 6.6642,
      "mean_token_accuracy": 0.10633082091808319,
      "num_tokens": 1863708.0,
      "step": 1070
    },
    {
      "entropy": 6.800260066986084,
      "epoch": 0.0836413149192764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999996260969748,
      "loss": 6.5127,
      "mean_token_accuracy": 0.11366265788674354,
      "num_tokens": 1871854.0,
      "step": 1075
    },
    {
      "entropy": 6.745605897903443,
      "epoch": 0.08403034429099397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999995738625475,
      "loss": 6.4832,
      "mean_token_accuracy": 0.11385836526751518,
      "num_tokens": 1880471.0,
      "step": 1080
    },
    {
      "entropy": 6.780485916137695,
      "epoch": 0.08441937366271153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999995182141099,
      "loss": 6.5533,
      "mean_token_accuracy": 0.11285655200481415,
      "num_tokens": 1889882.0,
      "step": 1085
    },
    {
      "entropy": 6.703962755203247,
      "epoch": 0.0848084030344291,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999994591516626,
      "loss": 6.4536,
      "mean_token_accuracy": 0.11884711608290673,
      "num_tokens": 1898701.0,
      "step": 1090
    },
    {
      "entropy": 6.733728122711182,
      "epoch": 0.08519743240614666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999993966752066,
      "loss": 6.5431,
      "mean_token_accuracy": 0.10564961209893227,
      "num_tokens": 1908165.0,
      "step": 1095
    },
    {
      "entropy": 6.80612211227417,
      "epoch": 0.08558646177786423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499999330784743,
      "loss": 6.6478,
      "mean_token_accuracy": 0.1123865082859993,
      "num_tokens": 1916295.0,
      "step": 1100
    },
    {
      "entropy": 6.840036296844483,
      "epoch": 0.08597549114958179,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999992614802725,
      "loss": 6.5403,
      "mean_token_accuracy": 0.1170975498855114,
      "num_tokens": 1924643.0,
      "step": 1105
    },
    {
      "entropy": 6.790605068206787,
      "epoch": 0.08636452052129936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999991887617966,
      "loss": 6.55,
      "mean_token_accuracy": 0.11264167055487632,
      "num_tokens": 1933822.0,
      "step": 1110
    },
    {
      "entropy": 6.68263874053955,
      "epoch": 0.08675354989301692,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999991126293158,
      "loss": 6.5535,
      "mean_token_accuracy": 0.11057107970118522,
      "num_tokens": 1943278.0,
      "step": 1115
    },
    {
      "entropy": 6.915115308761597,
      "epoch": 0.08714257926473448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999990330828318,
      "loss": 6.5695,
      "mean_token_accuracy": 0.1134694166481495,
      "num_tokens": 1951883.0,
      "step": 1120
    },
    {
      "entropy": 6.65633978843689,
      "epoch": 0.08753160863645205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999989501223456,
      "loss": 6.4789,
      "mean_token_accuracy": 0.120311389118433,
      "num_tokens": 1959567.0,
      "step": 1125
    },
    {
      "entropy": 6.808887338638305,
      "epoch": 0.08792063800816961,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999988637478584,
      "loss": 6.534,
      "mean_token_accuracy": 0.11390565186738968,
      "num_tokens": 1968190.0,
      "step": 1130
    },
    {
      "entropy": 6.660136461257935,
      "epoch": 0.08830966737988719,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999987739593716,
      "loss": 6.4539,
      "mean_token_accuracy": 0.1147844597697258,
      "num_tokens": 1977519.0,
      "step": 1135
    },
    {
      "entropy": 6.740057754516601,
      "epoch": 0.08869869675160474,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999986807568865,
      "loss": 6.4617,
      "mean_token_accuracy": 0.11601719185709954,
      "num_tokens": 1985796.0,
      "step": 1140
    },
    {
      "entropy": 6.815498018264771,
      "epoch": 0.08908772612332232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999985841404045,
      "loss": 6.59,
      "mean_token_accuracy": 0.10702956765890122,
      "num_tokens": 1993874.0,
      "step": 1145
    },
    {
      "entropy": 6.867396354675293,
      "epoch": 0.08947675549503988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999984841099271,
      "loss": 6.544,
      "mean_token_accuracy": 0.11298409327864647,
      "num_tokens": 2001663.0,
      "step": 1150
    },
    {
      "entropy": 6.687847232818603,
      "epoch": 0.08986578486675743,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499998380665456,
      "loss": 6.5284,
      "mean_token_accuracy": 0.11568290889263153,
      "num_tokens": 2010179.0,
      "step": 1155
    },
    {
      "entropy": 6.760761499404907,
      "epoch": 0.090254814238475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999982738069924,
      "loss": 6.434,
      "mean_token_accuracy": 0.11665870845317841,
      "num_tokens": 2020013.0,
      "step": 1160
    },
    {
      "entropy": 6.742422866821289,
      "epoch": 0.09064384361019256,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004999981635345382,
      "loss": 6.568,
      "mean_token_accuracy": 0.11969253122806549,
      "num_tokens": 2028396.0,
      "step": 1165
    },
    {
      "entropy": 6.825271940231323,
      "epoch": 0.09103287298191014,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999980498480949,
      "loss": 6.6335,
      "mean_token_accuracy": 0.11033851355314254,
      "num_tokens": 2036947.0,
      "step": 1170
    },
    {
      "entropy": 6.715671443939209,
      "epoch": 0.0914219023536277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999979327476644,
      "loss": 6.4834,
      "mean_token_accuracy": 0.11201950162649155,
      "num_tokens": 2045162.0,
      "step": 1175
    },
    {
      "entropy": 6.766329193115235,
      "epoch": 0.09181093172534527,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999978122332485,
      "loss": 6.5764,
      "mean_token_accuracy": 0.11277065798640251,
      "num_tokens": 2054198.0,
      "step": 1180
    },
    {
      "entropy": 6.686334037780762,
      "epoch": 0.09219996109706283,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999976883048487,
      "loss": 6.4753,
      "mean_token_accuracy": 0.11509751304984092,
      "num_tokens": 2062558.0,
      "step": 1185
    },
    {
      "entropy": 6.642856502532959,
      "epoch": 0.0925889904687804,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999975609624673,
      "loss": 6.3676,
      "mean_token_accuracy": 0.11287973672151566,
      "num_tokens": 2070816.0,
      "step": 1190
    },
    {
      "entropy": 6.751265525817871,
      "epoch": 0.09297801984049796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999974302061059,
      "loss": 6.4407,
      "mean_token_accuracy": 0.11461416780948638,
      "num_tokens": 2079058.0,
      "step": 1195
    },
    {
      "entropy": 6.6414995193481445,
      "epoch": 0.09336704921221552,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999972960357666,
      "loss": 6.3879,
      "mean_token_accuracy": 0.11782738789916039,
      "num_tokens": 2086679.0,
      "step": 1200
    },
    {
      "entropy": 6.725094270706177,
      "epoch": 0.09375607858393309,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999971584514516,
      "loss": 6.5033,
      "mean_token_accuracy": 0.11564282923936844,
      "num_tokens": 2095788.0,
      "step": 1205
    },
    {
      "entropy": 6.734344482421875,
      "epoch": 0.09414510795565065,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999970174531627,
      "loss": 6.5344,
      "mean_token_accuracy": 0.11681660562753678,
      "num_tokens": 2105435.0,
      "step": 1210
    },
    {
      "entropy": 6.622858190536499,
      "epoch": 0.09453413732736822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999968730409022,
      "loss": 6.4773,
      "mean_token_accuracy": 0.11257969439029694,
      "num_tokens": 2114318.0,
      "step": 1215
    },
    {
      "entropy": 6.707676362991333,
      "epoch": 0.09492316669908578,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999967252146723,
      "loss": 6.3901,
      "mean_token_accuracy": 0.1280004471540451,
      "num_tokens": 2122707.0,
      "step": 1220
    },
    {
      "entropy": 6.613141632080078,
      "epoch": 0.09531219607080335,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999965739744752,
      "loss": 6.4989,
      "mean_token_accuracy": 0.11614064946770668,
      "num_tokens": 2131176.0,
      "step": 1225
    },
    {
      "entropy": 6.743852138519287,
      "epoch": 0.09570122544252091,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999964193203132,
      "loss": 6.5268,
      "mean_token_accuracy": 0.1201147809624672,
      "num_tokens": 2139872.0,
      "step": 1230
    },
    {
      "entropy": 6.755846691131592,
      "epoch": 0.09609025481423847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999962612521888,
      "loss": 6.3904,
      "mean_token_accuracy": 0.12232213765382767,
      "num_tokens": 2148668.0,
      "step": 1235
    },
    {
      "entropy": 6.5976697444915775,
      "epoch": 0.09647928418595604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999960997701041,
      "loss": 6.3884,
      "mean_token_accuracy": 0.12410193756222725,
      "num_tokens": 2156628.0,
      "step": 1240
    },
    {
      "entropy": 6.723433208465576,
      "epoch": 0.0968683135576736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999959348740618,
      "loss": 6.4816,
      "mean_token_accuracy": 0.11155348718166351,
      "num_tokens": 2165620.0,
      "step": 1245
    },
    {
      "entropy": 6.685586643218994,
      "epoch": 0.09725734292939117,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999957665640641,
      "loss": 6.5553,
      "mean_token_accuracy": 0.11295825839042664,
      "num_tokens": 2174493.0,
      "step": 1250
    },
    {
      "entropy": 6.718245220184326,
      "epoch": 0.09764637230110873,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499995594840114,
      "loss": 6.4554,
      "mean_token_accuracy": 0.11334520950913429,
      "num_tokens": 2183258.0,
      "step": 1255
    },
    {
      "entropy": 6.7033568859100345,
      "epoch": 0.0980354016728263,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999954197022138,
      "loss": 6.4894,
      "mean_token_accuracy": 0.11108549013733864,
      "num_tokens": 2192263.0,
      "step": 1260
    },
    {
      "entropy": 6.724197483062744,
      "epoch": 0.09842443104454386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999952411503661,
      "loss": 6.5465,
      "mean_token_accuracy": 0.1062019944190979,
      "num_tokens": 2201797.0,
      "step": 1265
    },
    {
      "entropy": 6.554443073272705,
      "epoch": 0.09881346041626143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999950591845738,
      "loss": 6.2769,
      "mean_token_accuracy": 0.12360937595367431,
      "num_tokens": 2209454.0,
      "step": 1270
    },
    {
      "entropy": 6.728240442276001,
      "epoch": 0.09920248978797899,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999948738048397,
      "loss": 6.4715,
      "mean_token_accuracy": 0.12011136263608932,
      "num_tokens": 2217761.0,
      "step": 1275
    },
    {
      "entropy": 6.589418649673462,
      "epoch": 0.09959151915969655,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999946850111663,
      "loss": 6.4558,
      "mean_token_accuracy": 0.11745724231004714,
      "num_tokens": 2225793.0,
      "step": 1280
    },
    {
      "entropy": 6.725760078430175,
      "epoch": 0.09998054853141412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999944928035569,
      "loss": 6.3478,
      "mean_token_accuracy": 0.12596990913152695,
      "num_tokens": 2234620.0,
      "step": 1285
    },
    {
      "entropy": 6.576044940948487,
      "epoch": 0.10036957790313168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499994297182014,
      "loss": 6.4253,
      "mean_token_accuracy": 0.11709852069616318,
      "num_tokens": 2242829.0,
      "step": 1290
    },
    {
      "entropy": 6.768080854415894,
      "epoch": 0.10075860727484925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999940981465408,
      "loss": 6.4503,
      "mean_token_accuracy": 0.11201097890734672,
      "num_tokens": 2251356.0,
      "step": 1295
    },
    {
      "entropy": 6.582262754440308,
      "epoch": 0.10114763664656681,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999938956971404,
      "loss": 6.5095,
      "mean_token_accuracy": 0.11508250460028649,
      "num_tokens": 2259327.0,
      "step": 1300
    },
    {
      "entropy": 6.726982116699219,
      "epoch": 0.10153666601828439,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999936898338156,
      "loss": 6.523,
      "mean_token_accuracy": 0.11173656061291695,
      "num_tokens": 2267948.0,
      "step": 1305
    },
    {
      "entropy": 6.588463926315308,
      "epoch": 0.10192569539000194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999934805565698,
      "loss": 6.3409,
      "mean_token_accuracy": 0.12150438576936722,
      "num_tokens": 2275828.0,
      "step": 1310
    },
    {
      "entropy": 6.558595943450928,
      "epoch": 0.10231472476171952,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499993267865406,
      "loss": 6.2999,
      "mean_token_accuracy": 0.12410445511341095,
      "num_tokens": 2284246.0,
      "step": 1315
    },
    {
      "entropy": 6.591059923171997,
      "epoch": 0.10270375413343708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999930517603275,
      "loss": 6.2729,
      "mean_token_accuracy": 0.12803125903010368,
      "num_tokens": 2291837.0,
      "step": 1320
    },
    {
      "entropy": 6.616555118560791,
      "epoch": 0.10309278350515463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999928322413374,
      "loss": 6.3566,
      "mean_token_accuracy": 0.1179949127137661,
      "num_tokens": 2299853.0,
      "step": 1325
    },
    {
      "entropy": 6.700346422195435,
      "epoch": 0.1034818128768722,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999926093084394,
      "loss": 6.4513,
      "mean_token_accuracy": 0.11860990896821022,
      "num_tokens": 2308416.0,
      "step": 1330
    },
    {
      "entropy": 6.64585771560669,
      "epoch": 0.10387084224858976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999923829616365,
      "loss": 6.4687,
      "mean_token_accuracy": 0.11745152324438095,
      "num_tokens": 2318104.0,
      "step": 1335
    },
    {
      "entropy": 6.545508909225464,
      "epoch": 0.10425987162030734,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999921532009323,
      "loss": 6.3939,
      "mean_token_accuracy": 0.11805395260453225,
      "num_tokens": 2326807.0,
      "step": 1340
    },
    {
      "entropy": 6.693354845046997,
      "epoch": 0.1046489009920249,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999919200263304,
      "loss": 6.4063,
      "mean_token_accuracy": 0.11813065260648728,
      "num_tokens": 2335921.0,
      "step": 1345
    },
    {
      "entropy": 6.571820402145386,
      "epoch": 0.10503793036374247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999916834378342,
      "loss": 6.4246,
      "mean_token_accuracy": 0.11576211974024772,
      "num_tokens": 2344914.0,
      "step": 1350
    },
    {
      "entropy": 6.570094347000122,
      "epoch": 0.10542695973546003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999914434354472,
      "loss": 6.3649,
      "mean_token_accuracy": 0.11981132701039314,
      "num_tokens": 2353211.0,
      "step": 1355
    },
    {
      "entropy": 6.647089672088623,
      "epoch": 0.10581598910717759,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999912000191733,
      "loss": 6.4694,
      "mean_token_accuracy": 0.11573371812701225,
      "num_tokens": 2361717.0,
      "step": 1360
    },
    {
      "entropy": 6.617790842056275,
      "epoch": 0.10620501847889516,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499990953189016,
      "loss": 6.4211,
      "mean_token_accuracy": 0.12210528329014778,
      "num_tokens": 2370114.0,
      "step": 1365
    },
    {
      "entropy": 6.682390451431274,
      "epoch": 0.10659404785061272,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999907029449791,
      "loss": 6.4051,
      "mean_token_accuracy": 0.12277124375104904,
      "num_tokens": 2378820.0,
      "step": 1370
    },
    {
      "entropy": 6.514065790176391,
      "epoch": 0.10698307722233029,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999904492870663,
      "loss": 6.3213,
      "mean_token_accuracy": 0.12254384979605674,
      "num_tokens": 2387119.0,
      "step": 1375
    },
    {
      "entropy": 6.626542711257935,
      "epoch": 0.10737210659404785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999901922152816,
      "loss": 6.3765,
      "mean_token_accuracy": 0.1194961629807949,
      "num_tokens": 2396246.0,
      "step": 1380
    },
    {
      "entropy": 6.501721668243408,
      "epoch": 0.10776113596576542,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999899317296288,
      "loss": 6.4157,
      "mean_token_accuracy": 0.11856927499175071,
      "num_tokens": 2405428.0,
      "step": 1385
    },
    {
      "entropy": 6.596559715270996,
      "epoch": 0.10815016533748298,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999896678301121,
      "loss": 6.3741,
      "mean_token_accuracy": 0.1184548296034336,
      "num_tokens": 2414669.0,
      "step": 1390
    },
    {
      "entropy": 6.705186796188355,
      "epoch": 0.10853919470920055,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999894005167352,
      "loss": 6.4612,
      "mean_token_accuracy": 0.11556609496474265,
      "num_tokens": 2423193.0,
      "step": 1395
    },
    {
      "entropy": 6.6101438999176025,
      "epoch": 0.10892822408091811,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499989129789502,
      "loss": 6.4064,
      "mean_token_accuracy": 0.11643991023302078,
      "num_tokens": 2431150.0,
      "step": 1400
    },
    {
      "entropy": 6.540600776672363,
      "epoch": 0.10931725345263567,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999888556484172,
      "loss": 6.3807,
      "mean_token_accuracy": 0.11422878503799438,
      "num_tokens": 2440337.0,
      "step": 1405
    },
    {
      "entropy": 6.589718914031982,
      "epoch": 0.10970628282435324,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999885780934845,
      "loss": 6.3226,
      "mean_token_accuracy": 0.12455554753541946,
      "num_tokens": 2449246.0,
      "step": 1410
    },
    {
      "entropy": 6.561670303344727,
      "epoch": 0.1100953121960708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999882971247081,
      "loss": 6.3256,
      "mean_token_accuracy": 0.12393307089805602,
      "num_tokens": 2457982.0,
      "step": 1415
    },
    {
      "entropy": 6.534701871871948,
      "epoch": 0.11048434156778837,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999880127420926,
      "loss": 6.3142,
      "mean_token_accuracy": 0.12068029120564461,
      "num_tokens": 2466697.0,
      "step": 1420
    },
    {
      "entropy": 6.558716917037964,
      "epoch": 0.11087337093950593,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999877249456421,
      "loss": 6.2992,
      "mean_token_accuracy": 0.12362277731299401,
      "num_tokens": 2475457.0,
      "step": 1425
    },
    {
      "entropy": 6.464503622055053,
      "epoch": 0.1112624003112235,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999874337353609,
      "loss": 6.2568,
      "mean_token_accuracy": 0.12550607100129127,
      "num_tokens": 2484753.0,
      "step": 1430
    },
    {
      "entropy": 6.553883075714111,
      "epoch": 0.11165142968294106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999871391112535,
      "loss": 6.3796,
      "mean_token_accuracy": 0.12183060646057128,
      "num_tokens": 2492921.0,
      "step": 1435
    },
    {
      "entropy": 6.573325300216675,
      "epoch": 0.11204045905465862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999868410733244,
      "loss": 6.3569,
      "mean_token_accuracy": 0.12047385424375534,
      "num_tokens": 2501371.0,
      "step": 1440
    },
    {
      "entropy": 6.570412826538086,
      "epoch": 0.11242948842637619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499986539621578,
      "loss": 6.2245,
      "mean_token_accuracy": 0.12222100347280503,
      "num_tokens": 2509521.0,
      "step": 1445
    },
    {
      "entropy": 6.550269222259521,
      "epoch": 0.11281851779809375,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999862347560191,
      "loss": 6.3615,
      "mean_token_accuracy": 0.12241242602467536,
      "num_tokens": 2518316.0,
      "step": 1450
    },
    {
      "entropy": 6.55282735824585,
      "epoch": 0.11320754716981132,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999859264766521,
      "loss": 6.4293,
      "mean_token_accuracy": 0.11907708570361138,
      "num_tokens": 2527549.0,
      "step": 1455
    },
    {
      "entropy": 6.628605079650879,
      "epoch": 0.11359657654152888,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999856147834817,
      "loss": 6.4688,
      "mean_token_accuracy": 0.11421279311180114,
      "num_tokens": 2536341.0,
      "step": 1460
    },
    {
      "entropy": 6.505453729629517,
      "epoch": 0.11398560591324645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999852996765129,
      "loss": 6.3516,
      "mean_token_accuracy": 0.12754116728901863,
      "num_tokens": 2544869.0,
      "step": 1465
    },
    {
      "entropy": 6.604720211029052,
      "epoch": 0.11437463528496401,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999849811557502,
      "loss": 6.2474,
      "mean_token_accuracy": 0.13095576241612433,
      "num_tokens": 2553432.0,
      "step": 1470
    },
    {
      "entropy": 6.5074522495269775,
      "epoch": 0.11476366465668159,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999846592211984,
      "loss": 6.3718,
      "mean_token_accuracy": 0.1168175496160984,
      "num_tokens": 2561996.0,
      "step": 1475
    },
    {
      "entropy": 6.550441551208496,
      "epoch": 0.11515269402839914,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999843338728626,
      "loss": 6.3114,
      "mean_token_accuracy": 0.12313547134399414,
      "num_tokens": 2571315.0,
      "step": 1480
    },
    {
      "entropy": 6.485339689254761,
      "epoch": 0.1155417234001167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999840051107476,
      "loss": 6.4213,
      "mean_token_accuracy": 0.12167479395866394,
      "num_tokens": 2579589.0,
      "step": 1485
    },
    {
      "entropy": 6.56643123626709,
      "epoch": 0.11593075277183428,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999836729348584,
      "loss": 6.1891,
      "mean_token_accuracy": 0.12298046126961708,
      "num_tokens": 2588528.0,
      "step": 1490
    },
    {
      "entropy": 6.466246747970581,
      "epoch": 0.11631978214355183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999833373452,
      "loss": 6.3238,
      "mean_token_accuracy": 0.12104458585381508,
      "num_tokens": 2596820.0,
      "step": 1495
    },
    {
      "entropy": 6.466208410263062,
      "epoch": 0.1167088115152694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999829983417778,
      "loss": 6.2499,
      "mean_token_accuracy": 0.1204721324145794,
      "num_tokens": 2605502.0,
      "step": 1500
    },
    {
      "entropy": 6.4354174613952635,
      "epoch": 0.11709784088698696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999826559245965,
      "loss": 6.2334,
      "mean_token_accuracy": 0.13126121312379838,
      "num_tokens": 2613974.0,
      "step": 1505
    },
    {
      "entropy": 6.435812187194824,
      "epoch": 0.11748687025870454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999823100936615,
      "loss": 6.3333,
      "mean_token_accuracy": 0.12118990495800971,
      "num_tokens": 2622581.0,
      "step": 1510
    },
    {
      "entropy": 6.555751800537109,
      "epoch": 0.1178758996304221,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999819608489781,
      "loss": 6.2104,
      "mean_token_accuracy": 0.1205756813287735,
      "num_tokens": 2631868.0,
      "step": 1515
    },
    {
      "entropy": 6.42847900390625,
      "epoch": 0.11826492900213967,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999816081905515,
      "loss": 6.2463,
      "mean_token_accuracy": 0.12586833164095879,
      "num_tokens": 2640501.0,
      "step": 1520
    },
    {
      "entropy": 6.500116395950317,
      "epoch": 0.11865395837385723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999812521183872,
      "loss": 6.2471,
      "mean_token_accuracy": 0.12680162265896797,
      "num_tokens": 2648916.0,
      "step": 1525
    },
    {
      "entropy": 6.536718368530273,
      "epoch": 0.11904298774557479,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999808926324903,
      "loss": 6.3235,
      "mean_token_accuracy": 0.12868664413690567,
      "num_tokens": 2657725.0,
      "step": 1530
    },
    {
      "entropy": 6.425989866256714,
      "epoch": 0.11943201711729236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999805297328665,
      "loss": 6.2894,
      "mean_token_accuracy": 0.1203535608947277,
      "num_tokens": 2665983.0,
      "step": 1535
    },
    {
      "entropy": 6.5746509552001955,
      "epoch": 0.11982104648900992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999801634195214,
      "loss": 6.3172,
      "mean_token_accuracy": 0.1309082590043545,
      "num_tokens": 2674546.0,
      "step": 1540
    },
    {
      "entropy": 6.456331253051758,
      "epoch": 0.12021007586072749,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999797936924603,
      "loss": 6.2295,
      "mean_token_accuracy": 0.1303058944642544,
      "num_tokens": 2684088.0,
      "step": 1545
    },
    {
      "entropy": 6.505995798110962,
      "epoch": 0.12059910523244505,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499979420551689,
      "loss": 6.2917,
      "mean_token_accuracy": 0.121770840883255,
      "num_tokens": 2693289.0,
      "step": 1550
    },
    {
      "entropy": 6.439859294891358,
      "epoch": 0.12098813460416262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499979043997213,
      "loss": 6.2212,
      "mean_token_accuracy": 0.1267593041062355,
      "num_tokens": 2701744.0,
      "step": 1555
    },
    {
      "entropy": 6.472963476181031,
      "epoch": 0.12137716397588018,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999786640290381,
      "loss": 6.266,
      "mean_token_accuracy": 0.12355687320232392,
      "num_tokens": 2710584.0,
      "step": 1560
    },
    {
      "entropy": 6.534983539581299,
      "epoch": 0.12176619334759774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999782806471701,
      "loss": 6.287,
      "mean_token_accuracy": 0.1200690247118473,
      "num_tokens": 2719478.0,
      "step": 1565
    },
    {
      "entropy": 6.546087789535522,
      "epoch": 0.12215522271931531,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999778938516147,
      "loss": 6.3619,
      "mean_token_accuracy": 0.11858213022351265,
      "num_tokens": 2729105.0,
      "step": 1570
    },
    {
      "entropy": 6.497035884857178,
      "epoch": 0.12254425209103287,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499977503642378,
      "loss": 6.3199,
      "mean_token_accuracy": 0.1214341938495636,
      "num_tokens": 2738300.0,
      "step": 1575
    },
    {
      "entropy": 6.564480495452881,
      "epoch": 0.12293328146275044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999771100194656,
      "loss": 6.2592,
      "mean_token_accuracy": 0.12089609131217002,
      "num_tokens": 2747236.0,
      "step": 1580
    },
    {
      "entropy": 6.389290761947632,
      "epoch": 0.123322310834468,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999767129828837,
      "loss": 6.1946,
      "mean_token_accuracy": 0.12700149565935134,
      "num_tokens": 2755919.0,
      "step": 1585
    },
    {
      "entropy": 6.3935548782348635,
      "epoch": 0.12371134020618557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999763125326383,
      "loss": 6.1975,
      "mean_token_accuracy": 0.12940854504704474,
      "num_tokens": 2764469.0,
      "step": 1590
    },
    {
      "entropy": 6.486390113830566,
      "epoch": 0.12410036957790313,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999759086687355,
      "loss": 6.2067,
      "mean_token_accuracy": 0.13023579344153405,
      "num_tokens": 2772663.0,
      "step": 1595
    },
    {
      "entropy": 6.543991231918335,
      "epoch": 0.1244893989496207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999755013911813,
      "loss": 6.2463,
      "mean_token_accuracy": 0.12318727225065232,
      "num_tokens": 2780719.0,
      "step": 1600
    },
    {
      "entropy": 6.4445489883422855,
      "epoch": 0.12487842832133826,
      "grad_norm": 1.0,
      "learning_rate": 0.000499975090699982,
      "loss": 6.2958,
      "mean_token_accuracy": 0.12555501163005828,
      "num_tokens": 2789186.0,
      "step": 1605
    },
    {
      "entropy": 6.420124244689942,
      "epoch": 0.12526745769305583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999746765951438,
      "loss": 6.208,
      "mean_token_accuracy": 0.12907335609197618,
      "num_tokens": 2797993.0,
      "step": 1610
    },
    {
      "entropy": 6.432662439346314,
      "epoch": 0.1256564870647734,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999742590766729,
      "loss": 6.1669,
      "mean_token_accuracy": 0.12667878344655037,
      "num_tokens": 2806375.0,
      "step": 1615
    },
    {
      "entropy": 6.512080526351928,
      "epoch": 0.12604551643649095,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999738381445757,
      "loss": 6.3037,
      "mean_token_accuracy": 0.13135089427232743,
      "num_tokens": 2814772.0,
      "step": 1620
    },
    {
      "entropy": 6.298824071884155,
      "epoch": 0.1264345458082085,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999734137988586,
      "loss": 6.1753,
      "mean_token_accuracy": 0.13204587176442145,
      "num_tokens": 2824231.0,
      "step": 1625
    },
    {
      "entropy": 6.452016353607178,
      "epoch": 0.1268235751799261,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499972986039528,
      "loss": 6.2368,
      "mean_token_accuracy": 0.12344342246651649,
      "num_tokens": 2833470.0,
      "step": 1630
    },
    {
      "entropy": 6.50761456489563,
      "epoch": 0.12721260455164365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999725548665904,
      "loss": 6.25,
      "mean_token_accuracy": 0.12541535049676894,
      "num_tokens": 2843010.0,
      "step": 1635
    },
    {
      "entropy": 6.404108715057373,
      "epoch": 0.1276016339233612,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999721202800524,
      "loss": 6.1338,
      "mean_token_accuracy": 0.1386566400527954,
      "num_tokens": 2851545.0,
      "step": 1640
    },
    {
      "entropy": 6.3438740253448485,
      "epoch": 0.12799066329507877,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999716822799206,
      "loss": 6.2906,
      "mean_token_accuracy": 0.11413189694285393,
      "num_tokens": 2861230.0,
      "step": 1645
    },
    {
      "entropy": 6.47772855758667,
      "epoch": 0.12837969266679633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999712408662016,
      "loss": 6.2534,
      "mean_token_accuracy": 0.1262986846268177,
      "num_tokens": 2869436.0,
      "step": 1650
    },
    {
      "entropy": 6.392445182800293,
      "epoch": 0.12876872203851392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999707960389021,
      "loss": 6.122,
      "mean_token_accuracy": 0.13018414750695229,
      "num_tokens": 2877688.0,
      "step": 1655
    },
    {
      "entropy": 6.4603923797607425,
      "epoch": 0.12915775141023147,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999703477980288,
      "loss": 6.2708,
      "mean_token_accuracy": 0.1284830316901207,
      "num_tokens": 2887043.0,
      "step": 1660
    },
    {
      "entropy": 6.430318546295166,
      "epoch": 0.12954678078194903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999698961435885,
      "loss": 6.1854,
      "mean_token_accuracy": 0.12815706580877304,
      "num_tokens": 2895732.0,
      "step": 1665
    },
    {
      "entropy": 6.3375777244567875,
      "epoch": 0.1299358101536666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999694410755882,
      "loss": 6.1608,
      "mean_token_accuracy": 0.12703153863549232,
      "num_tokens": 2904162.0,
      "step": 1670
    },
    {
      "entropy": 6.449351215362549,
      "epoch": 0.13032483952538418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999689825940347,
      "loss": 6.3035,
      "mean_token_accuracy": 0.123183623701334,
      "num_tokens": 2912783.0,
      "step": 1675
    },
    {
      "entropy": 6.479056549072266,
      "epoch": 0.13071386889710174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999685206989349,
      "loss": 6.1934,
      "mean_token_accuracy": 0.1278615675866604,
      "num_tokens": 2921027.0,
      "step": 1680
    },
    {
      "entropy": 6.411128139495849,
      "epoch": 0.1311028982688193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999680553902959,
      "loss": 6.2066,
      "mean_token_accuracy": 0.13008249029517174,
      "num_tokens": 2930398.0,
      "step": 1685
    },
    {
      "entropy": 6.392834138870239,
      "epoch": 0.13149192764053685,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999675866681247,
      "loss": 6.2191,
      "mean_token_accuracy": 0.1267761617898941,
      "num_tokens": 2938703.0,
      "step": 1690
    },
    {
      "entropy": 6.381357955932617,
      "epoch": 0.1318809570122544,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999671145324286,
      "loss": 6.1658,
      "mean_token_accuracy": 0.12640352845191954,
      "num_tokens": 2947007.0,
      "step": 1695
    },
    {
      "entropy": 6.427257251739502,
      "epoch": 0.132269986383972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999666389832144,
      "loss": 6.2931,
      "mean_token_accuracy": 0.12982493937015532,
      "num_tokens": 2955485.0,
      "step": 1700
    },
    {
      "entropy": 6.446907711029053,
      "epoch": 0.13265901575568956,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999661600204897,
      "loss": 6.1536,
      "mean_token_accuracy": 0.13096046820282936,
      "num_tokens": 2964416.0,
      "step": 1705
    },
    {
      "entropy": 6.3575907230377195,
      "epoch": 0.13304804512740712,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999656776442615,
      "loss": 6.1576,
      "mean_token_accuracy": 0.1293207123875618,
      "num_tokens": 2972737.0,
      "step": 1710
    },
    {
      "entropy": 6.409917593002319,
      "epoch": 0.13343707449912467,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999651918545372,
      "loss": 6.1892,
      "mean_token_accuracy": 0.12660456970334052,
      "num_tokens": 2981166.0,
      "step": 1715
    },
    {
      "entropy": 6.444819402694702,
      "epoch": 0.13382610387084226,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999647026513243,
      "loss": 6.2508,
      "mean_token_accuracy": 0.12753460928797722,
      "num_tokens": 2989689.0,
      "step": 1720
    },
    {
      "entropy": 6.436755228042602,
      "epoch": 0.13421513324255982,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049996421003463,
      "loss": 6.1478,
      "mean_token_accuracy": 0.1300426684319973,
      "num_tokens": 2998793.0,
      "step": 1725
    },
    {
      "entropy": 6.365857887268066,
      "epoch": 0.13460416261427738,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999637140044619,
      "loss": 6.2926,
      "mean_token_accuracy": 0.12362960577011109,
      "num_tokens": 3007320.0,
      "step": 1730
    },
    {
      "entropy": 6.463358211517334,
      "epoch": 0.13499319198599494,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999632145608275,
      "loss": 6.196,
      "mean_token_accuracy": 0.12713374942541122,
      "num_tokens": 3015024.0,
      "step": 1735
    },
    {
      "entropy": 6.4128015518188475,
      "epoch": 0.1353822213577125,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999627117037343,
      "loss": 6.299,
      "mean_token_accuracy": 0.1275800235569477,
      "num_tokens": 3023058.0,
      "step": 1740
    },
    {
      "entropy": 6.471516561508179,
      "epoch": 0.13577125072943008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049996220543319,
      "loss": 6.1937,
      "mean_token_accuracy": 0.13457101881504058,
      "num_tokens": 3031308.0,
      "step": 1745
    },
    {
      "entropy": 6.40582799911499,
      "epoch": 0.13616028010114764,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999616957492025,
      "loss": 6.2654,
      "mean_token_accuracy": 0.12314189150929451,
      "num_tokens": 3039598.0,
      "step": 1750
    },
    {
      "entropy": 6.340579605102539,
      "epoch": 0.1365493094728652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999611826517793,
      "loss": 6.0849,
      "mean_token_accuracy": 0.13836994841694833,
      "num_tokens": 3047598.0,
      "step": 1755
    },
    {
      "entropy": 6.46547417640686,
      "epoch": 0.13693833884458276,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999606661409281,
      "loss": 6.2395,
      "mean_token_accuracy": 0.12904614359140396,
      "num_tokens": 3055702.0,
      "step": 1760
    },
    {
      "entropy": 6.346052122116089,
      "epoch": 0.13732736821630034,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999601462166569,
      "loss": 6.0973,
      "mean_token_accuracy": 0.13112309873104094,
      "num_tokens": 3063908.0,
      "step": 1765
    },
    {
      "entropy": 6.419245815277099,
      "epoch": 0.1377163975880179,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999596228789736,
      "loss": 6.1888,
      "mean_token_accuracy": 0.12716860547661782,
      "num_tokens": 3072077.0,
      "step": 1770
    },
    {
      "entropy": 6.381726551055908,
      "epoch": 0.13810542695973546,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999590961278859,
      "loss": 6.2046,
      "mean_token_accuracy": 0.12799116969108582,
      "num_tokens": 3080498.0,
      "step": 1775
    },
    {
      "entropy": 6.343014144897461,
      "epoch": 0.13849445633145302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999585659634021,
      "loss": 6.1964,
      "mean_token_accuracy": 0.13141690641641618,
      "num_tokens": 3089055.0,
      "step": 1780
    },
    {
      "entropy": 6.376903533935547,
      "epoch": 0.13888348570317058,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999580323855301,
      "loss": 6.1916,
      "mean_token_accuracy": 0.13284389600157737,
      "num_tokens": 3096944.0,
      "step": 1785
    },
    {
      "entropy": 6.361277532577515,
      "epoch": 0.13927251507488816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999574953942781,
      "loss": 6.0802,
      "mean_token_accuracy": 0.13013918548822404,
      "num_tokens": 3105596.0,
      "step": 1790
    },
    {
      "entropy": 6.29228663444519,
      "epoch": 0.13966154444660572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999569549896541,
      "loss": 6.0716,
      "mean_token_accuracy": 0.13826498091220857,
      "num_tokens": 3114349.0,
      "step": 1795
    },
    {
      "entropy": 6.349193048477173,
      "epoch": 0.14005057381832328,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999564111716665,
      "loss": 6.1805,
      "mean_token_accuracy": 0.12369971498847007,
      "num_tokens": 3122490.0,
      "step": 1800
    },
    {
      "entropy": 6.439586353302002,
      "epoch": 0.14043960319004084,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999558639403232,
      "loss": 6.2334,
      "mean_token_accuracy": 0.1289645791053772,
      "num_tokens": 3130988.0,
      "step": 1805
    },
    {
      "entropy": 6.373170948028564,
      "epoch": 0.1408286325617584,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999553132956328,
      "loss": 6.0905,
      "mean_token_accuracy": 0.12859657481312753,
      "num_tokens": 3139488.0,
      "step": 1810
    },
    {
      "entropy": 6.31482195854187,
      "epoch": 0.14121766193347599,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999547592376035,
      "loss": 6.1537,
      "mean_token_accuracy": 0.1342098221182823,
      "num_tokens": 3147955.0,
      "step": 1815
    },
    {
      "entropy": 6.337986850738526,
      "epoch": 0.14160669130519354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999542017662438,
      "loss": 6.1884,
      "mean_token_accuracy": 0.12379184886813163,
      "num_tokens": 3157608.0,
      "step": 1820
    },
    {
      "entropy": 6.413020277023316,
      "epoch": 0.1419957206769111,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999536408815623,
      "loss": 6.1462,
      "mean_token_accuracy": 0.12497628033161164,
      "num_tokens": 3166124.0,
      "step": 1825
    },
    {
      "entropy": 6.232908296585083,
      "epoch": 0.14238475004862866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999530765835672,
      "loss": 6.1422,
      "mean_token_accuracy": 0.12837766632437705,
      "num_tokens": 3175069.0,
      "step": 1830
    },
    {
      "entropy": 6.412486600875854,
      "epoch": 0.14277377942034625,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999525088722673,
      "loss": 6.1759,
      "mean_token_accuracy": 0.12886116057634353,
      "num_tokens": 3184259.0,
      "step": 1835
    },
    {
      "entropy": 6.393767547607422,
      "epoch": 0.1431628087920638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999519377476711,
      "loss": 6.0945,
      "mean_token_accuracy": 0.12900736853480338,
      "num_tokens": 3193015.0,
      "step": 1840
    },
    {
      "entropy": 6.337996959686279,
      "epoch": 0.14355183816378136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999513632097872,
      "loss": 6.008,
      "mean_token_accuracy": 0.13678519204258918,
      "num_tokens": 3201624.0,
      "step": 1845
    },
    {
      "entropy": 6.311671876907349,
      "epoch": 0.14394086753549892,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999507852586245,
      "loss": 6.1439,
      "mean_token_accuracy": 0.13350687846541404,
      "num_tokens": 3209474.0,
      "step": 1850
    },
    {
      "entropy": 6.269514083862305,
      "epoch": 0.14432989690721648,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999502038941917,
      "loss": 6.1012,
      "mean_token_accuracy": 0.1346833288669586,
      "num_tokens": 3218550.0,
      "step": 1855
    },
    {
      "entropy": 6.209073257446289,
      "epoch": 0.14471892627893407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999496191164975,
      "loss": 6.0819,
      "mean_token_accuracy": 0.13349551558494568,
      "num_tokens": 3227108.0,
      "step": 1860
    },
    {
      "entropy": 6.3549457550048825,
      "epoch": 0.14510795565065163,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499949030925551,
      "loss": 6.1167,
      "mean_token_accuracy": 0.13430884554982186,
      "num_tokens": 3235785.0,
      "step": 1865
    },
    {
      "entropy": 6.429311513900757,
      "epoch": 0.14549698502236919,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999484393213609,
      "loss": 6.1573,
      "mean_token_accuracy": 0.12800115197896958,
      "num_tokens": 3244180.0,
      "step": 1870
    },
    {
      "entropy": 6.257713270187378,
      "epoch": 0.14588601439408674,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999478443039363,
      "loss": 6.0194,
      "mean_token_accuracy": 0.13384133577346802,
      "num_tokens": 3253098.0,
      "step": 1875
    },
    {
      "entropy": 6.351937246322632,
      "epoch": 0.14627504376580433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999472458732862,
      "loss": 6.0853,
      "mean_token_accuracy": 0.13433253467082978,
      "num_tokens": 3261726.0,
      "step": 1880
    },
    {
      "entropy": 6.313936185836792,
      "epoch": 0.1466640731375219,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999466440294198,
      "loss": 6.029,
      "mean_token_accuracy": 0.13558507710695267,
      "num_tokens": 3270362.0,
      "step": 1885
    },
    {
      "entropy": 6.235980558395386,
      "epoch": 0.14705310250923945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499946038772346,
      "loss": 6.0079,
      "mean_token_accuracy": 0.13864972218871116,
      "num_tokens": 3278332.0,
      "step": 1890
    },
    {
      "entropy": 6.265728330612182,
      "epoch": 0.147442131880957,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999454301020741,
      "loss": 6.0011,
      "mean_token_accuracy": 0.13960243612527848,
      "num_tokens": 3286800.0,
      "step": 1895
    },
    {
      "entropy": 6.382432699203491,
      "epoch": 0.14783116125267456,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999448180186133,
      "loss": 6.2342,
      "mean_token_accuracy": 0.1212337888777256,
      "num_tokens": 3296186.0,
      "step": 1900
    },
    {
      "entropy": 6.223908090591431,
      "epoch": 0.14822019062439215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499944202521973,
      "loss": 6.0816,
      "mean_token_accuracy": 0.13997681587934493,
      "num_tokens": 3303973.0,
      "step": 1905
    },
    {
      "entropy": 6.313835668563843,
      "epoch": 0.1486092199961097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999435836121624,
      "loss": 6.0389,
      "mean_token_accuracy": 0.13727107942104338,
      "num_tokens": 3312550.0,
      "step": 1910
    },
    {
      "entropy": 6.271665716171265,
      "epoch": 0.14899824936782727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499942961289191,
      "loss": 6.1037,
      "mean_token_accuracy": 0.13144001960754395,
      "num_tokens": 3321342.0,
      "step": 1915
    },
    {
      "entropy": 6.258928823471069,
      "epoch": 0.14938727873954483,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999423355530683,
      "loss": 5.9791,
      "mean_token_accuracy": 0.14113496989011765,
      "num_tokens": 3329293.0,
      "step": 1920
    },
    {
      "entropy": 6.167897891998291,
      "epoch": 0.1497763081112624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999417064038035,
      "loss": 6.0517,
      "mean_token_accuracy": 0.13005898594856263,
      "num_tokens": 3338278.0,
      "step": 1925
    },
    {
      "entropy": 6.415949773788452,
      "epoch": 0.15016533748297997,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999410738414065,
      "loss": 6.1416,
      "mean_token_accuracy": 0.13462255373597146,
      "num_tokens": 3346757.0,
      "step": 1930
    },
    {
      "entropy": 6.1904011249542235,
      "epoch": 0.15055436685469753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999404378658866,
      "loss": 6.1487,
      "mean_token_accuracy": 0.12973785847425462,
      "num_tokens": 3355903.0,
      "step": 1935
    },
    {
      "entropy": 6.28576431274414,
      "epoch": 0.1509433962264151,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999397984772536,
      "loss": 6.1226,
      "mean_token_accuracy": 0.12523171529173852,
      "num_tokens": 3365440.0,
      "step": 1940
    },
    {
      "entropy": 6.344992446899414,
      "epoch": 0.15133242559813265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999391556755172,
      "loss": 6.0897,
      "mean_token_accuracy": 0.1339936561882496,
      "num_tokens": 3373672.0,
      "step": 1945
    },
    {
      "entropy": 6.230864191055298,
      "epoch": 0.15172145496985023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999385094606872,
      "loss": 6.0443,
      "mean_token_accuracy": 0.12970422506332396,
      "num_tokens": 3382733.0,
      "step": 1950
    },
    {
      "entropy": 6.247889184951783,
      "epoch": 0.1521104843415678,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999378598327733,
      "loss": 6.0138,
      "mean_token_accuracy": 0.13647327050566674,
      "num_tokens": 3390842.0,
      "step": 1955
    },
    {
      "entropy": 6.245198822021484,
      "epoch": 0.15249951371328535,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999372067917854,
      "loss": 6.1716,
      "mean_token_accuracy": 0.12445822656154633,
      "num_tokens": 3399761.0,
      "step": 1960
    },
    {
      "entropy": 6.363856220245362,
      "epoch": 0.1528885430850029,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999365503377335,
      "loss": 6.0931,
      "mean_token_accuracy": 0.13199654519557952,
      "num_tokens": 3408986.0,
      "step": 1965
    },
    {
      "entropy": 6.2719300270080565,
      "epoch": 0.1532775724567205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999358904706273,
      "loss": 6.0483,
      "mean_token_accuracy": 0.12709831818938255,
      "num_tokens": 3418051.0,
      "step": 1970
    },
    {
      "entropy": 6.36472110748291,
      "epoch": 0.15366660182843805,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999352271904771,
      "loss": 6.1339,
      "mean_token_accuracy": 0.13131385445594787,
      "num_tokens": 3427185.0,
      "step": 1975
    },
    {
      "entropy": 6.233695363998413,
      "epoch": 0.1540556312001556,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999345604972928,
      "loss": 6.052,
      "mean_token_accuracy": 0.1370918095111847,
      "num_tokens": 3435367.0,
      "step": 1980
    },
    {
      "entropy": 6.231957864761353,
      "epoch": 0.15444466057187317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999338903910846,
      "loss": 6.0959,
      "mean_token_accuracy": 0.12951285988092423,
      "num_tokens": 3443451.0,
      "step": 1985
    },
    {
      "entropy": 6.352871370315552,
      "epoch": 0.15483368994359073,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999332168718626,
      "loss": 6.0936,
      "mean_token_accuracy": 0.13427893146872522,
      "num_tokens": 3451038.0,
      "step": 1990
    },
    {
      "entropy": 6.450736808776855,
      "epoch": 0.15522271931530832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999325399396371,
      "loss": 6.237,
      "mean_token_accuracy": 0.11902682781219483,
      "num_tokens": 3459693.0,
      "step": 1995
    },
    {
      "entropy": 6.391193008422851,
      "epoch": 0.15561174868702587,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999318595944183,
      "loss": 6.1066,
      "mean_token_accuracy": 0.13078132271766663,
      "num_tokens": 3468347.0,
      "step": 2000
    },
    {
      "entropy": 6.26676607131958,
      "epoch": 0.15600077805874343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999311758362166,
      "loss": 6.0951,
      "mean_token_accuracy": 0.1329476974904537,
      "num_tokens": 3476030.0,
      "step": 2005
    },
    {
      "entropy": 6.2150873184204105,
      "epoch": 0.156389807430461,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999304886650422,
      "loss": 6.0683,
      "mean_token_accuracy": 0.13065901920199394,
      "num_tokens": 3484840.0,
      "step": 2010
    },
    {
      "entropy": 6.414690160751343,
      "epoch": 0.15677883680217858,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999297980809058,
      "loss": 6.1512,
      "mean_token_accuracy": 0.13070376738905906,
      "num_tokens": 3492662.0,
      "step": 2015
    },
    {
      "entropy": 6.281697416305542,
      "epoch": 0.15716786617389614,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999291040838176,
      "loss": 6.0634,
      "mean_token_accuracy": 0.1307703360915184,
      "num_tokens": 3501582.0,
      "step": 2020
    },
    {
      "entropy": 6.290822887420655,
      "epoch": 0.1575568955456137,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999284066737884,
      "loss": 6.0747,
      "mean_token_accuracy": 0.1364296093583107,
      "num_tokens": 3510141.0,
      "step": 2025
    },
    {
      "entropy": 6.335365724563599,
      "epoch": 0.15794592491733125,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999277058508285,
      "loss": 6.0953,
      "mean_token_accuracy": 0.1390254333615303,
      "num_tokens": 3518530.0,
      "step": 2030
    },
    {
      "entropy": 6.226547431945801,
      "epoch": 0.1583349542890488,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999270016149486,
      "loss": 6.0586,
      "mean_token_accuracy": 0.13643356561660766,
      "num_tokens": 3526815.0,
      "step": 2035
    },
    {
      "entropy": 6.253783798217773,
      "epoch": 0.1587239836607664,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999262939661596,
      "loss": 6.1464,
      "mean_token_accuracy": 0.12736352086067199,
      "num_tokens": 3535840.0,
      "step": 2040
    },
    {
      "entropy": 6.341825342178344,
      "epoch": 0.15911301303248396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999255829044721,
      "loss": 6.1546,
      "mean_token_accuracy": 0.12908357754349709,
      "num_tokens": 3544876.0,
      "step": 2045
    },
    {
      "entropy": 6.260779571533203,
      "epoch": 0.15950204240420152,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999248684298968,
      "loss": 5.9056,
      "mean_token_accuracy": 0.13816021084785463,
      "num_tokens": 3553018.0,
      "step": 2050
    },
    {
      "entropy": 6.160889387130737,
      "epoch": 0.15989107177591907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999241505424447,
      "loss": 5.922,
      "mean_token_accuracy": 0.14369428977370263,
      "num_tokens": 3560909.0,
      "step": 2055
    },
    {
      "entropy": 6.215629625320434,
      "epoch": 0.16028010114763663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999234292421265,
      "loss": 6.102,
      "mean_token_accuracy": 0.1345704235136509,
      "num_tokens": 3569235.0,
      "step": 2060
    },
    {
      "entropy": 6.185765171051026,
      "epoch": 0.16066913051935422,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999227045289535,
      "loss": 5.9789,
      "mean_token_accuracy": 0.13773323893547057,
      "num_tokens": 3577646.0,
      "step": 2065
    },
    {
      "entropy": 6.169071960449219,
      "epoch": 0.16105815989107178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999219764029363,
      "loss": 6.0667,
      "mean_token_accuracy": 0.13329384997487068,
      "num_tokens": 3586499.0,
      "step": 2070
    },
    {
      "entropy": 6.297885990142822,
      "epoch": 0.16144718926278934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999212448640861,
      "loss": 6.0712,
      "mean_token_accuracy": 0.13483210131525994,
      "num_tokens": 3595391.0,
      "step": 2075
    },
    {
      "entropy": 6.178416395187378,
      "epoch": 0.1618362186345069,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999205099124141,
      "loss": 6.0075,
      "mean_token_accuracy": 0.13845352903008462,
      "num_tokens": 3603774.0,
      "step": 2080
    },
    {
      "entropy": 6.350871467590332,
      "epoch": 0.16222524800622448,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999197715479313,
      "loss": 6.0964,
      "mean_token_accuracy": 0.13378985598683357,
      "num_tokens": 3611980.0,
      "step": 2085
    },
    {
      "entropy": 6.227251195907593,
      "epoch": 0.16261427737794204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999190297706489,
      "loss": 5.9931,
      "mean_token_accuracy": 0.1362592376768589,
      "num_tokens": 3620902.0,
      "step": 2090
    },
    {
      "entropy": 6.393834447860717,
      "epoch": 0.1630033067496596,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999182845805783,
      "loss": 6.274,
      "mean_token_accuracy": 0.12474248632788658,
      "num_tokens": 3628924.0,
      "step": 2095
    },
    {
      "entropy": 6.317929124832153,
      "epoch": 0.16339233612137716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999175359777306,
      "loss": 6.1134,
      "mean_token_accuracy": 0.13460347801446915,
      "num_tokens": 3637538.0,
      "step": 2100
    },
    {
      "entropy": 6.309267282485962,
      "epoch": 0.16378136549309472,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999167839621174,
      "loss": 6.1316,
      "mean_token_accuracy": 0.12891944050788878,
      "num_tokens": 3646309.0,
      "step": 2105
    },
    {
      "entropy": 6.248872947692871,
      "epoch": 0.1641703948648123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999160285337501,
      "loss": 5.9786,
      "mean_token_accuracy": 0.13321250006556512,
      "num_tokens": 3655204.0,
      "step": 2110
    },
    {
      "entropy": 6.269497966766357,
      "epoch": 0.16455942423652986,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999152696926399,
      "loss": 6.1226,
      "mean_token_accuracy": 0.13018277511000634,
      "num_tokens": 3663939.0,
      "step": 2115
    },
    {
      "entropy": 6.290651512145996,
      "epoch": 0.16494845360824742,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999145074387985,
      "loss": 6.0531,
      "mean_token_accuracy": 0.1347106009721756,
      "num_tokens": 3673167.0,
      "step": 2120
    },
    {
      "entropy": 6.206559896469116,
      "epoch": 0.16533748297996498,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999137417722374,
      "loss": 6.0324,
      "mean_token_accuracy": 0.13312025964260102,
      "num_tokens": 3682234.0,
      "step": 2125
    },
    {
      "entropy": 6.218487071990967,
      "epoch": 0.16572651235168256,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999129726929684,
      "loss": 5.9961,
      "mean_token_accuracy": 0.13849076852202416,
      "num_tokens": 3690835.0,
      "step": 2130
    },
    {
      "entropy": 6.239537239074707,
      "epoch": 0.16611554172340012,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999122002010029,
      "loss": 6.0597,
      "mean_token_accuracy": 0.1327889084815979,
      "num_tokens": 3699348.0,
      "step": 2135
    },
    {
      "entropy": 6.185144138336182,
      "epoch": 0.16650457109511768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999114242963527,
      "loss": 5.9676,
      "mean_token_accuracy": 0.13893314003944396,
      "num_tokens": 3708205.0,
      "step": 2140
    },
    {
      "entropy": 6.18302903175354,
      "epoch": 0.16689360046683524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999106449790298,
      "loss": 6.0267,
      "mean_token_accuracy": 0.13680716902017592,
      "num_tokens": 3717016.0,
      "step": 2145
    },
    {
      "entropy": 6.388441848754883,
      "epoch": 0.1672826298385528,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999098622490458,
      "loss": 6.0472,
      "mean_token_accuracy": 0.13382177725434302,
      "num_tokens": 3725509.0,
      "step": 2150
    },
    {
      "entropy": 6.247507572174072,
      "epoch": 0.16767165921027039,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999090761064125,
      "loss": 6.1223,
      "mean_token_accuracy": 0.132233477383852,
      "num_tokens": 3734674.0,
      "step": 2155
    },
    {
      "entropy": 6.184755468368531,
      "epoch": 0.16806068858198794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499908286551142,
      "loss": 5.9467,
      "mean_token_accuracy": 0.14066505506634713,
      "num_tokens": 3742930.0,
      "step": 2160
    },
    {
      "entropy": 6.189033365249633,
      "epoch": 0.1684497179537055,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999074935832463,
      "loss": 6.0435,
      "mean_token_accuracy": 0.1365055151283741,
      "num_tokens": 3751314.0,
      "step": 2165
    },
    {
      "entropy": 6.396620225906372,
      "epoch": 0.16883874732542306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999066972027373,
      "loss": 6.1547,
      "mean_token_accuracy": 0.13123885914683342,
      "num_tokens": 3760229.0,
      "step": 2170
    },
    {
      "entropy": 6.372897815704346,
      "epoch": 0.16922777669714065,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999058974096271,
      "loss": 6.1526,
      "mean_token_accuracy": 0.12955897599458693,
      "num_tokens": 3769037.0,
      "step": 2175
    },
    {
      "entropy": 6.131582260131836,
      "epoch": 0.1696168060688582,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499905094203928,
      "loss": 6.023,
      "mean_token_accuracy": 0.13808971717953683,
      "num_tokens": 3777686.0,
      "step": 2180
    },
    {
      "entropy": 6.375473928451538,
      "epoch": 0.17000583544057576,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499904287585652,
      "loss": 6.1105,
      "mean_token_accuracy": 0.12939185723662378,
      "num_tokens": 3786412.0,
      "step": 2185
    },
    {
      "entropy": 6.248593330383301,
      "epoch": 0.17039486481229332,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999034775548113,
      "loss": 6.1397,
      "mean_token_accuracy": 0.13407278731465339,
      "num_tokens": 3794981.0,
      "step": 2190
    },
    {
      "entropy": 6.278722524642944,
      "epoch": 0.17078389418401088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999026641114185,
      "loss": 6.0252,
      "mean_token_accuracy": 0.13181823194026948,
      "num_tokens": 3803069.0,
      "step": 2195
    },
    {
      "entropy": 6.28540301322937,
      "epoch": 0.17117292355572847,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999018472554857,
      "loss": 6.1039,
      "mean_token_accuracy": 0.1345643788576126,
      "num_tokens": 3811986.0,
      "step": 2200
    },
    {
      "entropy": 6.29768214225769,
      "epoch": 0.17156195292744603,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999010269870253,
      "loss": 6.0161,
      "mean_token_accuracy": 0.13579363152384757,
      "num_tokens": 3820626.0,
      "step": 2205
    },
    {
      "entropy": 6.200810480117798,
      "epoch": 0.17195098229916359,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999002033060498,
      "loss": 6.0268,
      "mean_token_accuracy": 0.13920362591743468,
      "num_tokens": 3828608.0,
      "step": 2210
    },
    {
      "entropy": 6.193022918701172,
      "epoch": 0.17234001167088114,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998993762125716,
      "loss": 5.9785,
      "mean_token_accuracy": 0.13870956525206565,
      "num_tokens": 3837059.0,
      "step": 2215
    },
    {
      "entropy": 6.207073402404785,
      "epoch": 0.17272904104259873,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998985457066034,
      "loss": 5.9823,
      "mean_token_accuracy": 0.1358140230178833,
      "num_tokens": 3845137.0,
      "step": 2220
    },
    {
      "entropy": 6.209107160568237,
      "epoch": 0.1731180704143163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998977117881576,
      "loss": 5.9817,
      "mean_token_accuracy": 0.14150308892130853,
      "num_tokens": 3854012.0,
      "step": 2225
    },
    {
      "entropy": 6.067155075073242,
      "epoch": 0.17350709978603385,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998968744572472,
      "loss": 5.9361,
      "mean_token_accuracy": 0.14006831347942353,
      "num_tokens": 3862308.0,
      "step": 2230
    },
    {
      "entropy": 6.171311616897583,
      "epoch": 0.1738961291577514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998960337138845,
      "loss": 5.8703,
      "mean_token_accuracy": 0.13672190457582473,
      "num_tokens": 3870975.0,
      "step": 2235
    },
    {
      "entropy": 6.223893642425537,
      "epoch": 0.17428515852946896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998951895580826,
      "loss": 6.031,
      "mean_token_accuracy": 0.13869215548038483,
      "num_tokens": 3879621.0,
      "step": 2240
    },
    {
      "entropy": 6.188286018371582,
      "epoch": 0.17467418790118655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998943419898542,
      "loss": 6.0074,
      "mean_token_accuracy": 0.1369040384888649,
      "num_tokens": 3889214.0,
      "step": 2245
    },
    {
      "entropy": 6.221569967269898,
      "epoch": 0.1750632172729041,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998934910092119,
      "loss": 6.0449,
      "mean_token_accuracy": 0.13841852992773057,
      "num_tokens": 3898250.0,
      "step": 2250
    },
    {
      "entropy": 6.1772825717926025,
      "epoch": 0.17545224664462167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998926366161689,
      "loss": 6.0791,
      "mean_token_accuracy": 0.1358688361942768,
      "num_tokens": 3906971.0,
      "step": 2255
    },
    {
      "entropy": 6.292248201370239,
      "epoch": 0.17584127601633923,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998917788107383,
      "loss": 6.1077,
      "mean_token_accuracy": 0.13412976861000062,
      "num_tokens": 3916663.0,
      "step": 2260
    },
    {
      "entropy": 6.169425630569458,
      "epoch": 0.17623030538805678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998909175929326,
      "loss": 5.9801,
      "mean_token_accuracy": 0.13465463891625404,
      "num_tokens": 3926848.0,
      "step": 2265
    },
    {
      "entropy": 6.196387386322021,
      "epoch": 0.17661933475977437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998900529627653,
      "loss": 5.9622,
      "mean_token_accuracy": 0.13440250605344772,
      "num_tokens": 3935729.0,
      "step": 2270
    },
    {
      "entropy": 6.222635746002197,
      "epoch": 0.17700836413149193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998891849202495,
      "loss": 5.9687,
      "mean_token_accuracy": 0.1374209851026535,
      "num_tokens": 3943408.0,
      "step": 2275
    },
    {
      "entropy": 6.188998603820801,
      "epoch": 0.1773973935032095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499888313465398,
      "loss": 5.9327,
      "mean_token_accuracy": 0.1366911180317402,
      "num_tokens": 3952473.0,
      "step": 2280
    },
    {
      "entropy": 6.274978828430176,
      "epoch": 0.17778642287492705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998874385982247,
      "loss": 6.1128,
      "mean_token_accuracy": 0.1279517851769924,
      "num_tokens": 3960878.0,
      "step": 2285
    },
    {
      "entropy": 6.270401048660278,
      "epoch": 0.17817545224664463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998865603187421,
      "loss": 6.0625,
      "mean_token_accuracy": 0.13058168217539787,
      "num_tokens": 3969282.0,
      "step": 2290
    },
    {
      "entropy": 6.220210218429566,
      "epoch": 0.1785644816183622,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499885678626964,
      "loss": 6.014,
      "mean_token_accuracy": 0.13722910657525061,
      "num_tokens": 3978516.0,
      "step": 2295
    },
    {
      "entropy": 6.137257432937622,
      "epoch": 0.17895351099007975,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998847935229038,
      "loss": 5.8965,
      "mean_token_accuracy": 0.1428119085729122,
      "num_tokens": 3986580.0,
      "step": 2300
    },
    {
      "entropy": 6.080727529525757,
      "epoch": 0.1793425403617973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998839050065746,
      "loss": 5.8953,
      "mean_token_accuracy": 0.14140310510993004,
      "num_tokens": 3995144.0,
      "step": 2305
    },
    {
      "entropy": 6.176934671401978,
      "epoch": 0.17973156973351487,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998830130779902,
      "loss": 5.8921,
      "mean_token_accuracy": 0.1366891786456108,
      "num_tokens": 4003887.0,
      "step": 2310
    },
    {
      "entropy": 6.2093428611755375,
      "epoch": 0.18012059910523245,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499882117737164,
      "loss": 6.0162,
      "mean_token_accuracy": 0.13656655922532082,
      "num_tokens": 4012607.0,
      "step": 2315
    },
    {
      "entropy": 6.150680923461914,
      "epoch": 0.18050962847695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998812189841096,
      "loss": 6.081,
      "mean_token_accuracy": 0.13370702415704727,
      "num_tokens": 4021950.0,
      "step": 2320
    },
    {
      "entropy": 6.2497515201568605,
      "epoch": 0.18089865784866757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998803168188407,
      "loss": 6.0404,
      "mean_token_accuracy": 0.13844593837857247,
      "num_tokens": 4031076.0,
      "step": 2325
    },
    {
      "entropy": 6.116616344451904,
      "epoch": 0.18128768722038513,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998794112413708,
      "loss": 5.8958,
      "mean_token_accuracy": 0.1410272441804409,
      "num_tokens": 4040414.0,
      "step": 2330
    },
    {
      "entropy": 6.26803183555603,
      "epoch": 0.18167671659210272,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998785022517137,
      "loss": 5.9815,
      "mean_token_accuracy": 0.14090483263134956,
      "num_tokens": 4048931.0,
      "step": 2335
    },
    {
      "entropy": 6.045533609390259,
      "epoch": 0.18206574596382027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998775898498835,
      "loss": 5.8712,
      "mean_token_accuracy": 0.14119707196950912,
      "num_tokens": 4057688.0,
      "step": 2340
    },
    {
      "entropy": 6.100311326980591,
      "epoch": 0.18245477533553783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998766740358936,
      "loss": 5.9949,
      "mean_token_accuracy": 0.13843753784894944,
      "num_tokens": 4066365.0,
      "step": 2345
    },
    {
      "entropy": 6.204766511917114,
      "epoch": 0.1828438047072554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998757548097582,
      "loss": 5.8852,
      "mean_token_accuracy": 0.14553529024124146,
      "num_tokens": 4075136.0,
      "step": 2350
    },
    {
      "entropy": 6.189942455291748,
      "epoch": 0.18323283407897295,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998748321714911,
      "loss": 5.9877,
      "mean_token_accuracy": 0.13881198838353156,
      "num_tokens": 4083337.0,
      "step": 2355
    },
    {
      "entropy": 6.132948780059815,
      "epoch": 0.18362186345069054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998739061211065,
      "loss": 5.9454,
      "mean_token_accuracy": 0.14238794520497322,
      "num_tokens": 4091201.0,
      "step": 2360
    },
    {
      "entropy": 6.194742584228516,
      "epoch": 0.1840108928224081,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998729766586181,
      "loss": 6.018,
      "mean_token_accuracy": 0.13472261428833007,
      "num_tokens": 4099480.0,
      "step": 2365
    },
    {
      "entropy": 6.089618396759033,
      "epoch": 0.18439992219412565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998720437840403,
      "loss": 5.9125,
      "mean_token_accuracy": 0.14144392758607865,
      "num_tokens": 4108151.0,
      "step": 2370
    },
    {
      "entropy": 6.022348308563233,
      "epoch": 0.1847889515658432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998711074973871,
      "loss": 5.9047,
      "mean_token_accuracy": 0.13933733105659485,
      "num_tokens": 4117380.0,
      "step": 2375
    },
    {
      "entropy": 6.1961781024932865,
      "epoch": 0.1851779809375608,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998701677986728,
      "loss": 5.9593,
      "mean_token_accuracy": 0.14379323944449424,
      "num_tokens": 4126757.0,
      "step": 2380
    },
    {
      "entropy": 6.117034149169922,
      "epoch": 0.18556701030927836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998692246879116,
      "loss": 5.8844,
      "mean_token_accuracy": 0.13709820285439492,
      "num_tokens": 4135326.0,
      "step": 2385
    },
    {
      "entropy": 6.139100790023804,
      "epoch": 0.18595603968099592,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998682781651178,
      "loss": 5.9112,
      "mean_token_accuracy": 0.14516514390707017,
      "num_tokens": 4143624.0,
      "step": 2390
    },
    {
      "entropy": 6.090607690811157,
      "epoch": 0.18634506905271347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998673282303059,
      "loss": 5.851,
      "mean_token_accuracy": 0.14274040311574937,
      "num_tokens": 4152193.0,
      "step": 2395
    },
    {
      "entropy": 6.108787965774536,
      "epoch": 0.18673409842443103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998663748834901,
      "loss": 5.9906,
      "mean_token_accuracy": 0.1388832874596119,
      "num_tokens": 4160614.0,
      "step": 2400
    },
    {
      "entropy": 6.251668167114258,
      "epoch": 0.18712312779614862,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998654181246851,
      "loss": 5.9976,
      "mean_token_accuracy": 0.14036377519369125,
      "num_tokens": 4168715.0,
      "step": 2405
    },
    {
      "entropy": 6.176876354217529,
      "epoch": 0.18751215716786618,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998644579539052,
      "loss": 6.0395,
      "mean_token_accuracy": 0.14050538837909698,
      "num_tokens": 4176676.0,
      "step": 2410
    },
    {
      "entropy": 6.176787853240967,
      "epoch": 0.18790118653958374,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499863494371165,
      "loss": 6.0212,
      "mean_token_accuracy": 0.13556102365255357,
      "num_tokens": 4185613.0,
      "step": 2415
    },
    {
      "entropy": 6.17087230682373,
      "epoch": 0.1882902159113013,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998625273764793,
      "loss": 6.0086,
      "mean_token_accuracy": 0.14141924679279327,
      "num_tokens": 4193985.0,
      "step": 2420
    },
    {
      "entropy": 6.127655124664306,
      "epoch": 0.18867924528301888,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998615569698626,
      "loss": 5.9067,
      "mean_token_accuracy": 0.1402697630226612,
      "num_tokens": 4202713.0,
      "step": 2425
    },
    {
      "entropy": 6.265314626693725,
      "epoch": 0.18906827465473644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998605831513296,
      "loss": 6.0448,
      "mean_token_accuracy": 0.13398367688059806,
      "num_tokens": 4210858.0,
      "step": 2430
    },
    {
      "entropy": 6.098290967941284,
      "epoch": 0.189457304026454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998596059208953,
      "loss": 5.9564,
      "mean_token_accuracy": 0.14138128459453583,
      "num_tokens": 4219570.0,
      "step": 2435
    },
    {
      "entropy": 6.1270262718200685,
      "epoch": 0.18984633339817156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998586252785743,
      "loss": 5.8816,
      "mean_token_accuracy": 0.14256712123751641,
      "num_tokens": 4228026.0,
      "step": 2440
    },
    {
      "entropy": 6.123672199249268,
      "epoch": 0.19023536276988912,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998576412243816,
      "loss": 5.8953,
      "mean_token_accuracy": 0.1403258338570595,
      "num_tokens": 4235993.0,
      "step": 2445
    },
    {
      "entropy": 6.104493427276611,
      "epoch": 0.1906243921416067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998566537583321,
      "loss": 5.8981,
      "mean_token_accuracy": 0.14673311710357667,
      "num_tokens": 4244225.0,
      "step": 2450
    },
    {
      "entropy": 6.084668493270874,
      "epoch": 0.19101342151332426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998556628804408,
      "loss": 5.8763,
      "mean_token_accuracy": 0.14110946282744408,
      "num_tokens": 4252989.0,
      "step": 2455
    },
    {
      "entropy": 6.1425386428833,
      "epoch": 0.19140245088504182,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998546685907225,
      "loss": 5.917,
      "mean_token_accuracy": 0.14222090542316437,
      "num_tokens": 4261297.0,
      "step": 2460
    },
    {
      "entropy": 6.189026689529419,
      "epoch": 0.19179148025675938,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998536708891927,
      "loss": 6.0593,
      "mean_token_accuracy": 0.13440425619482993,
      "num_tokens": 4270299.0,
      "step": 2465
    },
    {
      "entropy": 6.250662851333618,
      "epoch": 0.19218050962847694,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998526697758663,
      "loss": 6.0489,
      "mean_token_accuracy": 0.13370652571320535,
      "num_tokens": 4279662.0,
      "step": 2470
    },
    {
      "entropy": 6.1341468334198,
      "epoch": 0.19256953900019452,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998516652507585,
      "loss": 5.8811,
      "mean_token_accuracy": 0.14475788101553916,
      "num_tokens": 4287876.0,
      "step": 2475
    },
    {
      "entropy": 6.139857435226441,
      "epoch": 0.19295856837191208,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998506573138846,
      "loss": 6.0032,
      "mean_token_accuracy": 0.14181941896677017,
      "num_tokens": 4296415.0,
      "step": 2480
    },
    {
      "entropy": 6.04846830368042,
      "epoch": 0.19334759774362964,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998496459652598,
      "loss": 5.8458,
      "mean_token_accuracy": 0.14211475402116774,
      "num_tokens": 4304434.0,
      "step": 2485
    },
    {
      "entropy": 6.2117273807525635,
      "epoch": 0.1937366271153472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998486312048995,
      "loss": 6.0062,
      "mean_token_accuracy": 0.13499248623847962,
      "num_tokens": 4313266.0,
      "step": 2490
    },
    {
      "entropy": 6.064237928390503,
      "epoch": 0.19412565648706478,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998476130328193,
      "loss": 5.8892,
      "mean_token_accuracy": 0.14858543127775192,
      "num_tokens": 4322121.0,
      "step": 2495
    },
    {
      "entropy": 6.075144720077515,
      "epoch": 0.19451468585878234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998465914490343,
      "loss": 5.9025,
      "mean_token_accuracy": 0.14101851657032966,
      "num_tokens": 4330596.0,
      "step": 2500
    },
    {
      "entropy": 6.195386219024658,
      "epoch": 0.1949037152304999,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049984556645356,
      "loss": 6.0081,
      "mean_token_accuracy": 0.14117387384176255,
      "num_tokens": 4337779.0,
      "step": 2505
    },
    {
      "entropy": 6.147725248336792,
      "epoch": 0.19529274460221746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998445380464124,
      "loss": 5.8344,
      "mean_token_accuracy": 0.14705330207943917,
      "num_tokens": 4346623.0,
      "step": 2510
    },
    {
      "entropy": 6.002147150039673,
      "epoch": 0.19568177397393502,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998435062276066,
      "loss": 5.9455,
      "mean_token_accuracy": 0.13899349197745323,
      "num_tokens": 4355435.0,
      "step": 2515
    },
    {
      "entropy": 6.251942825317383,
      "epoch": 0.1960708033456526,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998424709971586,
      "loss": 5.9863,
      "mean_token_accuracy": 0.14054905995726585,
      "num_tokens": 4363636.0,
      "step": 2520
    },
    {
      "entropy": 6.0051055431365965,
      "epoch": 0.19645983271737016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998414323550839,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1456178233027458,
      "num_tokens": 4372337.0,
      "step": 2525
    },
    {
      "entropy": 6.116535997390747,
      "epoch": 0.19684886208908772,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998403903013984,
      "loss": 5.8476,
      "mean_token_accuracy": 0.1478264570236206,
      "num_tokens": 4380528.0,
      "step": 2530
    },
    {
      "entropy": 6.055860280990601,
      "epoch": 0.19723789146080528,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998393448361179,
      "loss": 5.9436,
      "mean_token_accuracy": 0.1379445366561413,
      "num_tokens": 4389464.0,
      "step": 2535
    },
    {
      "entropy": 6.178743028640747,
      "epoch": 0.19762692083252287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499838295959258,
      "loss": 6.0005,
      "mean_token_accuracy": 0.1344634510576725,
      "num_tokens": 4397879.0,
      "step": 2540
    },
    {
      "entropy": 6.147117757797242,
      "epoch": 0.19801595020424043,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499837243670835,
      "loss": 6.0062,
      "mean_token_accuracy": 0.13882223516702652,
      "num_tokens": 4406145.0,
      "step": 2545
    },
    {
      "entropy": 6.10461392402649,
      "epoch": 0.19840497957595798,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998361879708646,
      "loss": 5.9092,
      "mean_token_accuracy": 0.137804589420557,
      "num_tokens": 4415480.0,
      "step": 2550
    },
    {
      "entropy": 6.154189825057983,
      "epoch": 0.19879400894767554,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499835128859363,
      "loss": 5.9301,
      "mean_token_accuracy": 0.14363658130168916,
      "num_tokens": 4423355.0,
      "step": 2555
    },
    {
      "entropy": 6.089098072052002,
      "epoch": 0.1991830383193931,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998340663363461,
      "loss": 5.8279,
      "mean_token_accuracy": 0.1488031305372715,
      "num_tokens": 4432059.0,
      "step": 2560
    },
    {
      "entropy": 6.035713958740234,
      "epoch": 0.1995720676911107,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998330004018301,
      "loss": 5.9177,
      "mean_token_accuracy": 0.13287902176380156,
      "num_tokens": 4441146.0,
      "step": 2565
    },
    {
      "entropy": 6.091567516326904,
      "epoch": 0.19996109706282825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998319310558312,
      "loss": 5.7498,
      "mean_token_accuracy": 0.15268679708242416,
      "num_tokens": 4448553.0,
      "step": 2570
    },
    {
      "entropy": 5.924048852920532,
      "epoch": 0.2003501264345458,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998308582983657,
      "loss": 5.8311,
      "mean_token_accuracy": 0.1501924604177475,
      "num_tokens": 4456930.0,
      "step": 2575
    },
    {
      "entropy": 6.182810688018799,
      "epoch": 0.20073915580626336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998297821294497,
      "loss": 6.0141,
      "mean_token_accuracy": 0.14234298169612886,
      "num_tokens": 4464932.0,
      "step": 2580
    },
    {
      "entropy": 6.097385025024414,
      "epoch": 0.20112818517798095,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998287025490995,
      "loss": 5.7409,
      "mean_token_accuracy": 0.15158229768276216,
      "num_tokens": 4473695.0,
      "step": 2585
    },
    {
      "entropy": 5.993387842178345,
      "epoch": 0.2015172145496985,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998276195573317,
      "loss": 5.9849,
      "mean_token_accuracy": 0.1368507519364357,
      "num_tokens": 4483375.0,
      "step": 2590
    },
    {
      "entropy": 6.1823647022247314,
      "epoch": 0.20190624392141607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998265331541627,
      "loss": 6.0034,
      "mean_token_accuracy": 0.13301878720521926,
      "num_tokens": 4491695.0,
      "step": 2595
    },
    {
      "entropy": 6.069152975082398,
      "epoch": 0.20229527329313363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998254433396088,
      "loss": 5.9566,
      "mean_token_accuracy": 0.14346428364515304,
      "num_tokens": 4500361.0,
      "step": 2600
    },
    {
      "entropy": 6.137755489349365,
      "epoch": 0.20268430266485118,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998243501136867,
      "loss": 5.891,
      "mean_token_accuracy": 0.14259622395038604,
      "num_tokens": 4509368.0,
      "step": 2605
    },
    {
      "entropy": 6.1446033954620365,
      "epoch": 0.20307333203656877,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998232534764129,
      "loss": 5.8852,
      "mean_token_accuracy": 0.13794858306646346,
      "num_tokens": 4518146.0,
      "step": 2610
    },
    {
      "entropy": 6.1004749774932865,
      "epoch": 0.20346236140828633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998221534278041,
      "loss": 5.9193,
      "mean_token_accuracy": 0.1392908826470375,
      "num_tokens": 4526128.0,
      "step": 2615
    },
    {
      "entropy": 6.157025766372681,
      "epoch": 0.2038513907800039,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998210499678769,
      "loss": 5.9593,
      "mean_token_accuracy": 0.14478977620601655,
      "num_tokens": 4534724.0,
      "step": 2620
    },
    {
      "entropy": 6.023615980148316,
      "epoch": 0.20424042015172145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998199430966481,
      "loss": 5.821,
      "mean_token_accuracy": 0.1394574910402298,
      "num_tokens": 4543725.0,
      "step": 2625
    },
    {
      "entropy": 6.114223003387451,
      "epoch": 0.20462944952343903,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998188328141346,
      "loss": 5.9023,
      "mean_token_accuracy": 0.14211260229349137,
      "num_tokens": 4551758.0,
      "step": 2630
    },
    {
      "entropy": 6.020139360427857,
      "epoch": 0.2050184788951566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998177191203531,
      "loss": 5.7926,
      "mean_token_accuracy": 0.14570502564311028,
      "num_tokens": 4559990.0,
      "step": 2635
    },
    {
      "entropy": 6.072308111190796,
      "epoch": 0.20540750826687415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998166020153204,
      "loss": 5.8326,
      "mean_token_accuracy": 0.1404043950140476,
      "num_tokens": 4568651.0,
      "step": 2640
    },
    {
      "entropy": 5.98312201499939,
      "epoch": 0.2057965376385917,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998154814990538,
      "loss": 5.7583,
      "mean_token_accuracy": 0.14969710260629654,
      "num_tokens": 4577981.0,
      "step": 2645
    },
    {
      "entropy": 6.022286748886108,
      "epoch": 0.20618556701030927,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049981435757157,
      "loss": 5.9188,
      "mean_token_accuracy": 0.14773550853133202,
      "num_tokens": 4586480.0,
      "step": 2650
    },
    {
      "entropy": 6.1229582786560055,
      "epoch": 0.20657459638202685,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998132302328862,
      "loss": 5.8035,
      "mean_token_accuracy": 0.15050222277641295,
      "num_tokens": 4595554.0,
      "step": 2655
    },
    {
      "entropy": 6.027739095687866,
      "epoch": 0.2069636257537444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998120994830195,
      "loss": 5.8229,
      "mean_token_accuracy": 0.14623611718416213,
      "num_tokens": 4604091.0,
      "step": 2660
    },
    {
      "entropy": 6.164302587509155,
      "epoch": 0.20735265512546197,
      "grad_norm": 1.0,
      "learning_rate": 0.000499810965321987,
      "loss": 5.9514,
      "mean_token_accuracy": 0.14055586755275726,
      "num_tokens": 4612603.0,
      "step": 2665
    },
    {
      "entropy": 6.115392780303955,
      "epoch": 0.20774168449717953,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998098277498059,
      "loss": 5.8343,
      "mean_token_accuracy": 0.14225830659270286,
      "num_tokens": 4621041.0,
      "step": 2670
    },
    {
      "entropy": 5.985233879089355,
      "epoch": 0.2081307138688971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998086867664934,
      "loss": 5.8627,
      "mean_token_accuracy": 0.14746104627847673,
      "num_tokens": 4629849.0,
      "step": 2675
    },
    {
      "entropy": 6.099282693862915,
      "epoch": 0.20851974324061467,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499807542372067,
      "loss": 5.95,
      "mean_token_accuracy": 0.14111322686076164,
      "num_tokens": 4639426.0,
      "step": 2680
    },
    {
      "entropy": 6.099315929412842,
      "epoch": 0.20890877261233223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998063945665439,
      "loss": 5.8792,
      "mean_token_accuracy": 0.14589827060699462,
      "num_tokens": 4647257.0,
      "step": 2685
    },
    {
      "entropy": 6.0904378414154055,
      "epoch": 0.2092978019840498,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998052433499416,
      "loss": 5.8631,
      "mean_token_accuracy": 0.14348279163241387,
      "num_tokens": 4655773.0,
      "step": 2690
    },
    {
      "entropy": 6.052747917175293,
      "epoch": 0.20968683135576735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998040887222776,
      "loss": 5.8834,
      "mean_token_accuracy": 0.1426556721329689,
      "num_tokens": 4664292.0,
      "step": 2695
    },
    {
      "entropy": 6.178410577774048,
      "epoch": 0.21007586072748494,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998029306835693,
      "loss": 5.9237,
      "mean_token_accuracy": 0.1449746683239937,
      "num_tokens": 4672774.0,
      "step": 2700
    },
    {
      "entropy": 6.213541555404663,
      "epoch": 0.2104648900992025,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998017692338344,
      "loss": 6.0447,
      "mean_token_accuracy": 0.13644783496856688,
      "num_tokens": 4681855.0,
      "step": 2705
    },
    {
      "entropy": 6.082818508148193,
      "epoch": 0.21085391947092005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998006043730905,
      "loss": 5.814,
      "mean_token_accuracy": 0.14876434952020645,
      "num_tokens": 4689771.0,
      "step": 2710
    },
    {
      "entropy": 6.110875511169434,
      "epoch": 0.2112429488426376,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997994361013551,
      "loss": 5.9807,
      "mean_token_accuracy": 0.131246355175972,
      "num_tokens": 4698808.0,
      "step": 2715
    },
    {
      "entropy": 6.077038431167603,
      "epoch": 0.21163197821435517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997982644186461,
      "loss": 5.8661,
      "mean_token_accuracy": 0.14354205653071403,
      "num_tokens": 4707627.0,
      "step": 2720
    },
    {
      "entropy": 6.0479084014892575,
      "epoch": 0.21202100758607276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997970893249813,
      "loss": 5.8537,
      "mean_token_accuracy": 0.14187933951616288,
      "num_tokens": 4716091.0,
      "step": 2725
    },
    {
      "entropy": 6.063355493545532,
      "epoch": 0.21241003695779032,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997959108203785,
      "loss": 5.871,
      "mean_token_accuracy": 0.1473940022289753,
      "num_tokens": 4724042.0,
      "step": 2730
    },
    {
      "entropy": 5.985284852981567,
      "epoch": 0.21279906632950787,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997947289048554,
      "loss": 5.7784,
      "mean_token_accuracy": 0.1558399736881256,
      "num_tokens": 4731745.0,
      "step": 2735
    },
    {
      "entropy": 6.077976179122925,
      "epoch": 0.21318809570122543,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997935435784302,
      "loss": 5.9664,
      "mean_token_accuracy": 0.14258809015154839,
      "num_tokens": 4740386.0,
      "step": 2740
    },
    {
      "entropy": 6.121005344390869,
      "epoch": 0.21357712507294302,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997923548411208,
      "loss": 5.9529,
      "mean_token_accuracy": 0.135360998660326,
      "num_tokens": 4749809.0,
      "step": 2745
    },
    {
      "entropy": 6.177447128295898,
      "epoch": 0.21396615444466058,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499791162692945,
      "loss": 6.0766,
      "mean_token_accuracy": 0.1363483987748623,
      "num_tokens": 4759507.0,
      "step": 2750
    },
    {
      "entropy": 6.055140495300293,
      "epoch": 0.21435518381637814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997899671339214,
      "loss": 5.8378,
      "mean_token_accuracy": 0.14012878388166428,
      "num_tokens": 4768793.0,
      "step": 2755
    },
    {
      "entropy": 6.124060440063476,
      "epoch": 0.2147442131880957,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997887681640675,
      "loss": 5.9864,
      "mean_token_accuracy": 0.136216626316309,
      "num_tokens": 4777344.0,
      "step": 2760
    },
    {
      "entropy": 6.160050868988037,
      "epoch": 0.21513324255981325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997875657834021,
      "loss": 5.9176,
      "mean_token_accuracy": 0.14089114516973494,
      "num_tokens": 4785934.0,
      "step": 2765
    },
    {
      "entropy": 6.034284019470215,
      "epoch": 0.21552227193153084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997863599919432,
      "loss": 5.8573,
      "mean_token_accuracy": 0.1446565344929695,
      "num_tokens": 4794889.0,
      "step": 2770
    },
    {
      "entropy": 6.02419900894165,
      "epoch": 0.2159113013032484,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997851507897089,
      "loss": 5.9665,
      "mean_token_accuracy": 0.13921516314148902,
      "num_tokens": 4803597.0,
      "step": 2775
    },
    {
      "entropy": 6.091166543960571,
      "epoch": 0.21630033067496596,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997839381767178,
      "loss": 5.7469,
      "mean_token_accuracy": 0.15012719631195068,
      "num_tokens": 4812706.0,
      "step": 2780
    },
    {
      "entropy": 5.975925588607788,
      "epoch": 0.21668936004668352,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997827221529882,
      "loss": 5.8138,
      "mean_token_accuracy": 0.1410752348601818,
      "num_tokens": 4821790.0,
      "step": 2785
    },
    {
      "entropy": 6.095637559890747,
      "epoch": 0.2170783894184011,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997815027185386,
      "loss": 5.8651,
      "mean_token_accuracy": 0.14362238049507142,
      "num_tokens": 4830123.0,
      "step": 2790
    },
    {
      "entropy": 6.073716259002685,
      "epoch": 0.21746741879011866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997802798733874,
      "loss": 5.7984,
      "mean_token_accuracy": 0.14468253999948502,
      "num_tokens": 4838066.0,
      "step": 2795
    },
    {
      "entropy": 6.039735746383667,
      "epoch": 0.21785644816183622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997790536175534,
      "loss": 5.8614,
      "mean_token_accuracy": 0.1461779847741127,
      "num_tokens": 4846877.0,
      "step": 2800
    },
    {
      "entropy": 6.086436319351196,
      "epoch": 0.21824547753355378,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997778239510548,
      "loss": 5.9278,
      "mean_token_accuracy": 0.13898368775844575,
      "num_tokens": 4856960.0,
      "step": 2805
    },
    {
      "entropy": 6.088310766220093,
      "epoch": 0.21863450690527134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997765908739106,
      "loss": 5.9417,
      "mean_token_accuracy": 0.14190473333001136,
      "num_tokens": 4865440.0,
      "step": 2810
    },
    {
      "entropy": 6.019820642471314,
      "epoch": 0.21902353627698892,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997753543861395,
      "loss": 5.8224,
      "mean_token_accuracy": 0.14486733451485634,
      "num_tokens": 4874187.0,
      "step": 2815
    },
    {
      "entropy": 6.17541913986206,
      "epoch": 0.21941256564870648,
      "grad_norm": 0.875,
      "learning_rate": 0.00049977411448776,
      "loss": 5.9681,
      "mean_token_accuracy": 0.13802750185132026,
      "num_tokens": 4883151.0,
      "step": 2820
    },
    {
      "entropy": 6.097297096252442,
      "epoch": 0.21980159502042404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997728711787912,
      "loss": 5.8656,
      "mean_token_accuracy": 0.14226908087730408,
      "num_tokens": 4890968.0,
      "step": 2825
    },
    {
      "entropy": 6.005394792556762,
      "epoch": 0.2201906243921416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997716244592518,
      "loss": 5.7664,
      "mean_token_accuracy": 0.15000639855861664,
      "num_tokens": 4899389.0,
      "step": 2830
    },
    {
      "entropy": 6.016676950454712,
      "epoch": 0.22057965376385918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997703743291607,
      "loss": 5.8472,
      "mean_token_accuracy": 0.14968255087733268,
      "num_tokens": 4907622.0,
      "step": 2835
    },
    {
      "entropy": 6.053238010406494,
      "epoch": 0.22096868313557674,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499769120788537,
      "loss": 5.8416,
      "mean_token_accuracy": 0.14399737119674683,
      "num_tokens": 4916151.0,
      "step": 2840
    },
    {
      "entropy": 5.965752077102661,
      "epoch": 0.2213577125072943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997678638373995,
      "loss": 5.7334,
      "mean_token_accuracy": 0.1568694993853569,
      "num_tokens": 4923938.0,
      "step": 2845
    },
    {
      "entropy": 6.076515531539917,
      "epoch": 0.22174674187901186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997666034757676,
      "loss": 5.8339,
      "mean_token_accuracy": 0.15386861115694045,
      "num_tokens": 4932027.0,
      "step": 2850
    },
    {
      "entropy": 6.042121171951294,
      "epoch": 0.22213577125072942,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997653397036603,
      "loss": 5.8524,
      "mean_token_accuracy": 0.15097130537033082,
      "num_tokens": 4940229.0,
      "step": 2855
    },
    {
      "entropy": 5.992015314102173,
      "epoch": 0.222524800622447,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997640725210965,
      "loss": 5.9242,
      "mean_token_accuracy": 0.14181902557611464,
      "num_tokens": 4948662.0,
      "step": 2860
    },
    {
      "entropy": 6.072470712661743,
      "epoch": 0.22291382999416456,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997628019280958,
      "loss": 5.8773,
      "mean_token_accuracy": 0.14416748732328416,
      "num_tokens": 4956847.0,
      "step": 2865
    },
    {
      "entropy": 6.070145511627198,
      "epoch": 0.22330285936588212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997615279246773,
      "loss": 5.8031,
      "mean_token_accuracy": 0.14515076130628585,
      "num_tokens": 4965139.0,
      "step": 2870
    },
    {
      "entropy": 6.135627269744873,
      "epoch": 0.22369188873759968,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997602505108603,
      "loss": 5.9724,
      "mean_token_accuracy": 0.1393357440829277,
      "num_tokens": 4973671.0,
      "step": 2875
    },
    {
      "entropy": 6.075475406646729,
      "epoch": 0.22408091810931724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997589696866644,
      "loss": 5.8527,
      "mean_token_accuracy": 0.14160598665475846,
      "num_tokens": 4981970.0,
      "step": 2880
    },
    {
      "entropy": 6.048443412780761,
      "epoch": 0.22446994748103483,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997576854521088,
      "loss": 5.8601,
      "mean_token_accuracy": 0.14621118903160096,
      "num_tokens": 4990591.0,
      "step": 2885
    },
    {
      "entropy": 6.036798810958862,
      "epoch": 0.22485897685275238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499756397807213,
      "loss": 5.8737,
      "mean_token_accuracy": 0.1408968098461628,
      "num_tokens": 4999623.0,
      "step": 2890
    },
    {
      "entropy": 6.154255437850952,
      "epoch": 0.22524800622446994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997551067519966,
      "loss": 5.9326,
      "mean_token_accuracy": 0.13853658065199853,
      "num_tokens": 5008387.0,
      "step": 2895
    },
    {
      "entropy": 6.0091872215271,
      "epoch": 0.2256370355961875,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997538122864792,
      "loss": 5.8513,
      "mean_token_accuracy": 0.1488157957792282,
      "num_tokens": 5016788.0,
      "step": 2900
    },
    {
      "entropy": 6.0877045631408695,
      "epoch": 0.2260260649679051,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997525144106804,
      "loss": 5.8786,
      "mean_token_accuracy": 0.14177965149283409,
      "num_tokens": 5025508.0,
      "step": 2905
    },
    {
      "entropy": 6.003860330581665,
      "epoch": 0.22641509433962265,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049975121312462,
      "loss": 5.8675,
      "mean_token_accuracy": 0.13894690126180648,
      "num_tokens": 5035081.0,
      "step": 2910
    },
    {
      "entropy": 6.021999168395996,
      "epoch": 0.2268041237113402,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997499084283177,
      "loss": 5.7956,
      "mean_token_accuracy": 0.1498025491833687,
      "num_tokens": 5043655.0,
      "step": 2915
    },
    {
      "entropy": 6.045802927017212,
      "epoch": 0.22719315308305776,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997486003217932,
      "loss": 5.8186,
      "mean_token_accuracy": 0.1467895582318306,
      "num_tokens": 5052080.0,
      "step": 2920
    },
    {
      "entropy": 5.940340852737426,
      "epoch": 0.22758218245477532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997472888050664,
      "loss": 5.789,
      "mean_token_accuracy": 0.15500381588935852,
      "num_tokens": 5061325.0,
      "step": 2925
    },
    {
      "entropy": 6.0531168460845945,
      "epoch": 0.2279712118264929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997459738781573,
      "loss": 5.8874,
      "mean_token_accuracy": 0.1485513836145401,
      "num_tokens": 5069461.0,
      "step": 2930
    },
    {
      "entropy": 6.08068060874939,
      "epoch": 0.22836024119821047,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997446555410857,
      "loss": 5.8595,
      "mean_token_accuracy": 0.13977533280849458,
      "num_tokens": 5077747.0,
      "step": 2935
    },
    {
      "entropy": 6.086756801605224,
      "epoch": 0.22874927056992803,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997433337938716,
      "loss": 5.911,
      "mean_token_accuracy": 0.14513636901974677,
      "num_tokens": 5086300.0,
      "step": 2940
    },
    {
      "entropy": 5.952954530715942,
      "epoch": 0.22913829994164558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997420086365351,
      "loss": 5.6668,
      "mean_token_accuracy": 0.15248717218637467,
      "num_tokens": 5094762.0,
      "step": 2945
    },
    {
      "entropy": 5.985451412200928,
      "epoch": 0.22952732931336317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997406800690966,
      "loss": 5.856,
      "mean_token_accuracy": 0.14468831270933152,
      "num_tokens": 5103364.0,
      "step": 2950
    },
    {
      "entropy": 6.156504106521607,
      "epoch": 0.22991635868508073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997393480915758,
      "loss": 5.9605,
      "mean_token_accuracy": 0.14181874990463256,
      "num_tokens": 5111730.0,
      "step": 2955
    },
    {
      "entropy": 6.056686782836914,
      "epoch": 0.2303053880567983,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997380127039931,
      "loss": 5.7647,
      "mean_token_accuracy": 0.14477684944868088,
      "num_tokens": 5119878.0,
      "step": 2960
    },
    {
      "entropy": 6.014952659606934,
      "epoch": 0.23069441742851585,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997366739063688,
      "loss": 5.8686,
      "mean_token_accuracy": 0.14824746251106263,
      "num_tokens": 5129146.0,
      "step": 2965
    },
    {
      "entropy": 6.005827903747559,
      "epoch": 0.2310834468002334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997353316987231,
      "loss": 5.8798,
      "mean_token_accuracy": 0.14015825390815734,
      "num_tokens": 5137989.0,
      "step": 2970
    },
    {
      "entropy": 6.11515440940857,
      "epoch": 0.231472476171951,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997339860810765,
      "loss": 5.9271,
      "mean_token_accuracy": 0.14150395467877389,
      "num_tokens": 5146722.0,
      "step": 2975
    },
    {
      "entropy": 6.122695827484131,
      "epoch": 0.23186150554366855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997326370534495,
      "loss": 5.9296,
      "mean_token_accuracy": 0.14121474623680114,
      "num_tokens": 5154902.0,
      "step": 2980
    },
    {
      "entropy": 6.067652940750122,
      "epoch": 0.2322505349153861,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997312846158622,
      "loss": 5.9035,
      "mean_token_accuracy": 0.14448403418064118,
      "num_tokens": 5163594.0,
      "step": 2985
    },
    {
      "entropy": 6.061365270614624,
      "epoch": 0.23263956428710367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997299287683355,
      "loss": 5.7941,
      "mean_token_accuracy": 0.14882556200027466,
      "num_tokens": 5172139.0,
      "step": 2990
    },
    {
      "entropy": 6.065181398391724,
      "epoch": 0.23302859365882125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997285695108898,
      "loss": 5.9152,
      "mean_token_accuracy": 0.14413146376609803,
      "num_tokens": 5180077.0,
      "step": 2995
    },
    {
      "entropy": 5.950593709945679,
      "epoch": 0.2334176230305388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997272068435458,
      "loss": 5.7547,
      "mean_token_accuracy": 0.1455926775932312,
      "num_tokens": 5188500.0,
      "step": 3000
    },
    {
      "epoch": 0.2334176230305388,
      "eval_entropy": 5.776663673555082,
      "eval_loss": 5.887854099273682,
      "eval_mean_token_accuracy": 0.151562392559616,
      "eval_num_tokens": 5188500.0,
      "eval_runtime": 22.2562,
      "eval_samples_per_second": 1867.253,
      "eval_steps_per_second": 233.418,
      "step": 3000
    },
    {
      "entropy": 6.09710955619812,
      "epoch": 0.23380665240225637,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499725840766324,
      "loss": 5.8531,
      "mean_token_accuracy": 0.14021188840270044,
      "num_tokens": 5197214.0,
      "step": 3005
    },
    {
      "entropy": 6.07543306350708,
      "epoch": 0.23419568177397393,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997244712792453,
      "loss": 5.9323,
      "mean_token_accuracy": 0.1387889005243778,
      "num_tokens": 5206348.0,
      "step": 3010
    },
    {
      "entropy": 5.954164695739746,
      "epoch": 0.2345847111456915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997230983823305,
      "loss": 5.7521,
      "mean_token_accuracy": 0.14895113110542296,
      "num_tokens": 5215425.0,
      "step": 3015
    },
    {
      "entropy": 6.169281673431397,
      "epoch": 0.23497374051740907,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997217220756003,
      "loss": 5.9655,
      "mean_token_accuracy": 0.13963161259889603,
      "num_tokens": 5224899.0,
      "step": 3020
    },
    {
      "entropy": 6.051689672470093,
      "epoch": 0.23536276988912663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997203423590757,
      "loss": 5.8923,
      "mean_token_accuracy": 0.14200590774416924,
      "num_tokens": 5233829.0,
      "step": 3025
    },
    {
      "entropy": 6.0516040325164795,
      "epoch": 0.2357517992608442,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997189592327775,
      "loss": 6.0265,
      "mean_token_accuracy": 0.1386028565466404,
      "num_tokens": 5243144.0,
      "step": 3030
    },
    {
      "entropy": 6.124688720703125,
      "epoch": 0.23614082863256175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997175726967268,
      "loss": 5.829,
      "mean_token_accuracy": 0.1435486003756523,
      "num_tokens": 5251097.0,
      "step": 3035
    },
    {
      "entropy": 6.076294755935669,
      "epoch": 0.23652985800427934,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997161827509447,
      "loss": 5.9012,
      "mean_token_accuracy": 0.148855247348547,
      "num_tokens": 5260050.0,
      "step": 3040
    },
    {
      "entropy": 6.022133159637451,
      "epoch": 0.2369188873759969,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997147893954521,
      "loss": 5.738,
      "mean_token_accuracy": 0.14691296294331552,
      "num_tokens": 5268363.0,
      "step": 3045
    },
    {
      "entropy": 6.012548446655273,
      "epoch": 0.23730791674771445,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997133926302702,
      "loss": 5.9154,
      "mean_token_accuracy": 0.14275497198104858,
      "num_tokens": 5276821.0,
      "step": 3050
    },
    {
      "entropy": 6.063172245025635,
      "epoch": 0.237696946119432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997119924554204,
      "loss": 5.8067,
      "mean_token_accuracy": 0.14613002091646193,
      "num_tokens": 5285419.0,
      "step": 3055
    },
    {
      "entropy": 5.974448013305664,
      "epoch": 0.23808597549114957,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997105888709236,
      "loss": 5.7263,
      "mean_token_accuracy": 0.1492978572845459,
      "num_tokens": 5294004.0,
      "step": 3060
    },
    {
      "entropy": 5.97568645477295,
      "epoch": 0.23847500486286716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997091818768015,
      "loss": 5.9073,
      "mean_token_accuracy": 0.14691796600818635,
      "num_tokens": 5302699.0,
      "step": 3065
    },
    {
      "entropy": 6.065834379196167,
      "epoch": 0.23886403423458472,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499707771473075,
      "loss": 5.8291,
      "mean_token_accuracy": 0.14639508128166198,
      "num_tokens": 5311115.0,
      "step": 3070
    },
    {
      "entropy": 5.981069898605346,
      "epoch": 0.23925306360630227,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997063576597659,
      "loss": 5.8466,
      "mean_token_accuracy": 0.14814597368240356,
      "num_tokens": 5319795.0,
      "step": 3075
    },
    {
      "entropy": 6.080163955688477,
      "epoch": 0.23964209297801983,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997049404368954,
      "loss": 5.8745,
      "mean_token_accuracy": 0.14955340400338174,
      "num_tokens": 5328273.0,
      "step": 3080
    },
    {
      "entropy": 6.051873445510864,
      "epoch": 0.2400311223497374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499703519804485,
      "loss": 5.8622,
      "mean_token_accuracy": 0.1425059914588928,
      "num_tokens": 5336394.0,
      "step": 3085
    },
    {
      "entropy": 6.006466388702393,
      "epoch": 0.24042015172145498,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997020957625564,
      "loss": 5.8023,
      "mean_token_accuracy": 0.14917524605989457,
      "num_tokens": 5344480.0,
      "step": 3090
    },
    {
      "entropy": 6.016175365447998,
      "epoch": 0.24080918109317254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997006683111312,
      "loss": 5.8859,
      "mean_token_accuracy": 0.13614777624607086,
      "num_tokens": 5353445.0,
      "step": 3095
    },
    {
      "entropy": 6.001996326446533,
      "epoch": 0.2411982104648901,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499699237450231,
      "loss": 5.7481,
      "mean_token_accuracy": 0.1546046867966652,
      "num_tokens": 5362356.0,
      "step": 3100
    },
    {
      "entropy": 5.995383262634277,
      "epoch": 0.24158723983660765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996978031798774,
      "loss": 5.87,
      "mean_token_accuracy": 0.14354093372821808,
      "num_tokens": 5371467.0,
      "step": 3105
    },
    {
      "entropy": 5.926621675491333,
      "epoch": 0.24197626920832524,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996963655000924,
      "loss": 5.7448,
      "mean_token_accuracy": 0.15170810222625733,
      "num_tokens": 5380256.0,
      "step": 3110
    },
    {
      "entropy": 6.0773121356964115,
      "epoch": 0.2423652985800428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996949244108977,
      "loss": 5.7864,
      "mean_token_accuracy": 0.14706581383943557,
      "num_tokens": 5388884.0,
      "step": 3115
    },
    {
      "entropy": 6.019105672836304,
      "epoch": 0.24275432795176036,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499693479912315,
      "loss": 5.7611,
      "mean_token_accuracy": 0.14895851612091066,
      "num_tokens": 5397980.0,
      "step": 3120
    },
    {
      "entropy": 5.963564872741699,
      "epoch": 0.24314335732347792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996920320043666,
      "loss": 5.8691,
      "mean_token_accuracy": 0.14524483159184456,
      "num_tokens": 5406879.0,
      "step": 3125
    },
    {
      "entropy": 6.042530345916748,
      "epoch": 0.24353238669519547,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996905806870742,
      "loss": 5.8697,
      "mean_token_accuracy": 0.14106134623289107,
      "num_tokens": 5416161.0,
      "step": 3130
    },
    {
      "entropy": 5.972628784179688,
      "epoch": 0.24392141606691306,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996891259604598,
      "loss": 5.7354,
      "mean_token_accuracy": 0.15791234523057937,
      "num_tokens": 5424323.0,
      "step": 3135
    },
    {
      "entropy": 5.993604421615601,
      "epoch": 0.24431044543863062,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996876678245455,
      "loss": 5.8616,
      "mean_token_accuracy": 0.14106569737195968,
      "num_tokens": 5433750.0,
      "step": 3140
    },
    {
      "entropy": 6.093292856216431,
      "epoch": 0.24469947481034818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996862062793536,
      "loss": 5.88,
      "mean_token_accuracy": 0.1436523251235485,
      "num_tokens": 5442113.0,
      "step": 3145
    },
    {
      "entropy": 5.972458696365356,
      "epoch": 0.24508850418206574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996847413249061,
      "loss": 5.7793,
      "mean_token_accuracy": 0.14713010862469672,
      "num_tokens": 5450403.0,
      "step": 3150
    },
    {
      "entropy": 6.0378330707550045,
      "epoch": 0.24547753355378332,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996832729612252,
      "loss": 5.8463,
      "mean_token_accuracy": 0.14587439522147178,
      "num_tokens": 5459018.0,
      "step": 3155
    },
    {
      "entropy": 5.983373308181763,
      "epoch": 0.24586656292550088,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996818011883333,
      "loss": 5.8621,
      "mean_token_accuracy": 0.14378748834133148,
      "num_tokens": 5468314.0,
      "step": 3160
    },
    {
      "entropy": 6.051435470581055,
      "epoch": 0.24625559229721844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996803260062527,
      "loss": 5.8575,
      "mean_token_accuracy": 0.14111327528953552,
      "num_tokens": 5476950.0,
      "step": 3165
    },
    {
      "entropy": 6.037627172470093,
      "epoch": 0.246644621668936,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996788474150057,
      "loss": 5.8151,
      "mean_token_accuracy": 0.15004928931593894,
      "num_tokens": 5485201.0,
      "step": 3170
    },
    {
      "entropy": 6.066342830657959,
      "epoch": 0.24703365104065356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996773654146149,
      "loss": 5.9396,
      "mean_token_accuracy": 0.14466533660888672,
      "num_tokens": 5493096.0,
      "step": 3175
    },
    {
      "entropy": 5.986754989624023,
      "epoch": 0.24742268041237114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996758800051026,
      "loss": 5.7187,
      "mean_token_accuracy": 0.14505091235041617,
      "num_tokens": 5502413.0,
      "step": 3180
    },
    {
      "entropy": 6.032514381408691,
      "epoch": 0.2478117097840887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996743911864916,
      "loss": 5.8446,
      "mean_token_accuracy": 0.14924362301826477,
      "num_tokens": 5511670.0,
      "step": 3185
    },
    {
      "entropy": 6.015149879455566,
      "epoch": 0.24820073915580626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996728989588041,
      "loss": 5.7404,
      "mean_token_accuracy": 0.15484994351863862,
      "num_tokens": 5519344.0,
      "step": 3190
    },
    {
      "entropy": 5.893631410598755,
      "epoch": 0.24858976852752382,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996714033220632,
      "loss": 5.7525,
      "mean_token_accuracy": 0.14994811415672302,
      "num_tokens": 5528413.0,
      "step": 3195
    },
    {
      "entropy": 5.972471427917481,
      "epoch": 0.2489787978992414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996699042762911,
      "loss": 5.8435,
      "mean_token_accuracy": 0.14633680060505866,
      "num_tokens": 5537201.0,
      "step": 3200
    },
    {
      "entropy": 6.0863813877105715,
      "epoch": 0.24936782727095896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996684018215111,
      "loss": 5.847,
      "mean_token_accuracy": 0.1399237856268883,
      "num_tokens": 5545975.0,
      "step": 3205
    },
    {
      "entropy": 5.993616819381714,
      "epoch": 0.24975685664267652,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996668959577453,
      "loss": 5.8913,
      "mean_token_accuracy": 0.14371780753135682,
      "num_tokens": 5555237.0,
      "step": 3210
    },
    {
      "entropy": 6.0259552001953125,
      "epoch": 0.2501458860143941,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996653866850173,
      "loss": 5.7647,
      "mean_token_accuracy": 0.15066277533769606,
      "num_tokens": 5564048.0,
      "step": 3215
    },
    {
      "entropy": 5.993947887420655,
      "epoch": 0.25053491538611167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996638740033495,
      "loss": 5.809,
      "mean_token_accuracy": 0.1507560059428215,
      "num_tokens": 5573615.0,
      "step": 3220
    },
    {
      "entropy": 5.956318187713623,
      "epoch": 0.2509239447578292,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996623579127651,
      "loss": 5.6993,
      "mean_token_accuracy": 0.15559622645378113,
      "num_tokens": 5582211.0,
      "step": 3225
    },
    {
      "entropy": 6.015867280960083,
      "epoch": 0.2513129741295468,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996608384132868,
      "loss": 5.942,
      "mean_token_accuracy": 0.13800075203180312,
      "num_tokens": 5592542.0,
      "step": 3230
    },
    {
      "entropy": 6.056093454360962,
      "epoch": 0.25170200350126437,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499659315504938,
      "loss": 5.8988,
      "mean_token_accuracy": 0.15045691132545472,
      "num_tokens": 5601279.0,
      "step": 3235
    },
    {
      "entropy": 6.027131795883179,
      "epoch": 0.2520910328729819,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996577891877417,
      "loss": 5.8467,
      "mean_token_accuracy": 0.14176492244005204,
      "num_tokens": 5610266.0,
      "step": 3240
    },
    {
      "entropy": 6.020873641967773,
      "epoch": 0.2524800622446995,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996562594617209,
      "loss": 5.8752,
      "mean_token_accuracy": 0.14404108226299286,
      "num_tokens": 5618799.0,
      "step": 3245
    },
    {
      "entropy": 6.1522290229797365,
      "epoch": 0.252869091616417,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996547263268991,
      "loss": 5.8347,
      "mean_token_accuracy": 0.14960313588380814,
      "num_tokens": 5627861.0,
      "step": 3250
    },
    {
      "entropy": 6.022855186462403,
      "epoch": 0.2532581209881346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996531897832993,
      "loss": 5.8023,
      "mean_token_accuracy": 0.14954320192337037,
      "num_tokens": 5636229.0,
      "step": 3255
    },
    {
      "entropy": 6.063154315948486,
      "epoch": 0.2536471503598522,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499651649830945,
      "loss": 6.0249,
      "mean_token_accuracy": 0.13233352899551393,
      "num_tokens": 5645828.0,
      "step": 3260
    },
    {
      "entropy": 5.992672395706177,
      "epoch": 0.2540361797315697,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996501064698593,
      "loss": 5.6482,
      "mean_token_accuracy": 0.14677278697490692,
      "num_tokens": 5654750.0,
      "step": 3265
    },
    {
      "entropy": 5.949406385421753,
      "epoch": 0.2544252091032873,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499648559700066,
      "loss": 5.811,
      "mean_token_accuracy": 0.1472081035375595,
      "num_tokens": 5664017.0,
      "step": 3270
    },
    {
      "entropy": 6.023008060455322,
      "epoch": 0.25481423847500484,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996470095215884,
      "loss": 5.829,
      "mean_token_accuracy": 0.14243277683854103,
      "num_tokens": 5672220.0,
      "step": 3275
    },
    {
      "entropy": 5.904695510864258,
      "epoch": 0.2552032678467224,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996454559344498,
      "loss": 5.7538,
      "mean_token_accuracy": 0.14675017893314363,
      "num_tokens": 5680921.0,
      "step": 3280
    },
    {
      "entropy": 6.027059555053711,
      "epoch": 0.25559229721844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996438989386741,
      "loss": 5.7605,
      "mean_token_accuracy": 0.14695732966065406,
      "num_tokens": 5689226.0,
      "step": 3285
    },
    {
      "entropy": 6.055839252471924,
      "epoch": 0.25598132659015754,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996423385342847,
      "loss": 5.9056,
      "mean_token_accuracy": 0.14036790430545806,
      "num_tokens": 5698413.0,
      "step": 3290
    },
    {
      "entropy": 5.909748077392578,
      "epoch": 0.25637035596187513,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996407747213055,
      "loss": 5.7933,
      "mean_token_accuracy": 0.13998080119490625,
      "num_tokens": 5707724.0,
      "step": 3295
    },
    {
      "entropy": 6.027735948562622,
      "epoch": 0.25675938533359266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049963920749976,
      "loss": 5.8064,
      "mean_token_accuracy": 0.14243267476558685,
      "num_tokens": 5716566.0,
      "step": 3300
    },
    {
      "entropy": 5.964565181732178,
      "epoch": 0.25714841470531025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996376368696721,
      "loss": 5.7429,
      "mean_token_accuracy": 0.1470643550157547,
      "num_tokens": 5725463.0,
      "step": 3305
    },
    {
      "entropy": 5.96244068145752,
      "epoch": 0.25753744407702783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996360628310656,
      "loss": 5.7392,
      "mean_token_accuracy": 0.14964852184057237,
      "num_tokens": 5734143.0,
      "step": 3310
    },
    {
      "entropy": 6.0237079620361325,
      "epoch": 0.25792647344874536,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996344853839644,
      "loss": 5.8227,
      "mean_token_accuracy": 0.152262943983078,
      "num_tokens": 5742981.0,
      "step": 3315
    },
    {
      "entropy": 5.905067205429077,
      "epoch": 0.25831550282046295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996329045283924,
      "loss": 5.6996,
      "mean_token_accuracy": 0.157255619764328,
      "num_tokens": 5751294.0,
      "step": 3320
    },
    {
      "entropy": 5.940437126159668,
      "epoch": 0.25870453219218054,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996313202643737,
      "loss": 5.7977,
      "mean_token_accuracy": 0.14972240775823592,
      "num_tokens": 5760958.0,
      "step": 3325
    },
    {
      "entropy": 6.0460288524627686,
      "epoch": 0.25909356156389807,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996297325919321,
      "loss": 5.8221,
      "mean_token_accuracy": 0.15026443153619767,
      "num_tokens": 5769077.0,
      "step": 3330
    },
    {
      "entropy": 6.073997592926025,
      "epoch": 0.25948259093561565,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996281415110919,
      "loss": 5.8625,
      "mean_token_accuracy": 0.14942350834608079,
      "num_tokens": 5778093.0,
      "step": 3335
    },
    {
      "entropy": 5.951122236251831,
      "epoch": 0.2598716203073332,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996265470218773,
      "loss": 5.8174,
      "mean_token_accuracy": 0.14803160578012467,
      "num_tokens": 5786996.0,
      "step": 3340
    },
    {
      "entropy": 6.045619010925293,
      "epoch": 0.26026064967905077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996249491243122,
      "loss": 5.8187,
      "mean_token_accuracy": 0.14531699568033218,
      "num_tokens": 5794831.0,
      "step": 3345
    },
    {
      "entropy": 5.879084062576294,
      "epoch": 0.26064967905076836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996233478184211,
      "loss": 5.6585,
      "mean_token_accuracy": 0.1536806970834732,
      "num_tokens": 5802286.0,
      "step": 3350
    },
    {
      "entropy": 5.884199380874634,
      "epoch": 0.2610387084224859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996217431042282,
      "loss": 5.7945,
      "mean_token_accuracy": 0.14438097476959227,
      "num_tokens": 5810985.0,
      "step": 3355
    },
    {
      "entropy": 6.070495462417602,
      "epoch": 0.2614277377942035,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499620134981758,
      "loss": 5.8488,
      "mean_token_accuracy": 0.14648928120732307,
      "num_tokens": 5820037.0,
      "step": 3360
    },
    {
      "entropy": 5.866944599151611,
      "epoch": 0.261816767165921,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996185234510346,
      "loss": 5.7079,
      "mean_token_accuracy": 0.15443466156721114,
      "num_tokens": 5828232.0,
      "step": 3365
    },
    {
      "entropy": 6.007300472259521,
      "epoch": 0.2622057965376386,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996169085120828,
      "loss": 5.9279,
      "mean_token_accuracy": 0.14708923697471618,
      "num_tokens": 5838184.0,
      "step": 3370
    },
    {
      "entropy": 6.050236082077026,
      "epoch": 0.2625948259093562,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996152901649268,
      "loss": 5.7388,
      "mean_token_accuracy": 0.15235017836093903,
      "num_tokens": 5846705.0,
      "step": 3375
    },
    {
      "entropy": 5.947798109054565,
      "epoch": 0.2629838552810737,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996136684095913,
      "loss": 5.839,
      "mean_token_accuracy": 0.14474607408046722,
      "num_tokens": 5855959.0,
      "step": 3380
    },
    {
      "entropy": 5.985251426696777,
      "epoch": 0.2633728846527913,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996120432461009,
      "loss": 5.7722,
      "mean_token_accuracy": 0.1477278485894203,
      "num_tokens": 5865283.0,
      "step": 3385
    },
    {
      "entropy": 6.012517404556275,
      "epoch": 0.2637619140245088,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996104146744804,
      "loss": 5.8277,
      "mean_token_accuracy": 0.15219881534576415,
      "num_tokens": 5873833.0,
      "step": 3390
    },
    {
      "entropy": 6.023937559127807,
      "epoch": 0.2641509433962264,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996087826947541,
      "loss": 5.7365,
      "mean_token_accuracy": 0.1534780517220497,
      "num_tokens": 5882027.0,
      "step": 3395
    },
    {
      "entropy": 5.794362211227417,
      "epoch": 0.264539972767944,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996071473069471,
      "loss": 5.6159,
      "mean_token_accuracy": 0.15834907814860344,
      "num_tokens": 5890650.0,
      "step": 3400
    },
    {
      "entropy": 5.932468557357788,
      "epoch": 0.26492900213966153,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996055085110842,
      "loss": 5.6852,
      "mean_token_accuracy": 0.15304084569215776,
      "num_tokens": 5898595.0,
      "step": 3405
    },
    {
      "entropy": 6.016835165023804,
      "epoch": 0.2653180315113791,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996038663071902,
      "loss": 5.7838,
      "mean_token_accuracy": 0.14666232615709304,
      "num_tokens": 5907234.0,
      "step": 3410
    },
    {
      "entropy": 5.967805242538452,
      "epoch": 0.26570706088309665,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996022206952901,
      "loss": 5.7282,
      "mean_token_accuracy": 0.15326813012361526,
      "num_tokens": 5915690.0,
      "step": 3415
    },
    {
      "entropy": 5.928490114212036,
      "epoch": 0.26609609025481423,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996005716754086,
      "loss": 5.687,
      "mean_token_accuracy": 0.1530188575387001,
      "num_tokens": 5924481.0,
      "step": 3420
    },
    {
      "entropy": 5.895607328414917,
      "epoch": 0.2664851196265318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499598919247571,
      "loss": 5.7583,
      "mean_token_accuracy": 0.15381635576486588,
      "num_tokens": 5933059.0,
      "step": 3425
    },
    {
      "entropy": 6.008783197402954,
      "epoch": 0.26687414899824935,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995972634118023,
      "loss": 5.8795,
      "mean_token_accuracy": 0.14230265319347382,
      "num_tokens": 5941767.0,
      "step": 3430
    },
    {
      "entropy": 5.961049509048462,
      "epoch": 0.26726317836996694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995956041681277,
      "loss": 5.7632,
      "mean_token_accuracy": 0.15119194015860557,
      "num_tokens": 5950202.0,
      "step": 3435
    },
    {
      "entropy": 5.93410906791687,
      "epoch": 0.2676522077416845,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995939415165722,
      "loss": 5.7973,
      "mean_token_accuracy": 0.1453957512974739,
      "num_tokens": 5959242.0,
      "step": 3440
    },
    {
      "entropy": 5.938480091094971,
      "epoch": 0.26804123711340205,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995922754571611,
      "loss": 5.7665,
      "mean_token_accuracy": 0.15338513106107712,
      "num_tokens": 5968404.0,
      "step": 3445
    },
    {
      "entropy": 5.973701143264771,
      "epoch": 0.26843026648511964,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995906059899197,
      "loss": 5.8834,
      "mean_token_accuracy": 0.14646051377058028,
      "num_tokens": 5976711.0,
      "step": 3450
    },
    {
      "entropy": 5.987052536010742,
      "epoch": 0.26881929585683717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995889331148733,
      "loss": 5.7223,
      "mean_token_accuracy": 0.14493673294782639,
      "num_tokens": 5985049.0,
      "step": 3455
    },
    {
      "entropy": 5.914662218093872,
      "epoch": 0.26920832522855476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995872568320474,
      "loss": 5.7435,
      "mean_token_accuracy": 0.1537318557500839,
      "num_tokens": 5993904.0,
      "step": 3460
    },
    {
      "entropy": 5.991185903549194,
      "epoch": 0.26959735460027234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995855771414673,
      "loss": 5.8126,
      "mean_token_accuracy": 0.15051607936620712,
      "num_tokens": 6001832.0,
      "step": 3465
    },
    {
      "entropy": 5.910642814636231,
      "epoch": 0.2699863839719899,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995838940431585,
      "loss": 5.726,
      "mean_token_accuracy": 0.15267925411462785,
      "num_tokens": 6010609.0,
      "step": 3470
    },
    {
      "entropy": 5.965318393707276,
      "epoch": 0.27037541334370746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995822075371466,
      "loss": 5.793,
      "mean_token_accuracy": 0.14585788249969484,
      "num_tokens": 6019299.0,
      "step": 3475
    },
    {
      "entropy": 6.009939765930175,
      "epoch": 0.270764442715425,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995805176234571,
      "loss": 5.8277,
      "mean_token_accuracy": 0.14817733988165854,
      "num_tokens": 6028743.0,
      "step": 3480
    },
    {
      "entropy": 5.977865552902221,
      "epoch": 0.2711534720871426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995788243021158,
      "loss": 5.7351,
      "mean_token_accuracy": 0.14834726303815843,
      "num_tokens": 6038064.0,
      "step": 3485
    },
    {
      "entropy": 5.9414863109588625,
      "epoch": 0.27154250145886016,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995771275731483,
      "loss": 5.7411,
      "mean_token_accuracy": 0.14814434796571732,
      "num_tokens": 6047031.0,
      "step": 3490
    },
    {
      "entropy": 5.970782375335693,
      "epoch": 0.2719315308305777,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995754274365802,
      "loss": 5.8131,
      "mean_token_accuracy": 0.1453293301165104,
      "num_tokens": 6056147.0,
      "step": 3495
    },
    {
      "entropy": 5.9767937660217285,
      "epoch": 0.2723205602022953,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995737238924376,
      "loss": 5.7598,
      "mean_token_accuracy": 0.14954252243041993,
      "num_tokens": 6064641.0,
      "step": 3500
    },
    {
      "entropy": 5.881254386901856,
      "epoch": 0.2727095895740128,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995720169407461,
      "loss": 5.6509,
      "mean_token_accuracy": 0.14959076344966887,
      "num_tokens": 6073664.0,
      "step": 3505
    },
    {
      "entropy": 5.89289174079895,
      "epoch": 0.2730986189457304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995703065815317,
      "loss": 5.6485,
      "mean_token_accuracy": 0.15396696478128433,
      "num_tokens": 6081480.0,
      "step": 3510
    },
    {
      "entropy": 5.950675344467163,
      "epoch": 0.273487648317448,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995685928148203,
      "loss": 5.7711,
      "mean_token_accuracy": 0.1454616293311119,
      "num_tokens": 6090408.0,
      "step": 3515
    },
    {
      "entropy": 5.9502583026885985,
      "epoch": 0.2738766776891655,
      "grad_norm": 1.0,
      "learning_rate": 0.000499566875640638,
      "loss": 5.7766,
      "mean_token_accuracy": 0.14692020863294603,
      "num_tokens": 6099076.0,
      "step": 3520
    },
    {
      "entropy": 5.957012367248535,
      "epoch": 0.2742657070608831,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995651550590108,
      "loss": 5.807,
      "mean_token_accuracy": 0.14378090351819992,
      "num_tokens": 6108346.0,
      "step": 3525
    },
    {
      "entropy": 5.939613485336304,
      "epoch": 0.2746547364326007,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995634310699647,
      "loss": 5.7764,
      "mean_token_accuracy": 0.15250186771154403,
      "num_tokens": 6116974.0,
      "step": 3530
    },
    {
      "entropy": 5.999927234649658,
      "epoch": 0.2750437658043182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995617036735261,
      "loss": 5.7977,
      "mean_token_accuracy": 0.145340259373188,
      "num_tokens": 6125510.0,
      "step": 3535
    },
    {
      "entropy": 5.984943962097168,
      "epoch": 0.2754327951760358,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995599728697211,
      "loss": 5.8316,
      "mean_token_accuracy": 0.14638109505176544,
      "num_tokens": 6134959.0,
      "step": 3540
    },
    {
      "entropy": 6.03920316696167,
      "epoch": 0.27582182454775334,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995582386585759,
      "loss": 5.7725,
      "mean_token_accuracy": 0.14525687023997308,
      "num_tokens": 6144485.0,
      "step": 3545
    },
    {
      "entropy": 5.906405782699585,
      "epoch": 0.2762108539194709,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995565010401167,
      "loss": 5.7138,
      "mean_token_accuracy": 0.15489541739225388,
      "num_tokens": 6153991.0,
      "step": 3550
    },
    {
      "entropy": 5.823055219650269,
      "epoch": 0.2765998832911885,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995547600143702,
      "loss": 5.7147,
      "mean_token_accuracy": 0.14814141243696213,
      "num_tokens": 6162649.0,
      "step": 3555
    },
    {
      "entropy": 5.934897470474243,
      "epoch": 0.27698891266290604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995530155813625,
      "loss": 5.6178,
      "mean_token_accuracy": 0.15582334697246553,
      "num_tokens": 6171525.0,
      "step": 3560
    },
    {
      "entropy": 5.958289670944214,
      "epoch": 0.2773779420346236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995512677411203,
      "loss": 5.8577,
      "mean_token_accuracy": 0.14743831306695937,
      "num_tokens": 6180403.0,
      "step": 3565
    },
    {
      "entropy": 5.9663067817687985,
      "epoch": 0.27776697140634116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995495164936698,
      "loss": 5.6945,
      "mean_token_accuracy": 0.1548657938838005,
      "num_tokens": 6188361.0,
      "step": 3570
    },
    {
      "entropy": 5.893314409255981,
      "epoch": 0.27815600077805874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995477618390381,
      "loss": 5.737,
      "mean_token_accuracy": 0.14627340212464332,
      "num_tokens": 6197418.0,
      "step": 3575
    },
    {
      "entropy": 5.873566055297852,
      "epoch": 0.27854503014977633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995460037772513,
      "loss": 5.7557,
      "mean_token_accuracy": 0.15220080316066742,
      "num_tokens": 6206040.0,
      "step": 3580
    },
    {
      "entropy": 5.903412389755249,
      "epoch": 0.27893405952149386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995442423083365,
      "loss": 5.6771,
      "mean_token_accuracy": 0.1455555059015751,
      "num_tokens": 6214433.0,
      "step": 3585
    },
    {
      "entropy": 5.872088384628296,
      "epoch": 0.27932308889321145,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995424774323201,
      "loss": 5.6568,
      "mean_token_accuracy": 0.15279483944177627,
      "num_tokens": 6223375.0,
      "step": 3590
    },
    {
      "entropy": 5.886718320846557,
      "epoch": 0.279712118264929,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499540709149229,
      "loss": 5.7661,
      "mean_token_accuracy": 0.1504191666841507,
      "num_tokens": 6232597.0,
      "step": 3595
    },
    {
      "entropy": 5.890195322036743,
      "epoch": 0.28010114763664656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995389374590901,
      "loss": 5.6118,
      "mean_token_accuracy": 0.1608039543032646,
      "num_tokens": 6240634.0,
      "step": 3600
    },
    {
      "entropy": 5.85315260887146,
      "epoch": 0.28049017700836415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995371623619301,
      "loss": 5.5854,
      "mean_token_accuracy": 0.1523109719157219,
      "num_tokens": 6249098.0,
      "step": 3605
    },
    {
      "entropy": 5.825993442535401,
      "epoch": 0.2808792063800817,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499535383857776,
      "loss": 5.6156,
      "mean_token_accuracy": 0.15686782896518708,
      "num_tokens": 6257572.0,
      "step": 3610
    },
    {
      "entropy": 5.965735816955567,
      "epoch": 0.28126823575179927,
      "grad_norm": 1.0,
      "learning_rate": 0.000499533601946655,
      "loss": 5.8788,
      "mean_token_accuracy": 0.1445400133728981,
      "num_tokens": 6266492.0,
      "step": 3615
    },
    {
      "entropy": 5.960498523712158,
      "epoch": 0.2816572651235168,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995318166285938,
      "loss": 5.7507,
      "mean_token_accuracy": 0.14909357577562332,
      "num_tokens": 6275149.0,
      "step": 3620
    },
    {
      "entropy": 5.978596830368042,
      "epoch": 0.2820462944952344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995300279036199,
      "loss": 5.7612,
      "mean_token_accuracy": 0.14895476996898652,
      "num_tokens": 6283208.0,
      "step": 3625
    },
    {
      "entropy": 5.953970336914063,
      "epoch": 0.28243532386695197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049952823577176,
      "loss": 5.7027,
      "mean_token_accuracy": 0.15072022378444672,
      "num_tokens": 6292084.0,
      "step": 3630
    },
    {
      "entropy": 5.91438364982605,
      "epoch": 0.2828243532386695,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995264402330416,
      "loss": 5.7005,
      "mean_token_accuracy": 0.14980678856372834,
      "num_tokens": 6301855.0,
      "step": 3635
    },
    {
      "entropy": 5.934312725067139,
      "epoch": 0.2832133826103871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995246412874919,
      "loss": 5.7179,
      "mean_token_accuracy": 0.1545361801981926,
      "num_tokens": 6310166.0,
      "step": 3640
    },
    {
      "entropy": 5.900256109237671,
      "epoch": 0.2836024119821047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499522838935138,
      "loss": 5.7009,
      "mean_token_accuracy": 0.15615580528974532,
      "num_tokens": 6318928.0,
      "step": 3645
    },
    {
      "entropy": 5.930281686782837,
      "epoch": 0.2839914413538222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995210331760074,
      "loss": 5.7203,
      "mean_token_accuracy": 0.16006863862276077,
      "num_tokens": 6327790.0,
      "step": 3650
    },
    {
      "entropy": 5.968905830383301,
      "epoch": 0.2843804707255398,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995192240101275,
      "loss": 5.7851,
      "mean_token_accuracy": 0.15172517597675322,
      "num_tokens": 6335517.0,
      "step": 3655
    },
    {
      "entropy": 5.913986444473267,
      "epoch": 0.2847695000972573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995174114375258,
      "loss": 5.7823,
      "mean_token_accuracy": 0.15226266980171205,
      "num_tokens": 6344218.0,
      "step": 3660
    },
    {
      "entropy": 5.884621429443359,
      "epoch": 0.2851585294689749,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995155954582297,
      "loss": 5.5784,
      "mean_token_accuracy": 0.1623852953314781,
      "num_tokens": 6353125.0,
      "step": 3665
    },
    {
      "entropy": 5.889808320999146,
      "epoch": 0.2855475588406925,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995137760722668,
      "loss": 5.7274,
      "mean_token_accuracy": 0.15155290365219115,
      "num_tokens": 6362113.0,
      "step": 3670
    },
    {
      "entropy": 5.950609111785889,
      "epoch": 0.28593658821241,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995119532796646,
      "loss": 5.8001,
      "mean_token_accuracy": 0.15011403560638428,
      "num_tokens": 6370916.0,
      "step": 3675
    },
    {
      "entropy": 5.9118701934814455,
      "epoch": 0.2863256175841276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499510127080451,
      "loss": 5.6573,
      "mean_token_accuracy": 0.15422363579273224,
      "num_tokens": 6379256.0,
      "step": 3680
    },
    {
      "entropy": 5.909752559661865,
      "epoch": 0.28671464695584514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995082974746535,
      "loss": 5.7761,
      "mean_token_accuracy": 0.14952105581760405,
      "num_tokens": 6388098.0,
      "step": 3685
    },
    {
      "entropy": 5.841971826553345,
      "epoch": 0.28710367632756273,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995064644622999,
      "loss": 5.591,
      "mean_token_accuracy": 0.15846318304538726,
      "num_tokens": 6396292.0,
      "step": 3690
    },
    {
      "entropy": 5.888729095458984,
      "epoch": 0.2874927056992803,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995046280434181,
      "loss": 5.6134,
      "mean_token_accuracy": 0.16034238040447235,
      "num_tokens": 6404968.0,
      "step": 3695
    },
    {
      "entropy": 5.76981749534607,
      "epoch": 0.28788173507099785,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995027882180358,
      "loss": 5.7078,
      "mean_token_accuracy": 0.15313632637262345,
      "num_tokens": 6412839.0,
      "step": 3700
    },
    {
      "entropy": 5.88944640159607,
      "epoch": 0.28827076444271543,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995009449861812,
      "loss": 5.7299,
      "mean_token_accuracy": 0.14679443314671517,
      "num_tokens": 6421943.0,
      "step": 3705
    },
    {
      "entropy": 5.965779542922974,
      "epoch": 0.28865979381443296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499499098347882,
      "loss": 5.7994,
      "mean_token_accuracy": 0.14540371894836426,
      "num_tokens": 6430731.0,
      "step": 3710
    },
    {
      "entropy": 5.914122581481934,
      "epoch": 0.28904882318615055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994972483031662,
      "loss": 5.732,
      "mean_token_accuracy": 0.15313124060630798,
      "num_tokens": 6439365.0,
      "step": 3715
    },
    {
      "entropy": 5.8826902389526365,
      "epoch": 0.28943785255786814,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499495394852062,
      "loss": 5.6769,
      "mean_token_accuracy": 0.15879104733467103,
      "num_tokens": 6448085.0,
      "step": 3720
    },
    {
      "entropy": 5.810940217971802,
      "epoch": 0.28982688192958567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994935379945977,
      "loss": 5.7033,
      "mean_token_accuracy": 0.14827416092157364,
      "num_tokens": 6457630.0,
      "step": 3725
    },
    {
      "entropy": 5.945941162109375,
      "epoch": 0.29021591130130325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994916777308012,
      "loss": 5.6031,
      "mean_token_accuracy": 0.16021039783954621,
      "num_tokens": 6465781.0,
      "step": 3730
    },
    {
      "entropy": 5.911113929748535,
      "epoch": 0.29060494067302084,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994898140607007,
      "loss": 5.7692,
      "mean_token_accuracy": 0.15205744206905364,
      "num_tokens": 6473723.0,
      "step": 3735
    },
    {
      "entropy": 5.847357845306396,
      "epoch": 0.29099397004473837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994879469843247,
      "loss": 5.7222,
      "mean_token_accuracy": 0.15420670360326766,
      "num_tokens": 6482535.0,
      "step": 3740
    },
    {
      "entropy": 5.977249622344971,
      "epoch": 0.29138299941645596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994860765017013,
      "loss": 5.7479,
      "mean_token_accuracy": 0.159115831553936,
      "num_tokens": 6491297.0,
      "step": 3745
    },
    {
      "entropy": 5.8742663860321045,
      "epoch": 0.2917720287881735,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499484202612859,
      "loss": 5.7099,
      "mean_token_accuracy": 0.15262365341186523,
      "num_tokens": 6499809.0,
      "step": 3750
    },
    {
      "entropy": 5.892661190032959,
      "epoch": 0.2921610581598911,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994823253178264,
      "loss": 5.6757,
      "mean_token_accuracy": 0.15837236493825912,
      "num_tokens": 6507965.0,
      "step": 3755
    },
    {
      "entropy": 5.909889507293701,
      "epoch": 0.29255008753160866,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004994804446166317,
      "loss": 5.727,
      "mean_token_accuracy": 0.15141136050224305,
      "num_tokens": 6517106.0,
      "step": 3760
    },
    {
      "entropy": 5.9486315727233885,
      "epoch": 0.2929391169033262,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994785605093035,
      "loss": 5.792,
      "mean_token_accuracy": 0.1503278449177742,
      "num_tokens": 6526663.0,
      "step": 3765
    },
    {
      "entropy": 6.0040600299835205,
      "epoch": 0.2933281462750438,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994766729958705,
      "loss": 5.8254,
      "mean_token_accuracy": 0.14695937559008598,
      "num_tokens": 6535732.0,
      "step": 3770
    },
    {
      "entropy": 5.908503007888794,
      "epoch": 0.2937171756467613,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994747820763612,
      "loss": 5.7003,
      "mean_token_accuracy": 0.15465044006705284,
      "num_tokens": 6544938.0,
      "step": 3775
    },
    {
      "entropy": 5.900608539581299,
      "epoch": 0.2941062050184789,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994728877508046,
      "loss": 5.7173,
      "mean_token_accuracy": 0.14728575497865676,
      "num_tokens": 6554507.0,
      "step": 3780
    },
    {
      "entropy": 5.948572111129761,
      "epoch": 0.2944952343901965,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994709900192289,
      "loss": 5.7456,
      "mean_token_accuracy": 0.1493665762245655,
      "num_tokens": 6563285.0,
      "step": 3785
    },
    {
      "entropy": 5.848457145690918,
      "epoch": 0.294884263761914,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994690888816635,
      "loss": 5.6481,
      "mean_token_accuracy": 0.15300893634557725,
      "num_tokens": 6572223.0,
      "step": 3790
    },
    {
      "entropy": 5.931174421310425,
      "epoch": 0.2952732931336316,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994671843381367,
      "loss": 5.8046,
      "mean_token_accuracy": 0.14685370475053788,
      "num_tokens": 6580686.0,
      "step": 3795
    },
    {
      "entropy": 5.916421270370483,
      "epoch": 0.29566232250534913,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994652763886778,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1544279932975769,
      "num_tokens": 6589056.0,
      "step": 3800
    },
    {
      "entropy": 5.923410892486572,
      "epoch": 0.2960513518770667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994633650333155,
      "loss": 5.7378,
      "mean_token_accuracy": 0.1521855503320694,
      "num_tokens": 6598221.0,
      "step": 3805
    },
    {
      "entropy": 6.000279235839844,
      "epoch": 0.2964403812487843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994614502720792,
      "loss": 5.7899,
      "mean_token_accuracy": 0.15420753359794617,
      "num_tokens": 6606657.0,
      "step": 3810
    },
    {
      "entropy": 5.962191486358643,
      "epoch": 0.29682941062050183,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994595321049974,
      "loss": 5.8763,
      "mean_token_accuracy": 0.14296047687530516,
      "num_tokens": 6616203.0,
      "step": 3815
    },
    {
      "entropy": 6.01355504989624,
      "epoch": 0.2972184399922194,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994576105320994,
      "loss": 5.7268,
      "mean_token_accuracy": 0.1526578664779663,
      "num_tokens": 6624982.0,
      "step": 3820
    },
    {
      "entropy": 5.964144134521485,
      "epoch": 0.29760746936393695,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994556855534146,
      "loss": 5.6945,
      "mean_token_accuracy": 0.15711415261030198,
      "num_tokens": 6632548.0,
      "step": 3825
    },
    {
      "entropy": 5.877254629135132,
      "epoch": 0.29799649873565454,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994537571689719,
      "loss": 5.8253,
      "mean_token_accuracy": 0.14743569269776344,
      "num_tokens": 6641943.0,
      "step": 3830
    },
    {
      "entropy": 5.979266500473022,
      "epoch": 0.2983855281073721,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994518253788007,
      "loss": 5.674,
      "mean_token_accuracy": 0.15680679976940154,
      "num_tokens": 6650371.0,
      "step": 3835
    },
    {
      "entropy": 5.897865438461304,
      "epoch": 0.29877455747908965,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994498901829304,
      "loss": 5.7541,
      "mean_token_accuracy": 0.1545755073428154,
      "num_tokens": 6659324.0,
      "step": 3840
    },
    {
      "entropy": 5.886949491500855,
      "epoch": 0.29916358685080724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994479515813903,
      "loss": 5.7322,
      "mean_token_accuracy": 0.14822690486907958,
      "num_tokens": 6668237.0,
      "step": 3845
    },
    {
      "entropy": 5.942874813079834,
      "epoch": 0.2995526162225248,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994460095742096,
      "loss": 5.6296,
      "mean_token_accuracy": 0.1591558113694191,
      "num_tokens": 6677213.0,
      "step": 3850
    },
    {
      "entropy": 5.882146739959717,
      "epoch": 0.29994164559424236,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499444064161418,
      "loss": 5.6613,
      "mean_token_accuracy": 0.1546331986784935,
      "num_tokens": 6686850.0,
      "step": 3855
    },
    {
      "entropy": 5.899561023712158,
      "epoch": 0.30033067496595994,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499442115343045,
      "loss": 5.6233,
      "mean_token_accuracy": 0.16067153513431548,
      "num_tokens": 6695697.0,
      "step": 3860
    },
    {
      "entropy": 5.839262390136719,
      "epoch": 0.3007197043376775,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994401631191202,
      "loss": 5.69,
      "mean_token_accuracy": 0.15865806192159654,
      "num_tokens": 6704525.0,
      "step": 3865
    },
    {
      "entropy": 5.830469989776612,
      "epoch": 0.30110873370939506,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994382074896731,
      "loss": 5.6183,
      "mean_token_accuracy": 0.1629143700003624,
      "num_tokens": 6713562.0,
      "step": 3870
    },
    {
      "entropy": 5.961298418045044,
      "epoch": 0.30149776308111265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994362484547335,
      "loss": 5.7323,
      "mean_token_accuracy": 0.1521342873573303,
      "num_tokens": 6721817.0,
      "step": 3875
    },
    {
      "entropy": 5.929515409469604,
      "epoch": 0.3018867924528302,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994342860143309,
      "loss": 5.649,
      "mean_token_accuracy": 0.16174968481063842,
      "num_tokens": 6729757.0,
      "step": 3880
    },
    {
      "entropy": 5.865143585205078,
      "epoch": 0.30227582182454776,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994323201684953,
      "loss": 5.7268,
      "mean_token_accuracy": 0.1438479945063591,
      "num_tokens": 6738354.0,
      "step": 3885
    },
    {
      "entropy": 5.8890440464019775,
      "epoch": 0.3026648511962653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994303509172566,
      "loss": 5.7818,
      "mean_token_accuracy": 0.14600699692964553,
      "num_tokens": 6746505.0,
      "step": 3890
    },
    {
      "entropy": 5.968038892745971,
      "epoch": 0.3030538805679829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994283782606444,
      "loss": 5.6697,
      "mean_token_accuracy": 0.1525379940867424,
      "num_tokens": 6754547.0,
      "step": 3895
    },
    {
      "entropy": 5.833213520050049,
      "epoch": 0.30344290993970047,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994264021986888,
      "loss": 5.5961,
      "mean_token_accuracy": 0.1570156365633011,
      "num_tokens": 6763213.0,
      "step": 3900
    },
    {
      "entropy": 5.818766260147095,
      "epoch": 0.303831939311418,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994244227314197,
      "loss": 5.6512,
      "mean_token_accuracy": 0.15624647140502929,
      "num_tokens": 6771060.0,
      "step": 3905
    },
    {
      "entropy": 5.88864574432373,
      "epoch": 0.3042209686831356,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994224398588672,
      "loss": 5.7844,
      "mean_token_accuracy": 0.15062096118927001,
      "num_tokens": 6779819.0,
      "step": 3910
    },
    {
      "entropy": 5.910723733901977,
      "epoch": 0.3046099980548531,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994204535810615,
      "loss": 5.6995,
      "mean_token_accuracy": 0.1460881784558296,
      "num_tokens": 6788660.0,
      "step": 3915
    },
    {
      "entropy": 5.838244962692261,
      "epoch": 0.3049990274265707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994184638980326,
      "loss": 5.6322,
      "mean_token_accuracy": 0.15820442885160446,
      "num_tokens": 6797298.0,
      "step": 3920
    },
    {
      "entropy": 5.79624228477478,
      "epoch": 0.3053880567982883,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994164708098107,
      "loss": 5.6849,
      "mean_token_accuracy": 0.15061923563480378,
      "num_tokens": 6806148.0,
      "step": 3925
    },
    {
      "entropy": 5.885928583145142,
      "epoch": 0.3057770861700058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499414474316426,
      "loss": 5.7139,
      "mean_token_accuracy": 0.15625312477350234,
      "num_tokens": 6815494.0,
      "step": 3930
    },
    {
      "entropy": 5.979975891113281,
      "epoch": 0.3061661155417234,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994124744179088,
      "loss": 5.7769,
      "mean_token_accuracy": 0.14495111033320426,
      "num_tokens": 6824852.0,
      "step": 3935
    },
    {
      "entropy": 5.868782424926758,
      "epoch": 0.306555144913441,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994104711142896,
      "loss": 5.5771,
      "mean_token_accuracy": 0.1625044733285904,
      "num_tokens": 6834200.0,
      "step": 3940
    },
    {
      "entropy": 5.885298538208008,
      "epoch": 0.3069441742851585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994084644055986,
      "loss": 5.7219,
      "mean_token_accuracy": 0.15300173163414002,
      "num_tokens": 6842869.0,
      "step": 3945
    },
    {
      "entropy": 5.894306182861328,
      "epoch": 0.3073332036568761,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994064542918664,
      "loss": 5.6744,
      "mean_token_accuracy": 0.15615324527025223,
      "num_tokens": 6851668.0,
      "step": 3950
    },
    {
      "entropy": 5.926864671707153,
      "epoch": 0.30772223302859364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994044407731235,
      "loss": 5.7662,
      "mean_token_accuracy": 0.15563177019357682,
      "num_tokens": 6860710.0,
      "step": 3955
    },
    {
      "entropy": 5.8650439262390135,
      "epoch": 0.3081112624003112,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994024238494002,
      "loss": 5.6887,
      "mean_token_accuracy": 0.15064796358346938,
      "num_tokens": 6869608.0,
      "step": 3960
    },
    {
      "entropy": 5.949940347671509,
      "epoch": 0.3085002917720288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994004035207274,
      "loss": 5.7098,
      "mean_token_accuracy": 0.15246402025222777,
      "num_tokens": 6878108.0,
      "step": 3965
    },
    {
      "entropy": 5.8566819190979,
      "epoch": 0.30888932114374634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993983797871356,
      "loss": 5.6156,
      "mean_token_accuracy": 0.15605719834566117,
      "num_tokens": 6886369.0,
      "step": 3970
    },
    {
      "entropy": 5.817896604537964,
      "epoch": 0.30927835051546393,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993963526486555,
      "loss": 5.6814,
      "mean_token_accuracy": 0.15414216369390488,
      "num_tokens": 6895410.0,
      "step": 3975
    },
    {
      "entropy": 5.944117212295533,
      "epoch": 0.30966737988718146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499394322105318,
      "loss": 5.6537,
      "mean_token_accuracy": 0.15795325487852097,
      "num_tokens": 6904380.0,
      "step": 3980
    },
    {
      "entropy": 5.9510369300842285,
      "epoch": 0.31005640925889905,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993922881571538,
      "loss": 5.6804,
      "mean_token_accuracy": 0.15391244292259215,
      "num_tokens": 6913012.0,
      "step": 3985
    },
    {
      "entropy": 5.83461446762085,
      "epoch": 0.31044543863061663,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993902508041939,
      "loss": 5.7267,
      "mean_token_accuracy": 0.1524622529745102,
      "num_tokens": 6921696.0,
      "step": 3990
    },
    {
      "entropy": 5.881659746170044,
      "epoch": 0.31083446800233416,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499388210046469,
      "loss": 5.6363,
      "mean_token_accuracy": 0.15328868180513383,
      "num_tokens": 6930017.0,
      "step": 3995
    },
    {
      "entropy": 5.828969812393188,
      "epoch": 0.31122349737405175,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993861658840102,
      "loss": 5.6069,
      "mean_token_accuracy": 0.16110063344240189,
      "num_tokens": 6938410.0,
      "step": 4000
    },
    {
      "entropy": 5.869175863265991,
      "epoch": 0.3116125267457693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993841183168484,
      "loss": 5.6921,
      "mean_token_accuracy": 0.15099752247333526,
      "num_tokens": 6946227.0,
      "step": 4005
    },
    {
      "entropy": 5.917933177947998,
      "epoch": 0.31200155611748687,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993820673450148,
      "loss": 5.6613,
      "mean_token_accuracy": 0.15177705138921738,
      "num_tokens": 6954747.0,
      "step": 4010
    },
    {
      "entropy": 5.9147532939910885,
      "epoch": 0.31239058548920445,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993800129685404,
      "loss": 5.7437,
      "mean_token_accuracy": 0.15076399445533753,
      "num_tokens": 6962990.0,
      "step": 4015
    },
    {
      "entropy": 5.900157785415649,
      "epoch": 0.312779614860922,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993779551874565,
      "loss": 5.708,
      "mean_token_accuracy": 0.1561369016766548,
      "num_tokens": 6972351.0,
      "step": 4020
    },
    {
      "entropy": 5.833806705474854,
      "epoch": 0.31316864423263957,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993758940017943,
      "loss": 5.6374,
      "mean_token_accuracy": 0.15499215722084045,
      "num_tokens": 6980387.0,
      "step": 4025
    },
    {
      "entropy": 5.7456012725830075,
      "epoch": 0.31355767360435716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499373829411585,
      "loss": 5.6995,
      "mean_token_accuracy": 0.15577811747789383,
      "num_tokens": 6989388.0,
      "step": 4030
    },
    {
      "entropy": 5.9648643970489506,
      "epoch": 0.3139467029760747,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049937176141686,
      "loss": 5.6524,
      "mean_token_accuracy": 0.16123358607292176,
      "num_tokens": 6997714.0,
      "step": 4035
    },
    {
      "entropy": 5.904571390151977,
      "epoch": 0.3143357323477923,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993696900176506,
      "loss": 5.6425,
      "mean_token_accuracy": 0.15953611582517624,
      "num_tokens": 7006453.0,
      "step": 4040
    },
    {
      "entropy": 5.8469031810760494,
      "epoch": 0.3147247617195098,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993676152139883,
      "loss": 5.7245,
      "mean_token_accuracy": 0.15670867413282394,
      "num_tokens": 7015398.0,
      "step": 4045
    },
    {
      "entropy": 5.891272020339966,
      "epoch": 0.3151137910912274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993655370059046,
      "loss": 5.6789,
      "mean_token_accuracy": 0.15234642922878266,
      "num_tokens": 7023796.0,
      "step": 4050
    },
    {
      "entropy": 5.798313093185425,
      "epoch": 0.315502820462945,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993634553934309,
      "loss": 5.6087,
      "mean_token_accuracy": 0.15713260769844056,
      "num_tokens": 7032453.0,
      "step": 4055
    },
    {
      "entropy": 5.853315973281861,
      "epoch": 0.3158918498346625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993613703765989,
      "loss": 5.7517,
      "mean_token_accuracy": 0.147060726583004,
      "num_tokens": 7041034.0,
      "step": 4060
    },
    {
      "entropy": 5.992836666107178,
      "epoch": 0.3162808792063801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993592819554402,
      "loss": 5.8351,
      "mean_token_accuracy": 0.14717316254973412,
      "num_tokens": 7050289.0,
      "step": 4065
    },
    {
      "entropy": 5.859228849411011,
      "epoch": 0.3166699085780976,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993571901299866,
      "loss": 5.6727,
      "mean_token_accuracy": 0.163621486723423,
      "num_tokens": 7059638.0,
      "step": 4070
    },
    {
      "entropy": 5.871851444244385,
      "epoch": 0.3170589379498152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993550949002697,
      "loss": 5.6877,
      "mean_token_accuracy": 0.15704762786626816,
      "num_tokens": 7067763.0,
      "step": 4075
    },
    {
      "entropy": 5.825189065933228,
      "epoch": 0.3174479673215328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993529962663213,
      "loss": 5.664,
      "mean_token_accuracy": 0.15967883467674254,
      "num_tokens": 7076126.0,
      "step": 4080
    },
    {
      "entropy": 5.8473530292510985,
      "epoch": 0.31783699669325033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993508942281732,
      "loss": 5.5977,
      "mean_token_accuracy": 0.159432315826416,
      "num_tokens": 7084403.0,
      "step": 4085
    },
    {
      "entropy": 5.745470857620239,
      "epoch": 0.3182260260649679,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993487887858575,
      "loss": 5.5587,
      "mean_token_accuracy": 0.1603716179728508,
      "num_tokens": 7093450.0,
      "step": 4090
    },
    {
      "entropy": 5.778246736526489,
      "epoch": 0.31861505543668545,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499346679939406,
      "loss": 5.575,
      "mean_token_accuracy": 0.16252574771642686,
      "num_tokens": 7101722.0,
      "step": 4095
    },
    {
      "entropy": 5.819636249542237,
      "epoch": 0.31900408480840303,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993445676888507,
      "loss": 5.6656,
      "mean_token_accuracy": 0.15075717568397523,
      "num_tokens": 7111285.0,
      "step": 4100
    },
    {
      "entropy": 5.964729976654053,
      "epoch": 0.3193931141801206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993424520342236,
      "loss": 5.7198,
      "mean_token_accuracy": 0.15374993532896042,
      "num_tokens": 7120281.0,
      "step": 4105
    },
    {
      "entropy": 5.839518117904663,
      "epoch": 0.31978214355183815,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993403329755569,
      "loss": 5.599,
      "mean_token_accuracy": 0.1560365989804268,
      "num_tokens": 7128584.0,
      "step": 4110
    },
    {
      "entropy": 5.919183683395386,
      "epoch": 0.32017117292355574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993382105128828,
      "loss": 5.7146,
      "mean_token_accuracy": 0.15515556931495667,
      "num_tokens": 7137570.0,
      "step": 4115
    },
    {
      "entropy": 5.8676293849945065,
      "epoch": 0.32056020229527327,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993360846462333,
      "loss": 5.7457,
      "mean_token_accuracy": 0.15447864532470704,
      "num_tokens": 7146965.0,
      "step": 4120
    },
    {
      "entropy": 5.861276435852051,
      "epoch": 0.32094923166699085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993339553756408,
      "loss": 5.7006,
      "mean_token_accuracy": 0.14976990669965745,
      "num_tokens": 7155657.0,
      "step": 4125
    },
    {
      "entropy": 5.911796903610229,
      "epoch": 0.32133826103870844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993318227011378,
      "loss": 5.639,
      "mean_token_accuracy": 0.15447454154491425,
      "num_tokens": 7164369.0,
      "step": 4130
    },
    {
      "entropy": 5.9126640319824215,
      "epoch": 0.32172729041042597,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993296866227562,
      "loss": 5.6142,
      "mean_token_accuracy": 0.166330786049366,
      "num_tokens": 7172161.0,
      "step": 4135
    },
    {
      "entropy": 5.818078565597534,
      "epoch": 0.32211631978214356,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993275471405288,
      "loss": 5.6465,
      "mean_token_accuracy": 0.14939845651388167,
      "num_tokens": 7180606.0,
      "step": 4140
    },
    {
      "entropy": 5.8250974178314205,
      "epoch": 0.32250534915386114,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993254042544879,
      "loss": 5.6835,
      "mean_token_accuracy": 0.15783559530973434,
      "num_tokens": 7188912.0,
      "step": 4145
    },
    {
      "entropy": 5.869154214859009,
      "epoch": 0.3228943785255787,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499323257964666,
      "loss": 5.6958,
      "mean_token_accuracy": 0.15014004483819007,
      "num_tokens": 7197477.0,
      "step": 4150
    },
    {
      "entropy": 5.912593984603882,
      "epoch": 0.32328340789729626,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993211082710957,
      "loss": 5.6826,
      "mean_token_accuracy": 0.15499550700187684,
      "num_tokens": 7206571.0,
      "step": 4155
    },
    {
      "entropy": 5.819326686859131,
      "epoch": 0.3236724372690138,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993189551738097,
      "loss": 5.5213,
      "mean_token_accuracy": 0.15574385076761246,
      "num_tokens": 7215319.0,
      "step": 4160
    },
    {
      "entropy": 5.8085283756256105,
      "epoch": 0.3240614666407314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993167986728405,
      "loss": 5.6901,
      "mean_token_accuracy": 0.1536120668053627,
      "num_tokens": 7223389.0,
      "step": 4165
    },
    {
      "entropy": 5.89440016746521,
      "epoch": 0.32445049601244896,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499314638768221,
      "loss": 5.6772,
      "mean_token_accuracy": 0.1551726020872593,
      "num_tokens": 7231853.0,
      "step": 4170
    },
    {
      "entropy": 5.808723640441895,
      "epoch": 0.3248395253841665,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993124754599839,
      "loss": 5.5961,
      "mean_token_accuracy": 0.16090391278266908,
      "num_tokens": 7241081.0,
      "step": 4175
    },
    {
      "entropy": 5.803765726089478,
      "epoch": 0.3252285547558841,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993103087481619,
      "loss": 5.6173,
      "mean_token_accuracy": 0.1579311728477478,
      "num_tokens": 7249742.0,
      "step": 4180
    },
    {
      "entropy": 5.946992683410644,
      "epoch": 0.3256175841276016,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993081386327881,
      "loss": 5.7412,
      "mean_token_accuracy": 0.15369976907968522,
      "num_tokens": 7258285.0,
      "step": 4185
    },
    {
      "entropy": 5.836717367172241,
      "epoch": 0.3260066134993192,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993059651138953,
      "loss": 5.5644,
      "mean_token_accuracy": 0.15882256627082825,
      "num_tokens": 7266843.0,
      "step": 4190
    },
    {
      "entropy": 5.816893720626831,
      "epoch": 0.3263956428710368,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993037881915165,
      "loss": 5.6532,
      "mean_token_accuracy": 0.16056596040725707,
      "num_tokens": 7276041.0,
      "step": 4195
    },
    {
      "entropy": 5.843226528167724,
      "epoch": 0.3267846722427543,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993016078656847,
      "loss": 5.7217,
      "mean_token_accuracy": 0.14809716641902923,
      "num_tokens": 7285144.0,
      "step": 4200
    },
    {
      "entropy": 5.763007354736328,
      "epoch": 0.3271737016144719,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499299424136433,
      "loss": 5.5791,
      "mean_token_accuracy": 0.15921205282211304,
      "num_tokens": 7293976.0,
      "step": 4205
    },
    {
      "entropy": 5.877978038787842,
      "epoch": 0.32756273098618943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992972370037946,
      "loss": 5.6337,
      "mean_token_accuracy": 0.15276181995868682,
      "num_tokens": 7302365.0,
      "step": 4210
    },
    {
      "entropy": 5.869289970397949,
      "epoch": 0.327951760357907,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992950464678026,
      "loss": 5.7025,
      "mean_token_accuracy": 0.15320539623498916,
      "num_tokens": 7310909.0,
      "step": 4215
    },
    {
      "entropy": 5.847477769851684,
      "epoch": 0.3283407897296246,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992928525284903,
      "loss": 5.6565,
      "mean_token_accuracy": 0.15063217729330064,
      "num_tokens": 7319391.0,
      "step": 4220
    },
    {
      "entropy": 5.920811128616333,
      "epoch": 0.32872981910134214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499290655185891,
      "loss": 5.7399,
      "mean_token_accuracy": 0.15288178324699403,
      "num_tokens": 7328076.0,
      "step": 4225
    },
    {
      "entropy": 5.898917484283447,
      "epoch": 0.3291188484730597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499288454440038,
      "loss": 5.6853,
      "mean_token_accuracy": 0.1568422093987465,
      "num_tokens": 7336833.0,
      "step": 4230
    },
    {
      "entropy": 5.820904541015625,
      "epoch": 0.3295078778447773,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992862502909645,
      "loss": 5.5888,
      "mean_token_accuracy": 0.15529243350028993,
      "num_tokens": 7344862.0,
      "step": 4235
    },
    {
      "entropy": 5.883137321472168,
      "epoch": 0.32989690721649484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992840427387043,
      "loss": 5.631,
      "mean_token_accuracy": 0.15965203195810318,
      "num_tokens": 7354503.0,
      "step": 4240
    },
    {
      "entropy": 5.8461486339569095,
      "epoch": 0.3302859365882124,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992818317832908,
      "loss": 5.6215,
      "mean_token_accuracy": 0.15874299556016921,
      "num_tokens": 7363000.0,
      "step": 4245
    },
    {
      "entropy": 5.815576553344727,
      "epoch": 0.33067496595992996,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992796174247574,
      "loss": 5.6953,
      "mean_token_accuracy": 0.1515546426177025,
      "num_tokens": 7372619.0,
      "step": 4250
    },
    {
      "entropy": 5.854606962203979,
      "epoch": 0.33106399533164754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992773996631378,
      "loss": 5.5689,
      "mean_token_accuracy": 0.1614883705973625,
      "num_tokens": 7380531.0,
      "step": 4255
    },
    {
      "entropy": 5.827009010314941,
      "epoch": 0.33145302470336513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992751784984656,
      "loss": 5.6972,
      "mean_token_accuracy": 0.1566160276532173,
      "num_tokens": 7389153.0,
      "step": 4260
    },
    {
      "entropy": 5.8612353801727295,
      "epoch": 0.33184205407508266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992729539307745,
      "loss": 5.5867,
      "mean_token_accuracy": 0.16022619009017944,
      "num_tokens": 7397880.0,
      "step": 4265
    },
    {
      "entropy": 5.856357765197754,
      "epoch": 0.33223108344680025,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992707259600984,
      "loss": 5.6502,
      "mean_token_accuracy": 0.15121706277132035,
      "num_tokens": 7406180.0,
      "step": 4270
    },
    {
      "entropy": 5.796455955505371,
      "epoch": 0.3326201128185178,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992684945864709,
      "loss": 5.6584,
      "mean_token_accuracy": 0.1551528349518776,
      "num_tokens": 7415069.0,
      "step": 4275
    },
    {
      "entropy": 5.8814550876617435,
      "epoch": 0.33300914219023536,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992662598099261,
      "loss": 5.7172,
      "mean_token_accuracy": 0.1513809785246849,
      "num_tokens": 7423921.0,
      "step": 4280
    },
    {
      "entropy": 5.777252197265625,
      "epoch": 0.33339817156195295,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992640216304975,
      "loss": 5.4526,
      "mean_token_accuracy": 0.166193225979805,
      "num_tokens": 7432655.0,
      "step": 4285
    },
    {
      "entropy": 5.844664001464844,
      "epoch": 0.3337872009336705,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992617800482196,
      "loss": 5.6579,
      "mean_token_accuracy": 0.15427033901214598,
      "num_tokens": 7441858.0,
      "step": 4290
    },
    {
      "entropy": 5.807436227798462,
      "epoch": 0.33417623030538807,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499259535063126,
      "loss": 5.5975,
      "mean_token_accuracy": 0.15745315551757813,
      "num_tokens": 7450030.0,
      "step": 4295
    },
    {
      "entropy": 5.863565254211426,
      "epoch": 0.3345652596771056,
      "grad_norm": 1.0,
      "learning_rate": 0.000499257286675251,
      "loss": 5.6936,
      "mean_token_accuracy": 0.15200677663087844,
      "num_tokens": 7458256.0,
      "step": 4300
    },
    {
      "entropy": 5.848210716247559,
      "epoch": 0.3349542890488232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992550348846285,
      "loss": 5.6575,
      "mean_token_accuracy": 0.15124004632234572,
      "num_tokens": 7466710.0,
      "step": 4305
    },
    {
      "entropy": 5.7846517086029055,
      "epoch": 0.33534331842054077,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992527796912928,
      "loss": 5.612,
      "mean_token_accuracy": 0.16022450029850005,
      "num_tokens": 7475445.0,
      "step": 4310
    },
    {
      "entropy": 5.908424758911133,
      "epoch": 0.3357323477922583,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992505210952782,
      "loss": 5.6893,
      "mean_token_accuracy": 0.15276000201702117,
      "num_tokens": 7484326.0,
      "step": 4315
    },
    {
      "entropy": 5.804583883285522,
      "epoch": 0.3361213771639759,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992482590966188,
      "loss": 5.5834,
      "mean_token_accuracy": 0.15481097400188445,
      "num_tokens": 7493028.0,
      "step": 4320
    },
    {
      "entropy": 5.905902528762818,
      "epoch": 0.3365104065356934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992459936953489,
      "loss": 5.749,
      "mean_token_accuracy": 0.14618126153945923,
      "num_tokens": 7502284.0,
      "step": 4325
    },
    {
      "entropy": 5.892045593261718,
      "epoch": 0.336899435907411,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992437248915032,
      "loss": 5.6346,
      "mean_token_accuracy": 0.15248938798904418,
      "num_tokens": 7510728.0,
      "step": 4330
    },
    {
      "entropy": 5.792527294158935,
      "epoch": 0.3372884652791286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992414526851157,
      "loss": 5.5951,
      "mean_token_accuracy": 0.16316897571086883,
      "num_tokens": 7519578.0,
      "step": 4335
    },
    {
      "entropy": 5.806044721603394,
      "epoch": 0.3376774946508461,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992391770762212,
      "loss": 5.7128,
      "mean_token_accuracy": 0.15763205736875535,
      "num_tokens": 7528853.0,
      "step": 4340
    },
    {
      "entropy": 5.998908042907715,
      "epoch": 0.3380665240225637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499236898064854,
      "loss": 5.7671,
      "mean_token_accuracy": 0.15301490128040313,
      "num_tokens": 7537000.0,
      "step": 4345
    },
    {
      "entropy": 5.8523975849151615,
      "epoch": 0.3384555533942813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992346156510488,
      "loss": 5.7196,
      "mean_token_accuracy": 0.15149300843477248,
      "num_tokens": 7545439.0,
      "step": 4350
    },
    {
      "entropy": 5.856060743331909,
      "epoch": 0.3388445827659988,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992323298348403,
      "loss": 5.6033,
      "mean_token_accuracy": 0.15418616980314254,
      "num_tokens": 7553808.0,
      "step": 4355
    },
    {
      "entropy": 5.8948029518127445,
      "epoch": 0.3392336121377164,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499230040616263,
      "loss": 5.6446,
      "mean_token_accuracy": 0.16090449094772338,
      "num_tokens": 7562124.0,
      "step": 4360
    },
    {
      "entropy": 5.895649337768555,
      "epoch": 0.33962264150943394,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992277479953518,
      "loss": 5.688,
      "mean_token_accuracy": 0.15238938927650453,
      "num_tokens": 7570692.0,
      "step": 4365
    },
    {
      "entropy": 5.828374528884888,
      "epoch": 0.34001167088115153,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992254519721414,
      "loss": 5.5795,
      "mean_token_accuracy": 0.15819404125213624,
      "num_tokens": 7579380.0,
      "step": 4370
    },
    {
      "entropy": 5.795167303085327,
      "epoch": 0.3404007002528691,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992231525466666,
      "loss": 5.6407,
      "mean_token_accuracy": 0.15771155208349227,
      "num_tokens": 7587581.0,
      "step": 4375
    },
    {
      "entropy": 5.879871654510498,
      "epoch": 0.34078972962458665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992208497189624,
      "loss": 5.6951,
      "mean_token_accuracy": 0.1571160614490509,
      "num_tokens": 7596373.0,
      "step": 4380
    },
    {
      "entropy": 5.863206577301026,
      "epoch": 0.34117875899630423,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992185434890637,
      "loss": 5.7015,
      "mean_token_accuracy": 0.15472009479999543,
      "num_tokens": 7605698.0,
      "step": 4385
    },
    {
      "entropy": 5.922804832458496,
      "epoch": 0.34156778836802176,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992162338570055,
      "loss": 5.7139,
      "mean_token_accuracy": 0.14772787168622017,
      "num_tokens": 7614299.0,
      "step": 4390
    },
    {
      "entropy": 5.789683628082275,
      "epoch": 0.34195681773973935,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992139208228227,
      "loss": 5.6359,
      "mean_token_accuracy": 0.16233089715242385,
      "num_tokens": 7622244.0,
      "step": 4395
    },
    {
      "entropy": 5.862781143188476,
      "epoch": 0.34234584711145694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992116043865506,
      "loss": 5.7461,
      "mean_token_accuracy": 0.14375390112400055,
      "num_tokens": 7631421.0,
      "step": 4400
    },
    {
      "entropy": 5.8913651466369625,
      "epoch": 0.34273487648317447,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992092845482244,
      "loss": 5.6914,
      "mean_token_accuracy": 0.16382178217172622,
      "num_tokens": 7640483.0,
      "step": 4405
    },
    {
      "entropy": 5.813113498687744,
      "epoch": 0.34312390585489205,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992069613078791,
      "loss": 5.6053,
      "mean_token_accuracy": 0.1562155529856682,
      "num_tokens": 7648928.0,
      "step": 4410
    },
    {
      "entropy": 5.845783948898315,
      "epoch": 0.3435129352266096,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992046346655499,
      "loss": 5.6107,
      "mean_token_accuracy": 0.15283971428871154,
      "num_tokens": 7658158.0,
      "step": 4415
    },
    {
      "entropy": 5.907624340057373,
      "epoch": 0.34390196459832717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992023046212723,
      "loss": 5.6925,
      "mean_token_accuracy": 0.15578252226114273,
      "num_tokens": 7666657.0,
      "step": 4420
    },
    {
      "entropy": 5.849208879470825,
      "epoch": 0.34429099397004476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991999711750816,
      "loss": 5.6501,
      "mean_token_accuracy": 0.1562605082988739,
      "num_tokens": 7675182.0,
      "step": 4425
    },
    {
      "entropy": 5.807325410842895,
      "epoch": 0.3446800233417623,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991976343270133,
      "loss": 5.5959,
      "mean_token_accuracy": 0.15964483618736267,
      "num_tokens": 7684132.0,
      "step": 4430
    },
    {
      "entropy": 5.784048795700073,
      "epoch": 0.3450690527134799,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991952940771026,
      "loss": 5.5463,
      "mean_token_accuracy": 0.16304491460323334,
      "num_tokens": 7692574.0,
      "step": 4435
    },
    {
      "entropy": 5.784449100494385,
      "epoch": 0.34545808208519746,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991929504253852,
      "loss": 5.6104,
      "mean_token_accuracy": 0.15977637320756913,
      "num_tokens": 7700969.0,
      "step": 4440
    },
    {
      "entropy": 5.914915323257446,
      "epoch": 0.345847111456915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991906033718966,
      "loss": 5.7533,
      "mean_token_accuracy": 0.15570002794265747,
      "num_tokens": 7710688.0,
      "step": 4445
    },
    {
      "entropy": 5.914741516113281,
      "epoch": 0.3462361408286326,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991882529166724,
      "loss": 5.7238,
      "mean_token_accuracy": 0.15907996892929077,
      "num_tokens": 7720071.0,
      "step": 4450
    },
    {
      "entropy": 5.680905103683472,
      "epoch": 0.3466251702003501,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991858990597483,
      "loss": 5.5067,
      "mean_token_accuracy": 0.1640620693564415,
      "num_tokens": 7728418.0,
      "step": 4455
    },
    {
      "entropy": 5.833381080627442,
      "epoch": 0.3470141995720677,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991835418011601,
      "loss": 5.7421,
      "mean_token_accuracy": 0.15468787997961045,
      "num_tokens": 7736849.0,
      "step": 4460
    },
    {
      "entropy": 5.886175918579101,
      "epoch": 0.3474032289437853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991811811409434,
      "loss": 5.6654,
      "mean_token_accuracy": 0.15654261633753777,
      "num_tokens": 7745730.0,
      "step": 4465
    },
    {
      "entropy": 5.88368854522705,
      "epoch": 0.3477922583155028,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991788170791341,
      "loss": 5.5561,
      "mean_token_accuracy": 0.1647365629673004,
      "num_tokens": 7755040.0,
      "step": 4470
    },
    {
      "entropy": 5.8558320045471195,
      "epoch": 0.3481812876872204,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991764496157682,
      "loss": 5.6041,
      "mean_token_accuracy": 0.15823976844549179,
      "num_tokens": 7763765.0,
      "step": 4475
    },
    {
      "entropy": 5.704217386245728,
      "epoch": 0.34857031705893793,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991740787508814,
      "loss": 5.5711,
      "mean_token_accuracy": 0.15702741593122482,
      "num_tokens": 7772261.0,
      "step": 4480
    },
    {
      "entropy": 5.9202171802520756,
      "epoch": 0.3489593464306555,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991717044845097,
      "loss": 5.7251,
      "mean_token_accuracy": 0.1542724072933197,
      "num_tokens": 7781651.0,
      "step": 4485
    },
    {
      "entropy": 5.812010622024536,
      "epoch": 0.3493483758023731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991693268166892,
      "loss": 5.5888,
      "mean_token_accuracy": 0.15705888569355012,
      "num_tokens": 7789926.0,
      "step": 4490
    },
    {
      "entropy": 5.857488393783569,
      "epoch": 0.34973740517409063,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991669457474559,
      "loss": 5.6169,
      "mean_token_accuracy": 0.15753783285617828,
      "num_tokens": 7798435.0,
      "step": 4495
    },
    {
      "entropy": 5.712810802459717,
      "epoch": 0.3501264345458082,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991645612768461,
      "loss": 5.5488,
      "mean_token_accuracy": 0.1559806138277054,
      "num_tokens": 7806674.0,
      "step": 4500
    },
    {
      "entropy": 5.814238500595093,
      "epoch": 0.35051546391752575,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991621734048958,
      "loss": 5.6,
      "mean_token_accuracy": 0.16171200275421144,
      "num_tokens": 7814774.0,
      "step": 4505
    },
    {
      "entropy": 5.826798057556152,
      "epoch": 0.35090449328924334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991597821316415,
      "loss": 5.6531,
      "mean_token_accuracy": 0.16424768418073654,
      "num_tokens": 7822879.0,
      "step": 4510
    },
    {
      "entropy": 5.7536486148834225,
      "epoch": 0.3512935226609609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991573874571192,
      "loss": 5.588,
      "mean_token_accuracy": 0.16008912175893783,
      "num_tokens": 7831418.0,
      "step": 4515
    },
    {
      "entropy": 5.793634223937988,
      "epoch": 0.35168255203267845,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991549893813653,
      "loss": 5.5949,
      "mean_token_accuracy": 0.15561521500349046,
      "num_tokens": 7839513.0,
      "step": 4520
    },
    {
      "entropy": 5.831595373153687,
      "epoch": 0.35207158140439604,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991525879044162,
      "loss": 5.5878,
      "mean_token_accuracy": 0.1584882080554962,
      "num_tokens": 7847653.0,
      "step": 4525
    },
    {
      "entropy": 5.856019306182861,
      "epoch": 0.35246061077611357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991501830263084,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1556251361966133,
      "num_tokens": 7857093.0,
      "step": 4530
    },
    {
      "entropy": 5.83241229057312,
      "epoch": 0.35284964014783116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991477747470783,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1533559799194336,
      "num_tokens": 7866129.0,
      "step": 4535
    },
    {
      "entropy": 5.806477451324463,
      "epoch": 0.35323866951954874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991453630667625,
      "loss": 5.7329,
      "mean_token_accuracy": 0.15247348248958587,
      "num_tokens": 7874613.0,
      "step": 4540
    },
    {
      "entropy": 5.747210550308227,
      "epoch": 0.3536276988912663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991429479853976,
      "loss": 5.4863,
      "mean_token_accuracy": 0.16456296294927597,
      "num_tokens": 7882292.0,
      "step": 4545
    },
    {
      "entropy": 5.774030637741089,
      "epoch": 0.35401672826298386,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991405295030202,
      "loss": 5.4959,
      "mean_token_accuracy": 0.1663362666964531,
      "num_tokens": 7890810.0,
      "step": 4550
    },
    {
      "entropy": 5.854322004318237,
      "epoch": 0.35440575763470145,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991381076196671,
      "loss": 5.6354,
      "mean_token_accuracy": 0.15660456120967864,
      "num_tokens": 7899964.0,
      "step": 4555
    },
    {
      "entropy": 5.778811407089234,
      "epoch": 0.354794787006419,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991356823353748,
      "loss": 5.5401,
      "mean_token_accuracy": 0.15894368588924407,
      "num_tokens": 7908600.0,
      "step": 4560
    },
    {
      "entropy": 5.772751426696777,
      "epoch": 0.35518381637813656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991332536501804,
      "loss": 5.6431,
      "mean_token_accuracy": 0.15342240929603576,
      "num_tokens": 7916893.0,
      "step": 4565
    },
    {
      "entropy": 5.817650604248047,
      "epoch": 0.3555728457498541,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991308215641205,
      "loss": 5.5202,
      "mean_token_accuracy": 0.16820951849222182,
      "num_tokens": 7925230.0,
      "step": 4570
    },
    {
      "entropy": 5.798580551147461,
      "epoch": 0.3559618751215717,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991283860772323,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16215160638093948,
      "num_tokens": 7933487.0,
      "step": 4575
    },
    {
      "entropy": 5.771260070800781,
      "epoch": 0.35635090449328927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991259471895525,
      "loss": 5.6161,
      "mean_token_accuracy": 0.15994802117347717,
      "num_tokens": 7941957.0,
      "step": 4580
    },
    {
      "entropy": 5.832534074783325,
      "epoch": 0.3567399338650068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991235049011182,
      "loss": 5.5535,
      "mean_token_accuracy": 0.15915501564741136,
      "num_tokens": 7950912.0,
      "step": 4585
    },
    {
      "entropy": 5.753993844985962,
      "epoch": 0.3571289632367244,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991210592119663,
      "loss": 5.6032,
      "mean_token_accuracy": 0.1576017312705517,
      "num_tokens": 7959724.0,
      "step": 4590
    },
    {
      "entropy": 5.818193960189819,
      "epoch": 0.3575179926084419,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991186101221342,
      "loss": 5.588,
      "mean_token_accuracy": 0.16292066127061844,
      "num_tokens": 7968683.0,
      "step": 4595
    },
    {
      "entropy": 5.745058488845825,
      "epoch": 0.3579070219801595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499116157631659,
      "loss": 5.5467,
      "mean_token_accuracy": 0.1611933469772339,
      "num_tokens": 7977004.0,
      "step": 4600
    },
    {
      "entropy": 5.907850551605224,
      "epoch": 0.3582960513518771,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991137017405777,
      "loss": 5.7577,
      "mean_token_accuracy": 0.14255229756236076,
      "num_tokens": 7985537.0,
      "step": 4605
    },
    {
      "entropy": 5.8856831073760985,
      "epoch": 0.3586850807235946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991112424489277,
      "loss": 5.6558,
      "mean_token_accuracy": 0.1539052553474903,
      "num_tokens": 7994298.0,
      "step": 4610
    },
    {
      "entropy": 5.765837383270264,
      "epoch": 0.3590741100953122,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991087797567462,
      "loss": 5.6095,
      "mean_token_accuracy": 0.15761878043413163,
      "num_tokens": 8003145.0,
      "step": 4615
    },
    {
      "entropy": 5.780225086212158,
      "epoch": 0.35946313946702974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991063136640709,
      "loss": 5.5713,
      "mean_token_accuracy": 0.15816087424755096,
      "num_tokens": 8011839.0,
      "step": 4620
    },
    {
      "entropy": 5.883225059509277,
      "epoch": 0.3598521688387473,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991038441709388,
      "loss": 5.638,
      "mean_token_accuracy": 0.15164339542388916,
      "num_tokens": 8020248.0,
      "step": 4625
    },
    {
      "entropy": 5.853153276443481,
      "epoch": 0.3602411982104649,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991013712773877,
      "loss": 5.6816,
      "mean_token_accuracy": 0.15706008225679396,
      "num_tokens": 8029360.0,
      "step": 4630
    },
    {
      "entropy": 5.949721956253052,
      "epoch": 0.36063022758218244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990988949834549,
      "loss": 5.7427,
      "mean_token_accuracy": 0.1489767387509346,
      "num_tokens": 8037407.0,
      "step": 4635
    },
    {
      "entropy": 5.817428922653198,
      "epoch": 0.3610192569539,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499096415289178,
      "loss": 5.5554,
      "mean_token_accuracy": 0.16257950216531752,
      "num_tokens": 8045935.0,
      "step": 4640
    },
    {
      "entropy": 5.811090898513794,
      "epoch": 0.3614082863256176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990939321945948,
      "loss": 5.6986,
      "mean_token_accuracy": 0.14591270238161086,
      "num_tokens": 8054640.0,
      "step": 4645
    },
    {
      "entropy": 5.969038772583008,
      "epoch": 0.36179731569733514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990914456997427,
      "loss": 5.6927,
      "mean_token_accuracy": 0.15529640913009643,
      "num_tokens": 8063845.0,
      "step": 4650
    },
    {
      "entropy": 5.836845779418946,
      "epoch": 0.36218634506905273,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990889558046598,
      "loss": 5.6112,
      "mean_token_accuracy": 0.15886921882629396,
      "num_tokens": 8072389.0,
      "step": 4655
    },
    {
      "entropy": 5.754767656326294,
      "epoch": 0.36257537444077026,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990864625093836,
      "loss": 5.5007,
      "mean_token_accuracy": 0.157828189432621,
      "num_tokens": 8080683.0,
      "step": 4660
    },
    {
      "entropy": 5.697498416900634,
      "epoch": 0.36296440381248785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990839658139519,
      "loss": 5.5303,
      "mean_token_accuracy": 0.15996571630239487,
      "num_tokens": 8088945.0,
      "step": 4665
    },
    {
      "entropy": 5.781423330307007,
      "epoch": 0.36335343318420543,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990814657184028,
      "loss": 5.5986,
      "mean_token_accuracy": 0.1587238773703575,
      "num_tokens": 8097447.0,
      "step": 4670
    },
    {
      "entropy": 5.739903402328491,
      "epoch": 0.36374246255592296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990789622227741,
      "loss": 5.4449,
      "mean_token_accuracy": 0.1716950759291649,
      "num_tokens": 8105570.0,
      "step": 4675
    },
    {
      "entropy": 5.749712800979614,
      "epoch": 0.36413149192764055,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990764553271038,
      "loss": 5.5757,
      "mean_token_accuracy": 0.15827874094247818,
      "num_tokens": 8114842.0,
      "step": 4680
    },
    {
      "entropy": 5.777094793319702,
      "epoch": 0.3645205212993581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990739450314299,
      "loss": 5.5569,
      "mean_token_accuracy": 0.15483318269252777,
      "num_tokens": 8122960.0,
      "step": 4685
    },
    {
      "entropy": 5.901574230194091,
      "epoch": 0.36490955067107567,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990714313357906,
      "loss": 5.693,
      "mean_token_accuracy": 0.15768003612756729,
      "num_tokens": 8131850.0,
      "step": 4690
    },
    {
      "entropy": 5.799890947341919,
      "epoch": 0.36529858004279325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499068914240224,
      "loss": 5.5622,
      "mean_token_accuracy": 0.16395678371191025,
      "num_tokens": 8139818.0,
      "step": 4695
    },
    {
      "entropy": 5.717095994949341,
      "epoch": 0.3656876094145108,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990663937447682,
      "loss": 5.6505,
      "mean_token_accuracy": 0.15960523188114167,
      "num_tokens": 8147995.0,
      "step": 4700
    },
    {
      "entropy": 5.881393527984619,
      "epoch": 0.36607663878622837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990638698494615,
      "loss": 5.6229,
      "mean_token_accuracy": 0.1586279332637787,
      "num_tokens": 8156879.0,
      "step": 4705
    },
    {
      "entropy": 5.8190680027008055,
      "epoch": 0.3664656681579459,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990613425543423,
      "loss": 5.5931,
      "mean_token_accuracy": 0.15052010118961334,
      "num_tokens": 8164860.0,
      "step": 4710
    },
    {
      "entropy": 5.700157213211059,
      "epoch": 0.3668546975296635,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990588118594487,
      "loss": 5.4404,
      "mean_token_accuracy": 0.16611094921827316,
      "num_tokens": 8173319.0,
      "step": 4715
    },
    {
      "entropy": 5.810721349716187,
      "epoch": 0.3672437269013811,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990562777648194,
      "loss": 5.6415,
      "mean_token_accuracy": 0.15685734450817107,
      "num_tokens": 8182385.0,
      "step": 4720
    },
    {
      "entropy": 5.84443302154541,
      "epoch": 0.3676327562730986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990537402704928,
      "loss": 5.5731,
      "mean_token_accuracy": 0.16274489611387252,
      "num_tokens": 8190394.0,
      "step": 4725
    },
    {
      "entropy": 5.760211992263794,
      "epoch": 0.3680217856448162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990511993765071,
      "loss": 5.583,
      "mean_token_accuracy": 0.15801360756158828,
      "num_tokens": 8198532.0,
      "step": 4730
    },
    {
      "entropy": 5.691163206100464,
      "epoch": 0.3684108150165337,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990486550829011,
      "loss": 5.5359,
      "mean_token_accuracy": 0.16297680139541626,
      "num_tokens": 8207436.0,
      "step": 4735
    },
    {
      "entropy": 5.837288904190063,
      "epoch": 0.3687998443882513,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990461073897134,
      "loss": 5.5726,
      "mean_token_accuracy": 0.16301757246255874,
      "num_tokens": 8216316.0,
      "step": 4740
    },
    {
      "entropy": 5.766282176971435,
      "epoch": 0.3691888737599689,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990435562969827,
      "loss": 5.5246,
      "mean_token_accuracy": 0.1658928021788597,
      "num_tokens": 8224918.0,
      "step": 4745
    },
    {
      "entropy": 5.750837182998657,
      "epoch": 0.3695779031316864,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990410018047475,
      "loss": 5.5437,
      "mean_token_accuracy": 0.15735672414302826,
      "num_tokens": 8233506.0,
      "step": 4750
    },
    {
      "entropy": 5.795170307159424,
      "epoch": 0.369966932503404,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990384439130467,
      "loss": 5.5753,
      "mean_token_accuracy": 0.1599963992834091,
      "num_tokens": 8243122.0,
      "step": 4755
    },
    {
      "entropy": 5.768864107131958,
      "epoch": 0.3703559618751216,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990358826219192,
      "loss": 5.5704,
      "mean_token_accuracy": 0.16572270393371583,
      "num_tokens": 8251639.0,
      "step": 4760
    },
    {
      "entropy": 5.772831726074219,
      "epoch": 0.37074499124683913,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990333179314038,
      "loss": 5.6095,
      "mean_token_accuracy": 0.1563558891415596,
      "num_tokens": 8260487.0,
      "step": 4765
    },
    {
      "entropy": 5.807176303863526,
      "epoch": 0.3711340206185567,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990307498415393,
      "loss": 5.5948,
      "mean_token_accuracy": 0.16280901432037354,
      "num_tokens": 8270111.0,
      "step": 4770
    },
    {
      "entropy": 5.821304416656494,
      "epoch": 0.37152304999027425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990281783523648,
      "loss": 5.7295,
      "mean_token_accuracy": 0.15457207411527635,
      "num_tokens": 8278171.0,
      "step": 4775
    },
    {
      "entropy": 5.840618562698364,
      "epoch": 0.37191207936199183,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990256034639192,
      "loss": 5.6006,
      "mean_token_accuracy": 0.1598903015255928,
      "num_tokens": 8287278.0,
      "step": 4780
    },
    {
      "entropy": 5.860653972625732,
      "epoch": 0.3723011087337094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990230251762418,
      "loss": 5.726,
      "mean_token_accuracy": 0.15083621442317963,
      "num_tokens": 8296874.0,
      "step": 4785
    },
    {
      "entropy": 5.80598816871643,
      "epoch": 0.37269013810542695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990204434893713,
      "loss": 5.5853,
      "mean_token_accuracy": 0.16516731232404708,
      "num_tokens": 8305375.0,
      "step": 4790
    },
    {
      "entropy": 5.915070343017578,
      "epoch": 0.37307916747714454,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990178584033474,
      "loss": 5.6495,
      "mean_token_accuracy": 0.15535937994718552,
      "num_tokens": 8313960.0,
      "step": 4795
    },
    {
      "entropy": 5.777856874465942,
      "epoch": 0.37346819684886207,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990152699182089,
      "loss": 5.573,
      "mean_token_accuracy": 0.16572625935077667,
      "num_tokens": 8321838.0,
      "step": 4800
    },
    {
      "entropy": 5.722106504440307,
      "epoch": 0.37385722622057965,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990126780339953,
      "loss": 5.6074,
      "mean_token_accuracy": 0.15216947495937347,
      "num_tokens": 8331325.0,
      "step": 4805
    },
    {
      "entropy": 5.847458791732788,
      "epoch": 0.37424625559229724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990100827507459,
      "loss": 5.6519,
      "mean_token_accuracy": 0.1540816456079483,
      "num_tokens": 8340105.0,
      "step": 4810
    },
    {
      "entropy": 5.765224981307983,
      "epoch": 0.37463528496401477,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990074840684999,
      "loss": 5.4742,
      "mean_token_accuracy": 0.17038157880306243,
      "num_tokens": 8348256.0,
      "step": 4815
    },
    {
      "entropy": 5.759674692153931,
      "epoch": 0.37502431433573236,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990048819872969,
      "loss": 5.6683,
      "mean_token_accuracy": 0.15316962450742722,
      "num_tokens": 8357456.0,
      "step": 4820
    },
    {
      "entropy": 5.843084287643433,
      "epoch": 0.3754133437074499,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990022765071765,
      "loss": 5.5191,
      "mean_token_accuracy": 0.16007023751735688,
      "num_tokens": 8366939.0,
      "step": 4825
    },
    {
      "entropy": 5.919530153274536,
      "epoch": 0.3758023730791675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498999667628178,
      "loss": 5.6807,
      "mean_token_accuracy": 0.1519789770245552,
      "num_tokens": 8376062.0,
      "step": 4830
    },
    {
      "entropy": 5.824848794937134,
      "epoch": 0.37619140245088506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989970553503411,
      "loss": 5.5943,
      "mean_token_accuracy": 0.16140185594558715,
      "num_tokens": 8383803.0,
      "step": 4835
    },
    {
      "entropy": 5.799917411804199,
      "epoch": 0.3765804318226026,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989944396737054,
      "loss": 5.7424,
      "mean_token_accuracy": 0.15280164629220963,
      "num_tokens": 8392844.0,
      "step": 4840
    },
    {
      "entropy": 5.877451229095459,
      "epoch": 0.3769694611943202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989918205983106,
      "loss": 5.6404,
      "mean_token_accuracy": 0.16161811649799346,
      "num_tokens": 8401937.0,
      "step": 4845
    },
    {
      "entropy": 5.844389724731445,
      "epoch": 0.37735849056603776,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004989891981241964,
      "loss": 5.6497,
      "mean_token_accuracy": 0.15456714555621148,
      "num_tokens": 8410383.0,
      "step": 4850
    },
    {
      "entropy": 5.7435094833374025,
      "epoch": 0.3777475199377553,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989865722514027,
      "loss": 5.6158,
      "mean_token_accuracy": 0.15654363632202148,
      "num_tokens": 8419053.0,
      "step": 4855
    },
    {
      "entropy": 5.789836311340332,
      "epoch": 0.3781365493094729,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989839429799692,
      "loss": 5.5909,
      "mean_token_accuracy": 0.1606529399752617,
      "num_tokens": 8427311.0,
      "step": 4860
    },
    {
      "entropy": 5.791653251647949,
      "epoch": 0.3785255786811904,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498981310309936,
      "loss": 5.6292,
      "mean_token_accuracy": 0.15977821499109268,
      "num_tokens": 8436050.0,
      "step": 4865
    },
    {
      "entropy": 5.909187698364258,
      "epoch": 0.378914608052908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989786742413428,
      "loss": 5.5772,
      "mean_token_accuracy": 0.15534257888793945,
      "num_tokens": 8444579.0,
      "step": 4870
    },
    {
      "entropy": 5.708425283432007,
      "epoch": 0.3793036374246256,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989760347742298,
      "loss": 5.5942,
      "mean_token_accuracy": 0.16260575652122497,
      "num_tokens": 8454013.0,
      "step": 4875
    },
    {
      "entropy": 5.841428232192993,
      "epoch": 0.3796926667963431,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989733919086369,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1565110743045807,
      "num_tokens": 8462564.0,
      "step": 4880
    },
    {
      "entropy": 5.8611180782318115,
      "epoch": 0.3800816961680607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989707456446043,
      "loss": 5.5492,
      "mean_token_accuracy": 0.16574790924787522,
      "num_tokens": 8471035.0,
      "step": 4885
    },
    {
      "entropy": 5.75192437171936,
      "epoch": 0.38047072553977823,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989680959821721,
      "loss": 5.5417,
      "mean_token_accuracy": 0.16359886676073074,
      "num_tokens": 8479475.0,
      "step": 4890
    },
    {
      "entropy": 5.7016252994537355,
      "epoch": 0.3808597549114958,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989654429213805,
      "loss": 5.6019,
      "mean_token_accuracy": 0.1534335196018219,
      "num_tokens": 8488163.0,
      "step": 4895
    },
    {
      "entropy": 5.891068601608277,
      "epoch": 0.3812487842832134,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989627864622699,
      "loss": 5.6493,
      "mean_token_accuracy": 0.1583672896027565,
      "num_tokens": 8497735.0,
      "step": 4900
    },
    {
      "entropy": 5.8236159324646,
      "epoch": 0.38163781365493094,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989601266048804,
      "loss": 5.6098,
      "mean_token_accuracy": 0.15320331156253814,
      "num_tokens": 8506986.0,
      "step": 4905
    },
    {
      "entropy": 5.775536680221558,
      "epoch": 0.3820268430266485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989574633492525,
      "loss": 5.6405,
      "mean_token_accuracy": 0.1587393566966057,
      "num_tokens": 8515168.0,
      "step": 4910
    },
    {
      "entropy": 5.794271469116211,
      "epoch": 0.38241587239836605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989547966954266,
      "loss": 5.6051,
      "mean_token_accuracy": 0.1594834268093109,
      "num_tokens": 8523980.0,
      "step": 4915
    },
    {
      "entropy": 5.7487382888793945,
      "epoch": 0.38280490177008364,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498952126643443,
      "loss": 5.5639,
      "mean_token_accuracy": 0.1573392502963543,
      "num_tokens": 8532938.0,
      "step": 4920
    },
    {
      "entropy": 5.8110785484313965,
      "epoch": 0.3831939311418012,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989494531933424,
      "loss": 5.527,
      "mean_token_accuracy": 0.16117591559886932,
      "num_tokens": 8541951.0,
      "step": 4925
    },
    {
      "entropy": 5.818575191497803,
      "epoch": 0.38358296051351876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989467763451652,
      "loss": 5.6922,
      "mean_token_accuracy": 0.1557263121008873,
      "num_tokens": 8550981.0,
      "step": 4930
    },
    {
      "entropy": 5.767818212509155,
      "epoch": 0.38397198988523634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989440960989523,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1514250800013542,
      "num_tokens": 8559711.0,
      "step": 4935
    },
    {
      "entropy": 5.782817029953003,
      "epoch": 0.3843610192569539,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498941412454744,
      "loss": 5.5571,
      "mean_token_accuracy": 0.15684089809656143,
      "num_tokens": 8567797.0,
      "step": 4940
    },
    {
      "entropy": 5.776398611068726,
      "epoch": 0.38475004862867146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989387254125813,
      "loss": 5.6377,
      "mean_token_accuracy": 0.1555260881781578,
      "num_tokens": 8576719.0,
      "step": 4945
    },
    {
      "entropy": 5.820217752456665,
      "epoch": 0.38513907800038905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989360349725049,
      "loss": 5.6276,
      "mean_token_accuracy": 0.16372839659452437,
      "num_tokens": 8585871.0,
      "step": 4950
    },
    {
      "entropy": 5.802014875411987,
      "epoch": 0.3855281073721066,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989333411345555,
      "loss": 5.5297,
      "mean_token_accuracy": 0.15906736999750137,
      "num_tokens": 8594018.0,
      "step": 4955
    },
    {
      "entropy": 5.754960584640503,
      "epoch": 0.38591713674382416,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989306438987742,
      "loss": 5.6561,
      "mean_token_accuracy": 0.15382689237594604,
      "num_tokens": 8602788.0,
      "step": 4960
    },
    {
      "entropy": 5.839208936691284,
      "epoch": 0.38630616611554175,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989279432652018,
      "loss": 5.6141,
      "mean_token_accuracy": 0.1609216883778572,
      "num_tokens": 8611837.0,
      "step": 4965
    },
    {
      "entropy": 5.8078224658966064,
      "epoch": 0.3866951954872593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989252392338791,
      "loss": 5.5797,
      "mean_token_accuracy": 0.1679807111620903,
      "num_tokens": 8620114.0,
      "step": 4970
    },
    {
      "entropy": 5.905976915359497,
      "epoch": 0.38708422485897687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989225318048475,
      "loss": 5.7177,
      "mean_token_accuracy": 0.14695486649870873,
      "num_tokens": 8628777.0,
      "step": 4975
    },
    {
      "entropy": 5.82988543510437,
      "epoch": 0.3874732542306944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989198209781478,
      "loss": 5.6098,
      "mean_token_accuracy": 0.15751051902770996,
      "num_tokens": 8637088.0,
      "step": 4980
    },
    {
      "entropy": 5.868574142456055,
      "epoch": 0.387862283602412,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989171067538211,
      "loss": 5.6167,
      "mean_token_accuracy": 0.15895154774188996,
      "num_tokens": 8645261.0,
      "step": 4985
    },
    {
      "entropy": 5.833978366851807,
      "epoch": 0.38825131297412957,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498914389131909,
      "loss": 5.6268,
      "mean_token_accuracy": 0.16363714188337325,
      "num_tokens": 8653492.0,
      "step": 4990
    },
    {
      "entropy": 5.720838165283203,
      "epoch": 0.3886403423458471,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989116681124523,
      "loss": 5.5707,
      "mean_token_accuracy": 0.16599183827638625,
      "num_tokens": 8661909.0,
      "step": 4995
    },
    {
      "entropy": 5.814400243759155,
      "epoch": 0.3890293717175647,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989089436954924,
      "loss": 5.5613,
      "mean_token_accuracy": 0.15911221131682396,
      "num_tokens": 8670332.0,
      "step": 5000
    },
    {
      "entropy": 5.823642921447754,
      "epoch": 0.3894184010892822,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989062158810706,
      "loss": 5.594,
      "mean_token_accuracy": 0.16461922824382783,
      "num_tokens": 8678537.0,
      "step": 5005
    },
    {
      "entropy": 5.78374981880188,
      "epoch": 0.3898074304609998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989034846692284,
      "loss": 5.5506,
      "mean_token_accuracy": 0.15969980508089066,
      "num_tokens": 8686839.0,
      "step": 5010
    },
    {
      "entropy": 5.7792510986328125,
      "epoch": 0.3901964598327174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989007500600073,
      "loss": 5.6005,
      "mean_token_accuracy": 0.15860068202018737,
      "num_tokens": 8694407.0,
      "step": 5015
    },
    {
      "entropy": 5.754415273666382,
      "epoch": 0.3905854892044349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988980120534486,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1600363366305828,
      "num_tokens": 8703484.0,
      "step": 5020
    },
    {
      "entropy": 5.891473340988159,
      "epoch": 0.3909745185761525,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498895270649594,
      "loss": 5.6668,
      "mean_token_accuracy": 0.15318677425384522,
      "num_tokens": 8712634.0,
      "step": 5025
    },
    {
      "entropy": 5.756830167770386,
      "epoch": 0.39136354794787004,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498892525848485,
      "loss": 5.536,
      "mean_token_accuracy": 0.16174269914627076,
      "num_tokens": 8721229.0,
      "step": 5030
    },
    {
      "entropy": 5.766126585006714,
      "epoch": 0.3917525773195876,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988897776501633,
      "loss": 5.605,
      "mean_token_accuracy": 0.15656097382307052,
      "num_tokens": 8730010.0,
      "step": 5035
    },
    {
      "entropy": 5.770839738845825,
      "epoch": 0.3921416066913052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988870260546705,
      "loss": 5.5791,
      "mean_token_accuracy": 0.1527113988995552,
      "num_tokens": 8739046.0,
      "step": 5040
    },
    {
      "entropy": 5.774734878540039,
      "epoch": 0.39253063606302274,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988842710620486,
      "loss": 5.6733,
      "mean_token_accuracy": 0.16159862279891968,
      "num_tokens": 8747650.0,
      "step": 5045
    },
    {
      "entropy": 5.762890720367432,
      "epoch": 0.39291966543474033,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988815126723391,
      "loss": 5.587,
      "mean_token_accuracy": 0.16119773238897322,
      "num_tokens": 8756319.0,
      "step": 5050
    },
    {
      "entropy": 5.73527774810791,
      "epoch": 0.3933086948064579,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988787508855841,
      "loss": 5.6062,
      "mean_token_accuracy": 0.16382827013731002,
      "num_tokens": 8765448.0,
      "step": 5055
    },
    {
      "entropy": 5.772875070571899,
      "epoch": 0.39369772417817545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988759857018253,
      "loss": 5.5737,
      "mean_token_accuracy": 0.1571187347173691,
      "num_tokens": 8774146.0,
      "step": 5060
    },
    {
      "entropy": 5.706568622589112,
      "epoch": 0.39408675354989303,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988732171211048,
      "loss": 5.4516,
      "mean_token_accuracy": 0.16227682679891586,
      "num_tokens": 8782311.0,
      "step": 5065
    },
    {
      "entropy": 5.7286888599395756,
      "epoch": 0.39447578292161056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988704451434644,
      "loss": 5.4951,
      "mean_token_accuracy": 0.16643253117799758,
      "num_tokens": 8790196.0,
      "step": 5070
    },
    {
      "entropy": 5.7642089366912845,
      "epoch": 0.39486481229332815,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988676697689465,
      "loss": 5.5482,
      "mean_token_accuracy": 0.16135542541742326,
      "num_tokens": 8798836.0,
      "step": 5075
    },
    {
      "entropy": 5.753359842300415,
      "epoch": 0.39525384166504574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498864890997593,
      "loss": 5.5888,
      "mean_token_accuracy": 0.15956080853939056,
      "num_tokens": 8806942.0,
      "step": 5080
    },
    {
      "entropy": 5.863043117523193,
      "epoch": 0.39564287103676327,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988621088294461,
      "loss": 5.7959,
      "mean_token_accuracy": 0.15221165120601654,
      "num_tokens": 8816354.0,
      "step": 5085
    },
    {
      "entropy": 5.855939483642578,
      "epoch": 0.39603190040848085,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498859323264548,
      "loss": 5.5806,
      "mean_token_accuracy": 0.15516137778759004,
      "num_tokens": 8824802.0,
      "step": 5090
    },
    {
      "entropy": 5.760143566131592,
      "epoch": 0.3964209297801984,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498856534302941,
      "loss": 5.5476,
      "mean_token_accuracy": 0.1627080887556076,
      "num_tokens": 8833348.0,
      "step": 5095
    },
    {
      "entropy": 5.741815710067749,
      "epoch": 0.39680995915191597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988537419446673,
      "loss": 5.5485,
      "mean_token_accuracy": 0.15844069123268129,
      "num_tokens": 8841131.0,
      "step": 5100
    },
    {
      "entropy": 5.7614024639129635,
      "epoch": 0.39719898852363356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988509461897694,
      "loss": 5.5525,
      "mean_token_accuracy": 0.16660502105951308,
      "num_tokens": 8849768.0,
      "step": 5105
    },
    {
      "entropy": 5.814710569381714,
      "epoch": 0.3975880178953511,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988481470382897,
      "loss": 5.532,
      "mean_token_accuracy": 0.1703991949558258,
      "num_tokens": 8858170.0,
      "step": 5110
    },
    {
      "entropy": 5.753843307495117,
      "epoch": 0.3979770472670687,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988453444902708,
      "loss": 5.5563,
      "mean_token_accuracy": 0.1596818044781685,
      "num_tokens": 8866720.0,
      "step": 5115
    },
    {
      "entropy": 5.82450304031372,
      "epoch": 0.3983660766387862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988425385457549,
      "loss": 5.5772,
      "mean_token_accuracy": 0.16017207205295564,
      "num_tokens": 8875141.0,
      "step": 5120
    },
    {
      "entropy": 5.798159074783325,
      "epoch": 0.3987551060105038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988397292047848,
      "loss": 5.5773,
      "mean_token_accuracy": 0.16816945523023605,
      "num_tokens": 8884512.0,
      "step": 5125
    },
    {
      "entropy": 5.8259717464447025,
      "epoch": 0.3991441353822214,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988369164674032,
      "loss": 5.6611,
      "mean_token_accuracy": 0.1584048554301262,
      "num_tokens": 8893487.0,
      "step": 5130
    },
    {
      "entropy": 5.7687300682067875,
      "epoch": 0.3995331647539389,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988341003336526,
      "loss": 5.6013,
      "mean_token_accuracy": 0.1598537892103195,
      "num_tokens": 8902909.0,
      "step": 5135
    },
    {
      "entropy": 5.79081768989563,
      "epoch": 0.3999221941256565,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988312808035757,
      "loss": 5.5125,
      "mean_token_accuracy": 0.16256346851587294,
      "num_tokens": 8911311.0,
      "step": 5140
    },
    {
      "entropy": 5.678632736206055,
      "epoch": 0.400311223497374,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988284578772155,
      "loss": 5.5282,
      "mean_token_accuracy": 0.16317039728164673,
      "num_tokens": 8920222.0,
      "step": 5145
    },
    {
      "entropy": 5.743104028701782,
      "epoch": 0.4007002528690916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988256315546146,
      "loss": 5.5865,
      "mean_token_accuracy": 0.15737401098012924,
      "num_tokens": 8928763.0,
      "step": 5150
    },
    {
      "entropy": 5.911065292358399,
      "epoch": 0.4010892822408092,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988228018358161,
      "loss": 5.6368,
      "mean_token_accuracy": 0.16023626178503036,
      "num_tokens": 8937832.0,
      "step": 5155
    },
    {
      "entropy": 5.719410514831543,
      "epoch": 0.40147831161252673,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988199687208628,
      "loss": 5.4849,
      "mean_token_accuracy": 0.16412659734487534,
      "num_tokens": 8945662.0,
      "step": 5160
    },
    {
      "entropy": 5.727874660491944,
      "epoch": 0.4018673409842443,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988171322097977,
      "loss": 5.5932,
      "mean_token_accuracy": 0.15887476801872252,
      "num_tokens": 8954570.0,
      "step": 5165
    },
    {
      "entropy": 5.81026062965393,
      "epoch": 0.4022563703559619,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988142923026638,
      "loss": 5.5624,
      "mean_token_accuracy": 0.15488934963941575,
      "num_tokens": 8962827.0,
      "step": 5170
    },
    {
      "entropy": 5.819126462936401,
      "epoch": 0.40264539972767943,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988114489995044,
      "loss": 5.5872,
      "mean_token_accuracy": 0.16314277052879333,
      "num_tokens": 8972119.0,
      "step": 5175
    },
    {
      "entropy": 5.736156368255616,
      "epoch": 0.403034429099397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988086023003624,
      "loss": 5.5151,
      "mean_token_accuracy": 0.1629035845398903,
      "num_tokens": 8980498.0,
      "step": 5180
    },
    {
      "entropy": 5.796618366241455,
      "epoch": 0.40342345847111455,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988057522052811,
      "loss": 5.6636,
      "mean_token_accuracy": 0.15330852568149567,
      "num_tokens": 8989819.0,
      "step": 5185
    },
    {
      "entropy": 5.860735940933227,
      "epoch": 0.40381248784283214,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988028987143037,
      "loss": 5.6876,
      "mean_token_accuracy": 0.15596557408571243,
      "num_tokens": 8998884.0,
      "step": 5190
    },
    {
      "entropy": 5.783570384979248,
      "epoch": 0.4042015172145497,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988000418274736,
      "loss": 5.561,
      "mean_token_accuracy": 0.16359727084636688,
      "num_tokens": 9007806.0,
      "step": 5195
    },
    {
      "entropy": 5.853605699539185,
      "epoch": 0.40459054658626725,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987971815448341,
      "loss": 5.7196,
      "mean_token_accuracy": 0.15156864672899245,
      "num_tokens": 9016541.0,
      "step": 5200
    },
    {
      "entropy": 5.733155298233032,
      "epoch": 0.40497957595798484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987943178664285,
      "loss": 5.483,
      "mean_token_accuracy": 0.1652578666806221,
      "num_tokens": 9024503.0,
      "step": 5205
    },
    {
      "entropy": 5.784925317764282,
      "epoch": 0.40536860532970237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987914507923004,
      "loss": 5.5818,
      "mean_token_accuracy": 0.162842258810997,
      "num_tokens": 9033182.0,
      "step": 5210
    },
    {
      "entropy": 5.741500282287598,
      "epoch": 0.40575763470141996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987885803224931,
      "loss": 5.6002,
      "mean_token_accuracy": 0.16271113306283952,
      "num_tokens": 9042281.0,
      "step": 5215
    },
    {
      "entropy": 5.809188795089722,
      "epoch": 0.40614666407313754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987857064570505,
      "loss": 5.5079,
      "mean_token_accuracy": 0.16625353395938874,
      "num_tokens": 9050570.0,
      "step": 5220
    },
    {
      "entropy": 5.843684530258178,
      "epoch": 0.4065356934448551,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987828291960158,
      "loss": 5.6891,
      "mean_token_accuracy": 0.15593319088220597,
      "num_tokens": 9059782.0,
      "step": 5225
    },
    {
      "entropy": 5.702986431121826,
      "epoch": 0.40692472281657266,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498779948539433,
      "loss": 5.4231,
      "mean_token_accuracy": 0.17299046367406845,
      "num_tokens": 9067484.0,
      "step": 5230
    },
    {
      "entropy": 5.778748655319214,
      "epoch": 0.4073137521882902,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987770644873455,
      "loss": 5.6422,
      "mean_token_accuracy": 0.15873562246561052,
      "num_tokens": 9075740.0,
      "step": 5235
    },
    {
      "entropy": 5.8515716075897215,
      "epoch": 0.4077027815600078,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987741770397974,
      "loss": 5.6851,
      "mean_token_accuracy": 0.15533302426338197,
      "num_tokens": 9084137.0,
      "step": 5240
    },
    {
      "entropy": 5.719124126434326,
      "epoch": 0.40809181093172536,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987712861968323,
      "loss": 5.466,
      "mean_token_accuracy": 0.17165637165307998,
      "num_tokens": 9093245.0,
      "step": 5245
    },
    {
      "entropy": 5.678307723999024,
      "epoch": 0.4084808403034429,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498768391958494,
      "loss": 5.4749,
      "mean_token_accuracy": 0.16938251107931138,
      "num_tokens": 9102837.0,
      "step": 5250
    },
    {
      "entropy": 5.755242729187012,
      "epoch": 0.4088698696751605,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987654943248266,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16765681356191636,
      "num_tokens": 9111087.0,
      "step": 5255
    },
    {
      "entropy": 5.720174360275268,
      "epoch": 0.40925889904687807,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987625932958739,
      "loss": 5.5552,
      "mean_token_accuracy": 0.16979336589574814,
      "num_tokens": 9120506.0,
      "step": 5260
    },
    {
      "entropy": 5.8273317337036135,
      "epoch": 0.4096479284185956,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987596888716801,
      "loss": 5.7541,
      "mean_token_accuracy": 0.15222913324832915,
      "num_tokens": 9129637.0,
      "step": 5265
    },
    {
      "entropy": 5.870067453384399,
      "epoch": 0.4100369577903132,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987567810522892,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1620197355747223,
      "num_tokens": 9139338.0,
      "step": 5270
    },
    {
      "entropy": 5.704893350601196,
      "epoch": 0.4104259871620307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987538698377451,
      "loss": 5.5882,
      "mean_token_accuracy": 0.16080360859632492,
      "num_tokens": 9148034.0,
      "step": 5275
    },
    {
      "entropy": 5.789946460723877,
      "epoch": 0.4108150165337483,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987509552280924,
      "loss": 5.5285,
      "mean_token_accuracy": 0.1617719918489456,
      "num_tokens": 9156744.0,
      "step": 5280
    },
    {
      "entropy": 5.697597551345825,
      "epoch": 0.4112040459054659,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498748037223375,
      "loss": 5.5088,
      "mean_token_accuracy": 0.1672073483467102,
      "num_tokens": 9165174.0,
      "step": 5285
    },
    {
      "entropy": 5.749361371994018,
      "epoch": 0.4115930752771834,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987451158236372,
      "loss": 5.5088,
      "mean_token_accuracy": 0.16628395915031433,
      "num_tokens": 9173945.0,
      "step": 5290
    },
    {
      "entropy": 5.78261775970459,
      "epoch": 0.411982104648901,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987421910289234,
      "loss": 5.5233,
      "mean_token_accuracy": 0.16169533431529998,
      "num_tokens": 9182728.0,
      "step": 5295
    },
    {
      "entropy": 5.674267911911011,
      "epoch": 0.41237113402061853,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987392628392781,
      "loss": 5.5634,
      "mean_token_accuracy": 0.15644526481628418,
      "num_tokens": 9191494.0,
      "step": 5300
    },
    {
      "entropy": 5.766342067718506,
      "epoch": 0.4127601633923361,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987363312547456,
      "loss": 5.5669,
      "mean_token_accuracy": 0.15738924741744995,
      "num_tokens": 9199190.0,
      "step": 5305
    },
    {
      "entropy": 5.742343950271606,
      "epoch": 0.4131491927640537,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987333962753703,
      "loss": 5.501,
      "mean_token_accuracy": 0.1643590584397316,
      "num_tokens": 9208066.0,
      "step": 5310
    },
    {
      "entropy": 5.752305126190185,
      "epoch": 0.41353822213577124,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987304579011967,
      "loss": 5.517,
      "mean_token_accuracy": 0.16442365646362306,
      "num_tokens": 9217103.0,
      "step": 5315
    },
    {
      "entropy": 5.727795648574829,
      "epoch": 0.4139272515074888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987275161322697,
      "loss": 5.5545,
      "mean_token_accuracy": 0.15736429989337922,
      "num_tokens": 9225786.0,
      "step": 5320
    },
    {
      "entropy": 5.8252486228942875,
      "epoch": 0.41431628087920636,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987245709686337,
      "loss": 5.5403,
      "mean_token_accuracy": 0.1730848230421543,
      "num_tokens": 9234182.0,
      "step": 5325
    },
    {
      "entropy": 5.727452802658081,
      "epoch": 0.41470531025092394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987216224103334,
      "loss": 5.554,
      "mean_token_accuracy": 0.15562610179185868,
      "num_tokens": 9243844.0,
      "step": 5330
    },
    {
      "entropy": 5.7159261226654055,
      "epoch": 0.41509433962264153,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987186704574136,
      "loss": 5.5149,
      "mean_token_accuracy": 0.16753292083740234,
      "num_tokens": 9252901.0,
      "step": 5335
    },
    {
      "entropy": 5.682818651199341,
      "epoch": 0.41548336899435906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987157151099189,
      "loss": 5.5221,
      "mean_token_accuracy": 0.1638023689389229,
      "num_tokens": 9260947.0,
      "step": 5340
    },
    {
      "entropy": 5.787659597396851,
      "epoch": 0.41587239836607665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987127563678945,
      "loss": 5.5679,
      "mean_token_accuracy": 0.16228248178958893,
      "num_tokens": 9269249.0,
      "step": 5345
    },
    {
      "entropy": 5.746500825881958,
      "epoch": 0.4162614277377942,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987097942313852,
      "loss": 5.574,
      "mean_token_accuracy": 0.16198719888925553,
      "num_tokens": 9277668.0,
      "step": 5350
    },
    {
      "entropy": 5.731139993667602,
      "epoch": 0.41665045710951176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987068287004355,
      "loss": 5.5676,
      "mean_token_accuracy": 0.1581546276807785,
      "num_tokens": 9286000.0,
      "step": 5355
    },
    {
      "entropy": 5.7660356044769285,
      "epoch": 0.41703948648122935,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987038597750909,
      "loss": 5.5993,
      "mean_token_accuracy": 0.15479702204465867,
      "num_tokens": 9295090.0,
      "step": 5360
    },
    {
      "entropy": 5.742945861816406,
      "epoch": 0.4174285158529469,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987008874553964,
      "loss": 5.5078,
      "mean_token_accuracy": 0.1675591364502907,
      "num_tokens": 9302996.0,
      "step": 5365
    },
    {
      "entropy": 5.708066987991333,
      "epoch": 0.41781754522466447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986979117413969,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1674649015069008,
      "num_tokens": 9311636.0,
      "step": 5370
    },
    {
      "entropy": 5.762529182434082,
      "epoch": 0.41820657459638205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986949326331376,
      "loss": 5.5307,
      "mean_token_accuracy": 0.1603675067424774,
      "num_tokens": 9319884.0,
      "step": 5375
    },
    {
      "entropy": 5.669500494003296,
      "epoch": 0.4185956039680996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986919501306638,
      "loss": 5.4745,
      "mean_token_accuracy": 0.16405532956123353,
      "num_tokens": 9328273.0,
      "step": 5380
    },
    {
      "entropy": 5.723462247848511,
      "epoch": 0.41898463333981717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986889642340207,
      "loss": 5.6043,
      "mean_token_accuracy": 0.15512892007827758,
      "num_tokens": 9337696.0,
      "step": 5385
    },
    {
      "entropy": 5.819784021377563,
      "epoch": 0.4193736627115347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986859749432536,
      "loss": 5.5689,
      "mean_token_accuracy": 0.16415231674909592,
      "num_tokens": 9346596.0,
      "step": 5390
    },
    {
      "entropy": 5.7306143283844,
      "epoch": 0.4197626920832523,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986829822584078,
      "loss": 5.5143,
      "mean_token_accuracy": 0.1649942621588707,
      "num_tokens": 9355480.0,
      "step": 5395
    },
    {
      "entropy": 5.755883598327637,
      "epoch": 0.4201517214549699,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986799861795289,
      "loss": 5.557,
      "mean_token_accuracy": 0.16531360894441605,
      "num_tokens": 9364422.0,
      "step": 5400
    },
    {
      "entropy": 5.747479581832886,
      "epoch": 0.4205407508266874,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986769867066622,
      "loss": 5.4381,
      "mean_token_accuracy": 0.15953048020601274,
      "num_tokens": 9372311.0,
      "step": 5405
    },
    {
      "entropy": 5.70312237739563,
      "epoch": 0.420929780198405,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986739838398532,
      "loss": 5.5651,
      "mean_token_accuracy": 0.16256857365369798,
      "num_tokens": 9380936.0,
      "step": 5410
    },
    {
      "entropy": 5.7220171928405765,
      "epoch": 0.4213188095701225,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986709775791475,
      "loss": 5.5093,
      "mean_token_accuracy": 0.16373245269060135,
      "num_tokens": 9390008.0,
      "step": 5415
    },
    {
      "entropy": 5.767958498001098,
      "epoch": 0.4217078389418401,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986679679245907,
      "loss": 5.5894,
      "mean_token_accuracy": 0.16517327278852462,
      "num_tokens": 9398771.0,
      "step": 5420
    },
    {
      "entropy": 5.73075852394104,
      "epoch": 0.4220968683135577,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986649548762286,
      "loss": 5.5865,
      "mean_token_accuracy": 0.16206538826227188,
      "num_tokens": 9407105.0,
      "step": 5425
    },
    {
      "entropy": 5.781380558013916,
      "epoch": 0.4224858976852752,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986619384341066,
      "loss": 5.5738,
      "mean_token_accuracy": 0.16221633404493332,
      "num_tokens": 9415737.0,
      "step": 5430
    },
    {
      "entropy": 5.770466136932373,
      "epoch": 0.4228749270569928,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986589185982708,
      "loss": 5.5114,
      "mean_token_accuracy": 0.1658922851085663,
      "num_tokens": 9424661.0,
      "step": 5435
    },
    {
      "entropy": 5.838792228698731,
      "epoch": 0.42326395642871034,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986558953687671,
      "loss": 5.5888,
      "mean_token_accuracy": 0.15832666158676148,
      "num_tokens": 9433971.0,
      "step": 5440
    },
    {
      "entropy": 5.686177015304565,
      "epoch": 0.42365298580042793,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986528687456409,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1637871339917183,
      "num_tokens": 9442905.0,
      "step": 5445
    },
    {
      "entropy": 5.709370756149292,
      "epoch": 0.4240420151721455,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986498387289384,
      "loss": 5.5646,
      "mean_token_accuracy": 0.16772037446498872,
      "num_tokens": 9451258.0,
      "step": 5450
    },
    {
      "entropy": 5.747209644317627,
      "epoch": 0.42443104454386305,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986468053187058,
      "loss": 5.5211,
      "mean_token_accuracy": 0.16396936625242234,
      "num_tokens": 9460218.0,
      "step": 5455
    },
    {
      "entropy": 5.726138877868652,
      "epoch": 0.42482007391558063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986437685149887,
      "loss": 5.5649,
      "mean_token_accuracy": 0.16548483222723007,
      "num_tokens": 9468756.0,
      "step": 5460
    },
    {
      "entropy": 5.787696599960327,
      "epoch": 0.4252091032872982,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986407283178334,
      "loss": 5.5887,
      "mean_token_accuracy": 0.16122444719076157,
      "num_tokens": 9477622.0,
      "step": 5465
    },
    {
      "entropy": 5.7306321144104,
      "epoch": 0.42559813265901575,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986376847272861,
      "loss": 5.4465,
      "mean_token_accuracy": 0.16420955806970597,
      "num_tokens": 9486332.0,
      "step": 5470
    },
    {
      "entropy": 5.709982776641846,
      "epoch": 0.42598716203073334,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986346377433929,
      "loss": 5.4511,
      "mean_token_accuracy": 0.17362552434206008,
      "num_tokens": 9494703.0,
      "step": 5475
    },
    {
      "entropy": 5.711836528778076,
      "epoch": 0.42637619140245087,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986315873661999,
      "loss": 5.4233,
      "mean_token_accuracy": 0.17868884801864623,
      "num_tokens": 9503008.0,
      "step": 5480
    },
    {
      "entropy": 5.748386526107788,
      "epoch": 0.42676522077416845,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986285335957536,
      "loss": 5.5305,
      "mean_token_accuracy": 0.16153864860534667,
      "num_tokens": 9511752.0,
      "step": 5485
    },
    {
      "entropy": 5.707538223266601,
      "epoch": 0.42715425014588604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986254764321002,
      "loss": 5.4256,
      "mean_token_accuracy": 0.16900402903556824,
      "num_tokens": 9520490.0,
      "step": 5490
    },
    {
      "entropy": 5.716959381103516,
      "epoch": 0.42754327951760357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986224158752861,
      "loss": 5.5117,
      "mean_token_accuracy": 0.16020651012659073,
      "num_tokens": 9528474.0,
      "step": 5495
    },
    {
      "entropy": 5.75875678062439,
      "epoch": 0.42793230888932116,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986193519253578,
      "loss": 5.6025,
      "mean_token_accuracy": 0.15813741609454154,
      "num_tokens": 9537747.0,
      "step": 5500
    },
    {
      "entropy": 5.673283576965332,
      "epoch": 0.4283213382610387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986162845823618,
      "loss": 5.5013,
      "mean_token_accuracy": 0.16056323796510696,
      "num_tokens": 9546414.0,
      "step": 5505
    },
    {
      "entropy": 5.77592282295227,
      "epoch": 0.4287103676327563,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986132138463446,
      "loss": 5.5185,
      "mean_token_accuracy": 0.16403816938400267,
      "num_tokens": 9555166.0,
      "step": 5510
    },
    {
      "entropy": 5.775638008117676,
      "epoch": 0.42909939700447386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986101397173528,
      "loss": 5.577,
      "mean_token_accuracy": 0.16178260147571563,
      "num_tokens": 9563217.0,
      "step": 5515
    },
    {
      "entropy": 5.706945371627808,
      "epoch": 0.4294884263761914,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498607062195433,
      "loss": 5.6213,
      "mean_token_accuracy": 0.15517018884420394,
      "num_tokens": 9572837.0,
      "step": 5520
    },
    {
      "entropy": 5.780698585510254,
      "epoch": 0.429877455747909,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498603981280632,
      "loss": 5.506,
      "mean_token_accuracy": 0.15983346849679947,
      "num_tokens": 9581297.0,
      "step": 5525
    },
    {
      "entropy": 5.702202892303466,
      "epoch": 0.4302664851196265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986008969729964,
      "loss": 5.6021,
      "mean_token_accuracy": 0.15907532423734666,
      "num_tokens": 9590431.0,
      "step": 5530
    },
    {
      "entropy": 5.816613101959229,
      "epoch": 0.4306555144913441,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985978092725731,
      "loss": 5.6454,
      "mean_token_accuracy": 0.15682870224118234,
      "num_tokens": 9598719.0,
      "step": 5535
    },
    {
      "entropy": 5.826610517501831,
      "epoch": 0.4310445438630617,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498594718179409,
      "loss": 5.6334,
      "mean_token_accuracy": 0.15890460461378098,
      "num_tokens": 9606935.0,
      "step": 5540
    },
    {
      "entropy": 5.805395650863647,
      "epoch": 0.4314335732347792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985916236935508,
      "loss": 5.5825,
      "mean_token_accuracy": 0.16289237141609192,
      "num_tokens": 9615104.0,
      "step": 5545
    },
    {
      "entropy": 5.871425437927246,
      "epoch": 0.4318226026064968,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985885258150458,
      "loss": 5.6185,
      "mean_token_accuracy": 0.15472228229045867,
      "num_tokens": 9623332.0,
      "step": 5550
    },
    {
      "entropy": 5.6684393882751465,
      "epoch": 0.43221163197821433,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985854245439408,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1650062769651413,
      "num_tokens": 9631874.0,
      "step": 5555
    },
    {
      "entropy": 5.684298515319824,
      "epoch": 0.4326006613499319,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985823198802827,
      "loss": 5.5069,
      "mean_token_accuracy": 0.16083676666021346,
      "num_tokens": 9641150.0,
      "step": 5560
    },
    {
      "entropy": 5.681491708755493,
      "epoch": 0.4329896907216495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985792118241188,
      "loss": 5.3282,
      "mean_token_accuracy": 0.17461866438388823,
      "num_tokens": 9649116.0,
      "step": 5565
    },
    {
      "entropy": 5.725837230682373,
      "epoch": 0.43337872009336703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985761003754961,
      "loss": 5.5688,
      "mean_token_accuracy": 0.16638418585062026,
      "num_tokens": 9657579.0,
      "step": 5570
    },
    {
      "entropy": 5.729694890975952,
      "epoch": 0.4337677494650846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985729855344622,
      "loss": 5.5355,
      "mean_token_accuracy": 0.16389346569776536,
      "num_tokens": 9666242.0,
      "step": 5575
    },
    {
      "entropy": 5.737802982330322,
      "epoch": 0.4341567788368022,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498569867301064,
      "loss": 5.513,
      "mean_token_accuracy": 0.16292469799518586,
      "num_tokens": 9675196.0,
      "step": 5580
    },
    {
      "entropy": 5.725815010070801,
      "epoch": 0.43454580820851973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985667456753489,
      "loss": 5.4885,
      "mean_token_accuracy": 0.16520630717277526,
      "num_tokens": 9683681.0,
      "step": 5585
    },
    {
      "entropy": 5.718587112426758,
      "epoch": 0.4349348375802373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985636206573642,
      "loss": 5.5569,
      "mean_token_accuracy": 0.16270921379327774,
      "num_tokens": 9692445.0,
      "step": 5590
    },
    {
      "entropy": 5.699607849121094,
      "epoch": 0.43532386695195485,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985604922471575,
      "loss": 5.452,
      "mean_token_accuracy": 0.16585941165685653,
      "num_tokens": 9701256.0,
      "step": 5595
    },
    {
      "entropy": 5.706387710571289,
      "epoch": 0.43571289632367244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985573604447761,
      "loss": 5.4888,
      "mean_token_accuracy": 0.16422429382801057,
      "num_tokens": 9710351.0,
      "step": 5600
    },
    {
      "entropy": 5.788492059707641,
      "epoch": 0.43610192569539,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985542252502676,
      "loss": 5.5675,
      "mean_token_accuracy": 0.1557349070906639,
      "num_tokens": 9718795.0,
      "step": 5605
    },
    {
      "entropy": 5.734406137466431,
      "epoch": 0.43649095506710756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985510866636796,
      "loss": 5.5007,
      "mean_token_accuracy": 0.16339747607707977,
      "num_tokens": 9727136.0,
      "step": 5610
    },
    {
      "entropy": 5.670631456375122,
      "epoch": 0.43687998443882514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985479446850596,
      "loss": 5.4779,
      "mean_token_accuracy": 0.1683027356863022,
      "num_tokens": 9735712.0,
      "step": 5615
    },
    {
      "entropy": 5.76059398651123,
      "epoch": 0.4372690138105427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985447993144554,
      "loss": 5.5077,
      "mean_token_accuracy": 0.16206729263067246,
      "num_tokens": 9744027.0,
      "step": 5620
    },
    {
      "entropy": 5.80002646446228,
      "epoch": 0.43765804318226026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985416505519147,
      "loss": 5.6471,
      "mean_token_accuracy": 0.15733394920825958,
      "num_tokens": 9753083.0,
      "step": 5625
    },
    {
      "entropy": 5.822132205963134,
      "epoch": 0.43804707255397785,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985384983974853,
      "loss": 5.5773,
      "mean_token_accuracy": 0.1626378118991852,
      "num_tokens": 9762098.0,
      "step": 5630
    },
    {
      "entropy": 5.753451156616211,
      "epoch": 0.4384361019256954,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985353428512148,
      "loss": 5.5183,
      "mean_token_accuracy": 0.16592181026935576,
      "num_tokens": 9770956.0,
      "step": 5635
    },
    {
      "entropy": 5.679210948944092,
      "epoch": 0.43882513129741296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985321839131514,
      "loss": 5.4761,
      "mean_token_accuracy": 0.16773835718631744,
      "num_tokens": 9779427.0,
      "step": 5640
    },
    {
      "entropy": 5.724809980392456,
      "epoch": 0.4392141606691305,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985290215833428,
      "loss": 5.5573,
      "mean_token_accuracy": 0.16763297766447066,
      "num_tokens": 9787655.0,
      "step": 5645
    },
    {
      "entropy": 5.731773376464844,
      "epoch": 0.4396031900408481,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985258558618372,
      "loss": 5.4778,
      "mean_token_accuracy": 0.16903508752584456,
      "num_tokens": 9796850.0,
      "step": 5650
    },
    {
      "entropy": 5.732583665847779,
      "epoch": 0.43999221941256567,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985226867486825,
      "loss": 5.5562,
      "mean_token_accuracy": 0.16816309839487076,
      "num_tokens": 9805317.0,
      "step": 5655
    },
    {
      "entropy": 5.713742113113403,
      "epoch": 0.4403812487842832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985195142439267,
      "loss": 5.4973,
      "mean_token_accuracy": 0.16723138391971587,
      "num_tokens": 9814249.0,
      "step": 5660
    },
    {
      "entropy": 5.76673641204834,
      "epoch": 0.4407702781560008,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985163383476181,
      "loss": 5.5877,
      "mean_token_accuracy": 0.15856061577796937,
      "num_tokens": 9823122.0,
      "step": 5665
    },
    {
      "entropy": 5.682684278488159,
      "epoch": 0.44115930752771837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985131590598048,
      "loss": 5.4615,
      "mean_token_accuracy": 0.1660093516111374,
      "num_tokens": 9831470.0,
      "step": 5670
    },
    {
      "entropy": 5.7368285179138185,
      "epoch": 0.4415483368994359,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985099763805352,
      "loss": 5.5529,
      "mean_token_accuracy": 0.1614692747592926,
      "num_tokens": 9840096.0,
      "step": 5675
    },
    {
      "entropy": 5.77333779335022,
      "epoch": 0.4419373662711535,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985067903098574,
      "loss": 5.5176,
      "mean_token_accuracy": 0.16584004759788512,
      "num_tokens": 9848595.0,
      "step": 5680
    },
    {
      "entropy": 5.719666147232056,
      "epoch": 0.442326395642871,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985036008478197,
      "loss": 5.4589,
      "mean_token_accuracy": 0.16027969270944595,
      "num_tokens": 9857353.0,
      "step": 5685
    },
    {
      "entropy": 5.767026472091675,
      "epoch": 0.4427154250145886,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985004079944707,
      "loss": 5.5139,
      "mean_token_accuracy": 0.161882384121418,
      "num_tokens": 9865828.0,
      "step": 5690
    },
    {
      "entropy": 5.661985063552857,
      "epoch": 0.4431044543863062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984972117498587,
      "loss": 5.5389,
      "mean_token_accuracy": 0.16279342472553254,
      "num_tokens": 9874540.0,
      "step": 5695
    },
    {
      "entropy": 5.745433378219604,
      "epoch": 0.4434934837580237,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984940121140323,
      "loss": 5.5281,
      "mean_token_accuracy": 0.15967897921800614,
      "num_tokens": 9883046.0,
      "step": 5700
    },
    {
      "entropy": 5.7201752185821535,
      "epoch": 0.4438825131297413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049849080908704,
      "loss": 5.4766,
      "mean_token_accuracy": 0.16061441525816916,
      "num_tokens": 9892282.0,
      "step": 5705
    },
    {
      "entropy": 5.703311538696289,
      "epoch": 0.44427154250145884,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984876026689302,
      "loss": 5.4107,
      "mean_token_accuracy": 0.17046888768672944,
      "num_tokens": 9900653.0,
      "step": 5710
    },
    {
      "entropy": 5.66222562789917,
      "epoch": 0.4446605718731764,
      "grad_norm": 1.0,
      "learning_rate": 0.000498484392859752,
      "loss": 5.4283,
      "mean_token_accuracy": 0.16977153420448304,
      "num_tokens": 9909323.0,
      "step": 5715
    },
    {
      "entropy": 5.7286176681518555,
      "epoch": 0.445049601244894,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984811796595538,
      "loss": 5.5813,
      "mean_token_accuracy": 0.1572989210486412,
      "num_tokens": 9918155.0,
      "step": 5720
    },
    {
      "entropy": 5.8536262035369875,
      "epoch": 0.44543863061661154,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984779630683843,
      "loss": 5.5236,
      "mean_token_accuracy": 0.1601708397269249,
      "num_tokens": 9925977.0,
      "step": 5725
    },
    {
      "entropy": 5.723394155502319,
      "epoch": 0.44582765998832913,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984747430862924,
      "loss": 5.53,
      "mean_token_accuracy": 0.16191367208957672,
      "num_tokens": 9934606.0,
      "step": 5730
    },
    {
      "entropy": 5.670947790145874,
      "epoch": 0.44621668936004666,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984715197133271,
      "loss": 5.4895,
      "mean_token_accuracy": 0.16341402679681777,
      "num_tokens": 9943638.0,
      "step": 5735
    },
    {
      "entropy": 5.7193292617797855,
      "epoch": 0.44660571873176425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984682929495371,
      "loss": 5.5029,
      "mean_token_accuracy": 0.16260228008031846,
      "num_tokens": 9952686.0,
      "step": 5740
    },
    {
      "entropy": 5.7725379943847654,
      "epoch": 0.44699474810348183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984650627949715,
      "loss": 5.6854,
      "mean_token_accuracy": 0.14949709177017212,
      "num_tokens": 9961947.0,
      "step": 5745
    },
    {
      "entropy": 5.776320934295654,
      "epoch": 0.44738377747519936,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984618292496792,
      "loss": 5.5191,
      "mean_token_accuracy": 0.16245160549879073,
      "num_tokens": 9971218.0,
      "step": 5750
    },
    {
      "entropy": 5.731330871582031,
      "epoch": 0.44777280684691695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984585923137093,
      "loss": 5.537,
      "mean_token_accuracy": 0.16474678367376328,
      "num_tokens": 9980741.0,
      "step": 5755
    },
    {
      "entropy": 5.764980030059815,
      "epoch": 0.4481618362186345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498455351987111,
      "loss": 5.5275,
      "mean_token_accuracy": 0.16332996785640716,
      "num_tokens": 9989483.0,
      "step": 5760
    },
    {
      "entropy": 5.809801721572876,
      "epoch": 0.44855086559035207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984521082699333,
      "loss": 5.5421,
      "mean_token_accuracy": 0.16016984283924102,
      "num_tokens": 9998491.0,
      "step": 5765
    },
    {
      "entropy": 5.631897878646851,
      "epoch": 0.44893989496206965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984488611622256,
      "loss": 5.4374,
      "mean_token_accuracy": 0.17185879796743392,
      "num_tokens": 10006276.0,
      "step": 5770
    },
    {
      "entropy": 5.660624313354492,
      "epoch": 0.4493289243337872,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498445610664037,
      "loss": 5.5184,
      "mean_token_accuracy": 0.16512587666511536,
      "num_tokens": 10015088.0,
      "step": 5775
    },
    {
      "entropy": 5.750400733947754,
      "epoch": 0.44971795370550477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498442356775417,
      "loss": 5.4774,
      "mean_token_accuracy": 0.16268350481986998,
      "num_tokens": 10023305.0,
      "step": 5780
    },
    {
      "entropy": 5.666444969177246,
      "epoch": 0.45010698307722236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984390994964148,
      "loss": 5.4361,
      "mean_token_accuracy": 0.16807707846164704,
      "num_tokens": 10031540.0,
      "step": 5785
    },
    {
      "entropy": 5.7406463623046875,
      "epoch": 0.4504960124489399,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049843583882708,
      "loss": 5.5196,
      "mean_token_accuracy": 0.16563237011432647,
      "num_tokens": 10040025.0,
      "step": 5790
    },
    {
      "entropy": 5.720093536376953,
      "epoch": 0.4508850418206575,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498432574767462,
      "loss": 5.5349,
      "mean_token_accuracy": 0.1660214349627495,
      "num_tokens": 10048904.0,
      "step": 5795
    },
    {
      "entropy": 5.714317417144775,
      "epoch": 0.451274071192375,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984293073176103,
      "loss": 5.5693,
      "mean_token_accuracy": 0.1600137695670128,
      "num_tokens": 10057608.0,
      "step": 5800
    },
    {
      "entropy": 5.788346672058106,
      "epoch": 0.4516631005640926,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984260364775744,
      "loss": 5.5313,
      "mean_token_accuracy": 0.16886440366506578,
      "num_tokens": 10066587.0,
      "step": 5805
    },
    {
      "entropy": 5.68289008140564,
      "epoch": 0.4520521299358102,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498422762247404,
      "loss": 5.4926,
      "mean_token_accuracy": 0.1670097067952156,
      "num_tokens": 10075651.0,
      "step": 5810
    },
    {
      "entropy": 5.740344285964966,
      "epoch": 0.4524411593075277,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984194846271489,
      "loss": 5.6072,
      "mean_token_accuracy": 0.1580710768699646,
      "num_tokens": 10084158.0,
      "step": 5815
    },
    {
      "entropy": 5.800893640518188,
      "epoch": 0.4528301886792453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984162036168586,
      "loss": 5.5345,
      "mean_token_accuracy": 0.15725137144327164,
      "num_tokens": 10093730.0,
      "step": 5820
    },
    {
      "entropy": 5.7457794666290285,
      "epoch": 0.4532192180509628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984129192165831,
      "loss": 5.4392,
      "mean_token_accuracy": 0.1610160768032074,
      "num_tokens": 10102457.0,
      "step": 5825
    },
    {
      "entropy": 5.739707565307617,
      "epoch": 0.4536082474226804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984096314263722,
      "loss": 5.599,
      "mean_token_accuracy": 0.16421133130788804,
      "num_tokens": 10111188.0,
      "step": 5830
    },
    {
      "entropy": 5.763910245895386,
      "epoch": 0.453997276794398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984063402462757,
      "loss": 5.4815,
      "mean_token_accuracy": 0.17231282591819763,
      "num_tokens": 10120268.0,
      "step": 5835
    },
    {
      "entropy": 5.737266874313354,
      "epoch": 0.45438630616611553,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984030456763435,
      "loss": 5.5473,
      "mean_token_accuracy": 0.1597852438688278,
      "num_tokens": 10129262.0,
      "step": 5840
    },
    {
      "entropy": 5.682384490966797,
      "epoch": 0.4547753355378331,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983997477166257,
      "loss": 5.4998,
      "mean_token_accuracy": 0.1659440889954567,
      "num_tokens": 10137417.0,
      "step": 5845
    },
    {
      "entropy": 5.676105165481568,
      "epoch": 0.45516436490955064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983964463671723,
      "loss": 5.4386,
      "mean_token_accuracy": 0.16971610635519027,
      "num_tokens": 10145804.0,
      "step": 5850
    },
    {
      "entropy": 5.75293116569519,
      "epoch": 0.45555339428126823,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983931416280334,
      "loss": 5.575,
      "mean_token_accuracy": 0.16000862717628478,
      "num_tokens": 10154421.0,
      "step": 5855
    },
    {
      "entropy": 5.835421419143676,
      "epoch": 0.4559424236529858,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983898334992591,
      "loss": 5.5716,
      "mean_token_accuracy": 0.1663193479180336,
      "num_tokens": 10163631.0,
      "step": 5860
    },
    {
      "entropy": 5.737528228759766,
      "epoch": 0.45633145302470335,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983865219808998,
      "loss": 5.5026,
      "mean_token_accuracy": 0.16833841055631638,
      "num_tokens": 10172926.0,
      "step": 5865
    },
    {
      "entropy": 5.776432943344116,
      "epoch": 0.45672048239642093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004983832070730055,
      "loss": 5.6064,
      "mean_token_accuracy": 0.15432555228471756,
      "num_tokens": 10181671.0,
      "step": 5870
    },
    {
      "entropy": 5.67692461013794,
      "epoch": 0.4571095117681385,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983798887756265,
      "loss": 5.417,
      "mean_token_accuracy": 0.16950175762176514,
      "num_tokens": 10189626.0,
      "step": 5875
    },
    {
      "entropy": 5.775591325759888,
      "epoch": 0.45749854113985605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983765670888133,
      "loss": 5.58,
      "mean_token_accuracy": 0.16477911323308944,
      "num_tokens": 10198212.0,
      "step": 5880
    },
    {
      "entropy": 5.660595369338989,
      "epoch": 0.45788757051157364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983732420126163,
      "loss": 5.4723,
      "mean_token_accuracy": 0.16660816818475724,
      "num_tokens": 10207342.0,
      "step": 5885
    },
    {
      "entropy": 5.769796133041382,
      "epoch": 0.45827659988329117,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983699135470858,
      "loss": 5.5776,
      "mean_token_accuracy": 0.15508492141962052,
      "num_tokens": 10216413.0,
      "step": 5890
    },
    {
      "entropy": 5.741604661941528,
      "epoch": 0.45866562925500876,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983665816922723,
      "loss": 5.5158,
      "mean_token_accuracy": 0.16717685610055924,
      "num_tokens": 10225889.0,
      "step": 5895
    },
    {
      "entropy": 5.685088586807251,
      "epoch": 0.45905465862672634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983632464482267,
      "loss": 5.4684,
      "mean_token_accuracy": 0.16811624318361282,
      "num_tokens": 10234518.0,
      "step": 5900
    },
    {
      "entropy": 5.7592181205749515,
      "epoch": 0.4594436879984439,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983599078149991,
      "loss": 5.5918,
      "mean_token_accuracy": 0.15640947371721267,
      "num_tokens": 10243074.0,
      "step": 5905
    },
    {
      "entropy": 5.848401737213135,
      "epoch": 0.45983271737016146,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983565657926405,
      "loss": 5.5417,
      "mean_token_accuracy": 0.15850908905267716,
      "num_tokens": 10253473.0,
      "step": 5910
    },
    {
      "entropy": 5.770726490020752,
      "epoch": 0.460221746741879,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983532203812017,
      "loss": 5.618,
      "mean_token_accuracy": 0.16019521355628968,
      "num_tokens": 10261866.0,
      "step": 5915
    },
    {
      "entropy": 5.73525595664978,
      "epoch": 0.4606107761135966,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983498715807331,
      "loss": 5.4628,
      "mean_token_accuracy": 0.1711142748594284,
      "num_tokens": 10270350.0,
      "step": 5920
    },
    {
      "entropy": 5.749798345565796,
      "epoch": 0.46099980548531416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983465193912857,
      "loss": 5.6162,
      "mean_token_accuracy": 0.15639185309410095,
      "num_tokens": 10279556.0,
      "step": 5925
    },
    {
      "entropy": 5.654877853393555,
      "epoch": 0.4613888348570317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983431638129104,
      "loss": 5.3936,
      "mean_token_accuracy": 0.1686894953250885,
      "num_tokens": 10287489.0,
      "step": 5930
    },
    {
      "entropy": 5.668270206451416,
      "epoch": 0.4617778642287493,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983398048456581,
      "loss": 5.376,
      "mean_token_accuracy": 0.16921777576208114,
      "num_tokens": 10296280.0,
      "step": 5935
    },
    {
      "entropy": 5.697940444946289,
      "epoch": 0.4621668936004668,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983364424895796,
      "loss": 5.5136,
      "mean_token_accuracy": 0.15967504531145096,
      "num_tokens": 10304091.0,
      "step": 5940
    },
    {
      "entropy": 5.742635869979859,
      "epoch": 0.4625559229721844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983330767447262,
      "loss": 5.5646,
      "mean_token_accuracy": 0.1578671395778656,
      "num_tokens": 10313058.0,
      "step": 5945
    },
    {
      "entropy": 5.674269199371338,
      "epoch": 0.462944952343902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983297076111489,
      "loss": 5.5253,
      "mean_token_accuracy": 0.1637215420603752,
      "num_tokens": 10321642.0,
      "step": 5950
    },
    {
      "entropy": 5.679966878890991,
      "epoch": 0.4633339817156195,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983263350888987,
      "loss": 5.3551,
      "mean_token_accuracy": 0.17202294170856475,
      "num_tokens": 10329829.0,
      "step": 5955
    },
    {
      "entropy": 5.724870681762695,
      "epoch": 0.4637230110873371,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983229591780268,
      "loss": 5.5815,
      "mean_token_accuracy": 0.15950696021318436,
      "num_tokens": 10338182.0,
      "step": 5960
    },
    {
      "entropy": 5.75362229347229,
      "epoch": 0.46411204045905463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983195798785844,
      "loss": 5.4762,
      "mean_token_accuracy": 0.16557322591543197,
      "num_tokens": 10346377.0,
      "step": 5965
    },
    {
      "entropy": 5.720754766464234,
      "epoch": 0.4645010698307722,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983161971906228,
      "loss": 5.4618,
      "mean_token_accuracy": 0.16692903488874436,
      "num_tokens": 10354937.0,
      "step": 5970
    },
    {
      "entropy": 5.627158451080322,
      "epoch": 0.4648900992024898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983128111141935,
      "loss": 5.3571,
      "mean_token_accuracy": 0.17103014439344405,
      "num_tokens": 10363706.0,
      "step": 5975
    },
    {
      "entropy": 5.66728196144104,
      "epoch": 0.46527912857420733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983094216493475,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1635490983724594,
      "num_tokens": 10372292.0,
      "step": 5980
    },
    {
      "entropy": 5.715864467620849,
      "epoch": 0.4656681579459249,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983060287961367,
      "loss": 5.4854,
      "mean_token_accuracy": 0.17231816947460174,
      "num_tokens": 10381260.0,
      "step": 5985
    },
    {
      "entropy": 5.7313025951385494,
      "epoch": 0.4660571873176425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983026325546123,
      "loss": 5.5041,
      "mean_token_accuracy": 0.16354109942913056,
      "num_tokens": 10390646.0,
      "step": 5990
    },
    {
      "entropy": 5.736316871643067,
      "epoch": 0.46644621668936004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982992329248257,
      "loss": 5.576,
      "mean_token_accuracy": 0.1609186977148056,
      "num_tokens": 10399056.0,
      "step": 5995
    },
    {
      "entropy": 5.710217094421386,
      "epoch": 0.4668352460610776,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982958299068289,
      "loss": 5.5362,
      "mean_token_accuracy": 0.16015388816595078,
      "num_tokens": 10408746.0,
      "step": 6000
    },
    {
      "epoch": 0.4668352460610776,
      "eval_entropy": 5.515557648288848,
      "eval_loss": 5.536764144897461,
      "eval_mean_token_accuracy": 0.1716840387071923,
      "eval_num_tokens": 10408746.0,
      "eval_runtime": 21.9423,
      "eval_samples_per_second": 1893.971,
      "eval_steps_per_second": 236.758,
      "step": 6000
    }
  ],
  "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": 2095049433415680.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}