{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1705199492732417,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 8.189694690704346,
      "epoch": 0.0009755145839430299,
      "grad_norm": 1.1171875,
      "learning_rate": 2e-06,
      "loss": 8.1213,
      "mean_token_accuracy": 0.0928520105779171,
      "num_tokens": 17120.0,
      "step": 5
    },
    {
      "entropy": 8.088421869277955,
      "epoch": 0.0019510291678860598,
      "grad_norm": 1.328125,
      "learning_rate": 4.5e-06,
      "loss": 8.0338,
      "mean_token_accuracy": 0.09805552884936333,
      "num_tokens": 34182.0,
      "step": 10
    },
    {
      "entropy": 8.10444598197937,
      "epoch": 0.0029265437518290896,
      "grad_norm": 1.0625,
      "learning_rate": 7e-06,
      "loss": 8.0761,
      "mean_token_accuracy": 0.0990721695125103,
      "num_tokens": 51205.0,
      "step": 15
    },
    {
      "entropy": 8.136365795135498,
      "epoch": 0.0039020583357721197,
      "grad_norm": 0.9453125,
      "learning_rate": 9.5e-06,
      "loss": 8.1057,
      "mean_token_accuracy": 0.09148216098546982,
      "num_tokens": 67722.0,
      "step": 20
    },
    {
      "entropy": 8.127794647216797,
      "epoch": 0.00487757291971515,
      "grad_norm": 1.0078125,
      "learning_rate": 1.2e-05,
      "loss": 8.1015,
      "mean_token_accuracy": 0.09162097275257111,
      "num_tokens": 84230.0,
      "step": 25
    },
    {
      "entropy": 8.140947341918945,
      "epoch": 0.005853087503658179,
      "grad_norm": 0.82421875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 8.0101,
      "mean_token_accuracy": 0.10145255476236344,
      "num_tokens": 100967.0,
      "step": 30
    },
    {
      "entropy": 8.193094825744629,
      "epoch": 0.00682860208760121,
      "grad_norm": 0.88671875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 8.0798,
      "mean_token_accuracy": 0.09109266102313995,
      "num_tokens": 117450.0,
      "step": 35
    },
    {
      "entropy": 8.167135286331177,
      "epoch": 0.007804116671544239,
      "grad_norm": 0.8671875,
      "learning_rate": 1.95e-05,
      "loss": 7.9719,
      "mean_token_accuracy": 0.10474523082375527,
      "num_tokens": 135328.0,
      "step": 40
    },
    {
      "entropy": 8.237839794158935,
      "epoch": 0.00877963125548727,
      "grad_norm": 0.87890625,
      "learning_rate": 2.2e-05,
      "loss": 8.0049,
      "mean_token_accuracy": 0.09647766500711441,
      "num_tokens": 152571.0,
      "step": 45
    },
    {
      "entropy": 8.21173391342163,
      "epoch": 0.0097551458394303,
      "grad_norm": 0.91015625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.987,
      "mean_token_accuracy": 0.09857306927442551,
      "num_tokens": 170216.0,
      "step": 50
    },
    {
      "entropy": 8.188056182861327,
      "epoch": 0.010730660423373329,
      "grad_norm": 0.828125,
      "learning_rate": 2.7e-05,
      "loss": 8.0418,
      "mean_token_accuracy": 0.09847658723592759,
      "num_tokens": 186513.0,
      "step": 55
    },
    {
      "entropy": 8.209293651580811,
      "epoch": 0.011706175007316359,
      "grad_norm": 1.015625,
      "learning_rate": 2.95e-05,
      "loss": 8.0308,
      "mean_token_accuracy": 0.09393074885010719,
      "num_tokens": 202600.0,
      "step": 60
    },
    {
      "entropy": 8.16935420036316,
      "epoch": 0.01268168959125939,
      "grad_norm": 1.3125,
      "learning_rate": 3.2e-05,
      "loss": 7.9625,
      "mean_token_accuracy": 0.10780214741826058,
      "num_tokens": 220270.0,
      "step": 65
    },
    {
      "entropy": 8.189718246459961,
      "epoch": 0.01365720417520242,
      "grad_norm": 0.8203125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 7.9753,
      "mean_token_accuracy": 0.10615965947508813,
      "num_tokens": 236810.0,
      "step": 70
    },
    {
      "entropy": 8.23555679321289,
      "epoch": 0.014632718759145449,
      "grad_norm": 0.734375,
      "learning_rate": 3.7e-05,
      "loss": 8.0265,
      "mean_token_accuracy": 0.09320793077349662,
      "num_tokens": 252559.0,
      "step": 75
    },
    {
      "entropy": 8.191197776794434,
      "epoch": 0.015608233343088479,
      "grad_norm": 0.87890625,
      "learning_rate": 3.95e-05,
      "loss": 7.9595,
      "mean_token_accuracy": 0.10167472660541535,
      "num_tokens": 268507.0,
      "step": 80
    },
    {
      "entropy": 8.220345592498779,
      "epoch": 0.01658374792703151,
      "grad_norm": 0.79296875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 7.9799,
      "mean_token_accuracy": 0.09989922642707824,
      "num_tokens": 284776.0,
      "step": 85
    },
    {
      "entropy": 8.225536727905274,
      "epoch": 0.01755926251097454,
      "grad_norm": 0.9296875,
      "learning_rate": 4.45e-05,
      "loss": 8.0477,
      "mean_token_accuracy": 0.09435278922319412,
      "num_tokens": 300937.0,
      "step": 90
    },
    {
      "entropy": 8.22456979751587,
      "epoch": 0.01853477709491757,
      "grad_norm": 0.875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.0291,
      "mean_token_accuracy": 0.09500815868377685,
      "num_tokens": 317464.0,
      "step": 95
    },
    {
      "entropy": 8.196868324279786,
      "epoch": 0.0195102916788606,
      "grad_norm": 0.91796875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.9297,
      "mean_token_accuracy": 0.10513664558529853,
      "num_tokens": 334022.0,
      "step": 100
    },
    {
      "entropy": 8.174358654022218,
      "epoch": 0.02048580626280363,
      "grad_norm": 0.87109375,
      "learning_rate": 5.2e-05,
      "loss": 8.012,
      "mean_token_accuracy": 0.09847316145896912,
      "num_tokens": 350545.0,
      "step": 105
    },
    {
      "entropy": 8.142373085021973,
      "epoch": 0.021461320846746658,
      "grad_norm": 0.83984375,
      "learning_rate": 5.45e-05,
      "loss": 7.8485,
      "mean_token_accuracy": 0.11419763416051865,
      "num_tokens": 368878.0,
      "step": 110
    },
    {
      "entropy": 8.165599918365478,
      "epoch": 0.022436835430689687,
      "grad_norm": 0.86328125,
      "learning_rate": 5.7e-05,
      "loss": 8.0301,
      "mean_token_accuracy": 0.09600666835904122,
      "num_tokens": 388381.0,
      "step": 115
    },
    {
      "entropy": 8.214081382751464,
      "epoch": 0.023412350014632717,
      "grad_norm": 0.92578125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.048,
      "mean_token_accuracy": 0.09097032994031906,
      "num_tokens": 403727.0,
      "step": 120
    },
    {
      "entropy": 8.2991042137146,
      "epoch": 0.02438786459857575,
      "grad_norm": 0.921875,
      "learning_rate": 6.2e-05,
      "loss": 8.0491,
      "mean_token_accuracy": 0.0942738875746727,
      "num_tokens": 419046.0,
      "step": 125
    },
    {
      "entropy": 8.182039546966553,
      "epoch": 0.02536337918251878,
      "grad_norm": 1.640625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.9433,
      "mean_token_accuracy": 0.10728831514716149,
      "num_tokens": 435186.0,
      "step": 130
    },
    {
      "entropy": 8.232324981689453,
      "epoch": 0.02633889376646181,
      "grad_norm": 0.953125,
      "learning_rate": 6.7e-05,
      "loss": 8.0395,
      "mean_token_accuracy": 0.09328199326992034,
      "num_tokens": 451552.0,
      "step": 135
    },
    {
      "entropy": 8.199801969528199,
      "epoch": 0.02731440835040484,
      "grad_norm": 1.3125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.0061,
      "mean_token_accuracy": 0.09671647101640701,
      "num_tokens": 468006.0,
      "step": 140
    },
    {
      "entropy": 8.274113845825195,
      "epoch": 0.02828992293434787,
      "grad_norm": 0.984375,
      "learning_rate": 7.2e-05,
      "loss": 8.0112,
      "mean_token_accuracy": 0.09868691936135292,
      "num_tokens": 484381.0,
      "step": 145
    },
    {
      "entropy": 8.25528917312622,
      "epoch": 0.029265437518290898,
      "grad_norm": 1.421875,
      "learning_rate": 7.45e-05,
      "loss": 8.0257,
      "mean_token_accuracy": 0.0955542728304863,
      "num_tokens": 501309.0,
      "step": 150
    },
    {
      "entropy": 8.226487350463866,
      "epoch": 0.030240952102233928,
      "grad_norm": 0.99609375,
      "learning_rate": 7.7e-05,
      "loss": 7.9828,
      "mean_token_accuracy": 0.09746505990624428,
      "num_tokens": 518290.0,
      "step": 155
    },
    {
      "entropy": 8.181639671325684,
      "epoch": 0.031216466686176957,
      "grad_norm": 0.828125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.9965,
      "mean_token_accuracy": 0.09464227110147476,
      "num_tokens": 533700.0,
      "step": 160
    },
    {
      "entropy": 8.192685508728028,
      "epoch": 0.03219198127011999,
      "grad_norm": 1.0078125,
      "learning_rate": 8.2e-05,
      "loss": 7.8837,
      "mean_token_accuracy": 0.10680020377039909,
      "num_tokens": 548935.0,
      "step": 165
    },
    {
      "entropy": 8.122217988967895,
      "epoch": 0.03316749585406302,
      "grad_norm": 0.78515625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.9809,
      "mean_token_accuracy": 0.09729720130562783,
      "num_tokens": 565869.0,
      "step": 170
    },
    {
      "entropy": 8.255708599090577,
      "epoch": 0.03414301043800605,
      "grad_norm": 0.921875,
      "learning_rate": 8.7e-05,
      "loss": 7.9426,
      "mean_token_accuracy": 0.10057656094431877,
      "num_tokens": 583065.0,
      "step": 175
    },
    {
      "entropy": 8.147002172470092,
      "epoch": 0.03511852502194908,
      "grad_norm": 0.8359375,
      "learning_rate": 8.95e-05,
      "loss": 7.9597,
      "mean_token_accuracy": 0.10024290233850479,
      "num_tokens": 599875.0,
      "step": 180
    },
    {
      "entropy": 8.198067951202393,
      "epoch": 0.03609403960589211,
      "grad_norm": 0.890625,
      "learning_rate": 9.2e-05,
      "loss": 7.9192,
      "mean_token_accuracy": 0.1043715499341488,
      "num_tokens": 616296.0,
      "step": 185
    },
    {
      "entropy": 8.178347730636597,
      "epoch": 0.03706955418983514,
      "grad_norm": 0.8828125,
      "learning_rate": 9.45e-05,
      "loss": 7.8936,
      "mean_token_accuracy": 0.0991103172302246,
      "num_tokens": 632548.0,
      "step": 190
    },
    {
      "entropy": 8.182952737808227,
      "epoch": 0.03804506877377817,
      "grad_norm": 0.921875,
      "learning_rate": 9.7e-05,
      "loss": 7.9026,
      "mean_token_accuracy": 0.10161520764231682,
      "num_tokens": 650395.0,
      "step": 195
    },
    {
      "entropy": 8.191309070587158,
      "epoch": 0.0390205833577212,
      "grad_norm": 0.95703125,
      "learning_rate": 9.95e-05,
      "loss": 7.9511,
      "mean_token_accuracy": 0.10279864966869354,
      "num_tokens": 668453.0,
      "step": 200
    },
    {
      "entropy": 8.106355381011962,
      "epoch": 0.03999609794166423,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000102,
      "loss": 7.822,
      "mean_token_accuracy": 0.11003539934754372,
      "num_tokens": 683914.0,
      "step": 205
    },
    {
      "entropy": 8.108619880676269,
      "epoch": 0.04097161252560726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.9833,
      "mean_token_accuracy": 0.09929522946476936,
      "num_tokens": 700480.0,
      "step": 210
    },
    {
      "entropy": 8.311169815063476,
      "epoch": 0.041947127109550286,
      "grad_norm": 0.96875,
      "learning_rate": 0.000107,
      "loss": 7.9705,
      "mean_token_accuracy": 0.0956616073846817,
      "num_tokens": 716604.0,
      "step": 215
    },
    {
      "entropy": 8.151149940490722,
      "epoch": 0.042922641693493316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001095,
      "loss": 7.9283,
      "mean_token_accuracy": 0.09996144995093345,
      "num_tokens": 732697.0,
      "step": 220
    },
    {
      "entropy": 8.261075973510742,
      "epoch": 0.043898156277436345,
      "grad_norm": 0.953125,
      "learning_rate": 0.000112,
      "loss": 8.0403,
      "mean_token_accuracy": 0.09186563193798065,
      "num_tokens": 749072.0,
      "step": 225
    },
    {
      "entropy": 8.180721092224122,
      "epoch": 0.044873670861379375,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0001145,
      "loss": 7.9417,
      "mean_token_accuracy": 0.09489826485514641,
      "num_tokens": 767826.0,
      "step": 230
    },
    {
      "entropy": 8.116961860656739,
      "epoch": 0.045849185445322405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.8815,
      "mean_token_accuracy": 0.101999169588089,
      "num_tokens": 786863.0,
      "step": 235
    },
    {
      "entropy": 8.151587963104248,
      "epoch": 0.046824700029265434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.8594,
      "mean_token_accuracy": 0.11106476187705994,
      "num_tokens": 803925.0,
      "step": 240
    },
    {
      "entropy": 8.16553840637207,
      "epoch": 0.04780021461320847,
      "grad_norm": 1.046875,
      "learning_rate": 0.000122,
      "loss": 7.9056,
      "mean_token_accuracy": 0.09941518381237983,
      "num_tokens": 822088.0,
      "step": 245
    },
    {
      "entropy": 8.204985809326171,
      "epoch": 0.0487757291971515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0001245,
      "loss": 7.9578,
      "mean_token_accuracy": 0.09502575322985649,
      "num_tokens": 837991.0,
      "step": 250
    },
    {
      "entropy": 8.094996738433839,
      "epoch": 0.04975124378109453,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000127,
      "loss": 7.8384,
      "mean_token_accuracy": 0.11150547713041306,
      "num_tokens": 855593.0,
      "step": 255
    },
    {
      "entropy": 8.22648229598999,
      "epoch": 0.05072675836503756,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0001295,
      "loss": 7.8555,
      "mean_token_accuracy": 0.1016038067638874,
      "num_tokens": 873696.0,
      "step": 260
    },
    {
      "entropy": 8.093446922302245,
      "epoch": 0.05170227294898059,
      "grad_norm": 1.015625,
      "learning_rate": 0.000132,
      "loss": 7.7945,
      "mean_token_accuracy": 0.1079544596374035,
      "num_tokens": 890023.0,
      "step": 265
    },
    {
      "entropy": 8.219519519805909,
      "epoch": 0.05267778753292362,
      "grad_norm": 1.09375,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.0193,
      "mean_token_accuracy": 0.09408034533262252,
      "num_tokens": 906492.0,
      "step": 270
    },
    {
      "entropy": 8.144463109970093,
      "epoch": 0.05365330211686665,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.8985,
      "mean_token_accuracy": 0.09895867183804512,
      "num_tokens": 921837.0,
      "step": 275
    },
    {
      "entropy": 8.176286315917968,
      "epoch": 0.05462881670080968,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001395,
      "loss": 7.797,
      "mean_token_accuracy": 0.10809003040194512,
      "num_tokens": 938771.0,
      "step": 280
    },
    {
      "entropy": 8.145763206481934,
      "epoch": 0.05560433128475271,
      "grad_norm": 0.96875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.9095,
      "mean_token_accuracy": 0.09854145273566246,
      "num_tokens": 954562.0,
      "step": 285
    },
    {
      "entropy": 8.14554295539856,
      "epoch": 0.05657984586869574,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001445,
      "loss": 7.8002,
      "mean_token_accuracy": 0.10630740001797676,
      "num_tokens": 973700.0,
      "step": 290
    },
    {
      "entropy": 8.05322551727295,
      "epoch": 0.05755536045263877,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000147,
      "loss": 7.9073,
      "mean_token_accuracy": 0.09839207902550698,
      "num_tokens": 990688.0,
      "step": 295
    },
    {
      "entropy": 8.224873065948486,
      "epoch": 0.058530875036581796,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0001495,
      "loss": 8.0062,
      "mean_token_accuracy": 0.09405675381422043,
      "num_tokens": 1008144.0,
      "step": 300
    },
    {
      "entropy": 8.18083577156067,
      "epoch": 0.059506389620524826,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000152,
      "loss": 7.9383,
      "mean_token_accuracy": 0.09709960296750068,
      "num_tokens": 1027379.0,
      "step": 305
    },
    {
      "entropy": 8.178252029418946,
      "epoch": 0.060481904204467855,
      "grad_norm": 0.984375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.8793,
      "mean_token_accuracy": 0.10311919152736664,
      "num_tokens": 1043999.0,
      "step": 310
    },
    {
      "entropy": 8.098154592514039,
      "epoch": 0.061457418788410885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000157,
      "loss": 7.9274,
      "mean_token_accuracy": 0.09602248296141624,
      "num_tokens": 1062237.0,
      "step": 315
    },
    {
      "entropy": 8.244752025604248,
      "epoch": 0.062432933372353915,
      "grad_norm": 1.734375,
      "learning_rate": 0.0001595,
      "loss": 7.8576,
      "mean_token_accuracy": 0.10373107567429543,
      "num_tokens": 1078092.0,
      "step": 320
    },
    {
      "entropy": 8.097948360443116,
      "epoch": 0.06340844795629695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000162,
      "loss": 7.8987,
      "mean_token_accuracy": 0.10294966548681259,
      "num_tokens": 1095124.0,
      "step": 325
    },
    {
      "entropy": 8.17255973815918,
      "epoch": 0.06438396254023998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.8855,
      "mean_token_accuracy": 0.09729674383997917,
      "num_tokens": 1112775.0,
      "step": 330
    },
    {
      "entropy": 8.077517557144166,
      "epoch": 0.06535947712418301,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.7933,
      "mean_token_accuracy": 0.10203502476215362,
      "num_tokens": 1129726.0,
      "step": 335
    },
    {
      "entropy": 8.185189342498779,
      "epoch": 0.06633499170812604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.8883,
      "mean_token_accuracy": 0.10449600592255592,
      "num_tokens": 1144034.0,
      "step": 340
    },
    {
      "entropy": 8.129604816436768,
      "epoch": 0.06731050629206907,
      "grad_norm": 0.9375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.8179,
      "mean_token_accuracy": 0.10270989686250687,
      "num_tokens": 1159633.0,
      "step": 345
    },
    {
      "entropy": 8.0196063041687,
      "epoch": 0.0682860208760121,
      "grad_norm": 1.015625,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.8085,
      "mean_token_accuracy": 0.0996285319328308,
      "num_tokens": 1175350.0,
      "step": 350
    },
    {
      "entropy": 8.027859020233155,
      "epoch": 0.06926153545995513,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000177,
      "loss": 7.7379,
      "mean_token_accuracy": 0.10451692193746567,
      "num_tokens": 1193121.0,
      "step": 355
    },
    {
      "entropy": 8.143134546279907,
      "epoch": 0.07023705004389816,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001795,
      "loss": 7.8633,
      "mean_token_accuracy": 0.09846432060003281,
      "num_tokens": 1209896.0,
      "step": 360
    },
    {
      "entropy": 8.091192197799682,
      "epoch": 0.07121256462784119,
      "grad_norm": 1.015625,
      "learning_rate": 0.000182,
      "loss": 7.8953,
      "mean_token_accuracy": 0.09844174087047577,
      "num_tokens": 1227011.0,
      "step": 365
    },
    {
      "entropy": 8.028723001480103,
      "epoch": 0.07218807921178422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001845,
      "loss": 7.7943,
      "mean_token_accuracy": 0.09941637516021729,
      "num_tokens": 1241595.0,
      "step": 370
    },
    {
      "entropy": 8.113645601272584,
      "epoch": 0.07316359379572725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000187,
      "loss": 7.8044,
      "mean_token_accuracy": 0.09797000885009766,
      "num_tokens": 1256919.0,
      "step": 375
    },
    {
      "entropy": 8.04976544380188,
      "epoch": 0.07413910837967028,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001895,
      "loss": 7.7862,
      "mean_token_accuracy": 0.1137398436665535,
      "num_tokens": 1274344.0,
      "step": 380
    },
    {
      "entropy": 8.014974355697632,
      "epoch": 0.0751146229636133,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000192,
      "loss": 7.7822,
      "mean_token_accuracy": 0.10938268229365349,
      "num_tokens": 1290252.0,
      "step": 385
    },
    {
      "entropy": 8.139637756347657,
      "epoch": 0.07609013754755634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001945,
      "loss": 7.835,
      "mean_token_accuracy": 0.09761641547083855,
      "num_tokens": 1306987.0,
      "step": 390
    },
    {
      "entropy": 8.127484035491943,
      "epoch": 0.07706565213149937,
      "grad_norm": 1.09375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.8599,
      "mean_token_accuracy": 0.09425433501601219,
      "num_tokens": 1323031.0,
      "step": 395
    },
    {
      "entropy": 8.023318195343018,
      "epoch": 0.0780411667154424,
      "grad_norm": 0.875,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.8194,
      "mean_token_accuracy": 0.09804570376873016,
      "num_tokens": 1339592.0,
      "step": 400
    },
    {
      "entropy": 8.08116374015808,
      "epoch": 0.07901668129938542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000202,
      "loss": 7.7704,
      "mean_token_accuracy": 0.10749435946345329,
      "num_tokens": 1358131.0,
      "step": 405
    },
    {
      "entropy": 8.001975107192994,
      "epoch": 0.07999219588332845,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.6943,
      "mean_token_accuracy": 0.11095809042453766,
      "num_tokens": 1375705.0,
      "step": 410
    },
    {
      "entropy": 7.946876573562622,
      "epoch": 0.08096771046727148,
      "grad_norm": 1.1875,
      "learning_rate": 0.000207,
      "loss": 7.7367,
      "mean_token_accuracy": 0.10697664991021157,
      "num_tokens": 1392736.0,
      "step": 415
    },
    {
      "entropy": 8.051513481140137,
      "epoch": 0.08194322505121451,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0002095,
      "loss": 7.7993,
      "mean_token_accuracy": 0.10061158984899521,
      "num_tokens": 1408536.0,
      "step": 420
    },
    {
      "entropy": 7.932209205627442,
      "epoch": 0.08291873963515754,
      "grad_norm": 0.9375,
      "learning_rate": 0.000212,
      "loss": 7.7882,
      "mean_token_accuracy": 0.10027184262871743,
      "num_tokens": 1425423.0,
      "step": 425
    },
    {
      "entropy": 8.056283235549927,
      "epoch": 0.08389425421910057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002145,
      "loss": 7.7772,
      "mean_token_accuracy": 0.10296398922801017,
      "num_tokens": 1442420.0,
      "step": 430
    },
    {
      "entropy": 8.07236189842224,
      "epoch": 0.0848697688030436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.7303,
      "mean_token_accuracy": 0.10705514326691627,
      "num_tokens": 1458835.0,
      "step": 435
    },
    {
      "entropy": 8.004169511795045,
      "epoch": 0.08584528338698663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002195,
      "loss": 7.8228,
      "mean_token_accuracy": 0.09661487340927125,
      "num_tokens": 1475563.0,
      "step": 440
    },
    {
      "entropy": 8.09420862197876,
      "epoch": 0.08682079797092966,
      "grad_norm": 1.0,
      "learning_rate": 0.000222,
      "loss": 7.7797,
      "mean_token_accuracy": 0.10943613573908806,
      "num_tokens": 1492687.0,
      "step": 445
    },
    {
      "entropy": 8.086115980148316,
      "epoch": 0.08779631255487269,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002245,
      "loss": 7.7887,
      "mean_token_accuracy": 0.097627442330122,
      "num_tokens": 1508394.0,
      "step": 450
    },
    {
      "entropy": 7.934510850906372,
      "epoch": 0.08877182713881572,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.754,
      "mean_token_accuracy": 0.10472490787506103,
      "num_tokens": 1527112.0,
      "step": 455
    },
    {
      "entropy": 8.001490831375122,
      "epoch": 0.08974734172275875,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.798,
      "mean_token_accuracy": 0.10435827821493149,
      "num_tokens": 1545401.0,
      "step": 460
    },
    {
      "entropy": 7.947861576080323,
      "epoch": 0.09072285630670178,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.6915,
      "mean_token_accuracy": 0.10466418415307999,
      "num_tokens": 1561621.0,
      "step": 465
    },
    {
      "entropy": 8.044413089752197,
      "epoch": 0.09169837089064481,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.8059,
      "mean_token_accuracy": 0.09883959665894508,
      "num_tokens": 1579225.0,
      "step": 470
    },
    {
      "entropy": 8.022387886047364,
      "epoch": 0.09267388547458784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000237,
      "loss": 7.7096,
      "mean_token_accuracy": 0.10403750389814377,
      "num_tokens": 1596125.0,
      "step": 475
    },
    {
      "entropy": 7.9631647109985355,
      "epoch": 0.09364940005853087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002395,
      "loss": 7.8258,
      "mean_token_accuracy": 0.0952387198805809,
      "num_tokens": 1613451.0,
      "step": 480
    },
    {
      "entropy": 8.017077445983887,
      "epoch": 0.09462491464247391,
      "grad_norm": 0.953125,
      "learning_rate": 0.000242,
      "loss": 7.803,
      "mean_token_accuracy": 0.09880654960870743,
      "num_tokens": 1630929.0,
      "step": 485
    },
    {
      "entropy": 8.025622653961182,
      "epoch": 0.09560042922641694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002445,
      "loss": 7.7924,
      "mean_token_accuracy": 0.10142078548669815,
      "num_tokens": 1647379.0,
      "step": 490
    },
    {
      "entropy": 7.718832302093506,
      "epoch": 0.09657594381035997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000247,
      "loss": 7.5342,
      "mean_token_accuracy": 0.12226327732205391,
      "num_tokens": 1664415.0,
      "step": 495
    },
    {
      "entropy": 7.8884869575500485,
      "epoch": 0.097551458394303,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002495,
      "loss": 7.653,
      "mean_token_accuracy": 0.10562271177768708,
      "num_tokens": 1679556.0,
      "step": 500
    },
    {
      "entropy": 7.949688959121704,
      "epoch": 0.09852697297824603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000252,
      "loss": 7.6052,
      "mean_token_accuracy": 0.11016268655657768,
      "num_tokens": 1696727.0,
      "step": 505
    },
    {
      "entropy": 7.83725757598877,
      "epoch": 0.09950248756218906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002545,
      "loss": 7.6089,
      "mean_token_accuracy": 0.11472497954964637,
      "num_tokens": 1713918.0,
      "step": 510
    },
    {
      "entropy": 7.953091287612915,
      "epoch": 0.10047800214613209,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000257,
      "loss": 7.7951,
      "mean_token_accuracy": 0.0945059321820736,
      "num_tokens": 1729581.0,
      "step": 515
    },
    {
      "entropy": 7.921047019958496,
      "epoch": 0.10145351673007512,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002595,
      "loss": 7.6469,
      "mean_token_accuracy": 0.0992059625685215,
      "num_tokens": 1746367.0,
      "step": 520
    },
    {
      "entropy": 7.967546033859253,
      "epoch": 0.10242903131401815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000262,
      "loss": 7.7465,
      "mean_token_accuracy": 0.09730863869190216,
      "num_tokens": 1765009.0,
      "step": 525
    },
    {
      "entropy": 7.927223825454712,
      "epoch": 0.10340454589796118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.7539,
      "mean_token_accuracy": 0.09873774722218513,
      "num_tokens": 1782113.0,
      "step": 530
    },
    {
      "entropy": 7.883999395370483,
      "epoch": 0.10438006048190421,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.6519,
      "mean_token_accuracy": 0.10443588495254516,
      "num_tokens": 1798301.0,
      "step": 535
    },
    {
      "entropy": 7.911849689483643,
      "epoch": 0.10535557506584724,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.6931,
      "mean_token_accuracy": 0.10224330425262451,
      "num_tokens": 1815059.0,
      "step": 540
    },
    {
      "entropy": 7.769307279586792,
      "epoch": 0.10633108964979027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.6051,
      "mean_token_accuracy": 0.10878376364707946,
      "num_tokens": 1831699.0,
      "step": 545
    },
    {
      "entropy": 7.925337314605713,
      "epoch": 0.1073066042337333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002745,
      "loss": 7.6755,
      "mean_token_accuracy": 0.10547427088022232,
      "num_tokens": 1847993.0,
      "step": 550
    },
    {
      "entropy": 7.875313425064087,
      "epoch": 0.10828211881767633,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000277,
      "loss": 7.5326,
      "mean_token_accuracy": 0.10610557869076728,
      "num_tokens": 1864753.0,
      "step": 555
    },
    {
      "entropy": 7.869585037231445,
      "epoch": 0.10925763340161936,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002795,
      "loss": 7.6888,
      "mean_token_accuracy": 0.1044863447546959,
      "num_tokens": 1882154.0,
      "step": 560
    },
    {
      "entropy": 7.754839944839477,
      "epoch": 0.11023314798556239,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.5623,
      "mean_token_accuracy": 0.11065352857112884,
      "num_tokens": 1899683.0,
      "step": 565
    },
    {
      "entropy": 7.821217679977417,
      "epoch": 0.11120866256950541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002845,
      "loss": 7.5912,
      "mean_token_accuracy": 0.10635183900594711,
      "num_tokens": 1918489.0,
      "step": 570
    },
    {
      "entropy": 7.833246660232544,
      "epoch": 0.11218417715344844,
      "grad_norm": 1.03125,
      "learning_rate": 0.000287,
      "loss": 7.6345,
      "mean_token_accuracy": 0.10072939395904541,
      "num_tokens": 1935609.0,
      "step": 575
    },
    {
      "entropy": 7.7502540111541744,
      "epoch": 0.11315969173739147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002895,
      "loss": 7.6882,
      "mean_token_accuracy": 0.09968181252479554,
      "num_tokens": 1951311.0,
      "step": 580
    },
    {
      "entropy": 7.9223309516906735,
      "epoch": 0.1141352063213345,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000292,
      "loss": 7.7251,
      "mean_token_accuracy": 0.09712948650121689,
      "num_tokens": 1968908.0,
      "step": 585
    },
    {
      "entropy": 7.88884711265564,
      "epoch": 0.11511072090527753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002945,
      "loss": 7.6394,
      "mean_token_accuracy": 0.10047843605279923,
      "num_tokens": 1983807.0,
      "step": 590
    },
    {
      "entropy": 7.878595304489136,
      "epoch": 0.11608623548922056,
      "grad_norm": 1.109375,
      "learning_rate": 0.000297,
      "loss": 7.6755,
      "mean_token_accuracy": 0.1005166418850422,
      "num_tokens": 2000219.0,
      "step": 595
    },
    {
      "entropy": 7.826176500320434,
      "epoch": 0.11706175007316359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002995,
      "loss": 7.6359,
      "mean_token_accuracy": 0.10865101143717766,
      "num_tokens": 2017360.0,
      "step": 600
    },
    {
      "entropy": 7.724283885955811,
      "epoch": 0.11803726465710662,
      "grad_norm": 0.859375,
      "learning_rate": 0.000302,
      "loss": 7.6093,
      "mean_token_accuracy": 0.10661007612943649,
      "num_tokens": 2034953.0,
      "step": 605
    },
    {
      "entropy": 7.793875885009766,
      "epoch": 0.11901277924104965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003045,
      "loss": 7.5775,
      "mean_token_accuracy": 0.10660225003957749,
      "num_tokens": 2051551.0,
      "step": 610
    },
    {
      "entropy": 7.78004674911499,
      "epoch": 0.11998829382499268,
      "grad_norm": 1.0,
      "learning_rate": 0.000307,
      "loss": 7.5478,
      "mean_token_accuracy": 0.11335186585783959,
      "num_tokens": 2067188.0,
      "step": 615
    },
    {
      "entropy": 7.790237903594971,
      "epoch": 0.12096380840893571,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003095,
      "loss": 7.6232,
      "mean_token_accuracy": 0.10700822845101357,
      "num_tokens": 2082569.0,
      "step": 620
    },
    {
      "entropy": 7.847408294677734,
      "epoch": 0.12193932299287874,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000312,
      "loss": 7.6786,
      "mean_token_accuracy": 0.10037213563919067,
      "num_tokens": 2099446.0,
      "step": 625
    },
    {
      "entropy": 7.824612283706665,
      "epoch": 0.12291483757682177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003145,
      "loss": 7.6309,
      "mean_token_accuracy": 0.10579207465052605,
      "num_tokens": 2116526.0,
      "step": 630
    },
    {
      "entropy": 7.7642529010772705,
      "epoch": 0.1238903521607648,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000317,
      "loss": 7.6948,
      "mean_token_accuracy": 0.09671583101153373,
      "num_tokens": 2134624.0,
      "step": 635
    },
    {
      "entropy": 7.721237707138061,
      "epoch": 0.12486586674470783,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003195,
      "loss": 7.5155,
      "mean_token_accuracy": 0.11177115440368653,
      "num_tokens": 2152872.0,
      "step": 640
    },
    {
      "entropy": 7.853572702407837,
      "epoch": 0.12584138132865086,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000322,
      "loss": 7.6533,
      "mean_token_accuracy": 0.10234769731760025,
      "num_tokens": 2169480.0,
      "step": 645
    },
    {
      "entropy": 7.71003680229187,
      "epoch": 0.1268168959125939,
      "grad_norm": 0.859375,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.51,
      "mean_token_accuracy": 0.11612471118569374,
      "num_tokens": 2186054.0,
      "step": 650
    },
    {
      "entropy": 7.783114385604859,
      "epoch": 0.12779241049653692,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.6688,
      "mean_token_accuracy": 0.09879108145833015,
      "num_tokens": 2202320.0,
      "step": 655
    },
    {
      "entropy": 7.6497093677520756,
      "epoch": 0.12876792508047996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.5481,
      "mean_token_accuracy": 0.1086566023528576,
      "num_tokens": 2218817.0,
      "step": 660
    },
    {
      "entropy": 7.880639505386353,
      "epoch": 0.12974343966442298,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.7054,
      "mean_token_accuracy": 0.09768114015460014,
      "num_tokens": 2234845.0,
      "step": 665
    },
    {
      "entropy": 7.780311632156372,
      "epoch": 0.13071895424836602,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.6545,
      "mean_token_accuracy": 0.10292254686355591,
      "num_tokens": 2251904.0,
      "step": 670
    },
    {
      "entropy": 7.706472587585449,
      "epoch": 0.13169446883230904,
      "grad_norm": 1.0625,
      "learning_rate": 0.000337,
      "loss": 7.6584,
      "mean_token_accuracy": 0.10477369427680969,
      "num_tokens": 2268343.0,
      "step": 675
    },
    {
      "entropy": 7.830237913131714,
      "epoch": 0.13266998341625208,
      "grad_norm": 1.125,
      "learning_rate": 0.0003395,
      "loss": 7.651,
      "mean_token_accuracy": 0.10206437036395073,
      "num_tokens": 2284524.0,
      "step": 680
    },
    {
      "entropy": 7.651451110839844,
      "epoch": 0.1336454980001951,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000342,
      "loss": 7.5627,
      "mean_token_accuracy": 0.11217835396528245,
      "num_tokens": 2302350.0,
      "step": 685
    },
    {
      "entropy": 7.810202407836914,
      "epoch": 0.13462101258413814,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.6484,
      "mean_token_accuracy": 0.10006091371178627,
      "num_tokens": 2319953.0,
      "step": 690
    },
    {
      "entropy": 7.6707275867462155,
      "epoch": 0.13559652716808115,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000347,
      "loss": 7.5379,
      "mean_token_accuracy": 0.11016602963209152,
      "num_tokens": 2336698.0,
      "step": 695
    },
    {
      "entropy": 7.65397081375122,
      "epoch": 0.1365720417520242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 7.5069,
      "mean_token_accuracy": 0.11032541692256928,
      "num_tokens": 2353986.0,
      "step": 700
    },
    {
      "entropy": 7.720977354049682,
      "epoch": 0.1375475563359672,
      "grad_norm": 0.984375,
      "learning_rate": 0.000352,
      "loss": 7.5481,
      "mean_token_accuracy": 0.10689252316951751,
      "num_tokens": 2370320.0,
      "step": 705
    },
    {
      "entropy": 7.749428844451904,
      "epoch": 0.13852307091991026,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003545,
      "loss": 7.6172,
      "mean_token_accuracy": 0.10306816101074219,
      "num_tokens": 2385453.0,
      "step": 710
    },
    {
      "entropy": 7.7098816394805905,
      "epoch": 0.13949858550385327,
      "grad_norm": 0.9375,
      "learning_rate": 0.000357,
      "loss": 7.5977,
      "mean_token_accuracy": 0.10320264920592308,
      "num_tokens": 2401545.0,
      "step": 715
    },
    {
      "entropy": 7.686821317672729,
      "epoch": 0.14047410008779632,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003595,
      "loss": 7.6155,
      "mean_token_accuracy": 0.10371157601475715,
      "num_tokens": 2418414.0,
      "step": 720
    },
    {
      "entropy": 7.748217391967773,
      "epoch": 0.14144961467173933,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000362,
      "loss": 7.6492,
      "mean_token_accuracy": 0.10145975425839424,
      "num_tokens": 2434973.0,
      "step": 725
    },
    {
      "entropy": 7.769000911712647,
      "epoch": 0.14242512925568238,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003645,
      "loss": 7.6624,
      "mean_token_accuracy": 0.09807511419057846,
      "num_tokens": 2451053.0,
      "step": 730
    },
    {
      "entropy": 7.700602865219116,
      "epoch": 0.1434006438396254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000367,
      "loss": 7.616,
      "mean_token_accuracy": 0.09884371384978294,
      "num_tokens": 2467280.0,
      "step": 735
    },
    {
      "entropy": 7.650289630889892,
      "epoch": 0.14437615842356843,
      "grad_norm": 1.0,
      "learning_rate": 0.0003695,
      "loss": 7.5466,
      "mean_token_accuracy": 0.11135434210300446,
      "num_tokens": 2484588.0,
      "step": 740
    },
    {
      "entropy": 7.705961132049561,
      "epoch": 0.14535167300751145,
      "grad_norm": 1.109375,
      "learning_rate": 0.000372,
      "loss": 7.5657,
      "mean_token_accuracy": 0.10915652140974999,
      "num_tokens": 2501200.0,
      "step": 745
    },
    {
      "entropy": 7.66199746131897,
      "epoch": 0.1463271875914545,
      "grad_norm": 1.0,
      "learning_rate": 0.0003745,
      "loss": 7.5078,
      "mean_token_accuracy": 0.11285350546240806,
      "num_tokens": 2517199.0,
      "step": 750
    },
    {
      "entropy": 7.6683488368988035,
      "epoch": 0.1473027021753975,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000377,
      "loss": 7.5928,
      "mean_token_accuracy": 0.09981210231781006,
      "num_tokens": 2534824.0,
      "step": 755
    },
    {
      "entropy": 7.523636436462402,
      "epoch": 0.14827821675934055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003795,
      "loss": 7.4851,
      "mean_token_accuracy": 0.11066361516714096,
      "num_tokens": 2550842.0,
      "step": 760
    },
    {
      "entropy": 7.629826021194458,
      "epoch": 0.14925373134328357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000382,
      "loss": 7.4176,
      "mean_token_accuracy": 0.11151769161224365,
      "num_tokens": 2567878.0,
      "step": 765
    },
    {
      "entropy": 7.6675316333770756,
      "epoch": 0.1502292459272266,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003845,
      "loss": 7.5154,
      "mean_token_accuracy": 0.10651935636997223,
      "num_tokens": 2584148.0,
      "step": 770
    },
    {
      "entropy": 7.593663120269776,
      "epoch": 0.15120476051116966,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.556,
      "mean_token_accuracy": 0.10755399614572525,
      "num_tokens": 2600067.0,
      "step": 775
    },
    {
      "entropy": 7.689086866378784,
      "epoch": 0.15218027509511267,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.5272,
      "mean_token_accuracy": 0.10445193201303482,
      "num_tokens": 2615746.0,
      "step": 780
    },
    {
      "entropy": 7.613615942001343,
      "epoch": 0.15315578967905571,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.5604,
      "mean_token_accuracy": 0.10300727486610413,
      "num_tokens": 2631641.0,
      "step": 785
    },
    {
      "entropy": 7.590198183059693,
      "epoch": 0.15413130426299873,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.5502,
      "mean_token_accuracy": 0.10660960003733636,
      "num_tokens": 2647015.0,
      "step": 790
    },
    {
      "entropy": 7.6192991733551025,
      "epoch": 0.15510681884694177,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.4552,
      "mean_token_accuracy": 0.10671902745962143,
      "num_tokens": 2664045.0,
      "step": 795
    },
    {
      "entropy": 7.630334043502808,
      "epoch": 0.1560823334308848,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003995,
      "loss": 7.4689,
      "mean_token_accuracy": 0.10270929858088493,
      "num_tokens": 2680774.0,
      "step": 800
    },
    {
      "entropy": 7.6249841213226315,
      "epoch": 0.15705784801482783,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000402,
      "loss": 7.593,
      "mean_token_accuracy": 0.10395026803016663,
      "num_tokens": 2697719.0,
      "step": 805
    },
    {
      "entropy": 7.620457458496094,
      "epoch": 0.15803336259877085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004045,
      "loss": 7.5247,
      "mean_token_accuracy": 0.09749212935566902,
      "num_tokens": 2714518.0,
      "step": 810
    },
    {
      "entropy": 7.627707242965698,
      "epoch": 0.1590088771827139,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.5828,
      "mean_token_accuracy": 0.10642224773764611,
      "num_tokens": 2729288.0,
      "step": 815
    },
    {
      "entropy": 7.668026733398437,
      "epoch": 0.1599843917666569,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004095,
      "loss": 7.558,
      "mean_token_accuracy": 0.1009812332689762,
      "num_tokens": 2746782.0,
      "step": 820
    },
    {
      "entropy": 7.59692063331604,
      "epoch": 0.16095990635059995,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000412,
      "loss": 7.606,
      "mean_token_accuracy": 0.09608046188950539,
      "num_tokens": 2763159.0,
      "step": 825
    },
    {
      "entropy": 7.587530517578125,
      "epoch": 0.16193542093454297,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004145,
      "loss": 7.5677,
      "mean_token_accuracy": 0.10232978016138077,
      "num_tokens": 2780338.0,
      "step": 830
    },
    {
      "entropy": 7.5702805519104,
      "epoch": 0.162910935518486,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000417,
      "loss": 7.4619,
      "mean_token_accuracy": 0.1076440267264843,
      "num_tokens": 2796214.0,
      "step": 835
    },
    {
      "entropy": 7.637975788116455,
      "epoch": 0.16388645010242903,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004195,
      "loss": 7.584,
      "mean_token_accuracy": 0.09872226789593697,
      "num_tokens": 2812745.0,
      "step": 840
    },
    {
      "entropy": 7.524194765090942,
      "epoch": 0.16486196468637207,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000422,
      "loss": 7.4454,
      "mean_token_accuracy": 0.10416454449295998,
      "num_tokens": 2828288.0,
      "step": 845
    },
    {
      "entropy": 7.608716773986816,
      "epoch": 0.16583747927031509,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004245,
      "loss": 7.5874,
      "mean_token_accuracy": 0.0974365383386612,
      "num_tokens": 2843618.0,
      "step": 850
    },
    {
      "entropy": 7.5106102466583256,
      "epoch": 0.16681299385425813,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000427,
      "loss": 7.3893,
      "mean_token_accuracy": 0.10364697575569153,
      "num_tokens": 2860659.0,
      "step": 855
    },
    {
      "entropy": 7.55679759979248,
      "epoch": 0.16778850843820114,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004295,
      "loss": 7.4876,
      "mean_token_accuracy": 0.10175744146108627,
      "num_tokens": 2877482.0,
      "step": 860
    },
    {
      "entropy": 7.586041831970215,
      "epoch": 0.1687640230221442,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000432,
      "loss": 7.5667,
      "mean_token_accuracy": 0.09685270413756371,
      "num_tokens": 2895868.0,
      "step": 865
    },
    {
      "entropy": 7.574951267242431,
      "epoch": 0.1697395376060872,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004345,
      "loss": 7.5706,
      "mean_token_accuracy": 0.10406396687030792,
      "num_tokens": 2913103.0,
      "step": 870
    },
    {
      "entropy": 7.5868103981018065,
      "epoch": 0.17071505219003025,
      "grad_norm": 1.171875,
      "learning_rate": 0.000437,
      "loss": 7.4942,
      "mean_token_accuracy": 0.10660254284739494,
      "num_tokens": 2928291.0,
      "step": 875
    },
    {
      "entropy": 7.551671886444092,
      "epoch": 0.17169056677397326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004395,
      "loss": 7.5347,
      "mean_token_accuracy": 0.10479419305920601,
      "num_tokens": 2945115.0,
      "step": 880
    },
    {
      "entropy": 7.54897518157959,
      "epoch": 0.1726660813579163,
      "grad_norm": 0.953125,
      "learning_rate": 0.000442,
      "loss": 7.4282,
      "mean_token_accuracy": 0.10494303852319717,
      "num_tokens": 2962869.0,
      "step": 885
    },
    {
      "entropy": 7.5557914733886715,
      "epoch": 0.17364159594185932,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004445,
      "loss": 7.5715,
      "mean_token_accuracy": 0.09901110231876373,
      "num_tokens": 2979298.0,
      "step": 890
    },
    {
      "entropy": 7.518081378936768,
      "epoch": 0.17461711052580237,
      "grad_norm": 0.96875,
      "learning_rate": 0.000447,
      "loss": 7.4595,
      "mean_token_accuracy": 0.10685873478651046,
      "num_tokens": 2996146.0,
      "step": 895
    },
    {
      "entropy": 7.518190670013428,
      "epoch": 0.17559262510974538,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.5431,
      "mean_token_accuracy": 0.10059425905346871,
      "num_tokens": 3013075.0,
      "step": 900
    },
    {
      "entropy": 7.553641653060913,
      "epoch": 0.17656813969368843,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.4995,
      "mean_token_accuracy": 0.10074936151504517,
      "num_tokens": 3029494.0,
      "step": 905
    },
    {
      "entropy": 7.493304920196533,
      "epoch": 0.17754365427763144,
      "grad_norm": 0.859375,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.5081,
      "mean_token_accuracy": 0.10531372874975205,
      "num_tokens": 3046333.0,
      "step": 910
    },
    {
      "entropy": 7.5904456615448,
      "epoch": 0.17851916886157448,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.4867,
      "mean_token_accuracy": 0.10380415394902229,
      "num_tokens": 3062714.0,
      "step": 915
    },
    {
      "entropy": 7.521370792388916,
      "epoch": 0.1794946834455175,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.4754,
      "mean_token_accuracy": 0.10096871256828308,
      "num_tokens": 3080711.0,
      "step": 920
    },
    {
      "entropy": 7.505842781066894,
      "epoch": 0.18047019802946054,
      "grad_norm": 0.953125,
      "learning_rate": 0.000462,
      "loss": 7.5654,
      "mean_token_accuracy": 0.09803933352231979,
      "num_tokens": 3096885.0,
      "step": 925
    },
    {
      "entropy": 7.509501171112061,
      "epoch": 0.18144571261340356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004645,
      "loss": 7.4553,
      "mean_token_accuracy": 0.10695251300930977,
      "num_tokens": 3114536.0,
      "step": 930
    },
    {
      "entropy": 7.500088119506836,
      "epoch": 0.1824212271973466,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000467,
      "loss": 7.4054,
      "mean_token_accuracy": 0.11017186567187309,
      "num_tokens": 3131921.0,
      "step": 935
    },
    {
      "entropy": 7.524385738372803,
      "epoch": 0.18339674178128962,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004695,
      "loss": 7.5662,
      "mean_token_accuracy": 0.09947125390172004,
      "num_tokens": 3148064.0,
      "step": 940
    },
    {
      "entropy": 7.416093254089356,
      "epoch": 0.18437225636523266,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000472,
      "loss": 7.3813,
      "mean_token_accuracy": 0.11295887902379036,
      "num_tokens": 3165822.0,
      "step": 945
    },
    {
      "entropy": 7.466145849227905,
      "epoch": 0.18534777094917568,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004745,
      "loss": 7.4716,
      "mean_token_accuracy": 0.10708810836076736,
      "num_tokens": 3183286.0,
      "step": 950
    },
    {
      "entropy": 7.5944517135620115,
      "epoch": 0.18632328553311872,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 7.5058,
      "mean_token_accuracy": 0.10114045292139054,
      "num_tokens": 3198522.0,
      "step": 955
    },
    {
      "entropy": 7.4455090999603275,
      "epoch": 0.18729880011706174,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004795,
      "loss": 7.5287,
      "mean_token_accuracy": 0.10372934117913246,
      "num_tokens": 3218422.0,
      "step": 960
    },
    {
      "entropy": 7.534113931655884,
      "epoch": 0.18827431470100478,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000482,
      "loss": 7.5642,
      "mean_token_accuracy": 0.09645498842000962,
      "num_tokens": 3236407.0,
      "step": 965
    },
    {
      "entropy": 7.551708793640136,
      "epoch": 0.18924982928494782,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004845,
      "loss": 7.4922,
      "mean_token_accuracy": 0.10387791618704796,
      "num_tokens": 3252740.0,
      "step": 970
    },
    {
      "entropy": 7.470359468460083,
      "epoch": 0.19022534386889084,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000487,
      "loss": 7.4286,
      "mean_token_accuracy": 0.10193498730659485,
      "num_tokens": 3269874.0,
      "step": 975
    },
    {
      "entropy": 7.435076713562012,
      "epoch": 0.19120085845283388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895,
      "loss": 7.502,
      "mean_token_accuracy": 0.10077228620648385,
      "num_tokens": 3286515.0,
      "step": 980
    },
    {
      "entropy": 7.378189039230347,
      "epoch": 0.1921763730367769,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000492,
      "loss": 7.2799,
      "mean_token_accuracy": 0.12222360447049141,
      "num_tokens": 3303700.0,
      "step": 985
    },
    {
      "entropy": 7.442311334609985,
      "epoch": 0.19315188762071994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945,
      "loss": 7.4712,
      "mean_token_accuracy": 0.10699821785092353,
      "num_tokens": 3319768.0,
      "step": 990
    },
    {
      "entropy": 7.479348468780517,
      "epoch": 0.19412740220466296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497,
      "loss": 7.3723,
      "mean_token_accuracy": 0.1109627477824688,
      "num_tokens": 3336555.0,
      "step": 995
    },
    {
      "entropy": 7.386620044708252,
      "epoch": 0.195102916788606,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995,
      "loss": 7.4026,
      "mean_token_accuracy": 0.10794939920306205,
      "num_tokens": 3353273.0,
      "step": 1000
    },
    {
      "entropy": 7.3974672794342045,
      "epoch": 0.19607843137254902,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999999929644166,
      "loss": 7.2766,
      "mean_token_accuracy": 0.11775518953800201,
      "num_tokens": 3369408.0,
      "step": 1005
    },
    {
      "entropy": 7.376392459869384,
      "epoch": 0.19705394595649206,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999999643823598,
      "loss": 7.332,
      "mean_token_accuracy": 0.11445832625031471,
      "num_tokens": 3385834.0,
      "step": 1010
    },
    {
      "entropy": 7.386595392227173,
      "epoch": 0.19802946054043508,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999999138141084,
      "loss": 7.4388,
      "mean_token_accuracy": 0.10919031649827957,
      "num_tokens": 3401407.0,
      "step": 1015
    },
    {
      "entropy": 7.5447615623474125,
      "epoch": 0.19900497512437812,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999998412596673,
      "loss": 7.4617,
      "mean_token_accuracy": 0.10121053159236908,
      "num_tokens": 3418256.0,
      "step": 1020
    },
    {
      "entropy": 7.328778457641602,
      "epoch": 0.19998048970832114,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999997467190435,
      "loss": 7.4308,
      "mean_token_accuracy": 0.10942661762237549,
      "num_tokens": 3433932.0,
      "step": 1025
    },
    {
      "entropy": 7.4000791072845455,
      "epoch": 0.20095600429226418,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999996301922465,
      "loss": 7.464,
      "mean_token_accuracy": 0.09955221861600876,
      "num_tokens": 3451723.0,
      "step": 1030
    },
    {
      "entropy": 7.486557340621948,
      "epoch": 0.2019315188762072,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999994916792876,
      "loss": 7.4268,
      "mean_token_accuracy": 0.10579116493463517,
      "num_tokens": 3469170.0,
      "step": 1035
    },
    {
      "entropy": 7.439934206008911,
      "epoch": 0.20290703346015024,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999993311801802,
      "loss": 7.4632,
      "mean_token_accuracy": 0.10232352912425995,
      "num_tokens": 3485282.0,
      "step": 1040
    },
    {
      "entropy": 7.457415437698364,
      "epoch": 0.20388254804409325,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049999914869494,
      "loss": 7.449,
      "mean_token_accuracy": 0.09637642353773117,
      "num_tokens": 3502206.0,
      "step": 1045
    },
    {
      "entropy": 7.423295640945435,
      "epoch": 0.2048580626280363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499998944223585,
      "loss": 7.391,
      "mean_token_accuracy": 0.10561487749218941,
      "num_tokens": 3520384.0,
      "step": 1050
    },
    {
      "entropy": 7.305895376205444,
      "epoch": 0.2058335772119793,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999987177661351,
      "loss": 7.389,
      "mean_token_accuracy": 0.10617925822734833,
      "num_tokens": 3536314.0,
      "step": 1055
    },
    {
      "entropy": 7.391325521469116,
      "epoch": 0.20680909179592236,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999984693226124,
      "loss": 7.3376,
      "mean_token_accuracy": 0.11349749490618706,
      "num_tokens": 3551510.0,
      "step": 1060
    },
    {
      "entropy": 7.388658809661865,
      "epoch": 0.20778460637986537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999981988930411,
      "loss": 7.3434,
      "mean_token_accuracy": 0.11000111326575279,
      "num_tokens": 3568515.0,
      "step": 1065
    },
    {
      "entropy": 7.3375438213348385,
      "epoch": 0.20876012096380842,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999979064774478,
      "loss": 7.3821,
      "mean_token_accuracy": 0.11078001037240029,
      "num_tokens": 3585183.0,
      "step": 1070
    },
    {
      "entropy": 7.363852548599243,
      "epoch": 0.20973563554775143,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000499997592075861,
      "loss": 7.3389,
      "mean_token_accuracy": 0.11243807971477508,
      "num_tokens": 3601898.0,
      "step": 1075
    },
    {
      "entropy": 7.434067678451538,
      "epoch": 0.21071115013169447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999972556883113,
      "loss": 7.3884,
      "mean_token_accuracy": 0.10824741944670677,
      "num_tokens": 3616699.0,
      "step": 1080
    },
    {
      "entropy": 7.355287313461304,
      "epoch": 0.2116866647156375,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999968973148318,
      "loss": 7.4124,
      "mean_token_accuracy": 0.10511938706040383,
      "num_tokens": 3634030.0,
      "step": 1085
    },
    {
      "entropy": 7.44262728691101,
      "epoch": 0.21266217929958053,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999965169554573,
      "loss": 7.4147,
      "mean_token_accuracy": 0.10464555099606514,
      "num_tokens": 3650495.0,
      "step": 1090
    },
    {
      "entropy": 7.384248971939087,
      "epoch": 0.21363769388352355,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999961146102251,
      "loss": 7.3562,
      "mean_token_accuracy": 0.1107444167137146,
      "num_tokens": 3666926.0,
      "step": 1095
    },
    {
      "entropy": 7.320687627792358,
      "epoch": 0.2146132084674666,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999956902791746,
      "loss": 7.3405,
      "mean_token_accuracy": 0.11776096150279045,
      "num_tokens": 3683061.0,
      "step": 1100
    },
    {
      "entropy": 7.306242418289185,
      "epoch": 0.2155887230514096,
      "grad_norm": 0.75,
      "learning_rate": 0.0004999952439623469,
      "loss": 7.4162,
      "mean_token_accuracy": 0.10687667801976204,
      "num_tokens": 3703013.0,
      "step": 1105
    },
    {
      "entropy": 7.352738571166992,
      "epoch": 0.21656423763535265,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499994775659786,
      "loss": 7.2601,
      "mean_token_accuracy": 0.11846758648753167,
      "num_tokens": 3719885.0,
      "step": 1110
    },
    {
      "entropy": 7.383596754074096,
      "epoch": 0.21753975221929567,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999942853715375,
      "loss": 7.4022,
      "mean_token_accuracy": 0.10562850683927535,
      "num_tokens": 3735122.0,
      "step": 1115
    },
    {
      "entropy": 7.362527561187744,
      "epoch": 0.2185152668032387,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999937730976494,
      "loss": 7.3778,
      "mean_token_accuracy": 0.10143291056156159,
      "num_tokens": 3751627.0,
      "step": 1120
    },
    {
      "entropy": 7.2828888416290285,
      "epoch": 0.21949078138718173,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999932388381716,
      "loss": 7.3892,
      "mean_token_accuracy": 0.10415156185626984,
      "num_tokens": 3768165.0,
      "step": 1125
    },
    {
      "entropy": 7.505846929550171,
      "epoch": 0.22046629597112477,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999926825931564,
      "loss": 7.4244,
      "mean_token_accuracy": 0.10518527254462243,
      "num_tokens": 3784245.0,
      "step": 1130
    },
    {
      "entropy": 7.223800945281982,
      "epoch": 0.2214418105550678,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999921043626582,
      "loss": 7.1589,
      "mean_token_accuracy": 0.12127066403627396,
      "num_tokens": 3802057.0,
      "step": 1135
    },
    {
      "entropy": 7.311485624313354,
      "epoch": 0.22241732513901083,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999915041467333,
      "loss": 7.3624,
      "mean_token_accuracy": 0.10803040713071824,
      "num_tokens": 3817104.0,
      "step": 1140
    },
    {
      "entropy": 7.385139131546021,
      "epoch": 0.22339283972295385,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999908819454407,
      "loss": 7.3468,
      "mean_token_accuracy": 0.10780640989542008,
      "num_tokens": 3833037.0,
      "step": 1145
    },
    {
      "entropy": 7.259801578521729,
      "epoch": 0.2243683543068969,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999902377588408,
      "loss": 7.2262,
      "mean_token_accuracy": 0.11170127764344215,
      "num_tokens": 3849634.0,
      "step": 1150
    },
    {
      "entropy": 7.339283418655396,
      "epoch": 0.2253438688908399,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499989571586997,
      "loss": 7.4162,
      "mean_token_accuracy": 0.10458956733345985,
      "num_tokens": 3865949.0,
      "step": 1155
    },
    {
      "entropy": 7.189256191253662,
      "epoch": 0.22631938347478295,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999888834299739,
      "loss": 7.2902,
      "mean_token_accuracy": 0.11521978452801704,
      "num_tokens": 3884646.0,
      "step": 1160
    },
    {
      "entropy": 7.392727613449097,
      "epoch": 0.227294898058726,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999881732878391,
      "loss": 7.3759,
      "mean_token_accuracy": 0.10717562288045883,
      "num_tokens": 3900740.0,
      "step": 1165
    },
    {
      "entropy": 7.320269823074341,
      "epoch": 0.228270412642669,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999874411606618,
      "loss": 7.3326,
      "mean_token_accuracy": 0.11450612768530846,
      "num_tokens": 3918619.0,
      "step": 1170
    },
    {
      "entropy": 7.379632997512817,
      "epoch": 0.22924592722661205,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999866870485137,
      "loss": 7.3004,
      "mean_token_accuracy": 0.11032224223017692,
      "num_tokens": 3938161.0,
      "step": 1175
    },
    {
      "entropy": 7.34798526763916,
      "epoch": 0.23022144181055507,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999859109514684,
      "loss": 7.3542,
      "mean_token_accuracy": 0.10768297165632248,
      "num_tokens": 3953582.0,
      "step": 1180
    },
    {
      "entropy": 7.30758113861084,
      "epoch": 0.2311969563944981,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999851128696017,
      "loss": 7.3213,
      "mean_token_accuracy": 0.1020158052444458,
      "num_tokens": 3969747.0,
      "step": 1185
    },
    {
      "entropy": 7.3781046867370605,
      "epoch": 0.23217247097844113,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999842928029917,
      "loss": 7.3586,
      "mean_token_accuracy": 0.10794111490249633,
      "num_tokens": 3984646.0,
      "step": 1190
    },
    {
      "entropy": 7.299358987808228,
      "epoch": 0.23314798556238417,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999834507517184,
      "loss": 7.3486,
      "mean_token_accuracy": 0.10666156932711601,
      "num_tokens": 4003108.0,
      "step": 1195
    },
    {
      "entropy": 7.3120873928070065,
      "epoch": 0.23412350014632718,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999825867158642,
      "loss": 7.4151,
      "mean_token_accuracy": 0.1019670806825161,
      "num_tokens": 4019944.0,
      "step": 1200
    },
    {
      "entropy": 7.3608317375183105,
      "epoch": 0.23509901473027023,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999817006955134,
      "loss": 7.2993,
      "mean_token_accuracy": 0.1109683245420456,
      "num_tokens": 4037617.0,
      "step": 1205
    },
    {
      "entropy": 7.282232141494751,
      "epoch": 0.23607452931421324,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999807926907527,
      "loss": 7.2855,
      "mean_token_accuracy": 0.10396898314356803,
      "num_tokens": 4053483.0,
      "step": 1210
    },
    {
      "entropy": 7.354314088821411,
      "epoch": 0.2370500438981563,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999798627016707,
      "loss": 7.3621,
      "mean_token_accuracy": 0.10965256318449974,
      "num_tokens": 4070165.0,
      "step": 1215
    },
    {
      "entropy": 7.324768304824829,
      "epoch": 0.2380255584820993,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999789107283586,
      "loss": 7.3414,
      "mean_token_accuracy": 0.10487294420599938,
      "num_tokens": 4086208.0,
      "step": 1220
    },
    {
      "entropy": 7.26507248878479,
      "epoch": 0.23900107306604235,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499977936770909,
      "loss": 7.303,
      "mean_token_accuracy": 0.11123988702893257,
      "num_tokens": 4101790.0,
      "step": 1225
    },
    {
      "entropy": 7.320441770553589,
      "epoch": 0.23997658764998536,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999769408294173,
      "loss": 7.2266,
      "mean_token_accuracy": 0.11657417044043542,
      "num_tokens": 4117633.0,
      "step": 1230
    },
    {
      "entropy": 7.265048789978027,
      "epoch": 0.2409521022339284,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999759229039807,
      "loss": 7.3035,
      "mean_token_accuracy": 0.11276578083634377,
      "num_tokens": 4132634.0,
      "step": 1235
    },
    {
      "entropy": 7.320464944839477,
      "epoch": 0.24192761681787142,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004999748829946988,
      "loss": 7.3391,
      "mean_token_accuracy": 0.10908927693963051,
      "num_tokens": 4150702.0,
      "step": 1240
    },
    {
      "entropy": 7.26134147644043,
      "epoch": 0.24290313140181446,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999738211016732,
      "loss": 7.2057,
      "mean_token_accuracy": 0.12152880877256393,
      "num_tokens": 4168483.0,
      "step": 1245
    },
    {
      "entropy": 7.158234167098999,
      "epoch": 0.24387864598575748,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999727372250076,
      "loss": 7.1701,
      "mean_token_accuracy": 0.11431463360786438,
      "num_tokens": 4186199.0,
      "step": 1250
    },
    {
      "entropy": 7.278416919708252,
      "epoch": 0.24485416056970052,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999716313648079,
      "loss": 7.3169,
      "mean_token_accuracy": 0.1068569652736187,
      "num_tokens": 4202303.0,
      "step": 1255
    },
    {
      "entropy": 7.338901853561401,
      "epoch": 0.24582967515364354,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999705035211822,
      "loss": 7.3571,
      "mean_token_accuracy": 0.10634541511535645,
      "num_tokens": 4218541.0,
      "step": 1260
    },
    {
      "entropy": 7.319312906265258,
      "epoch": 0.24680518973758658,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999693536942406,
      "loss": 7.3192,
      "mean_token_accuracy": 0.10624103397130966,
      "num_tokens": 4234025.0,
      "step": 1265
    },
    {
      "entropy": 7.292185544967651,
      "epoch": 0.2477807043215296,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999681818840957,
      "loss": 7.3435,
      "mean_token_accuracy": 0.10807270258665085,
      "num_tokens": 4250798.0,
      "step": 1270
    },
    {
      "entropy": 7.286764860153198,
      "epoch": 0.24875621890547264,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499966988090862,
      "loss": 7.3177,
      "mean_token_accuracy": 0.10285230055451393,
      "num_tokens": 4266453.0,
      "step": 1275
    },
    {
      "entropy": 7.35591869354248,
      "epoch": 0.24973173348941566,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999657723146558,
      "loss": 7.2729,
      "mean_token_accuracy": 0.10762197226285934,
      "num_tokens": 4281825.0,
      "step": 1280
    },
    {
      "entropy": 7.098993587493896,
      "epoch": 0.2507072480733587,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999645345555962,
      "loss": 7.1453,
      "mean_token_accuracy": 0.12227144241333007,
      "num_tokens": 4298821.0,
      "step": 1285
    },
    {
      "entropy": 7.248935031890869,
      "epoch": 0.2516827626573017,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999632748138041,
      "loss": 7.1552,
      "mean_token_accuracy": 0.11829453259706497,
      "num_tokens": 4314076.0,
      "step": 1290
    },
    {
      "entropy": 7.217565107345581,
      "epoch": 0.25265827724124473,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999619930894025,
      "loss": 7.1797,
      "mean_token_accuracy": 0.11851233690977096,
      "num_tokens": 4330909.0,
      "step": 1295
    },
    {
      "entropy": 7.284217786788941,
      "epoch": 0.2536337918251878,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999606893825168,
      "loss": 7.2681,
      "mean_token_accuracy": 0.1124707892537117,
      "num_tokens": 4346927.0,
      "step": 1300
    },
    {
      "entropy": 7.240284633636475,
      "epoch": 0.2546093064091308,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999593636932742,
      "loss": 7.3215,
      "mean_token_accuracy": 0.10526167303323745,
      "num_tokens": 4364956.0,
      "step": 1305
    },
    {
      "entropy": 7.290760946273804,
      "epoch": 0.25558482099307384,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999580160218044,
      "loss": 7.2381,
      "mean_token_accuracy": 0.10958634614944458,
      "num_tokens": 4383014.0,
      "step": 1310
    },
    {
      "entropy": 7.144275045394897,
      "epoch": 0.25656033557701685,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999566463682391,
      "loss": 7.3101,
      "mean_token_accuracy": 0.10827136710286141,
      "num_tokens": 4401050.0,
      "step": 1315
    },
    {
      "entropy": 7.248860025405884,
      "epoch": 0.2575358501609599,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999552547327119,
      "loss": 7.2175,
      "mean_token_accuracy": 0.11122509092092514,
      "num_tokens": 4415453.0,
      "step": 1320
    },
    {
      "entropy": 7.1784752368927,
      "epoch": 0.25851136474490294,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999538411153591,
      "loss": 7.183,
      "mean_token_accuracy": 0.11985458731651306,
      "num_tokens": 4433107.0,
      "step": 1325
    },
    {
      "entropy": 7.25650372505188,
      "epoch": 0.25948687932884595,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999524055163187,
      "loss": 7.2758,
      "mean_token_accuracy": 0.10543364882469178,
      "num_tokens": 4449440.0,
      "step": 1330
    },
    {
      "entropy": 7.275032186508179,
      "epoch": 0.26046239391278897,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499950947935731,
      "loss": 7.2647,
      "mean_token_accuracy": 0.10925490632653237,
      "num_tokens": 4465327.0,
      "step": 1335
    },
    {
      "entropy": 7.2784374237060545,
      "epoch": 0.26143790849673204,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999494683737384,
      "loss": 7.2899,
      "mean_token_accuracy": 0.10701772198081017,
      "num_tokens": 4481518.0,
      "step": 1340
    },
    {
      "entropy": 7.292396831512451,
      "epoch": 0.26241342308067506,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999479668304854,
      "loss": 7.3115,
      "mean_token_accuracy": 0.10881960839033127,
      "num_tokens": 4497632.0,
      "step": 1345
    },
    {
      "entropy": 7.030213069915772,
      "epoch": 0.2633889376646181,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999464433061188,
      "loss": 7.0711,
      "mean_token_accuracy": 0.12342782318592072,
      "num_tokens": 4513790.0,
      "step": 1350
    },
    {
      "entropy": 7.235921001434326,
      "epoch": 0.26436445224856114,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999448978007876,
      "loss": 7.1959,
      "mean_token_accuracy": 0.11445838510990143,
      "num_tokens": 4530802.0,
      "step": 1355
    },
    {
      "entropy": 7.2668671131134035,
      "epoch": 0.26533996683250416,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999433303146426,
      "loss": 7.2485,
      "mean_token_accuracy": 0.10927580818533897,
      "num_tokens": 4546208.0,
      "step": 1360
    },
    {
      "entropy": 7.22170729637146,
      "epoch": 0.2663154814164472,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004999417408478371,
      "loss": 7.3095,
      "mean_token_accuracy": 0.11063284948468208,
      "num_tokens": 4563753.0,
      "step": 1365
    },
    {
      "entropy": 7.213341140747071,
      "epoch": 0.2672909960003902,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999401294005264,
      "loss": 7.1897,
      "mean_token_accuracy": 0.11640307679772377,
      "num_tokens": 4580274.0,
      "step": 1370
    },
    {
      "entropy": 7.195255708694458,
      "epoch": 0.26826651058433326,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499938495972868,
      "loss": 7.2656,
      "mean_token_accuracy": 0.10936984345316887,
      "num_tokens": 4596085.0,
      "step": 1375
    },
    {
      "entropy": 7.2880692958831785,
      "epoch": 0.2692420251682763,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999368405650214,
      "loss": 7.3116,
      "mean_token_accuracy": 0.10627636015415191,
      "num_tokens": 4613258.0,
      "step": 1380
    },
    {
      "entropy": 7.1051576137542725,
      "epoch": 0.2702175397522193,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999351631771485,
      "loss": 7.1188,
      "mean_token_accuracy": 0.11826564893126487,
      "num_tokens": 4630128.0,
      "step": 1385
    },
    {
      "entropy": 7.252414560317993,
      "epoch": 0.2711930543361623,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999334638094131,
      "loss": 7.1382,
      "mean_token_accuracy": 0.1179644674062729,
      "num_tokens": 4646181.0,
      "step": 1390
    },
    {
      "entropy": 7.185850667953491,
      "epoch": 0.2721685689201054,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999317424619812,
      "loss": 7.2119,
      "mean_token_accuracy": 0.10765682607889175,
      "num_tokens": 4661804.0,
      "step": 1395
    },
    {
      "entropy": 7.16576886177063,
      "epoch": 0.2731440835040484,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999299991350212,
      "loss": 7.1211,
      "mean_token_accuracy": 0.11724445223808289,
      "num_tokens": 4678163.0,
      "step": 1400
    },
    {
      "entropy": 7.2082353115081785,
      "epoch": 0.2741195980879914,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999282338287034,
      "loss": 7.1126,
      "mean_token_accuracy": 0.11289523839950562,
      "num_tokens": 4695469.0,
      "step": 1405
    },
    {
      "entropy": 7.2161016941070555,
      "epoch": 0.2750951126719344,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999264465432001,
      "loss": 7.2651,
      "mean_token_accuracy": 0.1120416909456253,
      "num_tokens": 4711599.0,
      "step": 1410
    },
    {
      "entropy": 7.22179012298584,
      "epoch": 0.2760706272558775,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999246372786862,
      "loss": 7.2664,
      "mean_token_accuracy": 0.10617031827569008,
      "num_tokens": 4728381.0,
      "step": 1415
    },
    {
      "entropy": 7.18099422454834,
      "epoch": 0.2770461418398205,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999228060353385,
      "loss": 7.249,
      "mean_token_accuracy": 0.10983582735061645,
      "num_tokens": 4745348.0,
      "step": 1420
    },
    {
      "entropy": 7.252281141281128,
      "epoch": 0.27802165642376353,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999209528133357,
      "loss": 7.2657,
      "mean_token_accuracy": 0.10274152681231499,
      "num_tokens": 4763011.0,
      "step": 1425
    },
    {
      "entropy": 7.224287223815918,
      "epoch": 0.27899717100770655,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499919077612859,
      "loss": 7.2037,
      "mean_token_accuracy": 0.1110018789768219,
      "num_tokens": 4779462.0,
      "step": 1430
    },
    {
      "entropy": 7.078258800506592,
      "epoch": 0.2799726855916496,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999171804340917,
      "loss": 7.1645,
      "mean_token_accuracy": 0.11882285252213479,
      "num_tokens": 4796844.0,
      "step": 1435
    },
    {
      "entropy": 7.212335729598999,
      "epoch": 0.28094820017559263,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999152612772191,
      "loss": 7.1761,
      "mean_token_accuracy": 0.11435465589165687,
      "num_tokens": 4812851.0,
      "step": 1440
    },
    {
      "entropy": 7.095452976226807,
      "epoch": 0.28192371475953565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999133201424288,
      "loss": 7.1163,
      "mean_token_accuracy": 0.11493284702301025,
      "num_tokens": 4830144.0,
      "step": 1445
    },
    {
      "entropy": 7.22363805770874,
      "epoch": 0.28289922934347866,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999113570299105,
      "loss": 7.2114,
      "mean_token_accuracy": 0.10577581152319908,
      "num_tokens": 4846981.0,
      "step": 1450
    },
    {
      "entropy": 7.309843254089356,
      "epoch": 0.28387474392742174,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004999093719398561,
      "loss": 7.3844,
      "mean_token_accuracy": 0.10269515588879585,
      "num_tokens": 4864404.0,
      "step": 1455
    },
    {
      "entropy": 7.181564807891846,
      "epoch": 0.28485025851136475,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999073648724592,
      "loss": 7.2608,
      "mean_token_accuracy": 0.10562214255332947,
      "num_tokens": 4882004.0,
      "step": 1460
    },
    {
      "entropy": 7.243942594528198,
      "epoch": 0.28582577309530777,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999053358279164,
      "loss": 7.1962,
      "mean_token_accuracy": 0.1058816634118557,
      "num_tokens": 4899444.0,
      "step": 1465
    },
    {
      "entropy": 7.221729326248169,
      "epoch": 0.2868012876792508,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999032848064256,
      "loss": 7.1964,
      "mean_token_accuracy": 0.10939600914716721,
      "num_tokens": 4917214.0,
      "step": 1470
    },
    {
      "entropy": 7.120475912094117,
      "epoch": 0.28777680226319385,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999012118081875,
      "loss": 7.2906,
      "mean_token_accuracy": 0.102950569242239,
      "num_tokens": 4934534.0,
      "step": 1475
    },
    {
      "entropy": 7.26613483428955,
      "epoch": 0.28875231684713687,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998991168334045,
      "loss": 7.1991,
      "mean_token_accuracy": 0.10897819027304649,
      "num_tokens": 4950740.0,
      "step": 1480
    },
    {
      "entropy": 7.185323762893677,
      "epoch": 0.2897278314310799,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998969998822814,
      "loss": 7.2858,
      "mean_token_accuracy": 0.1056202620267868,
      "num_tokens": 4966720.0,
      "step": 1485
    },
    {
      "entropy": 7.229295015335083,
      "epoch": 0.2907033460150229,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998948609550249,
      "loss": 7.1584,
      "mean_token_accuracy": 0.10480799153447151,
      "num_tokens": 4982195.0,
      "step": 1490
    },
    {
      "entropy": 7.172478818893433,
      "epoch": 0.29167886059896597,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998927000518442,
      "loss": 7.2806,
      "mean_token_accuracy": 0.10777373760938644,
      "num_tokens": 4997487.0,
      "step": 1495
    },
    {
      "entropy": 7.191869068145752,
      "epoch": 0.292654375182909,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998905171729505,
      "loss": 7.1667,
      "mean_token_accuracy": 0.10947065055370331,
      "num_tokens": 5012913.0,
      "step": 1500
    },
    {
      "entropy": 7.098269653320313,
      "epoch": 0.293629889766852,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998883123185568,
      "loss": 7.1303,
      "mean_token_accuracy": 0.11931669116020202,
      "num_tokens": 5031462.0,
      "step": 1505
    },
    {
      "entropy": 7.170979738235474,
      "epoch": 0.294605404350795,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998860854888788,
      "loss": 7.1056,
      "mean_token_accuracy": 0.11914367899298668,
      "num_tokens": 5047354.0,
      "step": 1510
    },
    {
      "entropy": 7.102825689315796,
      "epoch": 0.2955809189347381,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004998838366841341,
      "loss": 7.1882,
      "mean_token_accuracy": 0.11276983618736267,
      "num_tokens": 5065268.0,
      "step": 1515
    },
    {
      "entropy": 7.182042598724365,
      "epoch": 0.2965564335186811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998815659045423,
      "loss": 7.1841,
      "mean_token_accuracy": 0.10845357850193978,
      "num_tokens": 5081066.0,
      "step": 1520
    },
    {
      "entropy": 7.147319793701172,
      "epoch": 0.2975319481026241,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998792731503254,
      "loss": 7.2147,
      "mean_token_accuracy": 0.11445605456829071,
      "num_tokens": 5099440.0,
      "step": 1525
    },
    {
      "entropy": 7.120917463302613,
      "epoch": 0.29850746268656714,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998769584217074,
      "loss": 7.1576,
      "mean_token_accuracy": 0.11984050646424294,
      "num_tokens": 5116573.0,
      "step": 1530
    },
    {
      "entropy": 7.006188821792603,
      "epoch": 0.2994829772705102,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998746217189145,
      "loss": 6.9406,
      "mean_token_accuracy": 0.12909065559506416,
      "num_tokens": 5132920.0,
      "step": 1535
    },
    {
      "entropy": 7.300417137145996,
      "epoch": 0.3004584918544532,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499872263042175,
      "loss": 7.2107,
      "mean_token_accuracy": 0.10663213208317757,
      "num_tokens": 5150681.0,
      "step": 1540
    },
    {
      "entropy": 6.995895576477051,
      "epoch": 0.30143400643839624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998698823917196,
      "loss": 6.9672,
      "mean_token_accuracy": 0.12855525091290473,
      "num_tokens": 5167473.0,
      "step": 1545
    },
    {
      "entropy": 7.170604419708252,
      "epoch": 0.3024095210223393,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004998674797677805,
      "loss": 7.2533,
      "mean_token_accuracy": 0.1055234856903553,
      "num_tokens": 5183847.0,
      "step": 1550
    },
    {
      "entropy": 7.117762517929077,
      "epoch": 0.3033850356062823,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998650551705929,
      "loss": 7.2047,
      "mean_token_accuracy": 0.10641410052776337,
      "num_tokens": 5200019.0,
      "step": 1555
    },
    {
      "entropy": 7.073120164871216,
      "epoch": 0.30436055019022534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004998626086003934,
      "loss": 7.074,
      "mean_token_accuracy": 0.11620460152626037,
      "num_tokens": 5217245.0,
      "step": 1560
    },
    {
      "entropy": 7.112638330459594,
      "epoch": 0.30533606477416836,
      "grad_norm": 0.75,
      "learning_rate": 0.0004998601400574213,
      "loss": 6.9782,
      "mean_token_accuracy": 0.12163180261850357,
      "num_tokens": 5234822.0,
      "step": 1565
    },
    {
      "entropy": 7.115424585342407,
      "epoch": 0.30631157935811143,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998576495419177,
      "loss": 7.1305,
      "mean_token_accuracy": 0.11187227442860603,
      "num_tokens": 5251282.0,
      "step": 1570
    },
    {
      "entropy": 7.066088104248047,
      "epoch": 0.30728709394205445,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000499855137054126,
      "loss": 7.0935,
      "mean_token_accuracy": 0.11060698553919793,
      "num_tokens": 5269630.0,
      "step": 1575
    },
    {
      "entropy": 7.150989151000976,
      "epoch": 0.30826260852599746,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998526025942917,
      "loss": 7.1563,
      "mean_token_accuracy": 0.11012233942747116,
      "num_tokens": 5287106.0,
      "step": 1580
    },
    {
      "entropy": 7.177860069274902,
      "epoch": 0.3092381231099405,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998500461626625,
      "loss": 7.1918,
      "mean_token_accuracy": 0.10584679394960403,
      "num_tokens": 5303387.0,
      "step": 1585
    },
    {
      "entropy": 7.231128740310669,
      "epoch": 0.31021363769388355,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000499847467759488,
      "loss": 7.1797,
      "mean_token_accuracy": 0.10859874933958054,
      "num_tokens": 5320482.0,
      "step": 1590
    },
    {
      "entropy": 7.100463199615478,
      "epoch": 0.31118915227782656,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998448673850205,
      "loss": 7.1404,
      "mean_token_accuracy": 0.11622472479939461,
      "num_tokens": 5336420.0,
      "step": 1595
    },
    {
      "entropy": 7.01635103225708,
      "epoch": 0.3121646668617696,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998422450395139,
      "loss": 7.0872,
      "mean_token_accuracy": 0.11045913249254227,
      "num_tokens": 5353464.0,
      "step": 1600
    },
    {
      "entropy": 7.137714147567749,
      "epoch": 0.3131401814457126,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998396007232244,
      "loss": 7.1475,
      "mean_token_accuracy": 0.1154401995241642,
      "num_tokens": 5372607.0,
      "step": 1605
    },
    {
      "entropy": 7.122408390045166,
      "epoch": 0.31411569602965567,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998369344364105,
      "loss": 7.1723,
      "mean_token_accuracy": 0.11676934882998466,
      "num_tokens": 5390134.0,
      "step": 1610
    },
    {
      "entropy": 7.0988771438598635,
      "epoch": 0.3150912106135987,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998342461793328,
      "loss": 7.0821,
      "mean_token_accuracy": 0.1153003416955471,
      "num_tokens": 5407083.0,
      "step": 1615
    },
    {
      "entropy": 6.936686992645264,
      "epoch": 0.3160667251975417,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998315359522538,
      "loss": 6.9746,
      "mean_token_accuracy": 0.12812903001904488,
      "num_tokens": 5424617.0,
      "step": 1620
    },
    {
      "entropy": 7.118428945541382,
      "epoch": 0.3170422397814847,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998288037554383,
      "loss": 7.1248,
      "mean_token_accuracy": 0.12060828432440758,
      "num_tokens": 5440913.0,
      "step": 1625
    },
    {
      "entropy": 7.1336418151855465,
      "epoch": 0.3180177543654278,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998260495891535,
      "loss": 7.1488,
      "mean_token_accuracy": 0.11536478698253631,
      "num_tokens": 5457157.0,
      "step": 1630
    },
    {
      "entropy": 7.052104187011719,
      "epoch": 0.3189932689493708,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004998232734536685,
      "loss": 7.0581,
      "mean_token_accuracy": 0.11604453399777412,
      "num_tokens": 5474520.0,
      "step": 1635
    },
    {
      "entropy": 7.146209955215454,
      "epoch": 0.3199687835333138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998204753492544,
      "loss": 7.1398,
      "mean_token_accuracy": 0.11600466221570968,
      "num_tokens": 5490599.0,
      "step": 1640
    },
    {
      "entropy": 7.127297306060791,
      "epoch": 0.32094429811725683,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998176552761848,
      "loss": 7.2087,
      "mean_token_accuracy": 0.10203930661082268,
      "num_tokens": 5506714.0,
      "step": 1645
    },
    {
      "entropy": 7.160216569900513,
      "epoch": 0.3219198127011999,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998148132347352,
      "loss": 7.1485,
      "mean_token_accuracy": 0.10920300856232643,
      "num_tokens": 5522894.0,
      "step": 1650
    },
    {
      "entropy": 7.043737173080444,
      "epoch": 0.3228953272851429,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004998119492251832,
      "loss": 7.0243,
      "mean_token_accuracy": 0.12240474671125412,
      "num_tokens": 5540054.0,
      "step": 1655
    },
    {
      "entropy": 7.168790626525879,
      "epoch": 0.32387084186908593,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998090632478088,
      "loss": 7.1947,
      "mean_token_accuracy": 0.10093299672007561,
      "num_tokens": 5556189.0,
      "step": 1660
    },
    {
      "entropy": 7.159790563583374,
      "epoch": 0.32484635645302895,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499806155302894,
      "loss": 7.2081,
      "mean_token_accuracy": 0.10658285319805146,
      "num_tokens": 5572562.0,
      "step": 1665
    },
    {
      "entropy": 7.182238483428955,
      "epoch": 0.325821871036972,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499803225390723,
      "loss": 7.188,
      "mean_token_accuracy": 0.11000551506876946,
      "num_tokens": 5588992.0,
      "step": 1670
    },
    {
      "entropy": 7.151162958145141,
      "epoch": 0.32679738562091504,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998002735115819,
      "loss": 7.1148,
      "mean_token_accuracy": 0.11034396961331368,
      "num_tokens": 5604143.0,
      "step": 1675
    },
    {
      "entropy": 7.119655466079712,
      "epoch": 0.32777290020485805,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997972996657593,
      "loss": 7.2398,
      "mean_token_accuracy": 0.10105421096086502,
      "num_tokens": 5620887.0,
      "step": 1680
    },
    {
      "entropy": 7.158847618103027,
      "epoch": 0.32874841478880107,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997943038535459,
      "loss": 7.1709,
      "mean_token_accuracy": 0.1110181026160717,
      "num_tokens": 5637156.0,
      "step": 1685
    },
    {
      "entropy": 7.153251838684082,
      "epoch": 0.32972392937274414,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004997912860752341,
      "loss": 7.1217,
      "mean_token_accuracy": 0.10916642993688583,
      "num_tokens": 5654024.0,
      "step": 1690
    },
    {
      "entropy": 7.088912582397461,
      "epoch": 0.33069944395668716,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997882463311192,
      "loss": 7.062,
      "mean_token_accuracy": 0.11368629187345505,
      "num_tokens": 5671538.0,
      "step": 1695
    },
    {
      "entropy": 7.089573669433594,
      "epoch": 0.33167495854063017,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997851846214979,
      "loss": 7.1212,
      "mean_token_accuracy": 0.11628169491887093,
      "num_tokens": 5688788.0,
      "step": 1700
    },
    {
      "entropy": 7.056381750106811,
      "epoch": 0.3326504731245732,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004997821009466695,
      "loss": 7.0529,
      "mean_token_accuracy": 0.11595623269677162,
      "num_tokens": 5706066.0,
      "step": 1705
    },
    {
      "entropy": 7.03841462135315,
      "epoch": 0.33362598770851626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997789953069354,
      "loss": 7.0434,
      "mean_token_accuracy": 0.11439799889922142,
      "num_tokens": 5722100.0,
      "step": 1710
    },
    {
      "entropy": 7.124280977249145,
      "epoch": 0.3346015022924593,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004997758677025989,
      "loss": 7.1696,
      "mean_token_accuracy": 0.10830180943012238,
      "num_tokens": 5739068.0,
      "step": 1715
    },
    {
      "entropy": 7.088494777679443,
      "epoch": 0.3355770168764023,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004997727181339658,
      "loss": 7.1296,
      "mean_token_accuracy": 0.11222045645117759,
      "num_tokens": 5756147.0,
      "step": 1720
    },
    {
      "entropy": 7.158913946151733,
      "epoch": 0.3365525314603453,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997695466013438,
      "loss": 7.1013,
      "mean_token_accuracy": 0.11348066926002502,
      "num_tokens": 5772157.0,
      "step": 1725
    },
    {
      "entropy": 7.1085058689117435,
      "epoch": 0.3375280460442884,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997663531050427,
      "loss": 7.133,
      "mean_token_accuracy": 0.11211565360426903,
      "num_tokens": 5788128.0,
      "step": 1730
    },
    {
      "entropy": 7.048029518127441,
      "epoch": 0.3385035606282314,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997631376453749,
      "loss": 7.117,
      "mean_token_accuracy": 0.11303148567676544,
      "num_tokens": 5804793.0,
      "step": 1735
    },
    {
      "entropy": 7.091092491149903,
      "epoch": 0.3394790752121744,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997599002226541,
      "loss": 7.0445,
      "mean_token_accuracy": 0.11384404823184013,
      "num_tokens": 5822409.0,
      "step": 1740
    },
    {
      "entropy": 6.988621425628662,
      "epoch": 0.3404545897961175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499756640837197,
      "loss": 7.0837,
      "mean_token_accuracy": 0.11030225455760956,
      "num_tokens": 5839023.0,
      "step": 1745
    },
    {
      "entropy": 7.136508464813232,
      "epoch": 0.3414301043800605,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499753359489322,
      "loss": 7.1018,
      "mean_token_accuracy": 0.10952565595507621,
      "num_tokens": 5855804.0,
      "step": 1750
    },
    {
      "entropy": 7.064236307144165,
      "epoch": 0.3424056189640035,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997500561793498,
      "loss": 7.1143,
      "mean_token_accuracy": 0.11261064410209656,
      "num_tokens": 5872372.0,
      "step": 1755
    },
    {
      "entropy": 7.140656232833862,
      "epoch": 0.3433811335479465,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997467309076031,
      "loss": 7.1523,
      "mean_token_accuracy": 0.1071063719689846,
      "num_tokens": 5888666.0,
      "step": 1760
    },
    {
      "entropy": 6.9952013969421385,
      "epoch": 0.3443566481318896,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997433836744069,
      "loss": 7.0905,
      "mean_token_accuracy": 0.1121178537607193,
      "num_tokens": 5905577.0,
      "step": 1765
    },
    {
      "entropy": 7.113792610168457,
      "epoch": 0.3453321627158326,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997400144800882,
      "loss": 7.065,
      "mean_token_accuracy": 0.11290777698159218,
      "num_tokens": 5922570.0,
      "step": 1770
    },
    {
      "entropy": 6.972097253799438,
      "epoch": 0.34630767729977563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997366233249762,
      "loss": 7.0204,
      "mean_token_accuracy": 0.12438111305236817,
      "num_tokens": 5938995.0,
      "step": 1775
    },
    {
      "entropy": 7.042465877532959,
      "epoch": 0.34728319188371864,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997332102094024,
      "loss": 7.0303,
      "mean_token_accuracy": 0.1208010382950306,
      "num_tokens": 5954869.0,
      "step": 1780
    },
    {
      "entropy": 7.181672048568726,
      "epoch": 0.3482587064676617,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997297751337004,
      "loss": 7.1467,
      "mean_token_accuracy": 0.11480126529932022,
      "num_tokens": 5972750.0,
      "step": 1785
    },
    {
      "entropy": 7.021099853515625,
      "epoch": 0.34923422105160473,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997263180982055,
      "loss": 7.0622,
      "mean_token_accuracy": 0.11614978462457656,
      "num_tokens": 5989004.0,
      "step": 1790
    },
    {
      "entropy": 7.017993831634522,
      "epoch": 0.35020973563554775,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997228391032559,
      "loss": 7.0776,
      "mean_token_accuracy": 0.11161015629768371,
      "num_tokens": 6004749.0,
      "step": 1795
    },
    {
      "entropy": 6.965499401092529,
      "epoch": 0.35118525021949076,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997193381491913,
      "loss": 6.9504,
      "mean_token_accuracy": 0.11921020373702049,
      "num_tokens": 6022235.0,
      "step": 1800
    },
    {
      "entropy": 7.04525671005249,
      "epoch": 0.35216076480343383,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499715815236354,
      "loss": 7.1238,
      "mean_token_accuracy": 0.10555350333452225,
      "num_tokens": 6039194.0,
      "step": 1805
    },
    {
      "entropy": 7.0818140506744385,
      "epoch": 0.35313627938737685,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000499712270365088,
      "loss": 6.9903,
      "mean_token_accuracy": 0.11186902597546577,
      "num_tokens": 6055376.0,
      "step": 1810
    },
    {
      "entropy": 7.07976222038269,
      "epoch": 0.35411179397131987,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997087035357398,
      "loss": 7.0482,
      "mean_token_accuracy": 0.1155602365732193,
      "num_tokens": 6071105.0,
      "step": 1815
    },
    {
      "entropy": 6.994044351577759,
      "epoch": 0.3550873085552629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997051147486582,
      "loss": 6.9829,
      "mean_token_accuracy": 0.11727840378880501,
      "num_tokens": 6087964.0,
      "step": 1820
    },
    {
      "entropy": 6.997056674957276,
      "epoch": 0.35606282313920595,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997015040041935,
      "loss": 7.0576,
      "mean_token_accuracy": 0.10933208093047142,
      "num_tokens": 6105245.0,
      "step": 1825
    },
    {
      "entropy": 7.0379139423370365,
      "epoch": 0.35703833772314897,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996978713026986,
      "loss": 7.0079,
      "mean_token_accuracy": 0.12135875448584557,
      "num_tokens": 6121736.0,
      "step": 1830
    },
    {
      "entropy": 7.041876888275146,
      "epoch": 0.358013852307092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996942166445287,
      "loss": 7.0581,
      "mean_token_accuracy": 0.11438895165920257,
      "num_tokens": 6137751.0,
      "step": 1835
    },
    {
      "entropy": 7.019932746887207,
      "epoch": 0.358989366891035,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996905400300407,
      "loss": 7.1101,
      "mean_token_accuracy": 0.1149475984275341,
      "num_tokens": 6154655.0,
      "step": 1840
    },
    {
      "entropy": 7.147087574005127,
      "epoch": 0.35996488147497807,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499686841459594,
      "loss": 7.1113,
      "mean_token_accuracy": 0.10812506079673767,
      "num_tokens": 6172230.0,
      "step": 1845
    },
    {
      "entropy": 6.977477312088013,
      "epoch": 0.3609403960589211,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996831209335497,
      "loss": 7.072,
      "mean_token_accuracy": 0.10572083070874214,
      "num_tokens": 6188520.0,
      "step": 1850
    },
    {
      "entropy": 7.082860088348388,
      "epoch": 0.3619159106428641,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004996793784522719,
      "loss": 7.0966,
      "mean_token_accuracy": 0.11021933704614639,
      "num_tokens": 6205555.0,
      "step": 1855
    },
    {
      "entropy": 7.090164709091186,
      "epoch": 0.3628914252268071,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996756140161259,
      "loss": 7.0792,
      "mean_token_accuracy": 0.11376980021595955,
      "num_tokens": 6223548.0,
      "step": 1860
    },
    {
      "entropy": 7.03369836807251,
      "epoch": 0.3638669398107502,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996718276254797,
      "loss": 7.089,
      "mean_token_accuracy": 0.1082814022898674,
      "num_tokens": 6239266.0,
      "step": 1865
    },
    {
      "entropy": 7.072378301620484,
      "epoch": 0.3648424543946932,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996680192807032,
      "loss": 7.0513,
      "mean_token_accuracy": 0.11285683810710907,
      "num_tokens": 6255020.0,
      "step": 1870
    },
    {
      "entropy": 7.046410369873047,
      "epoch": 0.3658179689786362,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996641889821686,
      "loss": 6.9917,
      "mean_token_accuracy": 0.1103696033358574,
      "num_tokens": 6272774.0,
      "step": 1875
    },
    {
      "entropy": 6.994774389266968,
      "epoch": 0.36679348356257924,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996603367302502,
      "loss": 7.0055,
      "mean_token_accuracy": 0.11505129560828209,
      "num_tokens": 6290478.0,
      "step": 1880
    },
    {
      "entropy": 7.055514526367188,
      "epoch": 0.3677689981465223,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004996564625253244,
      "loss": 7.0872,
      "mean_token_accuracy": 0.11461991667747498,
      "num_tokens": 6308003.0,
      "step": 1885
    },
    {
      "entropy": 7.011760663986206,
      "epoch": 0.3687445127304653,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996525663677697,
      "loss": 7.0486,
      "mean_token_accuracy": 0.10762766152620315,
      "num_tokens": 6325748.0,
      "step": 1890
    },
    {
      "entropy": 7.06716947555542,
      "epoch": 0.36972002731440834,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499648648257967,
      "loss": 6.9936,
      "mean_token_accuracy": 0.12201232090592384,
      "num_tokens": 6344031.0,
      "step": 1895
    },
    {
      "entropy": 6.942547082901001,
      "epoch": 0.37069554189835136,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499644708196299,
      "loss": 7.0461,
      "mean_token_accuracy": 0.11835477128624916,
      "num_tokens": 6359821.0,
      "step": 1900
    },
    {
      "entropy": 6.970766019821167,
      "epoch": 0.3716710564822944,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996407461831507,
      "loss": 6.888,
      "mean_token_accuracy": 0.1269619345664978,
      "num_tokens": 6376464.0,
      "step": 1905
    },
    {
      "entropy": 7.020088863372803,
      "epoch": 0.37264657106623744,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996367622189094,
      "loss": 7.0657,
      "mean_token_accuracy": 0.1099945642054081,
      "num_tokens": 6392415.0,
      "step": 1910
    },
    {
      "entropy": 6.974881839752197,
      "epoch": 0.37362208565018046,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996327563039643,
      "loss": 7.0612,
      "mean_token_accuracy": 0.11171750500798225,
      "num_tokens": 6409151.0,
      "step": 1915
    },
    {
      "entropy": 7.1153641700744625,
      "epoch": 0.3745976002341235,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004996287284387068,
      "loss": 7.0241,
      "mean_token_accuracy": 0.11637166067957878,
      "num_tokens": 6426997.0,
      "step": 1920
    },
    {
      "entropy": 6.979948997497559,
      "epoch": 0.37557311481806654,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996246786235306,
      "loss": 7.0189,
      "mean_token_accuracy": 0.11474493220448494,
      "num_tokens": 6442871.0,
      "step": 1925
    },
    {
      "entropy": 6.961305904388428,
      "epoch": 0.37654862940200956,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996206068588314,
      "loss": 7.0048,
      "mean_token_accuracy": 0.11208005249500275,
      "num_tokens": 6459677.0,
      "step": 1930
    },
    {
      "entropy": 7.101429510116577,
      "epoch": 0.3775241439859526,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996165131450071,
      "loss": 7.1353,
      "mean_token_accuracy": 0.10877176895737647,
      "num_tokens": 6477239.0,
      "step": 1935
    },
    {
      "entropy": 7.059422874450684,
      "epoch": 0.37849965856989565,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996123974824576,
      "loss": 7.0037,
      "mean_token_accuracy": 0.11793602928519249,
      "num_tokens": 6493277.0,
      "step": 1940
    },
    {
      "entropy": 6.967101335525513,
      "epoch": 0.37947517315383866,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996082598715852,
      "loss": 7.0137,
      "mean_token_accuracy": 0.11420942023396492,
      "num_tokens": 6509204.0,
      "step": 1945
    },
    {
      "entropy": 6.933266544342041,
      "epoch": 0.3804506877377817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499604100312794,
      "loss": 6.8819,
      "mean_token_accuracy": 0.1260261982679367,
      "num_tokens": 6525486.0,
      "step": 1950
    },
    {
      "entropy": 6.971220445632935,
      "epoch": 0.3814262023217247,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004995999188064908,
      "loss": 7.0954,
      "mean_token_accuracy": 0.1095791406929493,
      "num_tokens": 6542227.0,
      "step": 1955
    },
    {
      "entropy": 7.0703741073608395,
      "epoch": 0.38240171690566777,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499595715353084,
      "loss": 7.1663,
      "mean_token_accuracy": 0.1052120178937912,
      "num_tokens": 6558951.0,
      "step": 1960
    },
    {
      "entropy": 7.107343959808349,
      "epoch": 0.3833772314896108,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995914899529844,
      "loss": 7.0602,
      "mean_token_accuracy": 0.11332785338163376,
      "num_tokens": 6574696.0,
      "step": 1965
    },
    {
      "entropy": 7.097244834899902,
      "epoch": 0.3843527460735538,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995872426066048,
      "loss": 7.0934,
      "mean_token_accuracy": 0.10992669835686683,
      "num_tokens": 6591628.0,
      "step": 1970
    },
    {
      "entropy": 7.00416316986084,
      "epoch": 0.3853282606574968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995829733143602,
      "loss": 6.9752,
      "mean_token_accuracy": 0.11769431084394455,
      "num_tokens": 6608107.0,
      "step": 1975
    },
    {
      "entropy": 7.009332847595215,
      "epoch": 0.3863037752414399,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499578682076668,
      "loss": 7.1112,
      "mean_token_accuracy": 0.11527311876416206,
      "num_tokens": 6626699.0,
      "step": 1980
    },
    {
      "entropy": 7.1108642578125,
      "epoch": 0.3872792898253829,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995743688939474,
      "loss": 7.1356,
      "mean_token_accuracy": 0.10339468270540238,
      "num_tokens": 6644540.0,
      "step": 1985
    },
    {
      "entropy": 7.082017087936402,
      "epoch": 0.3882548044093259,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004995700337666199,
      "loss": 7.0632,
      "mean_token_accuracy": 0.10787386372685433,
      "num_tokens": 6660653.0,
      "step": 1990
    },
    {
      "entropy": 7.05488920211792,
      "epoch": 0.38923031899326893,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499565676695109,
      "loss": 7.055,
      "mean_token_accuracy": 0.10950738564133644,
      "num_tokens": 6677553.0,
      "step": 1995
    },
    {
      "entropy": 7.056804180145264,
      "epoch": 0.390205833577212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995612976798406,
      "loss": 7.0546,
      "mean_token_accuracy": 0.10897849723696709,
      "num_tokens": 6695274.0,
      "step": 2000
    },
    {
      "entropy": 6.96084771156311,
      "epoch": 0.391181348161155,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995568967212425,
      "loss": 6.9898,
      "mean_token_accuracy": 0.1134484887123108,
      "num_tokens": 6711166.0,
      "step": 2005
    },
    {
      "entropy": 6.97249550819397,
      "epoch": 0.39215686274509803,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995524738197448,
      "loss": 6.9709,
      "mean_token_accuracy": 0.11566549465060234,
      "num_tokens": 6728827.0,
      "step": 2010
    },
    {
      "entropy": 6.880181360244751,
      "epoch": 0.39313237732904105,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995480289757796,
      "loss": 6.9491,
      "mean_token_accuracy": 0.12184445783495904,
      "num_tokens": 6745901.0,
      "step": 2015
    },
    {
      "entropy": 6.927216005325318,
      "epoch": 0.3941078919129841,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004995435621897815,
      "loss": 6.9157,
      "mean_token_accuracy": 0.12079904079437256,
      "num_tokens": 6762176.0,
      "step": 2020
    },
    {
      "entropy": 6.997888612747192,
      "epoch": 0.39508340649692714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995390734621867,
      "loss": 7.0721,
      "mean_token_accuracy": 0.10900324806571007,
      "num_tokens": 6779322.0,
      "step": 2025
    },
    {
      "entropy": 7.120794105529785,
      "epoch": 0.39605892108087015,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995345627934338,
      "loss": 7.0008,
      "mean_token_accuracy": 0.1168686531484127,
      "num_tokens": 6794293.0,
      "step": 2030
    },
    {
      "entropy": 6.7111622333526615,
      "epoch": 0.39703443566481317,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995300301839637,
      "loss": 6.8367,
      "mean_token_accuracy": 0.13236104398965837,
      "num_tokens": 6812880.0,
      "step": 2035
    },
    {
      "entropy": 7.063685464859009,
      "epoch": 0.39800995024875624,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995254756342194,
      "loss": 7.0484,
      "mean_token_accuracy": 0.1098434530198574,
      "num_tokens": 6829230.0,
      "step": 2040
    },
    {
      "entropy": 7.021308040618896,
      "epoch": 0.39898546483269925,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995208991446458,
      "loss": 7.0422,
      "mean_token_accuracy": 0.11086834147572518,
      "num_tokens": 6846787.0,
      "step": 2045
    },
    {
      "entropy": 6.921110916137695,
      "epoch": 0.39996097941664227,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995163007156902,
      "loss": 6.9141,
      "mean_token_accuracy": 0.11068061590194703,
      "num_tokens": 6862990.0,
      "step": 2050
    },
    {
      "entropy": 7.045805883407593,
      "epoch": 0.4009364940005853,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995116803478018,
      "loss": 7.0557,
      "mean_token_accuracy": 0.10982215702533722,
      "num_tokens": 6879170.0,
      "step": 2055
    },
    {
      "entropy": 6.93969578742981,
      "epoch": 0.40191200858452836,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995070380414321,
      "loss": 7.0471,
      "mean_token_accuracy": 0.11072775200009347,
      "num_tokens": 6895635.0,
      "step": 2060
    },
    {
      "entropy": 7.018460369110107,
      "epoch": 0.4028875231684714,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499502373797035,
      "loss": 6.9495,
      "mean_token_accuracy": 0.11848736256361007,
      "num_tokens": 6912836.0,
      "step": 2065
    },
    {
      "entropy": 6.938055992126465,
      "epoch": 0.4038630377524144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499497687615066,
      "loss": 6.9925,
      "mean_token_accuracy": 0.11958426535129547,
      "num_tokens": 6929507.0,
      "step": 2070
    },
    {
      "entropy": 7.012846374511719,
      "epoch": 0.4048385523363574,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499492979495983,
      "loss": 7.0108,
      "mean_token_accuracy": 0.11071802526712418,
      "num_tokens": 6946527.0,
      "step": 2075
    },
    {
      "entropy": 7.006974458694458,
      "epoch": 0.4058140669203005,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994882494402462,
      "loss": 7.0069,
      "mean_token_accuracy": 0.11493440046906471,
      "num_tokens": 6962582.0,
      "step": 2080
    },
    {
      "entropy": 7.00134916305542,
      "epoch": 0.4067895815042435,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994834974483178,
      "loss": 6.9966,
      "mean_token_accuracy": 0.1160830058157444,
      "num_tokens": 6979389.0,
      "step": 2085
    },
    {
      "entropy": 6.925145530700684,
      "epoch": 0.4077650960881865,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004994787235206621,
      "loss": 6.9717,
      "mean_token_accuracy": 0.11536258235573768,
      "num_tokens": 6996637.0,
      "step": 2090
    },
    {
      "entropy": 7.0829870223999025,
      "epoch": 0.4087406106721295,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004994739276577456,
      "loss": 7.0683,
      "mean_token_accuracy": 0.10761817693710327,
      "num_tokens": 7013013.0,
      "step": 2095
    },
    {
      "entropy": 6.928919935226441,
      "epoch": 0.4097161252560726,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004994691098600369,
      "loss": 7.0302,
      "mean_token_accuracy": 0.11743959337472916,
      "num_tokens": 7031476.0,
      "step": 2100
    },
    {
      "entropy": 7.063517665863037,
      "epoch": 0.4106916398400156,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994642701280068,
      "loss": 6.9326,
      "mean_token_accuracy": 0.11514554470777512,
      "num_tokens": 7047424.0,
      "step": 2105
    },
    {
      "entropy": 6.928591346740722,
      "epoch": 0.4116671544239586,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004994594084621282,
      "loss": 7.0055,
      "mean_token_accuracy": 0.110958231985569,
      "num_tokens": 7063793.0,
      "step": 2110
    },
    {
      "entropy": 7.087340021133423,
      "epoch": 0.41264266900790164,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994545248628762,
      "loss": 7.0733,
      "mean_token_accuracy": 0.10909361913800239,
      "num_tokens": 7079026.0,
      "step": 2115
    },
    {
      "entropy": 6.97515983581543,
      "epoch": 0.4136181835918447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499449619330728,
      "loss": 7.0004,
      "mean_token_accuracy": 0.1165304310619831,
      "num_tokens": 7097524.0,
      "step": 2120
    },
    {
      "entropy": 7.0495555877685545,
      "epoch": 0.41459369817578773,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499444691866163,
      "loss": 7.0737,
      "mean_token_accuracy": 0.11175737231969833,
      "num_tokens": 7115514.0,
      "step": 2125
    },
    {
      "entropy": 6.996109342575073,
      "epoch": 0.41556921275973074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994397424696625,
      "loss": 7.0165,
      "mean_token_accuracy": 0.1140232503414154,
      "num_tokens": 7132589.0,
      "step": 2130
    },
    {
      "entropy": 7.000941753387451,
      "epoch": 0.4165447273436738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994347711417105,
      "loss": 7.0172,
      "mean_token_accuracy": 0.1115983322262764,
      "num_tokens": 7149416.0,
      "step": 2135
    },
    {
      "entropy": 6.923533773422241,
      "epoch": 0.41752024192761683,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994297778827924,
      "loss": 6.9598,
      "mean_token_accuracy": 0.11849450021982193,
      "num_tokens": 7165763.0,
      "step": 2140
    },
    {
      "entropy": 6.934512424468994,
      "epoch": 0.41849575651155985,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994247626933964,
      "loss": 6.9401,
      "mean_token_accuracy": 0.12275434285402298,
      "num_tokens": 7183131.0,
      "step": 2145
    },
    {
      "entropy": 6.95603723526001,
      "epoch": 0.41947127109550286,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994197255740124,
      "loss": 6.968,
      "mean_token_accuracy": 0.1182591237127781,
      "num_tokens": 7200467.0,
      "step": 2150
    },
    {
      "entropy": 6.983357381820679,
      "epoch": 0.42044678567944593,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994146665251327,
      "loss": 6.9547,
      "mean_token_accuracy": 0.11873010322451591,
      "num_tokens": 7217311.0,
      "step": 2155
    },
    {
      "entropy": 6.967365360260009,
      "epoch": 0.42142230026338895,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004994095855472516,
      "loss": 6.9618,
      "mean_token_accuracy": 0.11772591844201089,
      "num_tokens": 7235244.0,
      "step": 2160
    },
    {
      "entropy": 6.99353609085083,
      "epoch": 0.42239781484733196,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004994044826408656,
      "loss": 6.9931,
      "mean_token_accuracy": 0.11206987127661705,
      "num_tokens": 7252235.0,
      "step": 2165
    },
    {
      "entropy": 6.952538728713989,
      "epoch": 0.423373329431275,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993993578064733,
      "loss": 7.0059,
      "mean_token_accuracy": 0.11217687204480171,
      "num_tokens": 7269367.0,
      "step": 2170
    },
    {
      "entropy": 7.075891733169556,
      "epoch": 0.42434884401521805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993942110445756,
      "loss": 6.9881,
      "mean_token_accuracy": 0.1137772336602211,
      "num_tokens": 7287198.0,
      "step": 2175
    },
    {
      "entropy": 6.9127795696258545,
      "epoch": 0.42532435859916107,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993890423556753,
      "loss": 6.9142,
      "mean_token_accuracy": 0.11559901610016823,
      "num_tokens": 7302356.0,
      "step": 2180
    },
    {
      "entropy": 6.960020351409912,
      "epoch": 0.4262998731831041,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993838517402775,
      "loss": 6.9235,
      "mean_token_accuracy": 0.12155768349766731,
      "num_tokens": 7318530.0,
      "step": 2185
    },
    {
      "entropy": 6.953242349624634,
      "epoch": 0.4272753877670471,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993786391988894,
      "loss": 6.9095,
      "mean_token_accuracy": 0.11496350169181824,
      "num_tokens": 7335195.0,
      "step": 2190
    },
    {
      "entropy": 6.883234643936158,
      "epoch": 0.42825090235099017,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993734047320206,
      "loss": 6.9147,
      "mean_token_accuracy": 0.11316282376646995,
      "num_tokens": 7352233.0,
      "step": 2195
    },
    {
      "entropy": 6.983979988098144,
      "epoch": 0.4292264169349332,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993681483401821,
      "loss": 6.9696,
      "mean_token_accuracy": 0.11637645065784455,
      "num_tokens": 7368569.0,
      "step": 2200
    },
    {
      "entropy": 6.933692598342896,
      "epoch": 0.4302019315188762,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993628700238879,
      "loss": 7.0137,
      "mean_token_accuracy": 0.114168231934309,
      "num_tokens": 7384669.0,
      "step": 2205
    },
    {
      "entropy": 7.095562601089478,
      "epoch": 0.4311774461028192,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993575697836537,
      "loss": 7.0289,
      "mean_token_accuracy": 0.11071271151304245,
      "num_tokens": 7401833.0,
      "step": 2210
    },
    {
      "entropy": 6.924860429763794,
      "epoch": 0.4321529606867623,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993522476199975,
      "loss": 6.9162,
      "mean_token_accuracy": 0.1127213828265667,
      "num_tokens": 7418663.0,
      "step": 2215
    },
    {
      "entropy": 6.971612501144409,
      "epoch": 0.4331284752707053,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993469035334392,
      "loss": 6.9014,
      "mean_token_accuracy": 0.1250322349369526,
      "num_tokens": 7434701.0,
      "step": 2220
    },
    {
      "entropy": 6.9419066429138185,
      "epoch": 0.4341039898546483,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499341537524501,
      "loss": 6.9828,
      "mean_token_accuracy": 0.10986461192369461,
      "num_tokens": 7450154.0,
      "step": 2225
    },
    {
      "entropy": 6.791154384613037,
      "epoch": 0.43507950443859134,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993361495937074,
      "loss": 6.7912,
      "mean_token_accuracy": 0.13519535660743714,
      "num_tokens": 7468110.0,
      "step": 2230
    },
    {
      "entropy": 6.939401054382325,
      "epoch": 0.4360550190225344,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993307397415848,
      "loss": 6.9107,
      "mean_token_accuracy": 0.11550685539841651,
      "num_tokens": 7482962.0,
      "step": 2235
    },
    {
      "entropy": 7.027690649032593,
      "epoch": 0.4370305336064774,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004993253079686619,
      "loss": 7.0536,
      "mean_token_accuracy": 0.1094849169254303,
      "num_tokens": 7499116.0,
      "step": 2240
    },
    {
      "entropy": 6.978180360794068,
      "epoch": 0.43800604819042044,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993198542754692,
      "loss": 6.9999,
      "mean_token_accuracy": 0.11379628702998161,
      "num_tokens": 7515765.0,
      "step": 2245
    },
    {
      "entropy": 6.975172281265259,
      "epoch": 0.43898156277436345,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993143786625401,
      "loss": 6.9918,
      "mean_token_accuracy": 0.11214097291231155,
      "num_tokens": 7532074.0,
      "step": 2250
    },
    {
      "entropy": 6.84090781211853,
      "epoch": 0.4399570773583065,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993088811304092,
      "loss": 6.8176,
      "mean_token_accuracy": 0.1292093053460121,
      "num_tokens": 7547704.0,
      "step": 2255
    },
    {
      "entropy": 6.949747610092163,
      "epoch": 0.44093259194224954,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993033616796139,
      "loss": 6.9434,
      "mean_token_accuracy": 0.11486721858382225,
      "num_tokens": 7563317.0,
      "step": 2260
    },
    {
      "entropy": 6.861780452728271,
      "epoch": 0.44190810652619256,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992978203106936,
      "loss": 6.9432,
      "mean_token_accuracy": 0.11190066412091255,
      "num_tokens": 7580770.0,
      "step": 2265
    },
    {
      "entropy": 6.96252031326294,
      "epoch": 0.4428836211101356,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004992922570241896,
      "loss": 6.9184,
      "mean_token_accuracy": 0.11765873730182648,
      "num_tokens": 7598181.0,
      "step": 2270
    },
    {
      "entropy": 6.973319005966187,
      "epoch": 0.44385913569407864,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992866718206457,
      "loss": 6.9181,
      "mean_token_accuracy": 0.11969833001494408,
      "num_tokens": 7613739.0,
      "step": 2275
    },
    {
      "entropy": 6.935301446914673,
      "epoch": 0.44483465027802166,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992810647006077,
      "loss": 7.0022,
      "mean_token_accuracy": 0.1112263709306717,
      "num_tokens": 7629297.0,
      "step": 2280
    },
    {
      "entropy": 7.030787324905395,
      "epoch": 0.4458101648619647,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992754356646234,
      "loss": 6.9824,
      "mean_token_accuracy": 0.11118089407682419,
      "num_tokens": 7645273.0,
      "step": 2285
    },
    {
      "entropy": 6.963811206817627,
      "epoch": 0.4467856794459077,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992697847132427,
      "loss": 6.9328,
      "mean_token_accuracy": 0.11047199815511703,
      "num_tokens": 7661296.0,
      "step": 2290
    },
    {
      "entropy": 6.7955968379974365,
      "epoch": 0.44776119402985076,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499264111847018,
      "loss": 6.9373,
      "mean_token_accuracy": 0.11633702144026756,
      "num_tokens": 7678658.0,
      "step": 2295
    },
    {
      "entropy": 7.02021017074585,
      "epoch": 0.4487367086137938,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004992584170665037,
      "loss": 6.9728,
      "mean_token_accuracy": 0.11074585020542145,
      "num_tokens": 7693678.0,
      "step": 2300
    },
    {
      "entropy": 6.931669616699219,
      "epoch": 0.4497122231977368,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499252700372256,
      "loss": 6.9621,
      "mean_token_accuracy": 0.11443416476249695,
      "num_tokens": 7709526.0,
      "step": 2305
    },
    {
      "entropy": 6.949428653717041,
      "epoch": 0.4506877377816798,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992469617648337,
      "loss": 6.8567,
      "mean_token_accuracy": 0.11352990642189979,
      "num_tokens": 7725055.0,
      "step": 2310
    },
    {
      "entropy": 6.787866640090942,
      "epoch": 0.4516632523656229,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992412012447976,
      "loss": 6.7685,
      "mean_token_accuracy": 0.13319482654333115,
      "num_tokens": 7742047.0,
      "step": 2315
    },
    {
      "entropy": 6.923736810684204,
      "epoch": 0.4526387669495659,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992354188127105,
      "loss": 6.9335,
      "mean_token_accuracy": 0.11316853016614914,
      "num_tokens": 7758283.0,
      "step": 2320
    },
    {
      "entropy": 6.89393982887268,
      "epoch": 0.4536142815335089,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992296144691375,
      "loss": 6.8739,
      "mean_token_accuracy": 0.11695389747619629,
      "num_tokens": 7774660.0,
      "step": 2325
    },
    {
      "entropy": 6.8337976932525635,
      "epoch": 0.454589796117452,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992237882146456,
      "loss": 6.8903,
      "mean_token_accuracy": 0.12075782865285874,
      "num_tokens": 7792147.0,
      "step": 2330
    },
    {
      "entropy": 6.886572599411011,
      "epoch": 0.455565310701395,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992179400498044,
      "loss": 6.8362,
      "mean_token_accuracy": 0.12014108300209045,
      "num_tokens": 7809021.0,
      "step": 2335
    },
    {
      "entropy": 6.939410781860351,
      "epoch": 0.456540825285338,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992120699751852,
      "loss": 6.9025,
      "mean_token_accuracy": 0.11456301808357239,
      "num_tokens": 7825074.0,
      "step": 2340
    },
    {
      "entropy": 6.76352219581604,
      "epoch": 0.45751633986928103,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992061779913616,
      "loss": 6.8407,
      "mean_token_accuracy": 0.11945309415459633,
      "num_tokens": 7843413.0,
      "step": 2345
    },
    {
      "entropy": 6.984707689285278,
      "epoch": 0.4584918544532241,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992002640989094,
      "loss": 6.9299,
      "mean_token_accuracy": 0.10952871888875962,
      "num_tokens": 7859874.0,
      "step": 2350
    },
    {
      "entropy": 6.80113639831543,
      "epoch": 0.4594673690371671,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004991943282984065,
      "loss": 6.865,
      "mean_token_accuracy": 0.12114636227488518,
      "num_tokens": 7877302.0,
      "step": 2355
    },
    {
      "entropy": 6.925329256057739,
      "epoch": 0.46044288362111013,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991883705904328,
      "loss": 6.976,
      "mean_token_accuracy": 0.11222620382905006,
      "num_tokens": 7893207.0,
      "step": 2360
    },
    {
      "entropy": 6.8837878704071045,
      "epoch": 0.46141839820505315,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991823909755705,
      "loss": 6.8151,
      "mean_token_accuracy": 0.12659919783473014,
      "num_tokens": 7910239.0,
      "step": 2365
    },
    {
      "entropy": 6.920702743530273,
      "epoch": 0.4623939127889962,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991763894544041,
      "loss": 6.958,
      "mean_token_accuracy": 0.11296103447675705,
      "num_tokens": 7928030.0,
      "step": 2370
    },
    {
      "entropy": 6.890843772888184,
      "epoch": 0.46336942737293924,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991703660275199,
      "loss": 6.8931,
      "mean_token_accuracy": 0.11770829334855079,
      "num_tokens": 7943653.0,
      "step": 2375
    },
    {
      "entropy": 6.974278211593628,
      "epoch": 0.46434494195688225,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991643206955064,
      "loss": 6.9705,
      "mean_token_accuracy": 0.11404236927628517,
      "num_tokens": 7960564.0,
      "step": 2380
    },
    {
      "entropy": 6.9140668392181395,
      "epoch": 0.46532045654082527,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004991582534589544,
      "loss": 6.9254,
      "mean_token_accuracy": 0.1193073645234108,
      "num_tokens": 7978045.0,
      "step": 2385
    },
    {
      "entropy": 6.957419967651367,
      "epoch": 0.46629597112476834,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991521643184567,
      "loss": 6.9591,
      "mean_token_accuracy": 0.11141033694148064,
      "num_tokens": 7995074.0,
      "step": 2390
    },
    {
      "entropy": 6.947557401657105,
      "epoch": 0.46727148570871135,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991460532746087,
      "loss": 6.9844,
      "mean_token_accuracy": 0.10783464983105659,
      "num_tokens": 8011800.0,
      "step": 2395
    },
    {
      "entropy": 6.9592297077178955,
      "epoch": 0.46824700029265437,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499139920328007,
      "loss": 6.9432,
      "mean_token_accuracy": 0.1144430011510849,
      "num_tokens": 8028090.0,
      "step": 2400
    },
    {
      "entropy": 6.912628507614135,
      "epoch": 0.4692225148765974,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991337654792512,
      "loss": 6.9,
      "mean_token_accuracy": 0.11696527898311615,
      "num_tokens": 8044140.0,
      "step": 2405
    },
    {
      "entropy": 6.872697257995606,
      "epoch": 0.47019802946054046,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991275887289427,
      "loss": 6.8495,
      "mean_token_accuracy": 0.1179678663611412,
      "num_tokens": 8060030.0,
      "step": 2410
    },
    {
      "entropy": 6.8621447563171385,
      "epoch": 0.47117354404448347,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991213900776849,
      "loss": 6.8854,
      "mean_token_accuracy": 0.11339947506785393,
      "num_tokens": 8075445.0,
      "step": 2415
    },
    {
      "entropy": 6.974141788482666,
      "epoch": 0.4721490586284265,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004991151695260838,
      "loss": 6.8922,
      "mean_token_accuracy": 0.11236568167805672,
      "num_tokens": 8093250.0,
      "step": 2420
    },
    {
      "entropy": 6.816811847686767,
      "epoch": 0.4731245732123695,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499108927074747,
      "loss": 6.837,
      "mean_token_accuracy": 0.12727442309260367,
      "num_tokens": 8109615.0,
      "step": 2425
    },
    {
      "entropy": 6.8544200420379635,
      "epoch": 0.4741000877963126,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991026627242846,
      "loss": 6.9403,
      "mean_token_accuracy": 0.11186167597770691,
      "num_tokens": 8126591.0,
      "step": 2430
    },
    {
      "entropy": 6.9048929691314695,
      "epoch": 0.4750756023802556,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990963764753088,
      "loss": 6.8668,
      "mean_token_accuracy": 0.11707579866051673,
      "num_tokens": 8141360.0,
      "step": 2435
    },
    {
      "entropy": 6.942644453048706,
      "epoch": 0.4760511169641986,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990900683284337,
      "loss": 6.9608,
      "mean_token_accuracy": 0.12016337439417839,
      "num_tokens": 8159688.0,
      "step": 2440
    },
    {
      "entropy": 6.867148971557617,
      "epoch": 0.4770266315481416,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990837382842758,
      "loss": 6.9146,
      "mean_token_accuracy": 0.11586871594190598,
      "num_tokens": 8176887.0,
      "step": 2445
    },
    {
      "entropy": 6.805676031112671,
      "epoch": 0.4780021461320847,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990773863434537,
      "loss": 6.7335,
      "mean_token_accuracy": 0.12699671015143393,
      "num_tokens": 8193372.0,
      "step": 2450
    },
    {
      "entropy": 6.888016700744629,
      "epoch": 0.4789776607160277,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990710125065879,
      "loss": 6.9435,
      "mean_token_accuracy": 0.1139527752995491,
      "num_tokens": 8209713.0,
      "step": 2455
    },
    {
      "entropy": 6.959731769561768,
      "epoch": 0.4799531752999707,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990646167743015,
      "loss": 6.9477,
      "mean_token_accuracy": 0.11611981242895127,
      "num_tokens": 8226359.0,
      "step": 2460
    },
    {
      "entropy": 6.841183757781982,
      "epoch": 0.48092868988391374,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004990581991472193,
      "loss": 6.9187,
      "mean_token_accuracy": 0.11466511264443398,
      "num_tokens": 8243182.0,
      "step": 2465
    },
    {
      "entropy": 6.923015356063843,
      "epoch": 0.4819042044678568,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004990517596259684,
      "loss": 6.9206,
      "mean_token_accuracy": 0.1125547505915165,
      "num_tokens": 8259185.0,
      "step": 2470
    },
    {
      "entropy": 6.927219390869141,
      "epoch": 0.4828797190517998,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990452982111781,
      "loss": 6.983,
      "mean_token_accuracy": 0.11010222509503365,
      "num_tokens": 8275941.0,
      "step": 2475
    },
    {
      "entropy": 6.935208940505982,
      "epoch": 0.48385523363574284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990388149034797,
      "loss": 6.9253,
      "mean_token_accuracy": 0.11687990576028824,
      "num_tokens": 8292535.0,
      "step": 2480
    },
    {
      "entropy": 6.887890434265136,
      "epoch": 0.48483074821968586,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004990323097035069,
      "loss": 6.8769,
      "mean_token_accuracy": 0.12070949226617814,
      "num_tokens": 8310562.0,
      "step": 2485
    },
    {
      "entropy": 6.870615386962891,
      "epoch": 0.48580626280362893,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004990257826118952,
      "loss": 6.9097,
      "mean_token_accuracy": 0.1069605253636837,
      "num_tokens": 8327241.0,
      "step": 2490
    },
    {
      "entropy": 6.865695381164551,
      "epoch": 0.48678177738757195,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990192336292824,
      "loss": 6.8423,
      "mean_token_accuracy": 0.12054572403430938,
      "num_tokens": 8343516.0,
      "step": 2495
    },
    {
      "entropy": 6.737391328811645,
      "epoch": 0.48775729197151496,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990126627563087,
      "loss": 6.709,
      "mean_token_accuracy": 0.12692616656422614,
      "num_tokens": 8361809.0,
      "step": 2500
    },
    {
      "entropy": 6.767363262176514,
      "epoch": 0.488732806555458,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990060699936159,
      "loss": 6.7707,
      "mean_token_accuracy": 0.13018333539366722,
      "num_tokens": 8378632.0,
      "step": 2505
    },
    {
      "entropy": 6.899938678741455,
      "epoch": 0.48970832113940105,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004989994553418483,
      "loss": 6.9353,
      "mean_token_accuracy": 0.110991932451725,
      "num_tokens": 8396201.0,
      "step": 2510
    },
    {
      "entropy": 6.8594200134277346,
      "epoch": 0.49068383572334406,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004989928188016522,
      "loss": 6.8072,
      "mean_token_accuracy": 0.12356401160359383,
      "num_tokens": 8412851.0,
      "step": 2515
    },
    {
      "entropy": 6.718408012390137,
      "epoch": 0.4916593503072871,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989861603736762,
      "loss": 6.6376,
      "mean_token_accuracy": 0.13155316412448884,
      "num_tokens": 8430026.0,
      "step": 2520
    },
    {
      "entropy": 6.828795909881592,
      "epoch": 0.49263486489123015,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989794800585711,
      "loss": 6.8795,
      "mean_token_accuracy": 0.11511812061071396,
      "num_tokens": 8446696.0,
      "step": 2525
    },
    {
      "entropy": 6.973077297210693,
      "epoch": 0.49361037947517317,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989727778569893,
      "loss": 6.9693,
      "mean_token_accuracy": 0.11310434341430664,
      "num_tokens": 8464827.0,
      "step": 2530
    },
    {
      "entropy": 6.9201208591461185,
      "epoch": 0.4945858940591162,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989660537695859,
      "loss": 6.9251,
      "mean_token_accuracy": 0.11566059663891792,
      "num_tokens": 8480185.0,
      "step": 2535
    },
    {
      "entropy": 6.761337852478027,
      "epoch": 0.4955614086430592,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498959307797018,
      "loss": 6.8128,
      "mean_token_accuracy": 0.12708098366856574,
      "num_tokens": 8497511.0,
      "step": 2540
    },
    {
      "entropy": 6.904820108413697,
      "epoch": 0.49653692322700227,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989525399399447,
      "loss": 6.8702,
      "mean_token_accuracy": 0.11620474085211754,
      "num_tokens": 8514394.0,
      "step": 2545
    },
    {
      "entropy": 6.879853820800781,
      "epoch": 0.4975124378109453,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004989457501990275,
      "loss": 6.8792,
      "mean_token_accuracy": 0.11609909012913704,
      "num_tokens": 8531659.0,
      "step": 2550
    },
    {
      "entropy": 6.903630208969116,
      "epoch": 0.4984879523948883,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004989389385749297,
      "loss": 6.9287,
      "mean_token_accuracy": 0.11620916053652763,
      "num_tokens": 8547986.0,
      "step": 2555
    },
    {
      "entropy": 6.872329998016357,
      "epoch": 0.4994634669788313,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004989321050683169,
      "loss": 6.8751,
      "mean_token_accuracy": 0.11848173439502716,
      "num_tokens": 8565021.0,
      "step": 2560
    },
    {
      "entropy": 6.9056610584259035,
      "epoch": 0.5004389815627743,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498925249679857,
      "loss": 6.8744,
      "mean_token_accuracy": 0.11459636986255646,
      "num_tokens": 8582135.0,
      "step": 2565
    },
    {
      "entropy": 6.792390298843384,
      "epoch": 0.5014144961467174,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989183724102196,
      "loss": 6.8993,
      "mean_token_accuracy": 0.11787593364715576,
      "num_tokens": 8599946.0,
      "step": 2570
    },
    {
      "entropy": 6.958661413192749,
      "epoch": 0.5023900107306605,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004989114732600771,
      "loss": 6.923,
      "mean_token_accuracy": 0.11336629018187523,
      "num_tokens": 8615831.0,
      "step": 2575
    },
    {
      "entropy": 6.779950428009033,
      "epoch": 0.5033655253146034,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989045522301033,
      "loss": 6.8112,
      "mean_token_accuracy": 0.11558634489774704,
      "num_tokens": 8631886.0,
      "step": 2580
    },
    {
      "entropy": 6.94764461517334,
      "epoch": 0.5043410398985465,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988976093209747,
      "loss": 6.8613,
      "mean_token_accuracy": 0.11486466601490974,
      "num_tokens": 8647520.0,
      "step": 2585
    },
    {
      "entropy": 6.8138525009155275,
      "epoch": 0.5053165544824895,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988906445333698,
      "loss": 6.7713,
      "mean_token_accuracy": 0.12794217690825463,
      "num_tokens": 8663995.0,
      "step": 2590
    },
    {
      "entropy": 6.822241306304932,
      "epoch": 0.5062920690664325,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988836578679691,
      "loss": 6.8511,
      "mean_token_accuracy": 0.11445828303694724,
      "num_tokens": 8680201.0,
      "step": 2595
    },
    {
      "entropy": 6.8948572158813475,
      "epoch": 0.5072675836503756,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004988766493254552,
      "loss": 6.8433,
      "mean_token_accuracy": 0.12341629490256309,
      "num_tokens": 8696978.0,
      "step": 2600
    },
    {
      "entropy": 6.804700136184692,
      "epoch": 0.5082430982343186,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498869618906513,
      "loss": 6.8946,
      "mean_token_accuracy": 0.11857298016548157,
      "num_tokens": 8713335.0,
      "step": 2605
    },
    {
      "entropy": 6.870307731628418,
      "epoch": 0.5092186128182616,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988625666118296,
      "loss": 6.8823,
      "mean_token_accuracy": 0.1196676455438137,
      "num_tokens": 8730722.0,
      "step": 2610
    },
    {
      "entropy": 6.774832820892334,
      "epoch": 0.5101941274022047,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498855492442094,
      "loss": 6.7952,
      "mean_token_accuracy": 0.12054291889071464,
      "num_tokens": 8747430.0,
      "step": 2615
    },
    {
      "entropy": 6.9103686809539795,
      "epoch": 0.5111696419861477,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004988483963979975,
      "loss": 6.8884,
      "mean_token_accuracy": 0.11560614630579949,
      "num_tokens": 8764817.0,
      "step": 2620
    },
    {
      "entropy": 6.927261114120483,
      "epoch": 0.5121451565700907,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988412784802336,
      "loss": 6.9015,
      "mean_token_accuracy": 0.11360720023512841,
      "num_tokens": 8780412.0,
      "step": 2625
    },
    {
      "entropy": 6.932803821563721,
      "epoch": 0.5131206711540337,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988341386894977,
      "loss": 6.9679,
      "mean_token_accuracy": 0.1152969904243946,
      "num_tokens": 8795526.0,
      "step": 2630
    },
    {
      "entropy": 6.7840962409973145,
      "epoch": 0.5140961857379768,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004988269770264876,
      "loss": 6.7672,
      "mean_token_accuracy": 0.12347180098295212,
      "num_tokens": 8813568.0,
      "step": 2635
    },
    {
      "entropy": 6.938270092010498,
      "epoch": 0.5150717003219198,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004988197934919029,
      "loss": 6.9232,
      "mean_token_accuracy": 0.1144092097878456,
      "num_tokens": 8831807.0,
      "step": 2640
    },
    {
      "entropy": 6.955628967285156,
      "epoch": 0.5160472149058628,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988125880864458,
      "loss": 6.9261,
      "mean_token_accuracy": 0.11702951192855834,
      "num_tokens": 8848875.0,
      "step": 2645
    },
    {
      "entropy": 6.676701450347901,
      "epoch": 0.5170227294898059,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988053608108203,
      "loss": 6.6355,
      "mean_token_accuracy": 0.12377685308456421,
      "num_tokens": 8866317.0,
      "step": 2650
    },
    {
      "entropy": 6.782362079620361,
      "epoch": 0.517998244073749,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987981116657325,
      "loss": 6.7683,
      "mean_token_accuracy": 0.11874385476112366,
      "num_tokens": 8882543.0,
      "step": 2655
    },
    {
      "entropy": 6.862992858886718,
      "epoch": 0.5189737586576919,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987908406518909,
      "loss": 6.7972,
      "mean_token_accuracy": 0.11631387770175934,
      "num_tokens": 8898640.0,
      "step": 2660
    },
    {
      "entropy": 6.851989984512329,
      "epoch": 0.519949273241635,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000498783547770006,
      "loss": 6.9017,
      "mean_token_accuracy": 0.11561635807156563,
      "num_tokens": 8914649.0,
      "step": 2665
    },
    {
      "entropy": 6.909957075119019,
      "epoch": 0.5209247878255779,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987762330207903,
      "loss": 6.9236,
      "mean_token_accuracy": 0.11679632440209389,
      "num_tokens": 8931585.0,
      "step": 2670
    },
    {
      "entropy": 6.771302366256714,
      "epoch": 0.521900302409521,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987688964049587,
      "loss": 6.7928,
      "mean_token_accuracy": 0.12180419787764549,
      "num_tokens": 8948827.0,
      "step": 2675
    },
    {
      "entropy": 6.790069627761841,
      "epoch": 0.5228758169934641,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000498761537923228,
      "loss": 6.7745,
      "mean_token_accuracy": 0.1269575834274292,
      "num_tokens": 8965680.0,
      "step": 2680
    },
    {
      "entropy": 6.952053117752075,
      "epoch": 0.523851331577407,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987541575763175,
      "loss": 6.9065,
      "mean_token_accuracy": 0.11049905940890312,
      "num_tokens": 8982813.0,
      "step": 2685
    },
    {
      "entropy": 6.867544031143188,
      "epoch": 0.5248268461613501,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498746755364948,
      "loss": 6.8833,
      "mean_token_accuracy": 0.11713560670614243,
      "num_tokens": 9000497.0,
      "step": 2690
    },
    {
      "entropy": 6.830367136001587,
      "epoch": 0.5258023607452932,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004987393312898431,
      "loss": 6.8102,
      "mean_token_accuracy": 0.11786004677414894,
      "num_tokens": 9016195.0,
      "step": 2695
    },
    {
      "entropy": 6.761811113357544,
      "epoch": 0.5267778753292361,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004987318853517282,
      "loss": 6.7723,
      "mean_token_accuracy": 0.12615545392036437,
      "num_tokens": 9033295.0,
      "step": 2700
    },
    {
      "entropy": 6.844594049453735,
      "epoch": 0.5277533899131792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987244175513308,
      "loss": 6.872,
      "mean_token_accuracy": 0.1141127772629261,
      "num_tokens": 9050725.0,
      "step": 2705
    },
    {
      "entropy": 6.833688688278198,
      "epoch": 0.5287289044971223,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987169278893807,
      "loss": 6.8168,
      "mean_token_accuracy": 0.11935415118932724,
      "num_tokens": 9067842.0,
      "step": 2710
    },
    {
      "entropy": 6.86349515914917,
      "epoch": 0.5297044190810652,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004987094163666098,
      "loss": 6.8351,
      "mean_token_accuracy": 0.11330854967236519,
      "num_tokens": 9084602.0,
      "step": 2715
    },
    {
      "entropy": 6.789287185668945,
      "epoch": 0.5306799336650083,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498701882983752,
      "loss": 6.7657,
      "mean_token_accuracy": 0.12709601148962973,
      "num_tokens": 9100530.0,
      "step": 2720
    },
    {
      "entropy": 6.867527198791504,
      "epoch": 0.5316554482489513,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986943277415435,
      "loss": 6.8614,
      "mean_token_accuracy": 0.12204445227980613,
      "num_tokens": 9117304.0,
      "step": 2725
    },
    {
      "entropy": 6.752386522293091,
      "epoch": 0.5326309628328944,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986867506407225,
      "loss": 6.806,
      "mean_token_accuracy": 0.13167238086462021,
      "num_tokens": 9133974.0,
      "step": 2730
    },
    {
      "entropy": 6.860329008102417,
      "epoch": 0.5336064774168374,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004986791516820295,
      "loss": 6.772,
      "mean_token_accuracy": 0.12175974771380424,
      "num_tokens": 9150600.0,
      "step": 2735
    },
    {
      "entropy": 6.737002849578857,
      "epoch": 0.5345819920007804,
      "grad_norm": 0.765625,
      "learning_rate": 0.000498671530866207,
      "loss": 6.7722,
      "mean_token_accuracy": 0.11789723262190818,
      "num_tokens": 9167449.0,
      "step": 2740
    },
    {
      "entropy": 6.818783092498779,
      "epoch": 0.5355575065847235,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986638881939997,
      "loss": 6.7219,
      "mean_token_accuracy": 0.12201179042458535,
      "num_tokens": 9184570.0,
      "step": 2745
    },
    {
      "entropy": 6.784221220016479,
      "epoch": 0.5365330211686665,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986562236661544,
      "loss": 6.9045,
      "mean_token_accuracy": 0.10912108793854713,
      "num_tokens": 9200537.0,
      "step": 2750
    },
    {
      "entropy": 6.902917003631591,
      "epoch": 0.5375085357526095,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00049864853728342,
      "loss": 6.8037,
      "mean_token_accuracy": 0.12473463490605355,
      "num_tokens": 9217252.0,
      "step": 2755
    },
    {
      "entropy": 6.864479970932007,
      "epoch": 0.5384840503365526,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004986408290465476,
      "loss": 6.8555,
      "mean_token_accuracy": 0.11590055152773857,
      "num_tokens": 9234108.0,
      "step": 2760
    },
    {
      "entropy": 6.812731981277466,
      "epoch": 0.5394595649204955,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986330989562906,
      "loss": 6.7575,
      "mean_token_accuracy": 0.12097472175955773,
      "num_tokens": 9251017.0,
      "step": 2765
    },
    {
      "entropy": 6.7629071235656735,
      "epoch": 0.5404350795044386,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004986253470134042,
      "loss": 6.7091,
      "mean_token_accuracy": 0.1275697208940983,
      "num_tokens": 9267922.0,
      "step": 2770
    },
    {
      "entropy": 6.700593709945679,
      "epoch": 0.5414105940883817,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004986175732186459,
      "loss": 6.703,
      "mean_token_accuracy": 0.13123294040560723,
      "num_tokens": 9284954.0,
      "step": 2775
    },
    {
      "entropy": 6.885139465332031,
      "epoch": 0.5423861086723246,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986097775727752,
      "loss": 6.8334,
      "mean_token_accuracy": 0.12190918251872063,
      "num_tokens": 9303306.0,
      "step": 2780
    },
    {
      "entropy": 6.733374404907226,
      "epoch": 0.5433616232562677,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986019600765541,
      "loss": 6.7219,
      "mean_token_accuracy": 0.12711875960230828,
      "num_tokens": 9319754.0,
      "step": 2785
    },
    {
      "entropy": 6.743921375274658,
      "epoch": 0.5443371378402108,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985941207307462,
      "loss": 6.8122,
      "mean_token_accuracy": 0.12274460345506669,
      "num_tokens": 9335736.0,
      "step": 2790
    },
    {
      "entropy": 6.8163683891296385,
      "epoch": 0.5453126524241537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985862595361179,
      "loss": 6.7784,
      "mean_token_accuracy": 0.12290315851569175,
      "num_tokens": 9350435.0,
      "step": 2795
    },
    {
      "entropy": 6.86185622215271,
      "epoch": 0.5462881670080968,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985783764934372,
      "loss": 6.8427,
      "mean_token_accuracy": 0.11779003739356994,
      "num_tokens": 9367613.0,
      "step": 2800
    },
    {
      "entropy": 6.787689065933227,
      "epoch": 0.5472636815920398,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985704716034744,
      "loss": 6.749,
      "mean_token_accuracy": 0.1232517659664154,
      "num_tokens": 9385590.0,
      "step": 2805
    },
    {
      "entropy": 6.836340427398682,
      "epoch": 0.5482391961759828,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004985625448670019,
      "loss": 6.8774,
      "mean_token_accuracy": 0.11052649319171906,
      "num_tokens": 9403286.0,
      "step": 2810
    },
    {
      "entropy": 6.910646963119507,
      "epoch": 0.5492147107599259,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985545962847943,
      "loss": 6.8702,
      "mean_token_accuracy": 0.1125256858766079,
      "num_tokens": 9419880.0,
      "step": 2815
    },
    {
      "entropy": 6.7993169784545895,
      "epoch": 0.5501902253438689,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985466258576283,
      "loss": 6.8319,
      "mean_token_accuracy": 0.11432699039578438,
      "num_tokens": 9436579.0,
      "step": 2820
    },
    {
      "entropy": 6.884487915039062,
      "epoch": 0.5511657399278119,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985386335862829,
      "loss": 6.8841,
      "mean_token_accuracy": 0.11265967264771462,
      "num_tokens": 9452802.0,
      "step": 2825
    },
    {
      "entropy": 6.851714277267456,
      "epoch": 0.552141254511755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985306194715387,
      "loss": 6.8927,
      "mean_token_accuracy": 0.11446214690804482,
      "num_tokens": 9468297.0,
      "step": 2830
    },
    {
      "entropy": 6.883763599395752,
      "epoch": 0.553116769095698,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985225835141792,
      "loss": 6.8999,
      "mean_token_accuracy": 0.11260369718074799,
      "num_tokens": 9484954.0,
      "step": 2835
    },
    {
      "entropy": 6.9504420280456545,
      "epoch": 0.554092283679641,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985145257149895,
      "loss": 6.9496,
      "mean_token_accuracy": 0.1100885510444641,
      "num_tokens": 9502211.0,
      "step": 2840
    },
    {
      "entropy": 6.860951280593872,
      "epoch": 0.555067798263584,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004985064460747569,
      "loss": 6.8337,
      "mean_token_accuracy": 0.11309531480073928,
      "num_tokens": 9518103.0,
      "step": 2845
    },
    {
      "entropy": 6.764395809173584,
      "epoch": 0.5560433128475271,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498498344594271,
      "loss": 6.7779,
      "mean_token_accuracy": 0.11637641638517379,
      "num_tokens": 9534155.0,
      "step": 2850
    },
    {
      "entropy": 6.738689994812011,
      "epoch": 0.5570188274314701,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984902212743234,
      "loss": 6.7507,
      "mean_token_accuracy": 0.12972750812768935,
      "num_tokens": 9549972.0,
      "step": 2855
    },
    {
      "entropy": 6.696974754333496,
      "epoch": 0.5579943420154131,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984820761157079,
      "loss": 6.6617,
      "mean_token_accuracy": 0.13014516234397888,
      "num_tokens": 9567306.0,
      "step": 2860
    },
    {
      "entropy": 6.883365678787231,
      "epoch": 0.5589698565993562,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984739091192205,
      "loss": 6.885,
      "mean_token_accuracy": 0.11657564714550972,
      "num_tokens": 9583488.0,
      "step": 2865
    },
    {
      "entropy": 6.694728088378906,
      "epoch": 0.5599453711832992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984657202856592,
      "loss": 6.6662,
      "mean_token_accuracy": 0.13098273277282715,
      "num_tokens": 9600030.0,
      "step": 2870
    },
    {
      "entropy": 6.725077056884766,
      "epoch": 0.5609208857672422,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498457509615824,
      "loss": 6.7896,
      "mean_token_accuracy": 0.12264834940433503,
      "num_tokens": 9615791.0,
      "step": 2875
    },
    {
      "entropy": 6.910763216018677,
      "epoch": 0.5618964003511853,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984492771105175,
      "loss": 6.8578,
      "mean_token_accuracy": 0.11508372128009796,
      "num_tokens": 9630955.0,
      "step": 2880
    },
    {
      "entropy": 6.71947979927063,
      "epoch": 0.5628719149351282,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984410227705439,
      "loss": 6.7604,
      "mean_token_accuracy": 0.1295547068119049,
      "num_tokens": 9647743.0,
      "step": 2885
    },
    {
      "entropy": 6.745148420333862,
      "epoch": 0.5638474295190713,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004984327465967101,
      "loss": 6.7171,
      "mean_token_accuracy": 0.12684635668992997,
      "num_tokens": 9665495.0,
      "step": 2890
    },
    {
      "entropy": 6.785642194747925,
      "epoch": 0.5648229441030144,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984244485898246,
      "loss": 6.7367,
      "mean_token_accuracy": 0.12310076728463173,
      "num_tokens": 9683489.0,
      "step": 2895
    },
    {
      "entropy": 6.579546022415161,
      "epoch": 0.5657984586869573,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004984161287506982,
      "loss": 6.6762,
      "mean_token_accuracy": 0.1427179493010044,
      "num_tokens": 9700491.0,
      "step": 2900
    },
    {
      "entropy": 6.925921773910522,
      "epoch": 0.5667739732709004,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498407787080144,
      "loss": 6.8309,
      "mean_token_accuracy": 0.11485328450798989,
      "num_tokens": 9715694.0,
      "step": 2905
    },
    {
      "entropy": 6.817408847808838,
      "epoch": 0.5677494878548435,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498399423578977,
      "loss": 6.8899,
      "mean_token_accuracy": 0.11257752552628517,
      "num_tokens": 9731317.0,
      "step": 2910
    },
    {
      "entropy": 6.870456409454346,
      "epoch": 0.5687250024387864,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983910382480146,
      "loss": 6.8671,
      "mean_token_accuracy": 0.11242569237947464,
      "num_tokens": 9747928.0,
      "step": 2915
    },
    {
      "entropy": 6.837937736511231,
      "epoch": 0.5697005170227295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983826310880762,
      "loss": 6.8039,
      "mean_token_accuracy": 0.11309616193175316,
      "num_tokens": 9765454.0,
      "step": 2920
    },
    {
      "entropy": 6.897300434112549,
      "epoch": 0.5706760316066726,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983742020999832,
      "loss": 6.8669,
      "mean_token_accuracy": 0.10900059267878533,
      "num_tokens": 9782255.0,
      "step": 2925
    },
    {
      "entropy": 6.696663665771484,
      "epoch": 0.5716515461906155,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004983657512845593,
      "loss": 6.6753,
      "mean_token_accuracy": 0.12878145053982734,
      "num_tokens": 9798680.0,
      "step": 2930
    },
    {
      "entropy": 6.806471347808838,
      "epoch": 0.5726270607745586,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983572786426302,
      "loss": 6.8175,
      "mean_token_accuracy": 0.11593940034508705,
      "num_tokens": 9813955.0,
      "step": 2935
    },
    {
      "entropy": 6.701576662063599,
      "epoch": 0.5736025753585016,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983487841750239,
      "loss": 6.711,
      "mean_token_accuracy": 0.129136098921299,
      "num_tokens": 9830692.0,
      "step": 2940
    },
    {
      "entropy": 6.839662647247314,
      "epoch": 0.5745780899424446,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004983402678825705,
      "loss": 6.8308,
      "mean_token_accuracy": 0.11594207361340522,
      "num_tokens": 9848825.0,
      "step": 2945
    },
    {
      "entropy": 6.809984397888184,
      "epoch": 0.5755536045263877,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983317297661021,
      "loss": 6.8305,
      "mean_token_accuracy": 0.12007160931825638,
      "num_tokens": 9865802.0,
      "step": 2950
    },
    {
      "entropy": 6.848187828063965,
      "epoch": 0.5765291191103307,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004983231698264531,
      "loss": 6.8632,
      "mean_token_accuracy": 0.11418323591351509,
      "num_tokens": 9883540.0,
      "step": 2955
    },
    {
      "entropy": 6.871307277679444,
      "epoch": 0.5775046336942737,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983145880644597,
      "loss": 6.8059,
      "mean_token_accuracy": 0.1159649781882763,
      "num_tokens": 9897724.0,
      "step": 2960
    },
    {
      "entropy": 6.807016468048095,
      "epoch": 0.5784801482782168,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983059844809608,
      "loss": 6.7525,
      "mean_token_accuracy": 0.12290950119495392,
      "num_tokens": 9914710.0,
      "step": 2965
    },
    {
      "entropy": 6.785981321334839,
      "epoch": 0.5794556628621598,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004982973590767969,
      "loss": 6.8696,
      "mean_token_accuracy": 0.12116017490625382,
      "num_tokens": 9932362.0,
      "step": 2970
    },
    {
      "entropy": 6.751751852035523,
      "epoch": 0.5804311774461028,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498288711852811,
      "loss": 6.6607,
      "mean_token_accuracy": 0.12567465156316757,
      "num_tokens": 9949409.0,
      "step": 2975
    },
    {
      "entropy": 6.827619361877441,
      "epoch": 0.5814066920300458,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000498280042809848,
      "loss": 6.8126,
      "mean_token_accuracy": 0.11727285012602806,
      "num_tokens": 9967467.0,
      "step": 2980
    },
    {
      "entropy": 6.7890321731567385,
      "epoch": 0.5823822066139889,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982713519487549,
      "loss": 6.7552,
      "mean_token_accuracy": 0.1225271537899971,
      "num_tokens": 9983082.0,
      "step": 2985
    },
    {
      "entropy": 6.790068006515503,
      "epoch": 0.5833577211979319,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004982626392703812,
      "loss": 6.8259,
      "mean_token_accuracy": 0.11854807063937187,
      "num_tokens": 10000046.0,
      "step": 2990
    },
    {
      "entropy": 6.802617406845092,
      "epoch": 0.5843332357818749,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498253904775578,
      "loss": 6.7577,
      "mean_token_accuracy": 0.11755326986312867,
      "num_tokens": 10014674.0,
      "step": 2995
    },
    {
      "entropy": 6.694525718688965,
      "epoch": 0.585308750365818,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982451484651988,
      "loss": 6.6491,
      "mean_token_accuracy": 0.12509795725345613,
      "num_tokens": 10030647.0,
      "step": 3000
    },
    {
      "epoch": 0.585308750365818,
      "eval_entropy": 6.631029434728553,
      "eval_loss": 6.8121867179870605,
      "eval_mean_token_accuracy": 0.12184531028524896,
      "eval_num_tokens": 10030647.0,
      "eval_runtime": 15.3329,
      "eval_samples_per_second": 1081.139,
      "eval_steps_per_second": 135.199,
      "step": 3000
    },
    {
      "entropy": 6.831427907943725,
      "epoch": 0.586284264949761,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982363703400995,
      "loss": 6.7257,
      "mean_token_accuracy": 0.12281443700194358,
      "num_tokens": 10046468.0,
      "step": 3005
    },
    {
      "entropy": 6.7997771263122555,
      "epoch": 0.587259779533704,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004982275704011377,
      "loss": 6.846,
      "mean_token_accuracy": 0.11487425044178963,
      "num_tokens": 10063062.0,
      "step": 3010
    },
    {
      "entropy": 6.789147233963012,
      "epoch": 0.5882352941176471,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982187486491734,
      "loss": 6.7879,
      "mean_token_accuracy": 0.11969002038240432,
      "num_tokens": 10079616.0,
      "step": 3015
    },
    {
      "entropy": 6.757223892211914,
      "epoch": 0.58921080870159,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982099050850685,
      "loss": 6.7546,
      "mean_token_accuracy": 0.12014183327555657,
      "num_tokens": 10096830.0,
      "step": 3020
    },
    {
      "entropy": 6.7580506801605225,
      "epoch": 0.5901863232855331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982010397096871,
      "loss": 6.7953,
      "mean_token_accuracy": 0.12178919091820717,
      "num_tokens": 10114041.0,
      "step": 3025
    },
    {
      "entropy": 6.746286010742187,
      "epoch": 0.5911618378694762,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981921525238957,
      "loss": 6.7098,
      "mean_token_accuracy": 0.12440267056226731,
      "num_tokens": 10130075.0,
      "step": 3030
    },
    {
      "entropy": 6.799335193634033,
      "epoch": 0.5921373524534191,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004981832435285626,
      "loss": 6.8539,
      "mean_token_accuracy": 0.11827502772212029,
      "num_tokens": 10146716.0,
      "step": 3035
    },
    {
      "entropy": 6.807714986801147,
      "epoch": 0.5931128670373622,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981743127245584,
      "loss": 6.7681,
      "mean_token_accuracy": 0.12098332419991494,
      "num_tokens": 10162116.0,
      "step": 3040
    },
    {
      "entropy": 6.873652029037475,
      "epoch": 0.5940883816213053,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981653601127557,
      "loss": 6.8731,
      "mean_token_accuracy": 0.11577322036027908,
      "num_tokens": 10179326.0,
      "step": 3045
    },
    {
      "entropy": 6.81105785369873,
      "epoch": 0.5950638962052482,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004981563856940295,
      "loss": 6.7972,
      "mean_token_accuracy": 0.12193589136004448,
      "num_tokens": 10195278.0,
      "step": 3050
    },
    {
      "entropy": 6.740652751922608,
      "epoch": 0.5960394107891913,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981473894692565,
      "loss": 6.7315,
      "mean_token_accuracy": 0.12532268688082696,
      "num_tokens": 10210175.0,
      "step": 3055
    },
    {
      "entropy": 6.789959907531738,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498138371439316,
      "loss": 6.8322,
      "mean_token_accuracy": 0.11782603338360786,
      "num_tokens": 10225600.0,
      "step": 3060
    },
    {
      "entropy": 6.65930027961731,
      "epoch": 0.5979904399570773,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004981293316050891,
      "loss": 6.5408,
      "mean_token_accuracy": 0.1448441632091999,
      "num_tokens": 10243451.0,
      "step": 3065
    },
    {
      "entropy": 6.798199939727783,
      "epoch": 0.5989659545410204,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004981202699674592,
      "loss": 6.7971,
      "mean_token_accuracy": 0.1183489978313446,
      "num_tokens": 10260392.0,
      "step": 3070
    },
    {
      "entropy": 6.853112602233887,
      "epoch": 0.5999414691249634,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004981111865273117,
      "loss": 6.8457,
      "mean_token_accuracy": 0.11489525362849236,
      "num_tokens": 10277016.0,
      "step": 3075
    },
    {
      "entropy": 6.661908054351807,
      "epoch": 0.6009169837089064,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004981020812855342,
      "loss": 6.7094,
      "mean_token_accuracy": 0.13075108528137208,
      "num_tokens": 10295927.0,
      "step": 3080
    },
    {
      "entropy": 6.8182652473449705,
      "epoch": 0.6018924982928495,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980929542430165,
      "loss": 6.7862,
      "mean_token_accuracy": 0.11778303235769272,
      "num_tokens": 10311140.0,
      "step": 3085
    },
    {
      "entropy": 6.795322895050049,
      "epoch": 0.6028680128767925,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004980838054006504,
      "loss": 6.8328,
      "mean_token_accuracy": 0.12109225764870643,
      "num_tokens": 10328496.0,
      "step": 3090
    },
    {
      "entropy": 6.845468139648437,
      "epoch": 0.6038435274607356,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049807463475933,
      "loss": 6.7725,
      "mean_token_accuracy": 0.11572813838720322,
      "num_tokens": 10345232.0,
      "step": 3095
    },
    {
      "entropy": 6.8890955448150635,
      "epoch": 0.6048190420446786,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980654423199513,
      "loss": 6.9109,
      "mean_token_accuracy": 0.11818017661571503,
      "num_tokens": 10361777.0,
      "step": 3100
    },
    {
      "entropy": 6.775675249099732,
      "epoch": 0.6057945566286216,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980562280834126,
      "loss": 6.8211,
      "mean_token_accuracy": 0.1202052615582943,
      "num_tokens": 10377117.0,
      "step": 3105
    },
    {
      "entropy": 6.572078657150269,
      "epoch": 0.6067700712125647,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004980469920506142,
      "loss": 6.4599,
      "mean_token_accuracy": 0.14576496481895446,
      "num_tokens": 10394506.0,
      "step": 3110
    },
    {
      "entropy": 6.693164539337158,
      "epoch": 0.6077455857965076,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980377342224588,
      "loss": 6.6,
      "mean_token_accuracy": 0.13376055136322976,
      "num_tokens": 10410281.0,
      "step": 3115
    },
    {
      "entropy": 6.780955123901367,
      "epoch": 0.6087211003804507,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980284545998509,
      "loss": 6.8127,
      "mean_token_accuracy": 0.11629964932799339,
      "num_tokens": 10425165.0,
      "step": 3120
    },
    {
      "entropy": 6.8751438617706295,
      "epoch": 0.6096966149643938,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980191531836973,
      "loss": 6.7842,
      "mean_token_accuracy": 0.11518170237541199,
      "num_tokens": 10441643.0,
      "step": 3125
    },
    {
      "entropy": 6.769915437698364,
      "epoch": 0.6106721295483367,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980098299749068,
      "loss": 6.8082,
      "mean_token_accuracy": 0.1134133942425251,
      "num_tokens": 10457436.0,
      "step": 3130
    },
    {
      "entropy": 6.830953073501587,
      "epoch": 0.6116476441322798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980004849743908,
      "loss": 6.8179,
      "mean_token_accuracy": 0.11666478887200356,
      "num_tokens": 10475020.0,
      "step": 3135
    },
    {
      "entropy": 6.540660333633423,
      "epoch": 0.6126231587162229,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497991118183062,
      "loss": 6.5744,
      "mean_token_accuracy": 0.13344746008515357,
      "num_tokens": 10492029.0,
      "step": 3140
    },
    {
      "entropy": 6.696297931671142,
      "epoch": 0.6135986733001658,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004979817296018359,
      "loss": 6.6176,
      "mean_token_accuracy": 0.13196553364396096,
      "num_tokens": 10510145.0,
      "step": 3145
    },
    {
      "entropy": 6.795686435699463,
      "epoch": 0.6145741878841089,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004979723192316299,
      "loss": 6.866,
      "mean_token_accuracy": 0.12016671150922775,
      "num_tokens": 10527192.0,
      "step": 3150
    },
    {
      "entropy": 6.852078294754028,
      "epoch": 0.6155497024680519,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979628870733637,
      "loss": 6.815,
      "mean_token_accuracy": 0.11651301831007004,
      "num_tokens": 10542269.0,
      "step": 3155
    },
    {
      "entropy": 6.761844110488892,
      "epoch": 0.6165252170519949,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979534331279586,
      "loss": 6.81,
      "mean_token_accuracy": 0.11997768953442574,
      "num_tokens": 10559425.0,
      "step": 3160
    },
    {
      "entropy": 6.8231072425842285,
      "epoch": 0.617500731635938,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979439573963388,
      "loss": 6.8084,
      "mean_token_accuracy": 0.1194613479077816,
      "num_tokens": 10576818.0,
      "step": 3165
    },
    {
      "entropy": 6.648348093032837,
      "epoch": 0.618476246219881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979344598794299,
      "loss": 6.7055,
      "mean_token_accuracy": 0.12330605536699295,
      "num_tokens": 10593456.0,
      "step": 3170
    },
    {
      "entropy": 6.83094367980957,
      "epoch": 0.619451760803824,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979249405781602,
      "loss": 6.8872,
      "mean_token_accuracy": 0.11359044089913369,
      "num_tokens": 10609049.0,
      "step": 3175
    },
    {
      "entropy": 6.758004474639892,
      "epoch": 0.6204272753877671,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004979153994934599,
      "loss": 6.7246,
      "mean_token_accuracy": 0.12095132991671562,
      "num_tokens": 10626921.0,
      "step": 3180
    },
    {
      "entropy": 6.763813543319702,
      "epoch": 0.6214027899717101,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979058366262613,
      "loss": 6.772,
      "mean_token_accuracy": 0.11619191244244576,
      "num_tokens": 10643447.0,
      "step": 3185
    },
    {
      "entropy": 6.845756721496582,
      "epoch": 0.6223783045556531,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004978962519774987,
      "loss": 6.839,
      "mean_token_accuracy": 0.11674403473734855,
      "num_tokens": 10659896.0,
      "step": 3190
    },
    {
      "entropy": 6.755891180038452,
      "epoch": 0.6233538191395961,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004978866455481087,
      "loss": 6.6889,
      "mean_token_accuracy": 0.1229845218360424,
      "num_tokens": 10675331.0,
      "step": 3195
    },
    {
      "entropy": 6.801825952529907,
      "epoch": 0.6243293337235392,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004978770173390301,
      "loss": 6.8546,
      "mean_token_accuracy": 0.11560527086257935,
      "num_tokens": 10692678.0,
      "step": 3200
    },
    {
      "entropy": 6.8644414901733395,
      "epoch": 0.6253048483074822,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978673673512039,
      "loss": 6.7796,
      "mean_token_accuracy": 0.11576822251081467,
      "num_tokens": 10708500.0,
      "step": 3205
    },
    {
      "entropy": 6.689521837234497,
      "epoch": 0.6262803628914252,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978576955855727,
      "loss": 6.7028,
      "mean_token_accuracy": 0.12143276259303093,
      "num_tokens": 10725424.0,
      "step": 3210
    },
    {
      "entropy": 6.811701154708862,
      "epoch": 0.6272558774753683,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004978480020430818,
      "loss": 6.7337,
      "mean_token_accuracy": 0.12571845650672914,
      "num_tokens": 10740960.0,
      "step": 3215
    },
    {
      "entropy": 6.744243907928467,
      "epoch": 0.6282313920593113,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978382867246784,
      "loss": 6.7754,
      "mean_token_accuracy": 0.11489084213972092,
      "num_tokens": 10756426.0,
      "step": 3220
    },
    {
      "entropy": 6.717027282714843,
      "epoch": 0.6292069066432543,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978285496313119,
      "loss": 6.6466,
      "mean_token_accuracy": 0.12705423235893248,
      "num_tokens": 10772748.0,
      "step": 3225
    },
    {
      "entropy": 6.707432460784912,
      "epoch": 0.6301824212271974,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978187907639336,
      "loss": 6.6897,
      "mean_token_accuracy": 0.12716799452900887,
      "num_tokens": 10790024.0,
      "step": 3230
    },
    {
      "entropy": 6.807447862625122,
      "epoch": 0.6311579358111403,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978090101234972,
      "loss": 6.7919,
      "mean_token_accuracy": 0.11302840635180474,
      "num_tokens": 10806280.0,
      "step": 3235
    },
    {
      "entropy": 6.72634449005127,
      "epoch": 0.6321334503950834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004977992077109585,
      "loss": 6.6794,
      "mean_token_accuracy": 0.1255945421755314,
      "num_tokens": 10822166.0,
      "step": 3240
    },
    {
      "entropy": 6.792590188980102,
      "epoch": 0.6331089649790265,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004977893835272754,
      "loss": 6.833,
      "mean_token_accuracy": 0.11566323265433312,
      "num_tokens": 10839833.0,
      "step": 3245
    },
    {
      "entropy": 6.819169521331787,
      "epoch": 0.6340844795629694,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004977795375734075,
      "loss": 6.7841,
      "mean_token_accuracy": 0.11633172258734703,
      "num_tokens": 10856803.0,
      "step": 3250
    },
    {
      "entropy": 6.680344963073731,
      "epoch": 0.6350599941469125,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004977696698503174,
      "loss": 6.6578,
      "mean_token_accuracy": 0.12872755005955697,
      "num_tokens": 10874245.0,
      "step": 3255
    },
    {
      "entropy": 6.74847526550293,
      "epoch": 0.6360355087308556,
      "grad_norm": 0.875,
      "learning_rate": 0.000497759780358969,
      "loss": 6.6544,
      "mean_token_accuracy": 0.12475790232419967,
      "num_tokens": 10890380.0,
      "step": 3260
    },
    {
      "entropy": 6.693929147720337,
      "epoch": 0.6370110233147985,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977498691003288,
      "loss": 6.7306,
      "mean_token_accuracy": 0.11771895214915276,
      "num_tokens": 10906927.0,
      "step": 3265
    },
    {
      "entropy": 6.8399042129516605,
      "epoch": 0.6379865378987416,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004977399360753654,
      "loss": 6.7864,
      "mean_token_accuracy": 0.1158551074564457,
      "num_tokens": 10922922.0,
      "step": 3270
    },
    {
      "entropy": 6.761132431030274,
      "epoch": 0.6389620524826846,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004977299812850492,
      "loss": 6.8122,
      "mean_token_accuracy": 0.11658454686403275,
      "num_tokens": 10939607.0,
      "step": 3275
    },
    {
      "entropy": 6.692282581329346,
      "epoch": 0.6399375670666276,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977200047303532,
      "loss": 6.7229,
      "mean_token_accuracy": 0.12422453016042709,
      "num_tokens": 10956830.0,
      "step": 3280
    },
    {
      "entropy": 6.696627807617188,
      "epoch": 0.6409130816505707,
      "grad_norm": 0.8125,
      "learning_rate": 0.000497710006412252,
      "loss": 6.6727,
      "mean_token_accuracy": 0.1216223694384098,
      "num_tokens": 10974364.0,
      "step": 3285
    },
    {
      "entropy": 6.775454521179199,
      "epoch": 0.6418885962345137,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004976999863317228,
      "loss": 6.734,
      "mean_token_accuracy": 0.1225014053285122,
      "num_tokens": 10990645.0,
      "step": 3290
    },
    {
      "entropy": 6.753648185729981,
      "epoch": 0.6428641108184567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976899444897446,
      "loss": 6.7353,
      "mean_token_accuracy": 0.12722136974334716,
      "num_tokens": 11007952.0,
      "step": 3295
    },
    {
      "entropy": 6.572395467758179,
      "epoch": 0.6438396254023998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976798808872987,
      "loss": 6.6405,
      "mean_token_accuracy": 0.1309990368783474,
      "num_tokens": 11024708.0,
      "step": 3300
    },
    {
      "entropy": 6.799116039276123,
      "epoch": 0.6448151399863428,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976697955253685,
      "loss": 6.7597,
      "mean_token_accuracy": 0.12013562396168709,
      "num_tokens": 11040447.0,
      "step": 3305
    },
    {
      "entropy": 6.77669358253479,
      "epoch": 0.6457906545702858,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976596884049395,
      "loss": 6.7912,
      "mean_token_accuracy": 0.12012482434511185,
      "num_tokens": 11057359.0,
      "step": 3310
    },
    {
      "entropy": 6.712330007553101,
      "epoch": 0.6467661691542289,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976495595269994,
      "loss": 6.6402,
      "mean_token_accuracy": 0.12592253908514978,
      "num_tokens": 11072997.0,
      "step": 3315
    },
    {
      "entropy": 6.786079168319702,
      "epoch": 0.6477416837381719,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976394088925377,
      "loss": 6.8668,
      "mean_token_accuracy": 0.11176015958189964,
      "num_tokens": 11089951.0,
      "step": 3320
    },
    {
      "entropy": 6.691059207916259,
      "epoch": 0.6487171983221149,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976292365025465,
      "loss": 6.5647,
      "mean_token_accuracy": 0.13053592219948768,
      "num_tokens": 11107929.0,
      "step": 3325
    },
    {
      "entropy": 6.536382246017456,
      "epoch": 0.6496927129060579,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976190423580199,
      "loss": 6.531,
      "mean_token_accuracy": 0.13823597580194474,
      "num_tokens": 11125479.0,
      "step": 3330
    },
    {
      "entropy": 6.733331871032715,
      "epoch": 0.650668227490001,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976088264599538,
      "loss": 6.7789,
      "mean_token_accuracy": 0.11511240527033806,
      "num_tokens": 11142316.0,
      "step": 3335
    },
    {
      "entropy": 6.837195873260498,
      "epoch": 0.651643742073944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975985888093467,
      "loss": 6.8475,
      "mean_token_accuracy": 0.11407514736056328,
      "num_tokens": 11158664.0,
      "step": 3340
    },
    {
      "entropy": 6.745405197143555,
      "epoch": 0.652619256657887,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975883294071987,
      "loss": 6.7473,
      "mean_token_accuracy": 0.11854999139904976,
      "num_tokens": 11175732.0,
      "step": 3345
    },
    {
      "entropy": 6.802050542831421,
      "epoch": 0.6535947712418301,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004975780482545125,
      "loss": 6.8198,
      "mean_token_accuracy": 0.11144728809595109,
      "num_tokens": 11191921.0,
      "step": 3350
    },
    {
      "entropy": 6.655164432525635,
      "epoch": 0.6545702858257731,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975677453522928,
      "loss": 6.5951,
      "mean_token_accuracy": 0.13362498506903647,
      "num_tokens": 11208119.0,
      "step": 3355
    },
    {
      "entropy": 6.819881916046143,
      "epoch": 0.6555458004097161,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975574207015462,
      "loss": 6.7646,
      "mean_token_accuracy": 0.12287907153367997,
      "num_tokens": 11224501.0,
      "step": 3360
    },
    {
      "entropy": 6.770030546188354,
      "epoch": 0.6565213149936592,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975470743032817,
      "loss": 6.7925,
      "mean_token_accuracy": 0.12093217000365257,
      "num_tokens": 11240211.0,
      "step": 3365
    },
    {
      "entropy": 6.730515766143799,
      "epoch": 0.6574968295776021,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975367061585102,
      "loss": 6.725,
      "mean_token_accuracy": 0.11983788833022117,
      "num_tokens": 11257921.0,
      "step": 3370
    },
    {
      "entropy": 6.738095045089722,
      "epoch": 0.6584723441615452,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975263162682449,
      "loss": 6.6894,
      "mean_token_accuracy": 0.12706872299313546,
      "num_tokens": 11274526.0,
      "step": 3375
    },
    {
      "entropy": 6.758507633209229,
      "epoch": 0.6594478587454883,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497515904633501,
      "loss": 6.6921,
      "mean_token_accuracy": 0.1189438670873642,
      "num_tokens": 11290883.0,
      "step": 3380
    },
    {
      "entropy": 6.7598652839660645,
      "epoch": 0.6604233733294312,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975054712552963,
      "loss": 6.7486,
      "mean_token_accuracy": 0.12027148604393005,
      "num_tokens": 11307406.0,
      "step": 3385
    },
    {
      "entropy": 6.625966501235962,
      "epoch": 0.6613988879133743,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974950161346496,
      "loss": 6.5738,
      "mean_token_accuracy": 0.12963328287005424,
      "num_tokens": 11322728.0,
      "step": 3390
    },
    {
      "entropy": 6.6055093765258786,
      "epoch": 0.6623744024973174,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497484539272583,
      "loss": 6.5612,
      "mean_token_accuracy": 0.13234575614333152,
      "num_tokens": 11338995.0,
      "step": 3395
    },
    {
      "entropy": 6.609748697280883,
      "epoch": 0.6633499170812603,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004974740406701203,
      "loss": 6.5706,
      "mean_token_accuracy": 0.1349887102842331,
      "num_tokens": 11356154.0,
      "step": 3400
    },
    {
      "entropy": 6.675825500488282,
      "epoch": 0.6643254316652034,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974635203282873,
      "loss": 6.7158,
      "mean_token_accuracy": 0.12272663190960884,
      "num_tokens": 11372755.0,
      "step": 3405
    },
    {
      "entropy": 6.6821647644042965,
      "epoch": 0.6653009462491464,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974529782481119,
      "loss": 6.6678,
      "mean_token_accuracy": 0.12941721081733704,
      "num_tokens": 11389424.0,
      "step": 3410
    },
    {
      "entropy": 6.737720394134522,
      "epoch": 0.6662764608330894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974424144306241,
      "loss": 6.7359,
      "mean_token_accuracy": 0.11891632974147796,
      "num_tokens": 11406515.0,
      "step": 3415
    },
    {
      "entropy": 6.713939046859741,
      "epoch": 0.6672519754170325,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004974318288768566,
      "loss": 6.6583,
      "mean_token_accuracy": 0.13016465604305266,
      "num_tokens": 11423195.0,
      "step": 3420
    },
    {
      "entropy": 6.72865858078003,
      "epoch": 0.6682274900009755,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974212215878435,
      "loss": 6.7531,
      "mean_token_accuracy": 0.1167912319302559,
      "num_tokens": 11440011.0,
      "step": 3425
    },
    {
      "entropy": 6.772357320785522,
      "epoch": 0.6692030045849185,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974105925646213,
      "loss": 6.7513,
      "mean_token_accuracy": 0.1217641294002533,
      "num_tokens": 11456131.0,
      "step": 3430
    },
    {
      "entropy": 6.661199617385864,
      "epoch": 0.6701785191688616,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973999418082287,
      "loss": 6.6747,
      "mean_token_accuracy": 0.12135104089975357,
      "num_tokens": 11471420.0,
      "step": 3435
    },
    {
      "entropy": 6.694121503829956,
      "epoch": 0.6711540337528046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973892693197064,
      "loss": 6.6099,
      "mean_token_accuracy": 0.13167080879211426,
      "num_tokens": 11488453.0,
      "step": 3440
    },
    {
      "entropy": 6.791295194625855,
      "epoch": 0.6721295483367477,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973785751000974,
      "loss": 6.7957,
      "mean_token_accuracy": 0.11769185960292816,
      "num_tokens": 11505144.0,
      "step": 3445
    },
    {
      "entropy": 6.719905948638916,
      "epoch": 0.6731050629206906,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973678591504464,
      "loss": 6.7382,
      "mean_token_accuracy": 0.11878200769424438,
      "num_tokens": 11521249.0,
      "step": 3450
    },
    {
      "entropy": 6.674118423461914,
      "epoch": 0.6740805775046337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497357121471801,
      "loss": 6.6694,
      "mean_token_accuracy": 0.1245422288775444,
      "num_tokens": 11537977.0,
      "step": 3455
    },
    {
      "entropy": 6.7263831615448,
      "epoch": 0.6750560920885768,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049734636206521,
      "loss": 6.7144,
      "mean_token_accuracy": 0.1181706465780735,
      "num_tokens": 11556642.0,
      "step": 3460
    },
    {
      "entropy": 6.798703718185425,
      "epoch": 0.6760316066725197,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000497335580931725,
      "loss": 6.887,
      "mean_token_accuracy": 0.1129963904619217,
      "num_tokens": 11574727.0,
      "step": 3465
    },
    {
      "entropy": 6.670430755615234,
      "epoch": 0.6770071212564628,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004973247780723995,
      "loss": 6.5722,
      "mean_token_accuracy": 0.1312798611819744,
      "num_tokens": 11592982.0,
      "step": 3470
    },
    {
      "entropy": 6.649127674102783,
      "epoch": 0.6779826358404059,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004973139534882891,
      "loss": 6.6513,
      "mean_token_accuracy": 0.13335268571972847,
      "num_tokens": 11609250.0,
      "step": 3475
    },
    {
      "entropy": 6.7355064868927,
      "epoch": 0.6789581504243488,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004973031071804515,
      "loss": 6.7005,
      "mean_token_accuracy": 0.11697419062256813,
      "num_tokens": 11626269.0,
      "step": 3480
    },
    {
      "entropy": 6.731721448898315,
      "epoch": 0.6799336650082919,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004972922391499465,
      "loss": 6.696,
      "mean_token_accuracy": 0.11838576644659042,
      "num_tokens": 11642721.0,
      "step": 3485
    },
    {
      "entropy": 6.55760006904602,
      "epoch": 0.680909179592235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004972813493978361,
      "loss": 6.6276,
      "mean_token_accuracy": 0.1276634491980076,
      "num_tokens": 11659729.0,
      "step": 3490
    },
    {
      "entropy": 6.741422891616821,
      "epoch": 0.6818846941761779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972704379251845,
      "loss": 6.6408,
      "mean_token_accuracy": 0.12111200913786888,
      "num_tokens": 11675568.0,
      "step": 3495
    },
    {
      "entropy": 6.75585560798645,
      "epoch": 0.682860208760121,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000497259504733058,
      "loss": 6.7616,
      "mean_token_accuracy": 0.12588724344968796,
      "num_tokens": 11692827.0,
      "step": 3500
    },
    {
      "entropy": 6.74263105392456,
      "epoch": 0.683835723344064,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004972485498225247,
      "loss": 6.7268,
      "mean_token_accuracy": 0.11614907085895539,
      "num_tokens": 11708730.0,
      "step": 3505
    },
    {
      "entropy": 6.713316535949707,
      "epoch": 0.684811237928007,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972375731946554,
      "loss": 6.7054,
      "mean_token_accuracy": 0.11759109199047088,
      "num_tokens": 11725031.0,
      "step": 3510
    },
    {
      "entropy": 6.7959671974182125,
      "epoch": 0.6857867525119501,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972265748505224,
      "loss": 6.7972,
      "mean_token_accuracy": 0.11676568910479546,
      "num_tokens": 11742419.0,
      "step": 3515
    },
    {
      "entropy": 6.6736372947692875,
      "epoch": 0.686762267095893,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972155547912005,
      "loss": 6.6613,
      "mean_token_accuracy": 0.11876440420746803,
      "num_tokens": 11757474.0,
      "step": 3520
    },
    {
      "entropy": 6.845918607711792,
      "epoch": 0.6877377816798361,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972045130177666,
      "loss": 6.7665,
      "mean_token_accuracy": 0.11521601900458336,
      "num_tokens": 11774467.0,
      "step": 3525
    },
    {
      "entropy": 6.661625814437866,
      "epoch": 0.6887132962637792,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004971934495312996,
      "loss": 6.6483,
      "mean_token_accuracy": 0.12561773881316185,
      "num_tokens": 11790597.0,
      "step": 3530
    },
    {
      "entropy": 6.7188530445098875,
      "epoch": 0.6896888108477222,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971823643328807,
      "loss": 6.6483,
      "mean_token_accuracy": 0.1310520239174366,
      "num_tokens": 11806459.0,
      "step": 3535
    },
    {
      "entropy": 6.637306451797485,
      "epoch": 0.6906643254316652,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497171257423593,
      "loss": 6.7176,
      "mean_token_accuracy": 0.12206159979104995,
      "num_tokens": 11823410.0,
      "step": 3540
    },
    {
      "entropy": 6.745995569229126,
      "epoch": 0.6916398400156082,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971601288045219,
      "loss": 6.6983,
      "mean_token_accuracy": 0.11592186763882636,
      "num_tokens": 11838684.0,
      "step": 3545
    },
    {
      "entropy": 6.702980995178223,
      "epoch": 0.6926153545995513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971489784767547,
      "loss": 6.6954,
      "mean_token_accuracy": 0.1243364654481411,
      "num_tokens": 11854324.0,
      "step": 3550
    },
    {
      "entropy": 6.7165563106536865,
      "epoch": 0.6935908691834943,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971378064413813,
      "loss": 6.6925,
      "mean_token_accuracy": 0.11902788132429123,
      "num_tokens": 11871992.0,
      "step": 3555
    },
    {
      "entropy": 6.789147996902466,
      "epoch": 0.6945663837674373,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004971266126994929,
      "loss": 6.7937,
      "mean_token_accuracy": 0.11624943763017655,
      "num_tokens": 11889195.0,
      "step": 3560
    },
    {
      "entropy": 6.752399063110351,
      "epoch": 0.6955418983513804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971153972521837,
      "loss": 6.7836,
      "mean_token_accuracy": 0.11215122267603875,
      "num_tokens": 11906316.0,
      "step": 3565
    },
    {
      "entropy": 6.771729516983032,
      "epoch": 0.6965174129353234,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971041601005495,
      "loss": 6.7825,
      "mean_token_accuracy": 0.11611792147159576,
      "num_tokens": 11924097.0,
      "step": 3570
    },
    {
      "entropy": 6.822442293167114,
      "epoch": 0.6974929275192664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970929012456884,
      "loss": 6.7922,
      "mean_token_accuracy": 0.11896535158157348,
      "num_tokens": 11942344.0,
      "step": 3575
    },
    {
      "entropy": 6.711492919921875,
      "epoch": 0.6984684421032095,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970816206887005,
      "loss": 6.6269,
      "mean_token_accuracy": 0.1311124972999096,
      "num_tokens": 11959045.0,
      "step": 3580
    },
    {
      "entropy": 6.694948005676269,
      "epoch": 0.6994439566871524,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004970703184306881,
      "loss": 6.6591,
      "mean_token_accuracy": 0.12259809374809265,
      "num_tokens": 11976244.0,
      "step": 3585
    },
    {
      "entropy": 6.697491884231567,
      "epoch": 0.7004194712710955,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970589944727558,
      "loss": 6.7038,
      "mean_token_accuracy": 0.1152302086353302,
      "num_tokens": 11993644.0,
      "step": 3590
    },
    {
      "entropy": 6.65057406425476,
      "epoch": 0.7013949858550386,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970476488160098,
      "loss": 6.6067,
      "mean_token_accuracy": 0.12608892619609832,
      "num_tokens": 12010419.0,
      "step": 3595
    },
    {
      "entropy": 6.647776556015015,
      "epoch": 0.7023705004389815,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970362814615592,
      "loss": 6.6271,
      "mean_token_accuracy": 0.12677855864167215,
      "num_tokens": 12027199.0,
      "step": 3600
    },
    {
      "entropy": 6.546076250076294,
      "epoch": 0.7033460150229246,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970248924105144,
      "loss": 6.6291,
      "mean_token_accuracy": 0.12805647104978563,
      "num_tokens": 12044267.0,
      "step": 3605
    },
    {
      "entropy": 6.777874708175659,
      "epoch": 0.7043215296068677,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970134816639884,
      "loss": 6.6917,
      "mean_token_accuracy": 0.11776589900255204,
      "num_tokens": 12062838.0,
      "step": 3610
    },
    {
      "entropy": 6.7569458961486815,
      "epoch": 0.7052970441908106,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970020492230963,
      "loss": 6.7804,
      "mean_token_accuracy": 0.11397483125329018,
      "num_tokens": 12079144.0,
      "step": 3615
    },
    {
      "entropy": 6.668662786483765,
      "epoch": 0.7062725587747537,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496990595088955,
      "loss": 6.6253,
      "mean_token_accuracy": 0.13019204288721084,
      "num_tokens": 12096477.0,
      "step": 3620
    },
    {
      "entropy": 6.679457521438598,
      "epoch": 0.7072480733586967,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004969791192626841,
      "loss": 6.6644,
      "mean_token_accuracy": 0.12415519654750824,
      "num_tokens": 12113307.0,
      "step": 3625
    },
    {
      "entropy": 6.728259658813476,
      "epoch": 0.7082235879426397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004969676217454048,
      "loss": 6.8076,
      "mean_token_accuracy": 0.11824259832501412,
      "num_tokens": 12131814.0,
      "step": 3630
    },
    {
      "entropy": 6.799320793151855,
      "epoch": 0.7091991025265828,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969561025382405,
      "loss": 6.6168,
      "mean_token_accuracy": 0.11877759173512459,
      "num_tokens": 12147110.0,
      "step": 3635
    },
    {
      "entropy": 6.565794038772583,
      "epoch": 0.7101746171105258,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969445616423171,
      "loss": 6.609,
      "mean_token_accuracy": 0.12383425161242485,
      "num_tokens": 12162954.0,
      "step": 3640
    },
    {
      "entropy": 6.5895367622375485,
      "epoch": 0.7111501316944688,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496932999058762,
      "loss": 6.5123,
      "mean_token_accuracy": 0.12579986304044724,
      "num_tokens": 12181169.0,
      "step": 3645
    },
    {
      "entropy": 6.818288135528564,
      "epoch": 0.7121256462784119,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969214147887053,
      "loss": 6.8367,
      "mean_token_accuracy": 0.11916090771555901,
      "num_tokens": 12198367.0,
      "step": 3650
    },
    {
      "entropy": 6.664682626724243,
      "epoch": 0.7131011608623549,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004969098088332789,
      "loss": 6.6441,
      "mean_token_accuracy": 0.12440235242247581,
      "num_tokens": 12215614.0,
      "step": 3655
    },
    {
      "entropy": 6.713459396362305,
      "epoch": 0.7140766754462979,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004968981811936169,
      "loss": 6.629,
      "mean_token_accuracy": 0.12557236924767495,
      "num_tokens": 12232123.0,
      "step": 3660
    },
    {
      "entropy": 6.65505576133728,
      "epoch": 0.7150521900302409,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968865318708555,
      "loss": 6.7188,
      "mean_token_accuracy": 0.11919585838913918,
      "num_tokens": 12248417.0,
      "step": 3665
    },
    {
      "entropy": 6.772244834899903,
      "epoch": 0.716027704614184,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004968748608661328,
      "loss": 6.669,
      "mean_token_accuracy": 0.1190882995724678,
      "num_tokens": 12264388.0,
      "step": 3670
    },
    {
      "entropy": 6.682482957839966,
      "epoch": 0.717003219198127,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968631681805897,
      "loss": 6.7371,
      "mean_token_accuracy": 0.11321451365947724,
      "num_tokens": 12282157.0,
      "step": 3675
    },
    {
      "entropy": 6.756937170028687,
      "epoch": 0.71797873378207,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968514538153685,
      "loss": 6.7202,
      "mean_token_accuracy": 0.11545223891735076,
      "num_tokens": 12298368.0,
      "step": 3680
    },
    {
      "entropy": 6.589137077331543,
      "epoch": 0.7189542483660131,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000496839717771614,
      "loss": 6.633,
      "mean_token_accuracy": 0.13081422224640846,
      "num_tokens": 12314860.0,
      "step": 3685
    },
    {
      "entropy": 6.745722484588623,
      "epoch": 0.7199297629499561,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968279600504729,
      "loss": 6.6493,
      "mean_token_accuracy": 0.12393339946866036,
      "num_tokens": 12331424.0,
      "step": 3690
    },
    {
      "entropy": 6.689153671264648,
      "epoch": 0.7209052775338991,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496816180653094,
      "loss": 6.7726,
      "mean_token_accuracy": 0.11809690073132514,
      "num_tokens": 12348358.0,
      "step": 3695
    },
    {
      "entropy": 6.760160493850708,
      "epoch": 0.7218807921178422,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968043795806285,
      "loss": 6.6698,
      "mean_token_accuracy": 0.12361822053790092,
      "num_tokens": 12364932.0,
      "step": 3700
    },
    {
      "entropy": 6.697461318969727,
      "epoch": 0.7228563067017852,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967925568342297,
      "loss": 6.7272,
      "mean_token_accuracy": 0.12022890001535416,
      "num_tokens": 12381852.0,
      "step": 3705
    },
    {
      "entropy": 6.739781618118286,
      "epoch": 0.7238318212857282,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967807124150527,
      "loss": 6.7317,
      "mean_token_accuracy": 0.11519643291831017,
      "num_tokens": 12397164.0,
      "step": 3710
    },
    {
      "entropy": 6.703133583068848,
      "epoch": 0.7248073358696713,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967688463242549,
      "loss": 6.6919,
      "mean_token_accuracy": 0.1210801623761654,
      "num_tokens": 12412520.0,
      "step": 3715
    },
    {
      "entropy": 6.7145641326904295,
      "epoch": 0.7257828504536142,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967569585629958,
      "loss": 6.7012,
      "mean_token_accuracy": 0.12436113134026527,
      "num_tokens": 12427039.0,
      "step": 3720
    },
    {
      "entropy": 6.7157691478729244,
      "epoch": 0.7267583650375573,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967450491324371,
      "loss": 6.6776,
      "mean_token_accuracy": 0.11965187937021256,
      "num_tokens": 12441903.0,
      "step": 3725
    },
    {
      "entropy": 6.5718762397766115,
      "epoch": 0.7277338796215004,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967331180337425,
      "loss": 6.5639,
      "mean_token_accuracy": 0.1299804575741291,
      "num_tokens": 12458883.0,
      "step": 3730
    },
    {
      "entropy": 6.745790910720825,
      "epoch": 0.7287093942054433,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967211652680779,
      "loss": 6.749,
      "mean_token_accuracy": 0.11917295753955841,
      "num_tokens": 12475347.0,
      "step": 3735
    },
    {
      "entropy": 6.7147444725036625,
      "epoch": 0.7296849087893864,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967091908366113,
      "loss": 6.7543,
      "mean_token_accuracy": 0.1114665225148201,
      "num_tokens": 12490393.0,
      "step": 3740
    },
    {
      "entropy": 6.778819227218628,
      "epoch": 0.7306604233733295,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966971947405127,
      "loss": 6.7165,
      "mean_token_accuracy": 0.12327706515789032,
      "num_tokens": 12505679.0,
      "step": 3745
    },
    {
      "entropy": 6.6433021068573,
      "epoch": 0.7316359379572724,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966851769809543,
      "loss": 6.6792,
      "mean_token_accuracy": 0.12412982806563377,
      "num_tokens": 12522316.0,
      "step": 3750
    },
    {
      "entropy": 6.746396255493164,
      "epoch": 0.7326114525412155,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004966731375591106,
      "loss": 6.7593,
      "mean_token_accuracy": 0.12042892426252365,
      "num_tokens": 12539638.0,
      "step": 3755
    },
    {
      "entropy": 6.67816047668457,
      "epoch": 0.7335869671251585,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966610764761579,
      "loss": 6.6136,
      "mean_token_accuracy": 0.1223549522459507,
      "num_tokens": 12556296.0,
      "step": 3760
    },
    {
      "entropy": 6.7266603946685795,
      "epoch": 0.7345624817091015,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966489937332748,
      "loss": 6.7101,
      "mean_token_accuracy": 0.12138539850711823,
      "num_tokens": 12573248.0,
      "step": 3765
    },
    {
      "entropy": 6.604175186157226,
      "epoch": 0.7355379962930446,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966368893316421,
      "loss": 6.5181,
      "mean_token_accuracy": 0.13614045158028604,
      "num_tokens": 12590918.0,
      "step": 3770
    },
    {
      "entropy": 6.63151273727417,
      "epoch": 0.7365135108769876,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966247632724425,
      "loss": 6.6442,
      "mean_token_accuracy": 0.12555419132113457,
      "num_tokens": 12607037.0,
      "step": 3775
    },
    {
      "entropy": 6.571084594726562,
      "epoch": 0.7374890254609306,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004966126155568609,
      "loss": 6.605,
      "mean_token_accuracy": 0.128154519200325,
      "num_tokens": 12623934.0,
      "step": 3780
    },
    {
      "entropy": 6.660620594024659,
      "epoch": 0.7384645400448737,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004966004461860843,
      "loss": 6.5903,
      "mean_token_accuracy": 0.13442015200853347,
      "num_tokens": 12640607.0,
      "step": 3785
    },
    {
      "entropy": 6.724188232421875,
      "epoch": 0.7394400546288167,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965882551613019,
      "loss": 6.6411,
      "mean_token_accuracy": 0.11834411472082138,
      "num_tokens": 12656671.0,
      "step": 3790
    },
    {
      "entropy": 6.724361848831177,
      "epoch": 0.7404155692127597,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496576042483705,
      "loss": 6.732,
      "mean_token_accuracy": 0.10988990291953087,
      "num_tokens": 12673271.0,
      "step": 3795
    },
    {
      "entropy": 6.546861267089843,
      "epoch": 0.7413910837967027,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004965638081544869,
      "loss": 6.5766,
      "mean_token_accuracy": 0.13610197082161904,
      "num_tokens": 12691005.0,
      "step": 3800
    },
    {
      "entropy": 6.7173137187957765,
      "epoch": 0.7423665983806458,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965515521748432,
      "loss": 6.7007,
      "mean_token_accuracy": 0.12019814178347588,
      "num_tokens": 12706170.0,
      "step": 3805
    },
    {
      "entropy": 6.634069156646729,
      "epoch": 0.7433421129645889,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004965392745459714,
      "loss": 6.5928,
      "mean_token_accuracy": 0.1261111244559288,
      "num_tokens": 12722794.0,
      "step": 3810
    },
    {
      "entropy": 6.757253742218017,
      "epoch": 0.7443176275485318,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965269752690713,
      "loss": 6.723,
      "mean_token_accuracy": 0.11596281379461289,
      "num_tokens": 12739676.0,
      "step": 3815
    },
    {
      "entropy": 6.694004917144776,
      "epoch": 0.7452931421324749,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965146543453446,
      "loss": 6.7258,
      "mean_token_accuracy": 0.11730321049690247,
      "num_tokens": 12756231.0,
      "step": 3820
    },
    {
      "entropy": 6.533119440078735,
      "epoch": 0.746268656716418,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004965023117759954,
      "loss": 6.5597,
      "mean_token_accuracy": 0.1383904203772545,
      "num_tokens": 12771877.0,
      "step": 3825
    },
    {
      "entropy": 6.759338665008545,
      "epoch": 0.7472441713003609,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964899475622299,
      "loss": 6.697,
      "mean_token_accuracy": 0.11760756820440292,
      "num_tokens": 12790698.0,
      "step": 3830
    },
    {
      "entropy": 6.709034442901611,
      "epoch": 0.748219685884304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496477561705256,
      "loss": 6.6847,
      "mean_token_accuracy": 0.11892438307404518,
      "num_tokens": 12806160.0,
      "step": 3835
    },
    {
      "entropy": 6.767488098144531,
      "epoch": 0.749195200468247,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964651542062841,
      "loss": 6.7336,
      "mean_token_accuracy": 0.11275624260306358,
      "num_tokens": 12822382.0,
      "step": 3840
    },
    {
      "entropy": 6.7063889503479,
      "epoch": 0.75017071505219,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964527250665268,
      "loss": 6.6636,
      "mean_token_accuracy": 0.12037805542349815,
      "num_tokens": 12839421.0,
      "step": 3845
    },
    {
      "entropy": 6.615977954864502,
      "epoch": 0.7511462296361331,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004964402742871983,
      "loss": 6.5017,
      "mean_token_accuracy": 0.1351486399769783,
      "num_tokens": 12856609.0,
      "step": 3850
    },
    {
      "entropy": 6.656347227096558,
      "epoch": 0.752121744220076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964278018695156,
      "loss": 6.6852,
      "mean_token_accuracy": 0.11694247499108315,
      "num_tokens": 12871659.0,
      "step": 3855
    },
    {
      "entropy": 6.699874019622802,
      "epoch": 0.7530972588040191,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964153078146971,
      "loss": 6.6938,
      "mean_token_accuracy": 0.12055180370807647,
      "num_tokens": 12889029.0,
      "step": 3860
    },
    {
      "entropy": 6.709548234939575,
      "epoch": 0.7540727733879622,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004964027921239639,
      "loss": 6.6458,
      "mean_token_accuracy": 0.12903001680970191,
      "num_tokens": 12905735.0,
      "step": 3865
    },
    {
      "entropy": 6.637347459793091,
      "epoch": 0.7550482879719052,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000496390254798539,
      "loss": 6.6525,
      "mean_token_accuracy": 0.12128332108259202,
      "num_tokens": 12921886.0,
      "step": 3870
    },
    {
      "entropy": 6.612480640411377,
      "epoch": 0.7560238025558482,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004963776958396475,
      "loss": 6.575,
      "mean_token_accuracy": 0.13231956139206885,
      "num_tokens": 12938658.0,
      "step": 3875
    },
    {
      "entropy": 6.699436855316162,
      "epoch": 0.7569993171397913,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963651152485165,
      "loss": 6.7311,
      "mean_token_accuracy": 0.1198471188545227,
      "num_tokens": 12955788.0,
      "step": 3880
    },
    {
      "entropy": 6.614861345291137,
      "epoch": 0.7579748317237343,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004963525130263753,
      "loss": 6.5527,
      "mean_token_accuracy": 0.13004487827420236,
      "num_tokens": 12972590.0,
      "step": 3885
    },
    {
      "entropy": 6.740634965896606,
      "epoch": 0.7589503463076773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963398891744557,
      "loss": 6.6934,
      "mean_token_accuracy": 0.11818641796708107,
      "num_tokens": 12989560.0,
      "step": 3890
    },
    {
      "entropy": 6.691430473327637,
      "epoch": 0.7599258608916203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004963272436939908,
      "loss": 6.6465,
      "mean_token_accuracy": 0.12754231542348862,
      "num_tokens": 13005584.0,
      "step": 3895
    },
    {
      "entropy": 6.67079348564148,
      "epoch": 0.7609013754755634,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963145765862166,
      "loss": 6.7046,
      "mean_token_accuracy": 0.11682451069355011,
      "num_tokens": 13021635.0,
      "step": 3900
    },
    {
      "entropy": 6.66444616317749,
      "epoch": 0.7618768900595064,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963018878523706,
      "loss": 6.6238,
      "mean_token_accuracy": 0.12423735633492469,
      "num_tokens": 13037838.0,
      "step": 3905
    },
    {
      "entropy": 6.592610311508179,
      "epoch": 0.7628524046434494,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962891774936929,
      "loss": 6.5752,
      "mean_token_accuracy": 0.13338154107332229,
      "num_tokens": 13054596.0,
      "step": 3910
    },
    {
      "entropy": 6.669826602935791,
      "epoch": 0.7638279192273925,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962764455114255,
      "loss": 6.6344,
      "mean_token_accuracy": 0.12492086216807366,
      "num_tokens": 13070462.0,
      "step": 3915
    },
    {
      "entropy": 6.612007570266724,
      "epoch": 0.7648034338113355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004962636919068125,
      "loss": 6.582,
      "mean_token_accuracy": 0.12788456529378892,
      "num_tokens": 13086686.0,
      "step": 3920
    },
    {
      "entropy": 6.722412586212158,
      "epoch": 0.7657789483952785,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962509166811001,
      "loss": 6.7315,
      "mean_token_accuracy": 0.11623572707176208,
      "num_tokens": 13102430.0,
      "step": 3925
    },
    {
      "entropy": 6.619417524337768,
      "epoch": 0.7667544629792216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962381198355367,
      "loss": 6.609,
      "mean_token_accuracy": 0.13012885600328444,
      "num_tokens": 13118944.0,
      "step": 3930
    },
    {
      "entropy": 6.67401704788208,
      "epoch": 0.7677299775631645,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962253013713726,
      "loss": 6.6412,
      "mean_token_accuracy": 0.1213919073343277,
      "num_tokens": 13136393.0,
      "step": 3935
    },
    {
      "entropy": 6.71761121749878,
      "epoch": 0.7687054921471076,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962124612898607,
      "loss": 6.711,
      "mean_token_accuracy": 0.11814533397555352,
      "num_tokens": 13153834.0,
      "step": 3940
    },
    {
      "entropy": 6.6320624351501465,
      "epoch": 0.7696810067310507,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004961995995922553,
      "loss": 6.5765,
      "mean_token_accuracy": 0.12632264271378518,
      "num_tokens": 13170904.0,
      "step": 3945
    },
    {
      "entropy": 6.74267168045044,
      "epoch": 0.7706565213149936,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004961867162798136,
      "loss": 6.6693,
      "mean_token_accuracy": 0.12143611088395119,
      "num_tokens": 13187274.0,
      "step": 3950
    },
    {
      "entropy": 6.550368452072144,
      "epoch": 0.7716320358989367,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004961738113537943,
      "loss": 6.6436,
      "mean_token_accuracy": 0.1292413629591465,
      "num_tokens": 13203448.0,
      "step": 3955
    },
    {
      "entropy": 6.744801807403564,
      "epoch": 0.7726075504828798,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961608848154583,
      "loss": 6.6496,
      "mean_token_accuracy": 0.1222207322716713,
      "num_tokens": 13220460.0,
      "step": 3960
    },
    {
      "entropy": 6.64396858215332,
      "epoch": 0.7735830650668227,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000496147936666069,
      "loss": 6.5958,
      "mean_token_accuracy": 0.12869085296988486,
      "num_tokens": 13237502.0,
      "step": 3965
    },
    {
      "entropy": 6.65892276763916,
      "epoch": 0.7745585796507658,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004961349669068913,
      "loss": 6.6473,
      "mean_token_accuracy": 0.12053527534008027,
      "num_tokens": 13254314.0,
      "step": 3970
    },
    {
      "entropy": 6.708866786956787,
      "epoch": 0.7755340942347088,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496121975539193,
      "loss": 6.7263,
      "mean_token_accuracy": 0.1213606797158718,
      "num_tokens": 13271737.0,
      "step": 3975
    },
    {
      "entropy": 6.738278484344482,
      "epoch": 0.7765096088186518,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961089625642433,
      "loss": 6.6632,
      "mean_token_accuracy": 0.12192397862672806,
      "num_tokens": 13288546.0,
      "step": 3980
    },
    {
      "entropy": 6.621914720535278,
      "epoch": 0.7774851234025949,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004960959279833137,
      "loss": 6.5657,
      "mean_token_accuracy": 0.13136719390749932,
      "num_tokens": 13304309.0,
      "step": 3985
    },
    {
      "entropy": 6.658552122116089,
      "epoch": 0.7784606379865379,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960828717976781,
      "loss": 6.634,
      "mean_token_accuracy": 0.1235605925321579,
      "num_tokens": 13319797.0,
      "step": 3990
    },
    {
      "entropy": 6.550283575057984,
      "epoch": 0.7794361525704809,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960697940086121,
      "loss": 6.4594,
      "mean_token_accuracy": 0.1420056365430355,
      "num_tokens": 13336040.0,
      "step": 3995
    },
    {
      "entropy": 6.7304822444915775,
      "epoch": 0.780411667154424,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004960566946173939,
      "loss": 6.6453,
      "mean_token_accuracy": 0.12395601570606232,
      "num_tokens": 13352762.0,
      "step": 4000
    },
    {
      "entropy": 6.642738580703735,
      "epoch": 0.781387181738367,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004960435736253034,
      "loss": 6.6083,
      "mean_token_accuracy": 0.12401333004236222,
      "num_tokens": 13368577.0,
      "step": 4005
    },
    {
      "entropy": 6.697572183609009,
      "epoch": 0.78236269632231,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960304310336226,
      "loss": 6.73,
      "mean_token_accuracy": 0.11546489074826241,
      "num_tokens": 13384170.0,
      "step": 4010
    },
    {
      "entropy": 6.599906873703003,
      "epoch": 0.783338210906253,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004960172668436359,
      "loss": 6.5338,
      "mean_token_accuracy": 0.13510606214404106,
      "num_tokens": 13400143.0,
      "step": 4015
    },
    {
      "entropy": 6.635289669036865,
      "epoch": 0.7843137254901961,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960040810566295,
      "loss": 6.6431,
      "mean_token_accuracy": 0.12509827315807343,
      "num_tokens": 13416553.0,
      "step": 4020
    },
    {
      "entropy": 6.6459962844848635,
      "epoch": 0.7852892400741391,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959908736738921,
      "loss": 6.6531,
      "mean_token_accuracy": 0.11897451877593994,
      "num_tokens": 13432694.0,
      "step": 4025
    },
    {
      "entropy": 6.713592004776001,
      "epoch": 0.7862647546580821,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959776446967141,
      "loss": 6.6739,
      "mean_token_accuracy": 0.11996177285909652,
      "num_tokens": 13450333.0,
      "step": 4030
    },
    {
      "entropy": 6.49674277305603,
      "epoch": 0.7872402692420252,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004959643941263882,
      "loss": 6.4699,
      "mean_token_accuracy": 0.13179363310337067,
      "num_tokens": 13467150.0,
      "step": 4035
    },
    {
      "entropy": 6.670515441894532,
      "epoch": 0.7882157838259682,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959511219642092,
      "loss": 6.6415,
      "mean_token_accuracy": 0.12280526459217071,
      "num_tokens": 13484796.0,
      "step": 4040
    },
    {
      "entropy": 6.642370653152466,
      "epoch": 0.7891912984099112,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959378282114741,
      "loss": 6.5641,
      "mean_token_accuracy": 0.12964036390185357,
      "num_tokens": 13501533.0,
      "step": 4045
    },
    {
      "entropy": 6.697076606750488,
      "epoch": 0.7901668129938543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959245128694819,
      "loss": 6.7463,
      "mean_token_accuracy": 0.11339326873421669,
      "num_tokens": 13518270.0,
      "step": 4050
    },
    {
      "entropy": 6.609881114959717,
      "epoch": 0.7911423275777972,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004959111759395337,
      "loss": 6.5216,
      "mean_token_accuracy": 0.1300751380622387,
      "num_tokens": 13535356.0,
      "step": 4055
    },
    {
      "entropy": 6.586223554611206,
      "epoch": 0.7921178421617403,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958978174229326,
      "loss": 6.5709,
      "mean_token_accuracy": 0.12651573047041892,
      "num_tokens": 13550530.0,
      "step": 4060
    },
    {
      "entropy": 6.550932788848877,
      "epoch": 0.7930933567456834,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004958844373209843,
      "loss": 6.5437,
      "mean_token_accuracy": 0.13148362189531326,
      "num_tokens": 13569492.0,
      "step": 4065
    },
    {
      "entropy": 6.7623930931091305,
      "epoch": 0.7940688713296263,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958710356349958,
      "loss": 6.7399,
      "mean_token_accuracy": 0.1175881266593933,
      "num_tokens": 13585096.0,
      "step": 4070
    },
    {
      "entropy": 6.708743715286255,
      "epoch": 0.7950443859135694,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495857612366277,
      "loss": 6.7106,
      "mean_token_accuracy": 0.1192586027085781,
      "num_tokens": 13600667.0,
      "step": 4075
    },
    {
      "entropy": 6.74160590171814,
      "epoch": 0.7960199004975125,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004958441675161395,
      "loss": 6.7338,
      "mean_token_accuracy": 0.11570419669151306,
      "num_tokens": 13618212.0,
      "step": 4080
    },
    {
      "entropy": 6.645744848251343,
      "epoch": 0.7969954150814554,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000495830701085897,
      "loss": 6.6492,
      "mean_token_accuracy": 0.12572448700666428,
      "num_tokens": 13634067.0,
      "step": 4085
    },
    {
      "entropy": 6.634458208084107,
      "epoch": 0.7979709296653985,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004958172130768655,
      "loss": 6.5937,
      "mean_token_accuracy": 0.12669137269258499,
      "num_tokens": 13651306.0,
      "step": 4090
    },
    {
      "entropy": 6.648689794540405,
      "epoch": 0.7989464442493416,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958037034903628,
      "loss": 6.6399,
      "mean_token_accuracy": 0.11853336989879608,
      "num_tokens": 13667563.0,
      "step": 4095
    },
    {
      "entropy": 6.575715827941894,
      "epoch": 0.7999219588332845,
      "grad_norm": 0.75,
      "learning_rate": 0.0004957901723277093,
      "loss": 6.5121,
      "mean_token_accuracy": 0.13093874156475066,
      "num_tokens": 13684857.0,
      "step": 4100
    },
    {
      "entropy": 6.768743896484375,
      "epoch": 0.8008974734172276,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957766195902271,
      "loss": 6.7342,
      "mean_token_accuracy": 0.11563626825809478,
      "num_tokens": 13700356.0,
      "step": 4105
    },
    {
      "entropy": 6.640881967544556,
      "epoch": 0.8018729880011706,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004957630452792404,
      "loss": 6.5781,
      "mean_token_accuracy": 0.13152394592761993,
      "num_tokens": 13717146.0,
      "step": 4110
    },
    {
      "entropy": 6.626404714584351,
      "epoch": 0.8028485025851136,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957494493960757,
      "loss": 6.6518,
      "mean_token_accuracy": 0.11726271882653236,
      "num_tokens": 13732520.0,
      "step": 4115
    },
    {
      "entropy": 6.765890789031983,
      "epoch": 0.8038240171690567,
      "grad_norm": 0.875,
      "learning_rate": 0.0004957358319420617,
      "loss": 6.7712,
      "mean_token_accuracy": 0.11684209108352661,
      "num_tokens": 13749135.0,
      "step": 4120
    },
    {
      "entropy": 6.618774509429931,
      "epoch": 0.8047995317529997,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004957221929185288,
      "loss": 6.4843,
      "mean_token_accuracy": 0.13539427891373634,
      "num_tokens": 13766218.0,
      "step": 4125
    },
    {
      "entropy": 6.652585124969482,
      "epoch": 0.8057750463369427,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004957085323268099,
      "loss": 6.683,
      "mean_token_accuracy": 0.12168987542390823,
      "num_tokens": 13784421.0,
      "step": 4130
    },
    {
      "entropy": 6.648529767990112,
      "epoch": 0.8067505609208858,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956948501682399,
      "loss": 6.5901,
      "mean_token_accuracy": 0.13115977346897126,
      "num_tokens": 13800054.0,
      "step": 4135
    },
    {
      "entropy": 6.650304937362671,
      "epoch": 0.8077260755048288,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956811464441556,
      "loss": 6.5789,
      "mean_token_accuracy": 0.12490106970071793,
      "num_tokens": 13816433.0,
      "step": 4140
    },
    {
      "entropy": 6.71581654548645,
      "epoch": 0.8087015900887718,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956674211558963,
      "loss": 6.7018,
      "mean_token_accuracy": 0.11724149659276009,
      "num_tokens": 13832441.0,
      "step": 4145
    },
    {
      "entropy": 6.664876651763916,
      "epoch": 0.8096771046727148,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000495653674304803,
      "loss": 6.6533,
      "mean_token_accuracy": 0.12444441989064217,
      "num_tokens": 13849012.0,
      "step": 4150
    },
    {
      "entropy": 6.721843481063843,
      "epoch": 0.8106526192566579,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004956399058922192,
      "loss": 6.705,
      "mean_token_accuracy": 0.11393826752901078,
      "num_tokens": 13866486.0,
      "step": 4155
    },
    {
      "entropy": 6.655822658538819,
      "epoch": 0.811628133840601,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049562611591949,
      "loss": 6.5585,
      "mean_token_accuracy": 0.12736134231090546,
      "num_tokens": 13882453.0,
      "step": 4160
    },
    {
      "entropy": 6.604769515991211,
      "epoch": 0.8126036484245439,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956123043879632,
      "loss": 6.6352,
      "mean_token_accuracy": 0.12367864102125167,
      "num_tokens": 13898090.0,
      "step": 4165
    },
    {
      "entropy": 6.6478125095367435,
      "epoch": 0.813579163008487,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004955984712989882,
      "loss": 6.5835,
      "mean_token_accuracy": 0.12453107833862305,
      "num_tokens": 13914818.0,
      "step": 4170
    },
    {
      "entropy": 6.709457302093506,
      "epoch": 0.81455467759243,
      "grad_norm": 0.875,
      "learning_rate": 0.0004955846166539169,
      "loss": 6.6894,
      "mean_token_accuracy": 0.11977913379669189,
      "num_tokens": 13930690.0,
      "step": 4175
    },
    {
      "entropy": 6.696602821350098,
      "epoch": 0.815530192176373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495570740454103,
      "loss": 6.7001,
      "mean_token_accuracy": 0.12241925448179244,
      "num_tokens": 13946128.0,
      "step": 4180
    },
    {
      "entropy": 6.634654378890991,
      "epoch": 0.8165057067603161,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955568427009024,
      "loss": 6.6133,
      "mean_token_accuracy": 0.12358218356966973,
      "num_tokens": 13961766.0,
      "step": 4185
    },
    {
      "entropy": 6.552938747406006,
      "epoch": 0.817481221344259,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955429233956733,
      "loss": 6.5788,
      "mean_token_accuracy": 0.13002277612686158,
      "num_tokens": 13978632.0,
      "step": 4190
    },
    {
      "entropy": 6.737991189956665,
      "epoch": 0.8184567359282021,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004955289825397757,
      "loss": 6.6516,
      "mean_token_accuracy": 0.1193838432431221,
      "num_tokens": 13995297.0,
      "step": 4195
    },
    {
      "entropy": 6.638138008117676,
      "epoch": 0.8194322505121452,
      "grad_norm": 1.0,
      "learning_rate": 0.000495515020134572,
      "loss": 6.6573,
      "mean_token_accuracy": 0.12201353684067726,
      "num_tokens": 14010439.0,
      "step": 4200
    },
    {
      "entropy": 6.675152111053467,
      "epoch": 0.8204077650960881,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955010361814264,
      "loss": 6.6628,
      "mean_token_accuracy": 0.12020387053489685,
      "num_tokens": 14028538.0,
      "step": 4205
    },
    {
      "entropy": 6.711331796646118,
      "epoch": 0.8213832796800312,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954870306817054,
      "loss": 6.68,
      "mean_token_accuracy": 0.11885645985603333,
      "num_tokens": 14045290.0,
      "step": 4210
    },
    {
      "entropy": 6.608047819137573,
      "epoch": 0.8223587942639743,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004954730036367776,
      "loss": 6.6061,
      "mean_token_accuracy": 0.12900031208992005,
      "num_tokens": 14062247.0,
      "step": 4215
    },
    {
      "entropy": 6.5941126346588135,
      "epoch": 0.8233343088479173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954589550480137,
      "loss": 6.6223,
      "mean_token_accuracy": 0.1269232727587223,
      "num_tokens": 14078599.0,
      "step": 4220
    },
    {
      "entropy": 6.649496746063233,
      "epoch": 0.8243098234318603,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004954448849167864,
      "loss": 6.6359,
      "mean_token_accuracy": 0.12365191578865051,
      "num_tokens": 14094686.0,
      "step": 4225
    },
    {
      "entropy": 6.636897945404053,
      "epoch": 0.8252853380158033,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004954307932444707,
      "loss": 6.6407,
      "mean_token_accuracy": 0.12143810689449311,
      "num_tokens": 14110255.0,
      "step": 4230
    },
    {
      "entropy": 6.6688474178314205,
      "epoch": 0.8262608525997464,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954166800324434,
      "loss": 6.644,
      "mean_token_accuracy": 0.12024810090661049,
      "num_tokens": 14127160.0,
      "step": 4235
    },
    {
      "entropy": 6.610847902297974,
      "epoch": 0.8272363671836894,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004954025452820837,
      "loss": 6.5516,
      "mean_token_accuracy": 0.12928074970841408,
      "num_tokens": 14144704.0,
      "step": 4240
    },
    {
      "entropy": 6.671830558776856,
      "epoch": 0.8282118817676324,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000495388388994773,
      "loss": 6.6308,
      "mean_token_accuracy": 0.12269783765077591,
      "num_tokens": 14162125.0,
      "step": 4245
    },
    {
      "entropy": 6.632227659225464,
      "epoch": 0.8291873963515755,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004953742111718942,
      "loss": 6.5495,
      "mean_token_accuracy": 0.12188868597149849,
      "num_tokens": 14179735.0,
      "step": 4250
    },
    {
      "entropy": 6.5946413516998295,
      "epoch": 0.8301629109355185,
      "grad_norm": 1.515625,
      "learning_rate": 0.000495360011814833,
      "loss": 6.741,
      "mean_token_accuracy": 0.11980281546711921,
      "num_tokens": 14196964.0,
      "step": 4255
    },
    {
      "entropy": 6.647566604614258,
      "epoch": 0.8311384255194615,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004953457909249768,
      "loss": 6.5618,
      "mean_token_accuracy": 0.1271596759557724,
      "num_tokens": 14212802.0,
      "step": 4260
    },
    {
      "entropy": 6.614435720443725,
      "epoch": 0.8321139401034046,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953315485037152,
      "loss": 6.58,
      "mean_token_accuracy": 0.12615046873688698,
      "num_tokens": 14229183.0,
      "step": 4265
    },
    {
      "entropy": 6.674267339706421,
      "epoch": 0.8330894546873476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953172845524401,
      "loss": 6.6137,
      "mean_token_accuracy": 0.12175215259194375,
      "num_tokens": 14245859.0,
      "step": 4270
    },
    {
      "entropy": 6.671193695068359,
      "epoch": 0.8340649692712906,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953029990725452,
      "loss": 6.6703,
      "mean_token_accuracy": 0.12936388850212097,
      "num_tokens": 14261689.0,
      "step": 4275
    },
    {
      "entropy": 6.560730695724487,
      "epoch": 0.8350404838552337,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004952886920654262,
      "loss": 6.5099,
      "mean_token_accuracy": 0.13164868652820588,
      "num_tokens": 14280638.0,
      "step": 4280
    },
    {
      "entropy": 6.6188336372375485,
      "epoch": 0.8360159984391766,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952743635324813,
      "loss": 6.5266,
      "mean_token_accuracy": 0.13677036315202712,
      "num_tokens": 14296595.0,
      "step": 4285
    },
    {
      "entropy": 6.669117403030396,
      "epoch": 0.8369915130231197,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952600134751109,
      "loss": 6.6944,
      "mean_token_accuracy": 0.11806081533432007,
      "num_tokens": 14312384.0,
      "step": 4290
    },
    {
      "entropy": 6.660558414459229,
      "epoch": 0.8379670276070628,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952456418947168,
      "loss": 6.609,
      "mean_token_accuracy": 0.12207264006137848,
      "num_tokens": 14327852.0,
      "step": 4295
    },
    {
      "entropy": 6.625396823883056,
      "epoch": 0.8389425421910057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952312487927037,
      "loss": 6.6182,
      "mean_token_accuracy": 0.12151826694607734,
      "num_tokens": 14343666.0,
      "step": 4300
    },
    {
      "entropy": 6.618071126937866,
      "epoch": 0.8399180567749488,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004952168341704778,
      "loss": 6.5845,
      "mean_token_accuracy": 0.1320745401084423,
      "num_tokens": 14359819.0,
      "step": 4305
    },
    {
      "entropy": 6.752963733673096,
      "epoch": 0.8408935713588919,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004952023980294478,
      "loss": 6.7093,
      "mean_token_accuracy": 0.11721565201878548,
      "num_tokens": 14377762.0,
      "step": 4310
    },
    {
      "entropy": 6.582974910736084,
      "epoch": 0.8418690859428348,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951879403710242,
      "loss": 6.5623,
      "mean_token_accuracy": 0.12620503529906274,
      "num_tokens": 14393995.0,
      "step": 4315
    },
    {
      "entropy": 6.721200466156006,
      "epoch": 0.8428446005267779,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004951734611966199,
      "loss": 6.6109,
      "mean_token_accuracy": 0.12899211049079895,
      "num_tokens": 14411959.0,
      "step": 4320
    },
    {
      "entropy": 6.622717046737671,
      "epoch": 0.8438201151107209,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951589605076495,
      "loss": 6.6905,
      "mean_token_accuracy": 0.11969695091247559,
      "num_tokens": 14427659.0,
      "step": 4325
    },
    {
      "entropy": 6.4425036907196045,
      "epoch": 0.8447956296946639,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951444383055304,
      "loss": 6.4305,
      "mean_token_accuracy": 0.1422936774790287,
      "num_tokens": 14443887.0,
      "step": 4330
    },
    {
      "entropy": 6.671658372879028,
      "epoch": 0.845771144278607,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004951298945916812,
      "loss": 6.6274,
      "mean_token_accuracy": 0.12339943274855614,
      "num_tokens": 14461907.0,
      "step": 4335
    },
    {
      "entropy": 6.567209959030151,
      "epoch": 0.84674665886255,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951153293675235,
      "loss": 6.5913,
      "mean_token_accuracy": 0.12541626989841462,
      "num_tokens": 14476912.0,
      "step": 4340
    },
    {
      "entropy": 6.582379293441773,
      "epoch": 0.847722173446493,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951007426344801,
      "loss": 6.485,
      "mean_token_accuracy": 0.1323241949081421,
      "num_tokens": 14493499.0,
      "step": 4345
    },
    {
      "entropy": 6.693007469177246,
      "epoch": 0.8486976880304361,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950861343939767,
      "loss": 6.6421,
      "mean_token_accuracy": 0.12452356815338135,
      "num_tokens": 14511266.0,
      "step": 4350
    },
    {
      "entropy": 6.64261531829834,
      "epoch": 0.8496732026143791,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950715046474405,
      "loss": 6.6435,
      "mean_token_accuracy": 0.12210259214043617,
      "num_tokens": 14528930.0,
      "step": 4355
    },
    {
      "entropy": 6.678401041030884,
      "epoch": 0.8506487171983221,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950568533963013,
      "loss": 6.6048,
      "mean_token_accuracy": 0.1340159386396408,
      "num_tokens": 14546637.0,
      "step": 4360
    },
    {
      "entropy": 6.513246393203735,
      "epoch": 0.8516242317822651,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950421806419907,
      "loss": 6.6232,
      "mean_token_accuracy": 0.12296181991696357,
      "num_tokens": 14563332.0,
      "step": 4365
    },
    {
      "entropy": 6.5499779224395756,
      "epoch": 0.8525997463662082,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950274863859425,
      "loss": 6.4769,
      "mean_token_accuracy": 0.13273719176650048,
      "num_tokens": 14579806.0,
      "step": 4370
    },
    {
      "entropy": 6.697094821929932,
      "epoch": 0.8535752609501512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950127706295923,
      "loss": 6.6829,
      "mean_token_accuracy": 0.12042219489812851,
      "num_tokens": 14597129.0,
      "step": 4375
    },
    {
      "entropy": 6.540702819824219,
      "epoch": 0.8545507755340942,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004949980333743786,
      "loss": 6.4188,
      "mean_token_accuracy": 0.1423679493367672,
      "num_tokens": 14615000.0,
      "step": 4380
    },
    {
      "entropy": 6.6257054805755615,
      "epoch": 0.8555262901180373,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000494983274621741,
      "loss": 6.6051,
      "mean_token_accuracy": 0.13396110981702805,
      "num_tokens": 14631338.0,
      "step": 4385
    },
    {
      "entropy": 6.592878341674805,
      "epoch": 0.8565018047019803,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004949684943731219,
      "loss": 6.5883,
      "mean_token_accuracy": 0.12827592790126802,
      "num_tokens": 14647886.0,
      "step": 4390
    },
    {
      "entropy": 6.647903966903686,
      "epoch": 0.8574773192859233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004949536926299654,
      "loss": 6.5702,
      "mean_token_accuracy": 0.1283034399151802,
      "num_tokens": 14665054.0,
      "step": 4395
    },
    {
      "entropy": 6.586231470108032,
      "epoch": 0.8584528338698664,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494938869393718,
      "loss": 6.5891,
      "mean_token_accuracy": 0.12495568320155144,
      "num_tokens": 14679413.0,
      "step": 4400
    },
    {
      "entropy": 6.572491598129273,
      "epoch": 0.8594283484538093,
      "grad_norm": 0.875,
      "learning_rate": 0.0004949240246658283,
      "loss": 6.551,
      "mean_token_accuracy": 0.1299852855503559,
      "num_tokens": 14696562.0,
      "step": 4405
    },
    {
      "entropy": 6.610397386550903,
      "epoch": 0.8604038630377524,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004949091584477465,
      "loss": 6.4564,
      "mean_token_accuracy": 0.13466879427433015,
      "num_tokens": 14713782.0,
      "step": 4410
    },
    {
      "entropy": 6.692568159103393,
      "epoch": 0.8613793776216955,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004948942707409257,
      "loss": 6.6986,
      "mean_token_accuracy": 0.12129929736256599,
      "num_tokens": 14730139.0,
      "step": 4415
    },
    {
      "entropy": 6.664574003219604,
      "epoch": 0.8623548922056384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948793615468205,
      "loss": 6.6325,
      "mean_token_accuracy": 0.11430357322096825,
      "num_tokens": 14746830.0,
      "step": 4420
    },
    {
      "entropy": 6.680889844894409,
      "epoch": 0.8633304067895815,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948644308668877,
      "loss": 6.6744,
      "mean_token_accuracy": 0.12209271788597106,
      "num_tokens": 14762410.0,
      "step": 4425
    },
    {
      "entropy": 6.618428707122803,
      "epoch": 0.8643059213735246,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004948494787025863,
      "loss": 6.5709,
      "mean_token_accuracy": 0.11955504938960075,
      "num_tokens": 14778283.0,
      "step": 4430
    },
    {
      "entropy": 6.556702613830566,
      "epoch": 0.8652814359574675,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948345050553774,
      "loss": 6.5227,
      "mean_token_accuracy": 0.12419091388583184,
      "num_tokens": 14797005.0,
      "step": 4435
    },
    {
      "entropy": 6.673793935775757,
      "epoch": 0.8662569505414106,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948195099267242,
      "loss": 6.6651,
      "mean_token_accuracy": 0.11517257392406463,
      "num_tokens": 14814548.0,
      "step": 4440
    },
    {
      "entropy": 6.639867973327637,
      "epoch": 0.8672324651253536,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948044933180919,
      "loss": 6.5713,
      "mean_token_accuracy": 0.1293208993971348,
      "num_tokens": 14830630.0,
      "step": 4445
    },
    {
      "entropy": 6.587173795700073,
      "epoch": 0.8682079797092966,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000494789455230948,
      "loss": 6.6044,
      "mean_token_accuracy": 0.12240093275904655,
      "num_tokens": 14846247.0,
      "step": 4450
    },
    {
      "entropy": 6.738630151748657,
      "epoch": 0.8691834942932397,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004947743956667617,
      "loss": 6.6372,
      "mean_token_accuracy": 0.12133599743247032,
      "num_tokens": 14864480.0,
      "step": 4455
    },
    {
      "entropy": 6.657829332351684,
      "epoch": 0.8701590088771827,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947593146270049,
      "loss": 6.6823,
      "mean_token_accuracy": 0.12227668762207031,
      "num_tokens": 14881382.0,
      "step": 4460
    },
    {
      "entropy": 6.660969209671021,
      "epoch": 0.8711345234611257,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494744212113151,
      "loss": 6.6091,
      "mean_token_accuracy": 0.11966304332017899,
      "num_tokens": 14897546.0,
      "step": 4465
    },
    {
      "entropy": 6.599333047866821,
      "epoch": 0.8721100380450688,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947290881266759,
      "loss": 6.5907,
      "mean_token_accuracy": 0.12330922856926918,
      "num_tokens": 14913995.0,
      "step": 4470
    },
    {
      "entropy": 6.653117036819458,
      "epoch": 0.8730855526290118,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004947139426690576,
      "loss": 6.6387,
      "mean_token_accuracy": 0.11546454504132271,
      "num_tokens": 14931097.0,
      "step": 4475
    },
    {
      "entropy": 6.564696884155273,
      "epoch": 0.8740610672129548,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946987757417757,
      "loss": 6.5276,
      "mean_token_accuracy": 0.1279122032225132,
      "num_tokens": 14948010.0,
      "step": 4480
    },
    {
      "entropy": 6.4721455574035645,
      "epoch": 0.8750365817968979,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004946835873463126,
      "loss": 6.4927,
      "mean_token_accuracy": 0.1376678839325905,
      "num_tokens": 14966940.0,
      "step": 4485
    },
    {
      "entropy": 6.7201872825622555,
      "epoch": 0.8760120963808409,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946683774841522,
      "loss": 6.6832,
      "mean_token_accuracy": 0.11887902468442917,
      "num_tokens": 14983482.0,
      "step": 4490
    },
    {
      "entropy": 6.552084589004517,
      "epoch": 0.876987610964784,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000494653146156781,
      "loss": 6.5306,
      "mean_token_accuracy": 0.12892355620861054,
      "num_tokens": 15000709.0,
      "step": 4495
    },
    {
      "entropy": 6.593361854553223,
      "epoch": 0.8779631255487269,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004946378933656871,
      "loss": 6.4968,
      "mean_token_accuracy": 0.12811356112360955,
      "num_tokens": 15016584.0,
      "step": 4500
    },
    {
      "entropy": 6.559517431259155,
      "epoch": 0.87893864013267,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494622619112361,
      "loss": 6.6718,
      "mean_token_accuracy": 0.12362031936645508,
      "num_tokens": 15033598.0,
      "step": 4505
    },
    {
      "entropy": 6.725758504867554,
      "epoch": 0.879914154716613,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946073233982954,
      "loss": 6.6417,
      "mean_token_accuracy": 0.120847687125206,
      "num_tokens": 15049842.0,
      "step": 4510
    },
    {
      "entropy": 6.5198863506317135,
      "epoch": 0.880889669300556,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945920062249849,
      "loss": 6.3953,
      "mean_token_accuracy": 0.13653058409690857,
      "num_tokens": 15066278.0,
      "step": 4515
    },
    {
      "entropy": 6.62023286819458,
      "epoch": 0.8818651838844991,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945766675939261,
      "loss": 6.6197,
      "mean_token_accuracy": 0.13151776865124704,
      "num_tokens": 15083316.0,
      "step": 4520
    },
    {
      "entropy": 6.5788593769073485,
      "epoch": 0.8828406984684422,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945613075066181,
      "loss": 6.6351,
      "mean_token_accuracy": 0.12246594652533531,
      "num_tokens": 15103135.0,
      "step": 4525
    },
    {
      "entropy": 6.571533346176148,
      "epoch": 0.8838162130523851,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004945459259645614,
      "loss": 6.4669,
      "mean_token_accuracy": 0.14429314136505128,
      "num_tokens": 15122441.0,
      "step": 4530
    },
    {
      "entropy": 6.700776863098144,
      "epoch": 0.8847917276363282,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945305229692594,
      "loss": 6.5989,
      "mean_token_accuracy": 0.12502915859222413,
      "num_tokens": 15139526.0,
      "step": 4535
    },
    {
      "entropy": 6.569632339477539,
      "epoch": 0.8857672422202711,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004945150985222171,
      "loss": 6.5836,
      "mean_token_accuracy": 0.12300720289349557,
      "num_tokens": 15156139.0,
      "step": 4540
    },
    {
      "entropy": 6.571629333496094,
      "epoch": 0.8867427568042142,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944996526249419,
      "loss": 6.4963,
      "mean_token_accuracy": 0.13339080959558486,
      "num_tokens": 15172894.0,
      "step": 4545
    },
    {
      "entropy": 6.62003607749939,
      "epoch": 0.8877182713881573,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944841852789427,
      "loss": 6.6203,
      "mean_token_accuracy": 0.12274017632007599,
      "num_tokens": 15190900.0,
      "step": 4550
    },
    {
      "entropy": 6.668445348739624,
      "epoch": 0.8886937859721002,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944686964857314,
      "loss": 6.5748,
      "mean_token_accuracy": 0.12155651971697808,
      "num_tokens": 15208018.0,
      "step": 4555
    },
    {
      "entropy": 6.605136108398438,
      "epoch": 0.8896693005560433,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944531862468211,
      "loss": 6.604,
      "mean_token_accuracy": 0.12640310376882552,
      "num_tokens": 15222781.0,
      "step": 4560
    },
    {
      "entropy": 6.583558130264282,
      "epoch": 0.8906448151399864,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004944376545637277,
      "loss": 6.6263,
      "mean_token_accuracy": 0.12514247372746468,
      "num_tokens": 15241405.0,
      "step": 4565
    },
    {
      "entropy": 6.719800090789795,
      "epoch": 0.8916203297239294,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944221014379688,
      "loss": 6.6362,
      "mean_token_accuracy": 0.12874823436141014,
      "num_tokens": 15258340.0,
      "step": 4570
    },
    {
      "entropy": 6.591956615447998,
      "epoch": 0.8925958443078724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944065268710641,
      "loss": 6.6028,
      "mean_token_accuracy": 0.12491160035133361,
      "num_tokens": 15274655.0,
      "step": 4575
    },
    {
      "entropy": 6.6703699111938475,
      "epoch": 0.8935713588918154,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943909308645356,
      "loss": 6.6401,
      "mean_token_accuracy": 0.11682420894503594,
      "num_tokens": 15291197.0,
      "step": 4580
    },
    {
      "entropy": 6.692414140701294,
      "epoch": 0.8945468734757585,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943753134199072,
      "loss": 6.6234,
      "mean_token_accuracy": 0.12216764837503433,
      "num_tokens": 15306879.0,
      "step": 4585
    },
    {
      "entropy": 6.611741161346435,
      "epoch": 0.8955223880597015,
      "grad_norm": 0.875,
      "learning_rate": 0.0004943596745387052,
      "loss": 6.6513,
      "mean_token_accuracy": 0.1236627571284771,
      "num_tokens": 15322880.0,
      "step": 4590
    },
    {
      "entropy": 6.613602113723755,
      "epoch": 0.8964979026436445,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943440142224575,
      "loss": 6.5592,
      "mean_token_accuracy": 0.12928596436977385,
      "num_tokens": 15338282.0,
      "step": 4595
    },
    {
      "entropy": 6.618643093109131,
      "epoch": 0.8974734172275876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943283324726944,
      "loss": 6.5748,
      "mean_token_accuracy": 0.12585204020142554,
      "num_tokens": 15354507.0,
      "step": 4600
    },
    {
      "entropy": 6.657996082305909,
      "epoch": 0.8984489318115306,
      "grad_norm": 0.875,
      "learning_rate": 0.0004943126292909485,
      "loss": 6.6046,
      "mean_token_accuracy": 0.12051715776324272,
      "num_tokens": 15373204.0,
      "step": 4605
    },
    {
      "entropy": 6.557158517837524,
      "epoch": 0.8994244463954736,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004942969046787541,
      "loss": 6.4559,
      "mean_token_accuracy": 0.13683562874794006,
      "num_tokens": 15390743.0,
      "step": 4610
    },
    {
      "entropy": 6.58685073852539,
      "epoch": 0.9003999609794167,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004942811586376476,
      "loss": 6.5732,
      "mean_token_accuracy": 0.12078804150223732,
      "num_tokens": 15408565.0,
      "step": 4615
    },
    {
      "entropy": 6.550992393493653,
      "epoch": 0.9013754755633596,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004942653911691679,
      "loss": 6.5244,
      "mean_token_accuracy": 0.133377505838871,
      "num_tokens": 15425747.0,
      "step": 4620
    },
    {
      "entropy": 6.563457250595093,
      "epoch": 0.9023509901473027,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942496022748557,
      "loss": 6.5413,
      "mean_token_accuracy": 0.12837236151099204,
      "num_tokens": 15441882.0,
      "step": 4625
    },
    {
      "entropy": 6.60471363067627,
      "epoch": 0.9033265047312458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942337919562539,
      "loss": 6.5105,
      "mean_token_accuracy": 0.12303609773516655,
      "num_tokens": 15458003.0,
      "step": 4630
    },
    {
      "entropy": 6.429346942901612,
      "epoch": 0.9043020193151887,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942179602149071,
      "loss": 6.4316,
      "mean_token_accuracy": 0.1395994633436203,
      "num_tokens": 15473798.0,
      "step": 4635
    },
    {
      "entropy": 6.5211766242980955,
      "epoch": 0.9052775338991318,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004942021070523626,
      "loss": 6.4391,
      "mean_token_accuracy": 0.1377377599477768,
      "num_tokens": 15490620.0,
      "step": 4640
    },
    {
      "entropy": 6.522173452377319,
      "epoch": 0.9062530484830749,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941862324701695,
      "loss": 6.457,
      "mean_token_accuracy": 0.13480929806828498,
      "num_tokens": 15506777.0,
      "step": 4645
    },
    {
      "entropy": 6.580625104904175,
      "epoch": 0.9072285630670178,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941703364698789,
      "loss": 6.6934,
      "mean_token_accuracy": 0.11610293239355088,
      "num_tokens": 15523416.0,
      "step": 4650
    },
    {
      "entropy": 6.6603333950042725,
      "epoch": 0.9082040776509609,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941544190530442,
      "loss": 6.4406,
      "mean_token_accuracy": 0.13544537127017975,
      "num_tokens": 15541064.0,
      "step": 4655
    },
    {
      "entropy": 6.6254747867584225,
      "epoch": 0.909179592234904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004941384802212207,
      "loss": 6.7144,
      "mean_token_accuracy": 0.12020516768097878,
      "num_tokens": 15559391.0,
      "step": 4660
    },
    {
      "entropy": 6.619871616363525,
      "epoch": 0.9101551068188469,
      "grad_norm": 0.953125,
      "learning_rate": 0.000494122519975966,
      "loss": 6.5346,
      "mean_token_accuracy": 0.1259423702955246,
      "num_tokens": 15576842.0,
      "step": 4665
    },
    {
      "entropy": 6.547988080978394,
      "epoch": 0.91113062140279,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941065383188396,
      "loss": 6.4978,
      "mean_token_accuracy": 0.13416629284620285,
      "num_tokens": 15592442.0,
      "step": 4670
    },
    {
      "entropy": 6.582059335708618,
      "epoch": 0.912106135986733,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940905352514032,
      "loss": 6.5826,
      "mean_token_accuracy": 0.12384884804487228,
      "num_tokens": 15608402.0,
      "step": 4675
    },
    {
      "entropy": 6.638591384887695,
      "epoch": 0.913081650570676,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004940745107752205,
      "loss": 6.5431,
      "mean_token_accuracy": 0.12366040647029877,
      "num_tokens": 15624349.0,
      "step": 4680
    },
    {
      "entropy": 6.574081993103027,
      "epoch": 0.9140571651546191,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940584648918573,
      "loss": 6.622,
      "mean_token_accuracy": 0.12882452756166457,
      "num_tokens": 15640901.0,
      "step": 4685
    },
    {
      "entropy": 6.678586816787719,
      "epoch": 0.9150326797385621,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004940423976028819,
      "loss": 6.6653,
      "mean_token_accuracy": 0.12176755666732789,
      "num_tokens": 15658947.0,
      "step": 4690
    },
    {
      "entropy": 6.744279861450195,
      "epoch": 0.9160081943225051,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940263089098639,
      "loss": 6.6835,
      "mean_token_accuracy": 0.12291955500841141,
      "num_tokens": 15674459.0,
      "step": 4695
    },
    {
      "entropy": 6.530014228820801,
      "epoch": 0.9169837089064482,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004940101988143757,
      "loss": 6.587,
      "mean_token_accuracy": 0.1277773506939411,
      "num_tokens": 15691951.0,
      "step": 4700
    },
    {
      "entropy": 6.647747039794922,
      "epoch": 0.9179592234903912,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939940673179914,
      "loss": 6.6083,
      "mean_token_accuracy": 0.11856793984770775,
      "num_tokens": 15708539.0,
      "step": 4705
    },
    {
      "entropy": 6.6020372867584225,
      "epoch": 0.9189347380743342,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004939779144222873,
      "loss": 6.6115,
      "mean_token_accuracy": 0.1254153236746788,
      "num_tokens": 15727108.0,
      "step": 4710
    },
    {
      "entropy": 6.653774452209473,
      "epoch": 0.9199102526582772,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939617401288419,
      "loss": 6.5654,
      "mean_token_accuracy": 0.12312598973512649,
      "num_tokens": 15743436.0,
      "step": 4715
    },
    {
      "entropy": 6.657113409042358,
      "epoch": 0.9208857672422203,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004939455444392357,
      "loss": 6.672,
      "mean_token_accuracy": 0.12144109159708023,
      "num_tokens": 15761438.0,
      "step": 4720
    },
    {
      "entropy": 6.629937219619751,
      "epoch": 0.9218612818261633,
      "grad_norm": 0.875,
      "learning_rate": 0.0004939293273550512,
      "loss": 6.526,
      "mean_token_accuracy": 0.12699644565582274,
      "num_tokens": 15777064.0,
      "step": 4725
    },
    {
      "entropy": 6.634288167953491,
      "epoch": 0.9228367964101063,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939130888778731,
      "loss": 6.55,
      "mean_token_accuracy": 0.12582860738039017,
      "num_tokens": 15793346.0,
      "step": 4730
    },
    {
      "entropy": 6.565315818786621,
      "epoch": 0.9238123109940494,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938968290092883,
      "loss": 6.5722,
      "mean_token_accuracy": 0.12200249135494232,
      "num_tokens": 15809557.0,
      "step": 4735
    },
    {
      "entropy": 6.656667709350586,
      "epoch": 0.9247878255779924,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938805477508853,
      "loss": 6.6122,
      "mean_token_accuracy": 0.12024437934160233,
      "num_tokens": 15826717.0,
      "step": 4740
    },
    {
      "entropy": 6.65801773071289,
      "epoch": 0.9257633401619354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938642451042554,
      "loss": 6.5661,
      "mean_token_accuracy": 0.12313027083873748,
      "num_tokens": 15842472.0,
      "step": 4745
    },
    {
      "entropy": 6.441526317596436,
      "epoch": 0.9267388547458785,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004938479210709915,
      "loss": 6.492,
      "mean_token_accuracy": 0.13215720579028128,
      "num_tokens": 15860136.0,
      "step": 4750
    },
    {
      "entropy": 6.64937572479248,
      "epoch": 0.9277143693298214,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004938315756526886,
      "loss": 6.5353,
      "mean_token_accuracy": 0.13468293994665145,
      "num_tokens": 15878538.0,
      "step": 4755
    },
    {
      "entropy": 6.631177520751953,
      "epoch": 0.9286898839137645,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938152088509442,
      "loss": 6.5839,
      "mean_token_accuracy": 0.12353575006127357,
      "num_tokens": 15892875.0,
      "step": 4760
    },
    {
      "entropy": 6.521851301193237,
      "epoch": 0.9296653984977076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937988206673576,
      "loss": 6.4349,
      "mean_token_accuracy": 0.13442096263170242,
      "num_tokens": 15909366.0,
      "step": 4765
    },
    {
      "entropy": 6.554766988754272,
      "epoch": 0.9306409130816505,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937824111035298,
      "loss": 6.5248,
      "mean_token_accuracy": 0.128006062656641,
      "num_tokens": 15927712.0,
      "step": 4770
    },
    {
      "entropy": 6.668897867202759,
      "epoch": 0.9316164276655936,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004937659801610647,
      "loss": 6.6434,
      "mean_token_accuracy": 0.11991560384631157,
      "num_tokens": 15943474.0,
      "step": 4775
    },
    {
      "entropy": 6.64123764038086,
      "epoch": 0.9325919422495367,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937495278415677,
      "loss": 6.5924,
      "mean_token_accuracy": 0.11783003583550453,
      "num_tokens": 15959546.0,
      "step": 4780
    },
    {
      "entropy": 6.614575481414795,
      "epoch": 0.9335674568334796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937330541466465,
      "loss": 6.5174,
      "mean_token_accuracy": 0.12169638276100159,
      "num_tokens": 15975669.0,
      "step": 4785
    },
    {
      "entropy": 6.506360483169556,
      "epoch": 0.9345429714174227,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937165590779108,
      "loss": 6.5112,
      "mean_token_accuracy": 0.14014261662960054,
      "num_tokens": 15992898.0,
      "step": 4790
    },
    {
      "entropy": 6.595851612091065,
      "epoch": 0.9355184860013657,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937000426369724,
      "loss": 6.5318,
      "mean_token_accuracy": 0.1223097413778305,
      "num_tokens": 16010102.0,
      "step": 4795
    },
    {
      "entropy": 6.640869760513306,
      "epoch": 0.9364940005853087,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004936835048254453,
      "loss": 6.6105,
      "mean_token_accuracy": 0.11882219612598419,
      "num_tokens": 16027167.0,
      "step": 4800
    },
    {
      "entropy": 6.595710182189942,
      "epoch": 0.9374695151692518,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936669456449455,
      "loss": 6.5645,
      "mean_token_accuracy": 0.12285120859742164,
      "num_tokens": 16042873.0,
      "step": 4805
    },
    {
      "entropy": 6.597416734695434,
      "epoch": 0.9384450297531948,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936503650970912,
      "loss": 6.6074,
      "mean_token_accuracy": 0.12608925998210907,
      "num_tokens": 16060814.0,
      "step": 4810
    },
    {
      "entropy": 6.648591136932373,
      "epoch": 0.9394205443371378,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004936337631835024,
      "loss": 6.5679,
      "mean_token_accuracy": 0.11792688220739364,
      "num_tokens": 16078378.0,
      "step": 4815
    },
    {
      "entropy": 6.424598741531372,
      "epoch": 0.9403960589210809,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936171399058017,
      "loss": 6.3499,
      "mean_token_accuracy": 0.13799759671092032,
      "num_tokens": 16094601.0,
      "step": 4820
    },
    {
      "entropy": 6.445580101013183,
      "epoch": 0.9413715735050239,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936004952656131,
      "loss": 6.4434,
      "mean_token_accuracy": 0.13445015549659728,
      "num_tokens": 16113189.0,
      "step": 4825
    },
    {
      "entropy": 6.6991331577301025,
      "epoch": 0.9423470880889669,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935838292645632,
      "loss": 6.6613,
      "mean_token_accuracy": 0.12006801664829254,
      "num_tokens": 16131225.0,
      "step": 4830
    },
    {
      "entropy": 6.551232576370239,
      "epoch": 0.9433226026729099,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004935671419042805,
      "loss": 6.5305,
      "mean_token_accuracy": 0.12254934832453727,
      "num_tokens": 16149733.0,
      "step": 4835
    },
    {
      "entropy": 6.560597658157349,
      "epoch": 0.944298117256853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935504331863958,
      "loss": 6.4786,
      "mean_token_accuracy": 0.1280203305184841,
      "num_tokens": 16167499.0,
      "step": 4840
    },
    {
      "entropy": 6.429924821853637,
      "epoch": 0.945273631840796,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004935337031125416,
      "loss": 6.4065,
      "mean_token_accuracy": 0.1410142570734024,
      "num_tokens": 16184858.0,
      "step": 4845
    },
    {
      "entropy": 6.5907329559326175,
      "epoch": 0.946249146424739,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935169516843528,
      "loss": 6.5949,
      "mean_token_accuracy": 0.13123267516493797,
      "num_tokens": 16201824.0,
      "step": 4850
    },
    {
      "entropy": 6.513433504104614,
      "epoch": 0.9472246610086821,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004935001789034664,
      "loss": 6.3972,
      "mean_token_accuracy": 0.13601651042699814,
      "num_tokens": 16219551.0,
      "step": 4855
    },
    {
      "entropy": 6.567834568023682,
      "epoch": 0.9482001755926251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934833847715212,
      "loss": 6.5164,
      "mean_token_accuracy": 0.13442356064915656,
      "num_tokens": 16237447.0,
      "step": 4860
    },
    {
      "entropy": 6.638275384902954,
      "epoch": 0.9491756901765681,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934665692901583,
      "loss": 6.6792,
      "mean_token_accuracy": 0.11922683641314506,
      "num_tokens": 16254377.0,
      "step": 4865
    },
    {
      "entropy": 6.614830207824707,
      "epoch": 0.9501512047605112,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493449732461021,
      "loss": 6.4985,
      "mean_token_accuracy": 0.12872367426753045,
      "num_tokens": 16271669.0,
      "step": 4870
    },
    {
      "entropy": 6.501110982894898,
      "epoch": 0.9511267193444543,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934328742857543,
      "loss": 6.4796,
      "mean_token_accuracy": 0.13378025442361832,
      "num_tokens": 16287088.0,
      "step": 4875
    },
    {
      "entropy": 6.432848691940308,
      "epoch": 0.9521022339283972,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004934159947660057,
      "loss": 6.4155,
      "mean_token_accuracy": 0.1301620826125145,
      "num_tokens": 16303646.0,
      "step": 4880
    },
    {
      "entropy": 6.56607608795166,
      "epoch": 0.9530777485123403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004933990939034244,
      "loss": 6.4897,
      "mean_token_accuracy": 0.12970860078930854,
      "num_tokens": 16320610.0,
      "step": 4885
    },
    {
      "entropy": 6.535564517974853,
      "epoch": 0.9540532630962832,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933821716996621,
      "loss": 6.4407,
      "mean_token_accuracy": 0.13329432606697084,
      "num_tokens": 16337983.0,
      "step": 4890
    },
    {
      "entropy": 6.489419412612915,
      "epoch": 0.9550287776802263,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933652281563723,
      "loss": 6.4974,
      "mean_token_accuracy": 0.1242811068892479,
      "num_tokens": 16356645.0,
      "step": 4895
    },
    {
      "entropy": 6.563281106948852,
      "epoch": 0.9560042922641694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933482632752107,
      "loss": 6.5218,
      "mean_token_accuracy": 0.12523329108953477,
      "num_tokens": 16371619.0,
      "step": 4900
    },
    {
      "entropy": 6.558269929885864,
      "epoch": 0.9569798068481123,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004933312770578351,
      "loss": 6.51,
      "mean_token_accuracy": 0.13266440927982331,
      "num_tokens": 16389831.0,
      "step": 4905
    },
    {
      "entropy": 6.619788646697998,
      "epoch": 0.9579553214320554,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004933142695059052,
      "loss": 6.5688,
      "mean_token_accuracy": 0.11982048377394676,
      "num_tokens": 16406132.0,
      "step": 4910
    },
    {
      "entropy": 6.575911998748779,
      "epoch": 0.9589308360159985,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004932972406210829,
      "loss": 6.5729,
      "mean_token_accuracy": 0.12568789944052697,
      "num_tokens": 16422442.0,
      "step": 4915
    },
    {
      "entropy": 6.542263221740723,
      "epoch": 0.9599063505999414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004932801904050323,
      "loss": 6.4864,
      "mean_token_accuracy": 0.12693430185317994,
      "num_tokens": 16440290.0,
      "step": 4920
    },
    {
      "entropy": 6.626277017593384,
      "epoch": 0.9608818651838845,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004932631188594195,
      "loss": 6.6013,
      "mean_token_accuracy": 0.127669321000576,
      "num_tokens": 16456904.0,
      "step": 4925
    },
    {
      "entropy": 6.5769611358642575,
      "epoch": 0.9618573797678275,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004932460259859127,
      "loss": 6.5205,
      "mean_token_accuracy": 0.13189415782690048,
      "num_tokens": 16473777.0,
      "step": 4930
    },
    {
      "entropy": 6.487201404571533,
      "epoch": 0.9628328943517706,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004932289117861819,
      "loss": 6.4567,
      "mean_token_accuracy": 0.1306969203054905,
      "num_tokens": 16490401.0,
      "step": 4935
    },
    {
      "entropy": 6.578268146514892,
      "epoch": 0.9638084089357136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932117762618997,
      "loss": 6.5112,
      "mean_token_accuracy": 0.12470794841647148,
      "num_tokens": 16508582.0,
      "step": 4940
    },
    {
      "entropy": 6.6271144390106205,
      "epoch": 0.9647839235196566,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931946194147404,
      "loss": 6.5595,
      "mean_token_accuracy": 0.1251258708536625,
      "num_tokens": 16523939.0,
      "step": 4945
    },
    {
      "entropy": 6.502145290374756,
      "epoch": 0.9657594381035997,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004931774412463806,
      "loss": 6.5341,
      "mean_token_accuracy": 0.127706278860569,
      "num_tokens": 16541139.0,
      "step": 4950
    },
    {
      "entropy": 6.62952938079834,
      "epoch": 0.9667349526875427,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931602417584989,
      "loss": 6.5679,
      "mean_token_accuracy": 0.11984375789761544,
      "num_tokens": 16558351.0,
      "step": 4955
    },
    {
      "entropy": 6.601421165466308,
      "epoch": 0.9677104672714857,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004931430209527759,
      "loss": 6.62,
      "mean_token_accuracy": 0.11872599944472313,
      "num_tokens": 16575777.0,
      "step": 4960
    },
    {
      "entropy": 6.660050773620606,
      "epoch": 0.9686859818554288,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931257788308943,
      "loss": 6.5669,
      "mean_token_accuracy": 0.12532232329249382,
      "num_tokens": 16592490.0,
      "step": 4965
    },
    {
      "entropy": 6.56386890411377,
      "epoch": 0.9696614964393717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493108515394539,
      "loss": 6.563,
      "mean_token_accuracy": 0.1281309336423874,
      "num_tokens": 16610378.0,
      "step": 4970
    },
    {
      "entropy": 6.645177555084229,
      "epoch": 0.9706370110233148,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930912306453971,
      "loss": 6.6102,
      "mean_token_accuracy": 0.11900732815265655,
      "num_tokens": 16625445.0,
      "step": 4975
    },
    {
      "entropy": 6.68939208984375,
      "epoch": 0.9716125256072579,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004930739245851571,
      "loss": 6.6029,
      "mean_token_accuracy": 0.11635512039065361,
      "num_tokens": 16641821.0,
      "step": 4980
    },
    {
      "entropy": 6.559339714050293,
      "epoch": 0.9725880401912008,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930565972155106,
      "loss": 6.5685,
      "mean_token_accuracy": 0.120558150857687,
      "num_tokens": 16658418.0,
      "step": 4985
    },
    {
      "entropy": 6.617754697799683,
      "epoch": 0.9735635547751439,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930392485381506,
      "loss": 6.5913,
      "mean_token_accuracy": 0.1256350226700306,
      "num_tokens": 16674218.0,
      "step": 4990
    },
    {
      "entropy": 6.649224424362183,
      "epoch": 0.974539069359087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004930218785547724,
      "loss": 6.599,
      "mean_token_accuracy": 0.12946683913469315,
      "num_tokens": 16692171.0,
      "step": 4995
    },
    {
      "entropy": 6.6916101455688475,
      "epoch": 0.9755145839430299,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004930044872670733,
      "loss": 6.6575,
      "mean_token_accuracy": 0.11413551047444344,
      "num_tokens": 16709231.0,
      "step": 5000
    },
    {
      "entropy": 6.589885330200195,
      "epoch": 0.976490098526973,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929870746767525,
      "loss": 6.5542,
      "mean_token_accuracy": 0.1230994075536728,
      "num_tokens": 16726281.0,
      "step": 5005
    },
    {
      "entropy": 6.589848327636719,
      "epoch": 0.977465613110916,
      "grad_norm": 0.875,
      "learning_rate": 0.0004929696407855118,
      "loss": 6.6276,
      "mean_token_accuracy": 0.11735391840338708,
      "num_tokens": 16744821.0,
      "step": 5010
    },
    {
      "entropy": 6.580704593658448,
      "epoch": 0.978441127694859,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004929521855950546,
      "loss": 6.4352,
      "mean_token_accuracy": 0.12736937776207924,
      "num_tokens": 16762564.0,
      "step": 5015
    },
    {
      "entropy": 6.487126779556275,
      "epoch": 0.9794166422788021,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004929347091070866,
      "loss": 6.391,
      "mean_token_accuracy": 0.1384567655622959,
      "num_tokens": 16777681.0,
      "step": 5020
    },
    {
      "entropy": 6.541004991531372,
      "epoch": 0.9803921568627451,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929172113233157,
      "loss": 6.629,
      "mean_token_accuracy": 0.12214543595910073,
      "num_tokens": 16793820.0,
      "step": 5025
    },
    {
      "entropy": 6.671449851989746,
      "epoch": 0.9813676714466881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928996922454515,
      "loss": 6.6245,
      "mean_token_accuracy": 0.12370117157697677,
      "num_tokens": 16811445.0,
      "step": 5030
    },
    {
      "entropy": 6.601174688339233,
      "epoch": 0.9823431860306312,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928821518752059,
      "loss": 6.5733,
      "mean_token_accuracy": 0.12485999017953872,
      "num_tokens": 16827961.0,
      "step": 5035
    },
    {
      "entropy": 6.608020257949829,
      "epoch": 0.9833187006145742,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928645902142931,
      "loss": 6.4564,
      "mean_token_accuracy": 0.12581372261047363,
      "num_tokens": 16844074.0,
      "step": 5040
    },
    {
      "entropy": 6.540974044799805,
      "epoch": 0.9842942151985172,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000492847007264429,
      "loss": 6.5032,
      "mean_token_accuracy": 0.12328941598534585,
      "num_tokens": 16862004.0,
      "step": 5045
    },
    {
      "entropy": 6.5645670890808105,
      "epoch": 0.9852697297824603,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928294030273317,
      "loss": 6.5198,
      "mean_token_accuracy": 0.1223757542669773,
      "num_tokens": 16879738.0,
      "step": 5050
    },
    {
      "entropy": 6.473462438583374,
      "epoch": 0.9862452443664033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004928117775047215,
      "loss": 6.4729,
      "mean_token_accuracy": 0.13235039561986922,
      "num_tokens": 16896565.0,
      "step": 5055
    },
    {
      "entropy": 6.609834384918213,
      "epoch": 0.9872207589503463,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927941306983207,
      "loss": 6.5596,
      "mean_token_accuracy": 0.1282456822693348,
      "num_tokens": 16912443.0,
      "step": 5060
    },
    {
      "entropy": 6.663063001632691,
      "epoch": 0.9881962735342893,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927764626098537,
      "loss": 6.6016,
      "mean_token_accuracy": 0.12092144340276718,
      "num_tokens": 16928292.0,
      "step": 5065
    },
    {
      "entropy": 6.578760957717895,
      "epoch": 0.9891717881182324,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000492758773241047,
      "loss": 6.5591,
      "mean_token_accuracy": 0.1281025916337967,
      "num_tokens": 16945432.0,
      "step": 5070
    },
    {
      "entropy": 6.507167625427246,
      "epoch": 0.9901473027021754,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000492741062593629,
      "loss": 6.3911,
      "mean_token_accuracy": 0.13755975365638734,
      "num_tokens": 16962440.0,
      "step": 5075
    },
    {
      "entropy": 6.513244533538819,
      "epoch": 0.9911228172861184,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004927233306693305,
      "loss": 6.5027,
      "mean_token_accuracy": 0.13238978758454323,
      "num_tokens": 16979227.0,
      "step": 5080
    },
    {
      "entropy": 6.6143170356750485,
      "epoch": 0.9920983318700615,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927055774698839,
      "loss": 6.5753,
      "mean_token_accuracy": 0.12050328850746155,
      "num_tokens": 16994873.0,
      "step": 5085
    },
    {
      "entropy": 6.643030452728271,
      "epoch": 0.9930738464540045,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926878029970245,
      "loss": 6.6202,
      "mean_token_accuracy": 0.12106775492429733,
      "num_tokens": 17011516.0,
      "step": 5090
    },
    {
      "entropy": 6.6469810009002686,
      "epoch": 0.9940493610379475,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926700072524887,
      "loss": 6.5242,
      "mean_token_accuracy": 0.12547477632761,
      "num_tokens": 17029407.0,
      "step": 5095
    },
    {
      "entropy": 6.525854015350342,
      "epoch": 0.9950248756218906,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926521902380154,
      "loss": 6.5847,
      "mean_token_accuracy": 0.12023040428757667,
      "num_tokens": 17044451.0,
      "step": 5100
    },
    {
      "entropy": 6.486868953704834,
      "epoch": 0.9960003902058335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926343519553461,
      "loss": 6.4294,
      "mean_token_accuracy": 0.13615420907735826,
      "num_tokens": 17060420.0,
      "step": 5105
    },
    {
      "entropy": 6.604794454574585,
      "epoch": 0.9969759047897766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926164924062234,
      "loss": 6.5597,
      "mean_token_accuracy": 0.12576316446065902,
      "num_tokens": 17076562.0,
      "step": 5110
    },
    {
      "entropy": 6.4584331035614015,
      "epoch": 0.9979514193737197,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004925986115923927,
      "loss": 6.6106,
      "mean_token_accuracy": 0.13719551265239716,
      "num_tokens": 17092441.0,
      "step": 5115
    },
    {
      "entropy": 6.634578704833984,
      "epoch": 0.9989269339576626,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004925807095156013,
      "loss": 6.584,
      "mean_token_accuracy": 0.12233686745166779,
      "num_tokens": 17109447.0,
      "step": 5120
    },
    {
      "entropy": 6.547192096710205,
      "epoch": 0.9999024485416057,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925627861775983,
      "loss": 6.4616,
      "mean_token_accuracy": 0.1306936465203762,
      "num_tokens": 17125795.0,
      "step": 5125
    },
    {
      "entropy": 6.543901549445258,
      "epoch": 1.0007804116671544,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925448415801353,
      "loss": 6.486,
      "mean_token_accuracy": 0.1299235431684388,
      "num_tokens": 17139062.0,
      "step": 5130
    },
    {
      "entropy": 6.642700672149658,
      "epoch": 1.0017559262510976,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925268757249657,
      "loss": 6.5074,
      "mean_token_accuracy": 0.11787568256258965,
      "num_tokens": 17155465.0,
      "step": 5135
    },
    {
      "entropy": 6.580320119857788,
      "epoch": 1.0027314408350405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925088886138451,
      "loss": 6.3898,
      "mean_token_accuracy": 0.12574450001120568,
      "num_tokens": 17171284.0,
      "step": 5140
    },
    {
      "entropy": 6.4070985317230225,
      "epoch": 1.0037069554189835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924908802485311,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1364549957215786,
      "num_tokens": 17190823.0,
      "step": 5145
    },
    {
      "entropy": 6.5885176181793215,
      "epoch": 1.0046824700029267,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004924728506307835,
      "loss": 6.4439,
      "mean_token_accuracy": 0.12542033568024635,
      "num_tokens": 17208866.0,
      "step": 5150
    },
    {
      "entropy": 6.593297624588013,
      "epoch": 1.0056579845868696,
      "grad_norm": 1.09375,
      "learning_rate": 0.000492454799762364,
      "loss": 6.532,
      "mean_token_accuracy": 0.12006717696785926,
      "num_tokens": 17225609.0,
      "step": 5155
    },
    {
      "entropy": 6.655113458633423,
      "epoch": 1.0066334991708126,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004924367276450364,
      "loss": 6.5155,
      "mean_token_accuracy": 0.12401865050196648,
      "num_tokens": 17242198.0,
      "step": 5160
    },
    {
      "entropy": 6.520212268829345,
      "epoch": 1.0076090137547555,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924186342805669,
      "loss": 6.4768,
      "mean_token_accuracy": 0.12438175305724145,
      "num_tokens": 17258712.0,
      "step": 5165
    },
    {
      "entropy": 6.55370717048645,
      "epoch": 1.0085845283386987,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924005196707233,
      "loss": 6.466,
      "mean_token_accuracy": 0.12993448227643967,
      "num_tokens": 17274790.0,
      "step": 5170
    },
    {
      "entropy": 6.5888594627380375,
      "epoch": 1.0095600429226417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923823838172756,
      "loss": 6.4045,
      "mean_token_accuracy": 0.12780738547444342,
      "num_tokens": 17290526.0,
      "step": 5175
    },
    {
      "entropy": 6.554125738143921,
      "epoch": 1.0105355575065846,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923642267219962,
      "loss": 6.4925,
      "mean_token_accuracy": 0.12453892305493355,
      "num_tokens": 17306022.0,
      "step": 5180
    },
    {
      "entropy": 6.6284057140350345,
      "epoch": 1.0115110720905278,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923460483866593,
      "loss": 6.4749,
      "mean_token_accuracy": 0.1256857134401798,
      "num_tokens": 17321601.0,
      "step": 5185
    },
    {
      "entropy": 6.428729629516601,
      "epoch": 1.0124865866744708,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923278488130412,
      "loss": 6.3365,
      "mean_token_accuracy": 0.13379083201289177,
      "num_tokens": 17337997.0,
      "step": 5190
    },
    {
      "entropy": 6.551972913742065,
      "epoch": 1.0134621012584137,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004923096280029202,
      "loss": 6.4396,
      "mean_token_accuracy": 0.1262338303029537,
      "num_tokens": 17354809.0,
      "step": 5195
    },
    {
      "entropy": 6.499165678024292,
      "epoch": 1.014437615842357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492291385958077,
      "loss": 6.305,
      "mean_token_accuracy": 0.1373203881084919,
      "num_tokens": 17371548.0,
      "step": 5200
    },
    {
      "entropy": 6.483451509475708,
      "epoch": 1.0154131304262999,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004922731226802939,
      "loss": 6.4109,
      "mean_token_accuracy": 0.12878562808036803,
      "num_tokens": 17388891.0,
      "step": 5205
    },
    {
      "entropy": 6.480581188201905,
      "epoch": 1.0163886450102428,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004922548381713556,
      "loss": 6.3965,
      "mean_token_accuracy": 0.13317395523190498,
      "num_tokens": 17407501.0,
      "step": 5210
    },
    {
      "entropy": 6.5679765224456785,
      "epoch": 1.017364159594186,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922365324330489,
      "loss": 6.4511,
      "mean_token_accuracy": 0.12619701623916627,
      "num_tokens": 17425541.0,
      "step": 5215
    },
    {
      "entropy": 6.532534170150757,
      "epoch": 1.018339674178129,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922182054671625,
      "loss": 6.3728,
      "mean_token_accuracy": 0.12432115897536278,
      "num_tokens": 17441588.0,
      "step": 5220
    },
    {
      "entropy": 6.5338184356689455,
      "epoch": 1.019315188762072,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921998572754872,
      "loss": 6.3766,
      "mean_token_accuracy": 0.12910501882433892,
      "num_tokens": 17458874.0,
      "step": 5225
    },
    {
      "entropy": 6.45886197090149,
      "epoch": 1.0202907033460151,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921814878598159,
      "loss": 6.3513,
      "mean_token_accuracy": 0.13410900682210922,
      "num_tokens": 17475815.0,
      "step": 5230
    },
    {
      "entropy": 6.487808179855347,
      "epoch": 1.021266217929958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921630972219437,
      "loss": 6.3876,
      "mean_token_accuracy": 0.12828269004821777,
      "num_tokens": 17492068.0,
      "step": 5235
    },
    {
      "entropy": 6.498036575317383,
      "epoch": 1.022241732513901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921446853636676,
      "loss": 6.3715,
      "mean_token_accuracy": 0.12724538892507553,
      "num_tokens": 17508156.0,
      "step": 5240
    },
    {
      "entropy": 6.549976825714111,
      "epoch": 1.023217247097844,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004921262522867869,
      "loss": 6.4671,
      "mean_token_accuracy": 0.12247593328356743,
      "num_tokens": 17525374.0,
      "step": 5245
    },
    {
      "entropy": 6.5659102439880375,
      "epoch": 1.0241927616817872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921077979931026,
      "loss": 6.3464,
      "mean_token_accuracy": 0.1320713482797146,
      "num_tokens": 17541453.0,
      "step": 5250
    },
    {
      "entropy": 6.519372463226318,
      "epoch": 1.0251682762657301,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000492089322484418,
      "loss": 6.4091,
      "mean_token_accuracy": 0.12513896450400352,
      "num_tokens": 17557981.0,
      "step": 5255
    },
    {
      "entropy": 6.477697515487671,
      "epoch": 1.026143790849673,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920708257625386,
      "loss": 6.3154,
      "mean_token_accuracy": 0.1310434892773628,
      "num_tokens": 17573094.0,
      "step": 5260
    },
    {
      "entropy": 6.5890285015106205,
      "epoch": 1.0271193054336163,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920523078292718,
      "loss": 6.5426,
      "mean_token_accuracy": 0.12401133254170418,
      "num_tokens": 17589933.0,
      "step": 5265
    },
    {
      "entropy": 6.610663414001465,
      "epoch": 1.0280948200175593,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920337686864269,
      "loss": 6.4706,
      "mean_token_accuracy": 0.12492510154843331,
      "num_tokens": 17607008.0,
      "step": 5270
    },
    {
      "entropy": 6.593128871917725,
      "epoch": 1.0290703346015022,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920152083358158,
      "loss": 6.426,
      "mean_token_accuracy": 0.13316763043403626,
      "num_tokens": 17624160.0,
      "step": 5275
    },
    {
      "entropy": 6.51133246421814,
      "epoch": 1.0300458491854454,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000491996626779252,
      "loss": 6.3832,
      "mean_token_accuracy": 0.1318292073905468,
      "num_tokens": 17640067.0,
      "step": 5280
    },
    {
      "entropy": 6.584697628021241,
      "epoch": 1.0310213637693884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004919780240185512,
      "loss": 6.4751,
      "mean_token_accuracy": 0.12637813314795493,
      "num_tokens": 17658278.0,
      "step": 5285
    },
    {
      "entropy": 6.608877277374267,
      "epoch": 1.0319968783533313,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004919594000555311,
      "loss": 6.4655,
      "mean_token_accuracy": 0.12451366484165191,
      "num_tokens": 17673878.0,
      "step": 5290
    },
    {
      "entropy": 6.488671541213989,
      "epoch": 1.0329723929372745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919407548920118,
      "loss": 6.3915,
      "mean_token_accuracy": 0.12885335385799407,
      "num_tokens": 17688872.0,
      "step": 5295
    },
    {
      "entropy": 6.570520305633545,
      "epoch": 1.0339479075212175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000491922088529815,
      "loss": 6.4222,
      "mean_token_accuracy": 0.12663409486413002,
      "num_tokens": 17704284.0,
      "step": 5300
    },
    {
      "entropy": 6.441459703445434,
      "epoch": 1.0349234221051604,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004919034009707651,
      "loss": 6.2912,
      "mean_token_accuracy": 0.13701564073562622,
      "num_tokens": 17721296.0,
      "step": 5305
    },
    {
      "entropy": 6.562744283676148,
      "epoch": 1.0358989366891036,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918846922166877,
      "loss": 6.4985,
      "mean_token_accuracy": 0.12668490931391715,
      "num_tokens": 17738549.0,
      "step": 5310
    },
    {
      "entropy": 6.525688791275025,
      "epoch": 1.0368744512730466,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004918659622694111,
      "loss": 6.3775,
      "mean_token_accuracy": 0.13631529957056046,
      "num_tokens": 17755297.0,
      "step": 5315
    },
    {
      "entropy": 6.476410818099976,
      "epoch": 1.0378499658569895,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004918472111307658,
      "loss": 6.4181,
      "mean_token_accuracy": 0.1274620771408081,
      "num_tokens": 17771459.0,
      "step": 5320
    },
    {
      "entropy": 6.494926452636719,
      "epoch": 1.0388254804409325,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918284388025837,
      "loss": 6.4222,
      "mean_token_accuracy": 0.13316541090607642,
      "num_tokens": 17787009.0,
      "step": 5325
    },
    {
      "entropy": 6.542292928695678,
      "epoch": 1.0398009950248757,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004918096452866995,
      "loss": 6.4418,
      "mean_token_accuracy": 0.12500547543168067,
      "num_tokens": 17804643.0,
      "step": 5330
    },
    {
      "entropy": 6.5983319759368895,
      "epoch": 1.0407765096088186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917908305849493,
      "loss": 6.4917,
      "mean_token_accuracy": 0.1235867828130722,
      "num_tokens": 17821104.0,
      "step": 5335
    },
    {
      "entropy": 6.630991125106812,
      "epoch": 1.0417520241927616,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491771994699172,
      "loss": 6.4894,
      "mean_token_accuracy": 0.12842014953494071,
      "num_tokens": 17838094.0,
      "step": 5340
    },
    {
      "entropy": 6.389151096343994,
      "epoch": 1.0427275387767048,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917531376312079,
      "loss": 6.3312,
      "mean_token_accuracy": 0.13209505304694175,
      "num_tokens": 17853856.0,
      "step": 5345
    },
    {
      "entropy": 6.558684825897217,
      "epoch": 1.0437030533606477,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004917342593828997,
      "loss": 6.4536,
      "mean_token_accuracy": 0.11728858128190041,
      "num_tokens": 17870032.0,
      "step": 5350
    },
    {
      "entropy": 6.483453321456909,
      "epoch": 1.0446785679445907,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491715359956092,
      "loss": 6.3387,
      "mean_token_accuracy": 0.13639483228325844,
      "num_tokens": 17886384.0,
      "step": 5355
    },
    {
      "entropy": 6.68730936050415,
      "epoch": 1.0456540825285339,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916964393526319,
      "loss": 6.5521,
      "mean_token_accuracy": 0.12626202329993247,
      "num_tokens": 17903444.0,
      "step": 5360
    },
    {
      "entropy": 6.589110660552978,
      "epoch": 1.0466295971124768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004916774975743679,
      "loss": 6.3992,
      "mean_token_accuracy": 0.12684224173426628,
      "num_tokens": 17920042.0,
      "step": 5365
    },
    {
      "entropy": 6.424794054031372,
      "epoch": 1.0476051116964198,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004916585346231513,
      "loss": 6.3912,
      "mean_token_accuracy": 0.14277585595846176,
      "num_tokens": 17937115.0,
      "step": 5370
    },
    {
      "entropy": 6.543014430999756,
      "epoch": 1.048580626280363,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004916395505008347,
      "loss": 6.4379,
      "mean_token_accuracy": 0.12608346790075303,
      "num_tokens": 17953582.0,
      "step": 5375
    },
    {
      "entropy": 6.462535381317139,
      "epoch": 1.049556140864306,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916205452092735,
      "loss": 6.4421,
      "mean_token_accuracy": 0.12468311265110969,
      "num_tokens": 17970457.0,
      "step": 5380
    },
    {
      "entropy": 6.552306222915649,
      "epoch": 1.050531655448249,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004916015187503245,
      "loss": 6.4127,
      "mean_token_accuracy": 0.13060517907142638,
      "num_tokens": 17987168.0,
      "step": 5385
    },
    {
      "entropy": 6.4050170421600345,
      "epoch": 1.051507170032192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915824711258472,
      "loss": 6.3547,
      "mean_token_accuracy": 0.13541659563779831,
      "num_tokens": 18004964.0,
      "step": 5390
    },
    {
      "entropy": 6.534900856018067,
      "epoch": 1.052482684616135,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004915634023377026,
      "loss": 6.4337,
      "mean_token_accuracy": 0.1288573294878006,
      "num_tokens": 18021778.0,
      "step": 5395
    },
    {
      "entropy": 6.385112047195435,
      "epoch": 1.053458199200078,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004915443123877543,
      "loss": 6.2787,
      "mean_token_accuracy": 0.1336907796561718,
      "num_tokens": 18037620.0,
      "step": 5400
    },
    {
      "entropy": 6.552325487136841,
      "epoch": 1.0544337137840212,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915252012778676,
      "loss": 6.394,
      "mean_token_accuracy": 0.13195589184761047,
      "num_tokens": 18054623.0,
      "step": 5405
    },
    {
      "entropy": 6.4744292259216305,
      "epoch": 1.0554092283679641,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915060690099098,
      "loss": 6.3933,
      "mean_token_accuracy": 0.12510302737355233,
      "num_tokens": 18071271.0,
      "step": 5410
    },
    {
      "entropy": 6.571962642669678,
      "epoch": 1.056384742951907,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004914869155857506,
      "loss": 6.4719,
      "mean_token_accuracy": 0.12047100514173507,
      "num_tokens": 18087661.0,
      "step": 5415
    },
    {
      "entropy": 6.478859949111938,
      "epoch": 1.05736025753585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914677410072616,
      "loss": 6.3695,
      "mean_token_accuracy": 0.1329769879579544,
      "num_tokens": 18106116.0,
      "step": 5420
    },
    {
      "entropy": 6.552979183197022,
      "epoch": 1.0583357721197932,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004914485452763165,
      "loss": 6.4803,
      "mean_token_accuracy": 0.12470565959811211,
      "num_tokens": 18122420.0,
      "step": 5425
    },
    {
      "entropy": 6.59436993598938,
      "epoch": 1.0593112867037362,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491429328394791,
      "loss": 6.4897,
      "mean_token_accuracy": 0.12055835202336311,
      "num_tokens": 18137974.0,
      "step": 5430
    },
    {
      "entropy": 6.425874805450439,
      "epoch": 1.0602868012876792,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914100903645629,
      "loss": 6.3083,
      "mean_token_accuracy": 0.13964463025331497,
      "num_tokens": 18154711.0,
      "step": 5435
    },
    {
      "entropy": 6.5224708080291744,
      "epoch": 1.0612623158716223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491390831187512,
      "loss": 6.4956,
      "mean_token_accuracy": 0.12689333856105806,
      "num_tokens": 18172422.0,
      "step": 5440
    },
    {
      "entropy": 6.472756052017212,
      "epoch": 1.0622378304555653,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004913715508655203,
      "loss": 6.3615,
      "mean_token_accuracy": 0.1397356577217579,
      "num_tokens": 18189517.0,
      "step": 5445
    },
    {
      "entropy": 6.516772222518921,
      "epoch": 1.0632133450395083,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913522494004719,
      "loss": 6.427,
      "mean_token_accuracy": 0.11843429729342461,
      "num_tokens": 18206339.0,
      "step": 5450
    },
    {
      "entropy": 6.433005952835083,
      "epoch": 1.0641888596234514,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913329267942527,
      "loss": 6.2785,
      "mean_token_accuracy": 0.1416793182492256,
      "num_tokens": 18223953.0,
      "step": 5455
    },
    {
      "entropy": 6.61129240989685,
      "epoch": 1.0651643742073944,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491313583048751,
      "loss": 6.4058,
      "mean_token_accuracy": 0.1288385085761547,
      "num_tokens": 18241361.0,
      "step": 5460
    },
    {
      "entropy": 6.431782293319702,
      "epoch": 1.0661398887913374,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004912942181658569,
      "loss": 6.4317,
      "mean_token_accuracy": 0.1297380320727825,
      "num_tokens": 18257897.0,
      "step": 5465
    },
    {
      "entropy": 6.580384254455566,
      "epoch": 1.0671154033752805,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004912748321474627,
      "loss": 6.4405,
      "mean_token_accuracy": 0.12627588510513305,
      "num_tokens": 18274935.0,
      "step": 5470
    },
    {
      "entropy": 6.470012903213501,
      "epoch": 1.0680909179592235,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004912554249954626,
      "loss": 6.4855,
      "mean_token_accuracy": 0.1255176343023777,
      "num_tokens": 18292529.0,
      "step": 5475
    },
    {
      "entropy": 6.610748672485352,
      "epoch": 1.0690664325431665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912359967117532,
      "loss": 6.4318,
      "mean_token_accuracy": 0.1265476994216442,
      "num_tokens": 18310278.0,
      "step": 5480
    },
    {
      "entropy": 6.5340207576751705,
      "epoch": 1.0700419471271096,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004912165472982329,
      "loss": 6.4403,
      "mean_token_accuracy": 0.13245197534561157,
      "num_tokens": 18326972.0,
      "step": 5485
    },
    {
      "entropy": 6.516624927520752,
      "epoch": 1.0710174617110526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004911970767568022,
      "loss": 6.5307,
      "mean_token_accuracy": 0.12481719180941582,
      "num_tokens": 18345065.0,
      "step": 5490
    },
    {
      "entropy": 6.539786100387573,
      "epoch": 1.0719929762949956,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004911775850893637,
      "loss": 6.4737,
      "mean_token_accuracy": 0.1282710961997509,
      "num_tokens": 18363251.0,
      "step": 5495
    },
    {
      "entropy": 6.470887851715088,
      "epoch": 1.0729684908789388,
      "grad_norm": 0.953125,
      "learning_rate": 0.000491158072297822,
      "loss": 6.2992,
      "mean_token_accuracy": 0.13582689315080643,
      "num_tokens": 18378650.0,
      "step": 5500
    },
    {
      "entropy": 6.524680328369141,
      "epoch": 1.0739440054628817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004911385383840839,
      "loss": 6.449,
      "mean_token_accuracy": 0.1297501429915428,
      "num_tokens": 18394533.0,
      "step": 5505
    },
    {
      "entropy": 6.597027158737182,
      "epoch": 1.0749195200468247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911189833500582,
      "loss": 6.4884,
      "mean_token_accuracy": 0.1234996847808361,
      "num_tokens": 18410752.0,
      "step": 5510
    },
    {
      "entropy": 6.391909265518189,
      "epoch": 1.0758950346307676,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910994071976556,
      "loss": 6.2779,
      "mean_token_accuracy": 0.13792186751961708,
      "num_tokens": 18428027.0,
      "step": 5515
    },
    {
      "entropy": 6.483095407485962,
      "epoch": 1.0768705492147108,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910798099287894,
      "loss": 6.3619,
      "mean_token_accuracy": 0.13373625725507737,
      "num_tokens": 18446380.0,
      "step": 5520
    },
    {
      "entropy": 6.395686626434326,
      "epoch": 1.0778460637986538,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491060191545374,
      "loss": 6.2976,
      "mean_token_accuracy": 0.14429001659154891,
      "num_tokens": 18464628.0,
      "step": 5525
    },
    {
      "entropy": 6.484730672836304,
      "epoch": 1.0788215783825967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910405520493268,
      "loss": 6.395,
      "mean_token_accuracy": 0.12930625081062316,
      "num_tokens": 18479995.0,
      "step": 5530
    },
    {
      "entropy": 6.549951362609863,
      "epoch": 1.07979709296654,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004910208914425669,
      "loss": 6.4664,
      "mean_token_accuracy": 0.1281146414577961,
      "num_tokens": 18497143.0,
      "step": 5535
    },
    {
      "entropy": 6.524411582946778,
      "epoch": 1.0807726075504829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910012097270153,
      "loss": 6.4284,
      "mean_token_accuracy": 0.1245824545621872,
      "num_tokens": 18513428.0,
      "step": 5540
    },
    {
      "entropy": 6.609167385101318,
      "epoch": 1.0817481221344258,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909815069045955,
      "loss": 6.5365,
      "mean_token_accuracy": 0.12252979725599289,
      "num_tokens": 18529475.0,
      "step": 5545
    },
    {
      "entropy": 6.636512899398804,
      "epoch": 1.082723636718369,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909617829772325,
      "loss": 6.4689,
      "mean_token_accuracy": 0.12756646499037744,
      "num_tokens": 18546156.0,
      "step": 5550
    },
    {
      "entropy": 6.335690450668335,
      "epoch": 1.083699151302312,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909420379468537,
      "loss": 6.2429,
      "mean_token_accuracy": 0.15434979870915413,
      "num_tokens": 18562719.0,
      "step": 5555
    },
    {
      "entropy": 6.515286493301391,
      "epoch": 1.084674665886255,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909222718153887,
      "loss": 6.4212,
      "mean_token_accuracy": 0.1350042961537838,
      "num_tokens": 18580123.0,
      "step": 5560
    },
    {
      "entropy": 6.452692365646362,
      "epoch": 1.0856501804701981,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909024845847689,
      "loss": 6.316,
      "mean_token_accuracy": 0.13494626954197883,
      "num_tokens": 18597003.0,
      "step": 5565
    },
    {
      "entropy": 6.510738658905029,
      "epoch": 1.086625695054141,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908826762569276,
      "loss": 6.4224,
      "mean_token_accuracy": 0.12684929892420768,
      "num_tokens": 18613812.0,
      "step": 5570
    },
    {
      "entropy": 6.427829456329346,
      "epoch": 1.087601209638084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004908628468338008,
      "loss": 6.3548,
      "mean_token_accuracy": 0.1417312055826187,
      "num_tokens": 18630333.0,
      "step": 5575
    },
    {
      "entropy": 6.5562749862670895,
      "epoch": 1.088576724222027,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000490842996317326,
      "loss": 6.4478,
      "mean_token_accuracy": 0.12321656420826912,
      "num_tokens": 18646784.0,
      "step": 5580
    },
    {
      "entropy": 6.524008798599243,
      "epoch": 1.0895522388059702,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004908231247094427,
      "loss": 6.3993,
      "mean_token_accuracy": 0.13203469887375832,
      "num_tokens": 18664358.0,
      "step": 5585
    },
    {
      "entropy": 6.451200199127197,
      "epoch": 1.0905277533899131,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908032320120929,
      "loss": 6.3222,
      "mean_token_accuracy": 0.1336761973798275,
      "num_tokens": 18681849.0,
      "step": 5590
    },
    {
      "entropy": 6.5338733196258545,
      "epoch": 1.091503267973856,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907833182272204,
      "loss": 6.4373,
      "mean_token_accuracy": 0.1272146612405777,
      "num_tokens": 18697127.0,
      "step": 5595
    },
    {
      "entropy": 6.422943639755249,
      "epoch": 1.0924787825577993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907633833567711,
      "loss": 6.3815,
      "mean_token_accuracy": 0.13286206424236296,
      "num_tokens": 18714859.0,
      "step": 5600
    },
    {
      "entropy": 6.52388710975647,
      "epoch": 1.0934542971417422,
      "grad_norm": 1.328125,
      "learning_rate": 0.000490743427402693,
      "loss": 6.3952,
      "mean_token_accuracy": 0.12756307274103165,
      "num_tokens": 18730083.0,
      "step": 5605
    },
    {
      "entropy": 6.452530241012573,
      "epoch": 1.0944298117256852,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004907234503669361,
      "loss": 6.3451,
      "mean_token_accuracy": 0.12977044135332108,
      "num_tokens": 18746694.0,
      "step": 5610
    },
    {
      "entropy": 6.575695753097534,
      "epoch": 1.0954053263096284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004907034522514524,
      "loss": 6.4208,
      "mean_token_accuracy": 0.13124295845627784,
      "num_tokens": 18764333.0,
      "step": 5615
    },
    {
      "entropy": 6.518064689636231,
      "epoch": 1.0963808408935714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906834330581963,
      "loss": 6.4296,
      "mean_token_accuracy": 0.12336959093809127,
      "num_tokens": 18780309.0,
      "step": 5620
    },
    {
      "entropy": 6.492573833465576,
      "epoch": 1.0973563554775143,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004906633927891237,
      "loss": 6.4434,
      "mean_token_accuracy": 0.1296766772866249,
      "num_tokens": 18797322.0,
      "step": 5625
    },
    {
      "entropy": 6.507973909378052,
      "epoch": 1.0983318700614575,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490643331446193,
      "loss": 6.4701,
      "mean_token_accuracy": 0.12462025359272957,
      "num_tokens": 18813185.0,
      "step": 5630
    },
    {
      "entropy": 6.434871196746826,
      "epoch": 1.0993073846454005,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906232490313646,
      "loss": 6.3477,
      "mean_token_accuracy": 0.13458036184310912,
      "num_tokens": 18829445.0,
      "step": 5635
    },
    {
      "entropy": 6.539690923690796,
      "epoch": 1.1002828992293434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004906031455466008,
      "loss": 6.5127,
      "mean_token_accuracy": 0.12800995633006096,
      "num_tokens": 18846229.0,
      "step": 5640
    },
    {
      "entropy": 6.542237615585327,
      "epoch": 1.1012584138132866,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004905830209938659,
      "loss": 6.46,
      "mean_token_accuracy": 0.12293026074767113,
      "num_tokens": 18861904.0,
      "step": 5645
    },
    {
      "entropy": 6.532463550567627,
      "epoch": 1.1022339283972296,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905628753751267,
      "loss": 6.4902,
      "mean_token_accuracy": 0.12506885156035424,
      "num_tokens": 18879160.0,
      "step": 5650
    },
    {
      "entropy": 6.601658058166504,
      "epoch": 1.1032094429811725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905427086923515,
      "loss": 6.479,
      "mean_token_accuracy": 0.12098010703921318,
      "num_tokens": 18895388.0,
      "step": 5655
    },
    {
      "entropy": 6.50784330368042,
      "epoch": 1.1041849575651157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490522520947511,
      "loss": 6.4236,
      "mean_token_accuracy": 0.12711116075515747,
      "num_tokens": 18912392.0,
      "step": 5660
    },
    {
      "entropy": 6.552887487411499,
      "epoch": 1.1051604721490587,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905023121425779,
      "loss": 6.4131,
      "mean_token_accuracy": 0.12565437257289885,
      "num_tokens": 18928721.0,
      "step": 5665
    },
    {
      "entropy": 6.571647596359253,
      "epoch": 1.1061359867330016,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490482082279527,
      "loss": 6.4737,
      "mean_token_accuracy": 0.12290273308753967,
      "num_tokens": 18947144.0,
      "step": 5670
    },
    {
      "entropy": 6.495156383514404,
      "epoch": 1.1071115013169446,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490461831360335,
      "loss": 6.4029,
      "mean_token_accuracy": 0.12450420036911965,
      "num_tokens": 18963682.0,
      "step": 5675
    },
    {
      "entropy": 6.502926683425903,
      "epoch": 1.1080870159008878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904415593869806,
      "loss": 6.4359,
      "mean_token_accuracy": 0.12691038623452186,
      "num_tokens": 18980812.0,
      "step": 5680
    },
    {
      "entropy": 6.520251846313476,
      "epoch": 1.1090625304848307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904212663614452,
      "loss": 6.4051,
      "mean_token_accuracy": 0.1274564929306507,
      "num_tokens": 18996816.0,
      "step": 5685
    },
    {
      "entropy": 6.502163696289062,
      "epoch": 1.1100380450687737,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004904009522857111,
      "loss": 6.4236,
      "mean_token_accuracy": 0.12663414478302001,
      "num_tokens": 19012324.0,
      "step": 5690
    },
    {
      "entropy": 6.369348859786987,
      "epoch": 1.1110135596527169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903806171617638,
      "loss": 6.2713,
      "mean_token_accuracy": 0.1388439081609249,
      "num_tokens": 19030195.0,
      "step": 5695
    },
    {
      "entropy": 6.471956586837768,
      "epoch": 1.1119890742366598,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004903602609915901,
      "loss": 6.2779,
      "mean_token_accuracy": 0.13918960466980934,
      "num_tokens": 19048467.0,
      "step": 5700
    },
    {
      "entropy": 6.437216901779175,
      "epoch": 1.1129645888206028,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903398837771794,
      "loss": 6.3929,
      "mean_token_accuracy": 0.13159190937876702,
      "num_tokens": 19066290.0,
      "step": 5705
    },
    {
      "entropy": 6.39988341331482,
      "epoch": 1.113940103404546,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004903194855205227,
      "loss": 6.3191,
      "mean_token_accuracy": 0.1365206614136696,
      "num_tokens": 19083455.0,
      "step": 5710
    },
    {
      "entropy": 6.543653345108032,
      "epoch": 1.114915617988489,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004902990662236132,
      "loss": 6.3714,
      "mean_token_accuracy": 0.12700596079230309,
      "num_tokens": 19100347.0,
      "step": 5715
    },
    {
      "entropy": 6.42630820274353,
      "epoch": 1.1158911325724319,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902786258884463,
      "loss": 6.427,
      "mean_token_accuracy": 0.1294461965560913,
      "num_tokens": 19117379.0,
      "step": 5720
    },
    {
      "entropy": 6.526592493057251,
      "epoch": 1.116866647156375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902581645170194,
      "loss": 6.3781,
      "mean_token_accuracy": 0.1336417481303215,
      "num_tokens": 19134751.0,
      "step": 5725
    },
    {
      "entropy": 6.4938384056091305,
      "epoch": 1.117842161740318,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902376821113319,
      "loss": 6.4188,
      "mean_token_accuracy": 0.1278844356536865,
      "num_tokens": 19150647.0,
      "step": 5730
    },
    {
      "entropy": 6.460244989395141,
      "epoch": 1.118817676324261,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902171786733852,
      "loss": 6.3908,
      "mean_token_accuracy": 0.12736958041787147,
      "num_tokens": 19166654.0,
      "step": 5735
    },
    {
      "entropy": 6.472190856933594,
      "epoch": 1.1197931909082042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901966542051828,
      "loss": 6.3906,
      "mean_token_accuracy": 0.12907447591423987,
      "num_tokens": 19182787.0,
      "step": 5740
    },
    {
      "entropy": 6.492206287384033,
      "epoch": 1.1207687054921471,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004901761087087303,
      "loss": 6.3853,
      "mean_token_accuracy": 0.132924634963274,
      "num_tokens": 19200218.0,
      "step": 5745
    },
    {
      "entropy": 6.55646276473999,
      "epoch": 1.12174422007609,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901555421860354,
      "loss": 6.4706,
      "mean_token_accuracy": 0.13060002252459527,
      "num_tokens": 19217310.0,
      "step": 5750
    },
    {
      "entropy": 6.574576377868652,
      "epoch": 1.1227197346600333,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901349546391078,
      "loss": 6.4299,
      "mean_token_accuracy": 0.1237056277692318,
      "num_tokens": 19233293.0,
      "step": 5755
    },
    {
      "entropy": 6.437239456176758,
      "epoch": 1.1236952492439762,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901143460699592,
      "loss": 6.3212,
      "mean_token_accuracy": 0.14128868877887726,
      "num_tokens": 19250606.0,
      "step": 5760
    },
    {
      "entropy": 6.400079774856567,
      "epoch": 1.1246707638279192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004900937164806033,
      "loss": 6.2896,
      "mean_token_accuracy": 0.1401648662984371,
      "num_tokens": 19267848.0,
      "step": 5765
    },
    {
      "entropy": 6.499271249771118,
      "epoch": 1.1256462784118622,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004900730658730561,
      "loss": 6.3952,
      "mean_token_accuracy": 0.13200172856450082,
      "num_tokens": 19284317.0,
      "step": 5770
    },
    {
      "entropy": 6.438356065750122,
      "epoch": 1.1266217929958053,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004900523942493353,
      "loss": 6.3576,
      "mean_token_accuracy": 0.13344767391681672,
      "num_tokens": 19300766.0,
      "step": 5775
    },
    {
      "entropy": 6.520315551757813,
      "epoch": 1.1275973075797483,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004900317016114611,
      "loss": 6.3403,
      "mean_token_accuracy": 0.1313906006515026,
      "num_tokens": 19316966.0,
      "step": 5780
    },
    {
      "entropy": 6.449410009384155,
      "epoch": 1.1285728221636913,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900109879614554,
      "loss": 6.4313,
      "mean_token_accuracy": 0.1263804905116558,
      "num_tokens": 19334389.0,
      "step": 5785
    },
    {
      "entropy": 6.572518444061279,
      "epoch": 1.1295483367476344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004899902533013423,
      "loss": 6.4437,
      "mean_token_accuracy": 0.12664312422275542,
      "num_tokens": 19350238.0,
      "step": 5790
    },
    {
      "entropy": 6.5427642345428465,
      "epoch": 1.1305238513315774,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899694976331478,
      "loss": 6.3899,
      "mean_token_accuracy": 0.12375639081001281,
      "num_tokens": 19367244.0,
      "step": 5795
    },
    {
      "entropy": 6.482330274581909,
      "epoch": 1.1314993659155204,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004899487209589001,
      "loss": 6.4,
      "mean_token_accuracy": 0.12929678186774254,
      "num_tokens": 19384949.0,
      "step": 5800
    },
    {
      "entropy": 6.5406207084655765,
      "epoch": 1.1324748804994635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899279232806297,
      "loss": 6.501,
      "mean_token_accuracy": 0.12479613572359086,
      "num_tokens": 19402149.0,
      "step": 5805
    },
    {
      "entropy": 6.469014501571655,
      "epoch": 1.1334503950834065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899071046003686,
      "loss": 6.3655,
      "mean_token_accuracy": 0.1272161290049553,
      "num_tokens": 19419032.0,
      "step": 5810
    },
    {
      "entropy": 6.528951740264892,
      "epoch": 1.1344259096673495,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898862649201511,
      "loss": 6.3967,
      "mean_token_accuracy": 0.13093573823571206,
      "num_tokens": 19435656.0,
      "step": 5815
    },
    {
      "entropy": 6.4022545337677,
      "epoch": 1.1354014242512926,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898654042420137,
      "loss": 6.3225,
      "mean_token_accuracy": 0.13853389546275138,
      "num_tokens": 19451482.0,
      "step": 5820
    },
    {
      "entropy": 6.569735908508301,
      "epoch": 1.1363769388352356,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004898445225679948,
      "loss": 6.4593,
      "mean_token_accuracy": 0.1269934207201004,
      "num_tokens": 19467758.0,
      "step": 5825
    },
    {
      "entropy": 6.557801723480225,
      "epoch": 1.1373524534191786,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898236199001347,
      "loss": 6.4628,
      "mean_token_accuracy": 0.12197138965129853,
      "num_tokens": 19483673.0,
      "step": 5830
    },
    {
      "entropy": 6.5787523746490475,
      "epoch": 1.1383279680031215,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898026962404762,
      "loss": 6.5308,
      "mean_token_accuracy": 0.12353024780750274,
      "num_tokens": 19500171.0,
      "step": 5835
    },
    {
      "entropy": 6.49024224281311,
      "epoch": 1.1393034825870647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897817515910639,
      "loss": 6.347,
      "mean_token_accuracy": 0.1376006379723549,
      "num_tokens": 19516531.0,
      "step": 5840
    },
    {
      "entropy": 6.513188123703003,
      "epoch": 1.1402789971710077,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004897607859539443,
      "loss": 6.4501,
      "mean_token_accuracy": 0.12228616252541542,
      "num_tokens": 19534289.0,
      "step": 5845
    },
    {
      "entropy": 6.344129991531372,
      "epoch": 1.1412545117549509,
      "grad_norm": 1.109375,
      "learning_rate": 0.000489739799331166,
      "loss": 6.2616,
      "mean_token_accuracy": 0.13857797905802727,
      "num_tokens": 19550190.0,
      "step": 5850
    },
    {
      "entropy": 6.391070795059204,
      "epoch": 1.1422300263388938,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048971879172478,
      "loss": 6.196,
      "mean_token_accuracy": 0.14407580196857453,
      "num_tokens": 19567774.0,
      "step": 5855
    },
    {
      "entropy": 6.548099184036255,
      "epoch": 1.1432055409228368,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004896977631368389,
      "loss": 6.47,
      "mean_token_accuracy": 0.12213794216513633,
      "num_tokens": 19583690.0,
      "step": 5860
    },
    {
      "entropy": 6.487272882461548,
      "epoch": 1.1441810555067797,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004896767135693976,
      "loss": 6.3095,
      "mean_token_accuracy": 0.1404709152877331,
      "num_tokens": 19600563.0,
      "step": 5865
    },
    {
      "entropy": 6.461253118515015,
      "epoch": 1.145156570090723,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489655643024513,
      "loss": 6.3653,
      "mean_token_accuracy": 0.13340827748179435,
      "num_tokens": 19618043.0,
      "step": 5870
    },
    {
      "entropy": 6.4762125492095945,
      "epoch": 1.1461320846746659,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004896345515042439,
      "loss": 6.4284,
      "mean_token_accuracy": 0.1327342212200165,
      "num_tokens": 19634807.0,
      "step": 5875
    },
    {
      "entropy": 6.447623538970947,
      "epoch": 1.1471075992586088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896134390106515,
      "loss": 6.2818,
      "mean_token_accuracy": 0.13807629346847533,
      "num_tokens": 19650814.0,
      "step": 5880
    },
    {
      "entropy": 6.3012604236602785,
      "epoch": 1.148083113842552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004895923055457987,
      "loss": 6.2246,
      "mean_token_accuracy": 0.1486947812139988,
      "num_tokens": 19666293.0,
      "step": 5885
    },
    {
      "entropy": 6.380581569671631,
      "epoch": 1.149058628426495,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895711511117506,
      "loss": 6.296,
      "mean_token_accuracy": 0.13551393449306487,
      "num_tokens": 19682549.0,
      "step": 5890
    },
    {
      "entropy": 6.474105262756348,
      "epoch": 1.150034143010438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895499757105743,
      "loss": 6.4044,
      "mean_token_accuracy": 0.12615978270769118,
      "num_tokens": 19697776.0,
      "step": 5895
    },
    {
      "entropy": 6.470328617095947,
      "epoch": 1.1510096575943811,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895287793443391,
      "loss": 6.3347,
      "mean_token_accuracy": 0.13672792837023734,
      "num_tokens": 19715219.0,
      "step": 5900
    },
    {
      "entropy": 6.423890542984009,
      "epoch": 1.151985172178324,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004895075620151162,
      "loss": 6.3626,
      "mean_token_accuracy": 0.13388260155916215,
      "num_tokens": 19731934.0,
      "step": 5905
    },
    {
      "entropy": 6.43372220993042,
      "epoch": 1.152960686762267,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004894863237249789,
      "loss": 6.368,
      "mean_token_accuracy": 0.1364196851849556,
      "num_tokens": 19748933.0,
      "step": 5910
    },
    {
      "entropy": 6.4860907077789305,
      "epoch": 1.1539362013462102,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894650644760024,
      "loss": 6.3982,
      "mean_token_accuracy": 0.130922332406044,
      "num_tokens": 19765430.0,
      "step": 5915
    },
    {
      "entropy": 6.501723051071167,
      "epoch": 1.1549117159301532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894437842702642,
      "loss": 6.4473,
      "mean_token_accuracy": 0.12613650932908058,
      "num_tokens": 19783626.0,
      "step": 5920
    },
    {
      "entropy": 6.516066217422486,
      "epoch": 1.1558872305140961,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894224831098437,
      "loss": 6.4334,
      "mean_token_accuracy": 0.12587256282567977,
      "num_tokens": 19800269.0,
      "step": 5925
    },
    {
      "entropy": 6.531294631958008,
      "epoch": 1.156862745098039,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894011609968224,
      "loss": 6.3792,
      "mean_token_accuracy": 0.12718194276094436,
      "num_tokens": 19816837.0,
      "step": 5930
    },
    {
      "entropy": 6.560290098190308,
      "epoch": 1.1578382596819823,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893798179332837,
      "loss": 6.5371,
      "mean_token_accuracy": 0.12006494030356407,
      "num_tokens": 19833029.0,
      "step": 5935
    },
    {
      "entropy": 6.404336309432983,
      "epoch": 1.1588137742659252,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004893584539213134,
      "loss": 6.2893,
      "mean_token_accuracy": 0.142698435485363,
      "num_tokens": 19851600.0,
      "step": 5940
    },
    {
      "entropy": 6.480049800872803,
      "epoch": 1.1597892888498684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893370689629988,
      "loss": 6.3953,
      "mean_token_accuracy": 0.12810298949480056,
      "num_tokens": 19867152.0,
      "step": 5945
    },
    {
      "entropy": 6.398224878311157,
      "epoch": 1.1607648034338114,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004893156630604298,
      "loss": 6.2426,
      "mean_token_accuracy": 0.14025613367557527,
      "num_tokens": 19884035.0,
      "step": 5950
    },
    {
      "entropy": 6.511564064025879,
      "epoch": 1.1617403180177543,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489294236215698,
      "loss": 6.3752,
      "mean_token_accuracy": 0.13071327805519103,
      "num_tokens": 19899186.0,
      "step": 5955
    },
    {
      "entropy": 6.488106918334961,
      "epoch": 1.1627158326016973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892727884308972,
      "loss": 6.4076,
      "mean_token_accuracy": 0.1283169336616993,
      "num_tokens": 19914162.0,
      "step": 5960
    },
    {
      "entropy": 6.479491567611694,
      "epoch": 1.1636913471856405,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892513197081233,
      "loss": 6.3603,
      "mean_token_accuracy": 0.12555169239640235,
      "num_tokens": 19931758.0,
      "step": 5965
    },
    {
      "entropy": 6.465991449356079,
      "epoch": 1.1646668617695835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000489229830049474,
      "loss": 6.4396,
      "mean_token_accuracy": 0.12887430116534232,
      "num_tokens": 19949397.0,
      "step": 5970
    },
    {
      "entropy": 6.448368167877197,
      "epoch": 1.1656423763535264,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004892083194570492,
      "loss": 6.3359,
      "mean_token_accuracy": 0.13140734061598777,
      "num_tokens": 19967356.0,
      "step": 5975
    },
    {
      "entropy": 6.456110048294067,
      "epoch": 1.1666178909374696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004891867879329509,
      "loss": 6.2947,
      "mean_token_accuracy": 0.13778769299387933,
      "num_tokens": 19984343.0,
      "step": 5980
    },
    {
      "entropy": 6.563321542739868,
      "epoch": 1.1675934055214126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004891652354792831,
      "loss": 6.4598,
      "mean_token_accuracy": 0.12788356244564056,
      "num_tokens": 20001600.0,
      "step": 5985
    },
    {
      "entropy": 6.518469285964966,
      "epoch": 1.1685689201053555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891436620981517,
      "loss": 6.3924,
      "mean_token_accuracy": 0.1274819128215313,
      "num_tokens": 20016816.0,
      "step": 5990
    },
    {
      "entropy": 6.514881420135498,
      "epoch": 1.1695444346892987,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004891220677916649,
      "loss": 6.4103,
      "mean_token_accuracy": 0.12748466655611992,
      "num_tokens": 20033754.0,
      "step": 5995
    },
    {
      "entropy": 6.3635094165802,
      "epoch": 1.1705199492732417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004891004525619327,
      "loss": 6.3523,
      "mean_token_accuracy": 0.14000822231173515,
      "num_tokens": 20051899.0,
      "step": 6000
    },
    {
      "epoch": 1.1705199492732417,
      "eval_entropy": 6.291171086334803,
      "eval_loss": 6.503299236297607,
      "eval_mean_token_accuracy": 0.1310783804749672,
      "eval_num_tokens": 20051899.0,
      "eval_runtime": 15.4373,
      "eval_samples_per_second": 1073.827,
      "eval_steps_per_second": 134.285,
      "step": 6000
    }
  ],
  "logging_steps": 5,
  "max_steps": 51250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.7034114664448e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}