zho-hans-100mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.096771046727149,
  "eval_steps": 3000,
  "global_step": 21000,
  "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
    },
    {
      "entropy": 6.438280487060547,
      "epoch": 1.1714954638571846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890788164110674,
      "loss": 6.3453,
      "mean_token_accuracy": 0.13287174627184867,
      "num_tokens": 20069066.0,
      "step": 6005
    },
    {
      "entropy": 6.521517610549926,
      "epoch": 1.1724709784411278,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890571593411833,
      "loss": 6.3872,
      "mean_token_accuracy": 0.12669881135225297,
      "num_tokens": 20085482.0,
      "step": 6010
    },
    {
      "entropy": 6.536331033706665,
      "epoch": 1.1734464930250708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004890354813543963,
      "loss": 6.4422,
      "mean_token_accuracy": 0.12886352092027664,
      "num_tokens": 20101591.0,
      "step": 6015
    },
    {
      "entropy": 6.4531354904174805,
      "epoch": 1.1744220076090137,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000489013782452825,
      "loss": 6.3878,
      "mean_token_accuracy": 0.12797778397798537,
      "num_tokens": 20118261.0,
      "step": 6020
    },
    {
      "entropy": 6.442244243621826,
      "epoch": 1.1753975221929567,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004889920626385896,
      "loss": 6.36,
      "mean_token_accuracy": 0.13415333703160287,
      "num_tokens": 20133552.0,
      "step": 6025
    },
    {
      "entropy": 6.332036066055298,
      "epoch": 1.1763730367768999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004889703219138125,
      "loss": 6.192,
      "mean_token_accuracy": 0.14723025113344193,
      "num_tokens": 20150237.0,
      "step": 6030
    },
    {
      "entropy": 6.483112859725952,
      "epoch": 1.1773485513608428,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889485602806181,
      "loss": 6.4049,
      "mean_token_accuracy": 0.1308147169649601,
      "num_tokens": 20166926.0,
      "step": 6035
    },
    {
      "entropy": 6.502467632293701,
      "epoch": 1.1783240659447858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889267777411331,
      "loss": 6.439,
      "mean_token_accuracy": 0.13117024153470994,
      "num_tokens": 20184507.0,
      "step": 6040
    },
    {
      "entropy": 6.4535845756530765,
      "epoch": 1.179299580528729,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004889049742974858,
      "loss": 6.362,
      "mean_token_accuracy": 0.13900133445858956,
      "num_tokens": 20200647.0,
      "step": 6045
    },
    {
      "entropy": 6.468359088897705,
      "epoch": 1.180275095112672,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004888831499518066,
      "loss": 6.412,
      "mean_token_accuracy": 0.13279227018356324,
      "num_tokens": 20218634.0,
      "step": 6050
    },
    {
      "entropy": 6.477481985092163,
      "epoch": 1.1812506096966149,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004888613047062284,
      "loss": 6.4457,
      "mean_token_accuracy": 0.12565266266465186,
      "num_tokens": 20234641.0,
      "step": 6055
    },
    {
      "entropy": 6.486688232421875,
      "epoch": 1.182226124280558,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004888394385628857,
      "loss": 6.4295,
      "mean_token_accuracy": 0.12311068177223206,
      "num_tokens": 20251973.0,
      "step": 6060
    },
    {
      "entropy": 6.47110390663147,
      "epoch": 1.183201638864501,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888175515239152,
      "loss": 6.3421,
      "mean_token_accuracy": 0.13398097679018975,
      "num_tokens": 20270300.0,
      "step": 6065
    },
    {
      "entropy": 6.450536727905273,
      "epoch": 1.184177153448444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004887956435914556,
      "loss": 6.3748,
      "mean_token_accuracy": 0.13443400487303733,
      "num_tokens": 20286887.0,
      "step": 6070
    },
    {
      "entropy": 6.500335550308227,
      "epoch": 1.1851526680323872,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887737147676477,
      "loss": 6.442,
      "mean_token_accuracy": 0.12693579643964767,
      "num_tokens": 20302859.0,
      "step": 6075
    },
    {
      "entropy": 6.481752586364746,
      "epoch": 1.1861281826163301,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887517650546344,
      "loss": 6.3678,
      "mean_token_accuracy": 0.13718537613749504,
      "num_tokens": 20319863.0,
      "step": 6080
    },
    {
      "entropy": 6.46849045753479,
      "epoch": 1.187103697200273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887297944545603,
      "loss": 6.3853,
      "mean_token_accuracy": 0.13058708161115645,
      "num_tokens": 20336134.0,
      "step": 6085
    },
    {
      "entropy": 6.341468715667725,
      "epoch": 1.188079211784216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004887078029695725,
      "loss": 6.2557,
      "mean_token_accuracy": 0.14786370024085044,
      "num_tokens": 20353171.0,
      "step": 6090
    },
    {
      "entropy": 6.467123556137085,
      "epoch": 1.1890547263681592,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886857906018198,
      "loss": 6.3429,
      "mean_token_accuracy": 0.1336496502161026,
      "num_tokens": 20368464.0,
      "step": 6095
    },
    {
      "entropy": 6.450306463241577,
      "epoch": 1.1900302409521022,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886637573534534,
      "loss": 6.3765,
      "mean_token_accuracy": 0.12948411628603934,
      "num_tokens": 20383842.0,
      "step": 6100
    },
    {
      "entropy": 6.421968460083008,
      "epoch": 1.1910057555360454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000488641703226626,
      "loss": 6.3433,
      "mean_token_accuracy": 0.13810566291213036,
      "num_tokens": 20399503.0,
      "step": 6105
    },
    {
      "entropy": 6.425341939926147,
      "epoch": 1.1919812701199883,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886196282234927,
      "loss": 6.2753,
      "mean_token_accuracy": 0.13032771721482278,
      "num_tokens": 20416012.0,
      "step": 6110
    },
    {
      "entropy": 6.51609320640564,
      "epoch": 1.1929567847039313,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004885975323462109,
      "loss": 6.4092,
      "mean_token_accuracy": 0.12382623106241226,
      "num_tokens": 20432092.0,
      "step": 6115
    },
    {
      "entropy": 6.450653505325318,
      "epoch": 1.1939322992878743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885754155969393,
      "loss": 6.3747,
      "mean_token_accuracy": 0.13994504436850547,
      "num_tokens": 20449388.0,
      "step": 6120
    },
    {
      "entropy": 6.389624547958374,
      "epoch": 1.1949078138718174,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004885532779778394,
      "loss": 6.3277,
      "mean_token_accuracy": 0.13783815056085585,
      "num_tokens": 20464851.0,
      "step": 6125
    },
    {
      "entropy": 6.483812761306763,
      "epoch": 1.1958833284557604,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885311194910743,
      "loss": 6.387,
      "mean_token_accuracy": 0.130421943962574,
      "num_tokens": 20481346.0,
      "step": 6130
    },
    {
      "entropy": 6.478879022598266,
      "epoch": 1.1968588430397034,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004885089401388093,
      "loss": 6.4446,
      "mean_token_accuracy": 0.12645828798413278,
      "num_tokens": 20498373.0,
      "step": 6135
    },
    {
      "entropy": 6.467660570144654,
      "epoch": 1.1978343576236465,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004884867399232116,
      "loss": 6.3595,
      "mean_token_accuracy": 0.12550423592329024,
      "num_tokens": 20515375.0,
      "step": 6140
    },
    {
      "entropy": 6.457030344009399,
      "epoch": 1.1988098722075895,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004884645188464505,
      "loss": 6.374,
      "mean_token_accuracy": 0.1273153193295002,
      "num_tokens": 20531593.0,
      "step": 6145
    },
    {
      "entropy": 6.57569432258606,
      "epoch": 1.1997853867915325,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004884422769106975,
      "loss": 6.3663,
      "mean_token_accuracy": 0.13325131237506865,
      "num_tokens": 20548668.0,
      "step": 6150
    },
    {
      "entropy": 6.559942197799683,
      "epoch": 1.2007609013754756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004884200141181258,
      "loss": 6.4642,
      "mean_token_accuracy": 0.1240674987435341,
      "num_tokens": 20566772.0,
      "step": 6155
    },
    {
      "entropy": 6.414913368225098,
      "epoch": 1.2017364159594186,
      "grad_norm": 1.046875,
      "learning_rate": 0.000488397730470911,
      "loss": 6.3334,
      "mean_token_accuracy": 0.1334030508995056,
      "num_tokens": 20582557.0,
      "step": 6160
    },
    {
      "entropy": 6.454658031463623,
      "epoch": 1.2027119305433616,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004883754259712305,
      "loss": 6.3841,
      "mean_token_accuracy": 0.12764096185564994,
      "num_tokens": 20598516.0,
      "step": 6165
    },
    {
      "entropy": 6.478035593032837,
      "epoch": 1.2036874451273047,
      "grad_norm": 1.25,
      "learning_rate": 0.000488353100621264,
      "loss": 6.3638,
      "mean_token_accuracy": 0.12734246999025345,
      "num_tokens": 20613773.0,
      "step": 6170
    },
    {
      "entropy": 6.400567102432251,
      "epoch": 1.2046629597112477,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004883307544231928,
      "loss": 6.3849,
      "mean_token_accuracy": 0.13135465681552888,
      "num_tokens": 20631141.0,
      "step": 6175
    },
    {
      "entropy": 6.505388689041138,
      "epoch": 1.2056384742951907,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004883083873792007,
      "loss": 6.3694,
      "mean_token_accuracy": 0.13406961038708687,
      "num_tokens": 20648408.0,
      "step": 6180
    },
    {
      "entropy": 6.46686635017395,
      "epoch": 1.2066139888791336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882859994914731,
      "loss": 6.4002,
      "mean_token_accuracy": 0.13488613292574883,
      "num_tokens": 20666418.0,
      "step": 6185
    },
    {
      "entropy": 6.463993930816651,
      "epoch": 1.2075895034630768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048826359076219785,
      "loss": 6.3232,
      "mean_token_accuracy": 0.13257390782237052,
      "num_tokens": 20682795.0,
      "step": 6190
    },
    {
      "entropy": 6.472417879104614,
      "epoch": 1.2085650180470198,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004882411611935646,
      "loss": 6.464,
      "mean_token_accuracy": 0.12984257638454438,
      "num_tokens": 20700964.0,
      "step": 6195
    },
    {
      "entropy": 6.4479334354400635,
      "epoch": 1.209540532630963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004882187107877651,
      "loss": 6.312,
      "mean_token_accuracy": 0.13208060041069986,
      "num_tokens": 20717288.0,
      "step": 6200
    },
    {
      "entropy": 6.4256064891815186,
      "epoch": 1.210516047214906,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048819623954699307,
      "loss": 6.3846,
      "mean_token_accuracy": 0.12734122276306153,
      "num_tokens": 20733859.0,
      "step": 6205
    },
    {
      "entropy": 6.467871570587159,
      "epoch": 1.2114915617988489,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004881737474734444,
      "loss": 6.4341,
      "mean_token_accuracy": 0.1309346340596676,
      "num_tokens": 20751467.0,
      "step": 6210
    },
    {
      "entropy": 6.542931842803955,
      "epoch": 1.2124670763827918,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048815123456931686,
      "loss": 6.4495,
      "mean_token_accuracy": 0.12728480920195578,
      "num_tokens": 20768019.0,
      "step": 6215
    },
    {
      "entropy": 6.518567848205566,
      "epoch": 1.213442590966735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004881287008368104,
      "loss": 6.4504,
      "mean_token_accuracy": 0.12411093711853027,
      "num_tokens": 20785007.0,
      "step": 6220
    },
    {
      "entropy": 6.441523933410645,
      "epoch": 1.214418105550678,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048810614627812685,
      "loss": 6.3504,
      "mean_token_accuracy": 0.1284393087029457,
      "num_tokens": 20801693.0,
      "step": 6225
    },
    {
      "entropy": 6.306365585327148,
      "epoch": 1.215393620134621,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004880835708954702,
      "loss": 6.1894,
      "mean_token_accuracy": 0.14394301623106004,
      "num_tokens": 20819866.0,
      "step": 6230
    },
    {
      "entropy": 6.466395235061645,
      "epoch": 1.2163691347185641,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880609746910465,
      "loss": 6.4102,
      "mean_token_accuracy": 0.13187225237488748,
      "num_tokens": 20836228.0,
      "step": 6235
    },
    {
      "entropy": 6.493359851837158,
      "epoch": 1.217344649302507,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048803835766706366,
      "loss": 6.338,
      "mean_token_accuracy": 0.12426519393920898,
      "num_tokens": 20853489.0,
      "step": 6240
    },
    {
      "entropy": 6.176209878921509,
      "epoch": 1.21832016388645,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048801571982573185,
      "loss": 6.1105,
      "mean_token_accuracy": 0.15951610431075097,
      "num_tokens": 20870627.0,
      "step": 6245
    },
    {
      "entropy": 6.488164949417114,
      "epoch": 1.2192956784703932,
      "grad_norm": 1.0,
      "learning_rate": 0.00048799306116926306,
      "loss": 6.393,
      "mean_token_accuracy": 0.13407148495316507,
      "num_tokens": 20887277.0,
      "step": 6250
    },
    {
      "entropy": 6.524453783035279,
      "epoch": 1.2202711930543362,
      "grad_norm": 1.0,
      "learning_rate": 0.00048797038169987153,
      "loss": 6.367,
      "mean_token_accuracy": 0.12355743870139121,
      "num_tokens": 20902603.0,
      "step": 6255
    },
    {
      "entropy": 6.449866914749146,
      "epoch": 1.2212467076382791,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048794768141977317,
      "loss": 6.3658,
      "mean_token_accuracy": 0.13299671858549117,
      "num_tokens": 20918700.0,
      "step": 6260
    },
    {
      "entropy": 6.426352500915527,
      "epoch": 1.2222222222222223,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004879249603311865,
      "loss": 6.2691,
      "mean_token_accuracy": 0.13894441574811936,
      "num_tokens": 20936437.0,
      "step": 6265
    },
    {
      "entropy": 6.390188121795655,
      "epoch": 1.2231977368061653,
      "grad_norm": 1.0,
      "learning_rate": 0.00048790221843633154,
      "loss": 6.2854,
      "mean_token_accuracy": 0.14287915602326393,
      "num_tokens": 20954984.0,
      "step": 6270
    },
    {
      "entropy": 6.355689430236817,
      "epoch": 1.2241732513901082,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004878794557374306,
      "loss": 6.3027,
      "mean_token_accuracy": 0.14551016166806222,
      "num_tokens": 20971119.0,
      "step": 6275
    },
    {
      "entropy": 6.46147928237915,
      "epoch": 1.2251487659740512,
      "grad_norm": 0.875,
      "learning_rate": 0.0004878566722367079,
      "loss": 6.3441,
      "mean_token_accuracy": 0.13906254842877389,
      "num_tokens": 20988211.0,
      "step": 6280
    },
    {
      "entropy": 6.415366697311401,
      "epoch": 1.2261242805579944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004878338679363899,
      "loss": 6.337,
      "mean_token_accuracy": 0.1338707521557808,
      "num_tokens": 21004240.0,
      "step": 6285
    },
    {
      "entropy": 6.500727796554566,
      "epoch": 1.2270997951419373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004878110428387048,
      "loss": 6.4362,
      "mean_token_accuracy": 0.1257238656282425,
      "num_tokens": 21021977.0,
      "step": 6290
    },
    {
      "entropy": 6.405777215957642,
      "epoch": 1.2280753097258805,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048778819694588317,
      "loss": 6.3011,
      "mean_token_accuracy": 0.1370128720998764,
      "num_tokens": 21038422.0,
      "step": 6295
    },
    {
      "entropy": 6.251122236251831,
      "epoch": 1.2290508243098235,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048776533026015725,
      "loss": 6.149,
      "mean_token_accuracy": 0.15784930884838105,
      "num_tokens": 21055148.0,
      "step": 6300
    },
    {
      "entropy": 6.458118867874146,
      "epoch": 1.2300263388937664,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048774244278376164,
      "loss": 6.392,
      "mean_token_accuracy": 0.13304935097694398,
      "num_tokens": 21072137.0,
      "step": 6305
    },
    {
      "entropy": 6.467202806472779,
      "epoch": 1.2310018534777094,
      "grad_norm": 1.0,
      "learning_rate": 0.0004877195345189326,
      "loss": 6.326,
      "mean_token_accuracy": 0.1336950771510601,
      "num_tokens": 21089012.0,
      "step": 6310
    },
    {
      "entropy": 6.475249004364014,
      "epoch": 1.2319773680616526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048769660546790896,
      "loss": 6.3369,
      "mean_token_accuracy": 0.12961943596601486,
      "num_tokens": 21105274.0,
      "step": 6315
    },
    {
      "entropy": 6.466490316390991,
      "epoch": 1.2329528826455955,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048767365563293097,
      "loss": 6.4344,
      "mean_token_accuracy": 0.1328072652220726,
      "num_tokens": 21122784.0,
      "step": 6320
    },
    {
      "entropy": 6.459510707855225,
      "epoch": 1.2339283972295385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004876506850162414,
      "loss": 6.2827,
      "mean_token_accuracy": 0.13549240231513976,
      "num_tokens": 21139829.0,
      "step": 6325
    },
    {
      "entropy": 6.378039836883545,
      "epoch": 1.2349039118134817,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048762769362008473,
      "loss": 6.2631,
      "mean_token_accuracy": 0.13462610468268393,
      "num_tokens": 21155095.0,
      "step": 6330
    },
    {
      "entropy": 6.437053203582764,
      "epoch": 1.2358794263974247,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004876046814467077,
      "loss": 6.4165,
      "mean_token_accuracy": 0.129904293268919,
      "num_tokens": 21171865.0,
      "step": 6335
    },
    {
      "entropy": 6.558261680603027,
      "epoch": 1.2368549409813676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000487581648498359,
      "loss": 6.402,
      "mean_token_accuracy": 0.12491710484027863,
      "num_tokens": 21188757.0,
      "step": 6340
    },
    {
      "entropy": 6.480385494232178,
      "epoch": 1.2378304555653108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004875585947772892,
      "loss": 6.3787,
      "mean_token_accuracy": 0.1348903626203537,
      "num_tokens": 21204324.0,
      "step": 6345
    },
    {
      "entropy": 6.395302152633667,
      "epoch": 1.2388059701492538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048753552028575105,
      "loss": 6.3058,
      "mean_token_accuracy": 0.1336951047182083,
      "num_tokens": 21219836.0,
      "step": 6350
    },
    {
      "entropy": 6.4256867408752445,
      "epoch": 1.2397814847331967,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004875124250259994,
      "loss": 6.3949,
      "mean_token_accuracy": 0.12555817067623137,
      "num_tokens": 21235848.0,
      "step": 6355
    },
    {
      "entropy": 6.426709365844727,
      "epoch": 1.24075699931714,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048748930900029103,
      "loss": 6.3481,
      "mean_token_accuracy": 0.14378245919942856,
      "num_tokens": 21252779.0,
      "step": 6360
    },
    {
      "entropy": 6.42932653427124,
      "epoch": 1.2417325139010829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048746617221088464,
      "loss": 6.3119,
      "mean_token_accuracy": 0.13536582365632058,
      "num_tokens": 21269004.0,
      "step": 6365
    },
    {
      "entropy": 6.486655235290527,
      "epoch": 1.2427080284850258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048744301466004114,
      "loss": 6.4038,
      "mean_token_accuracy": 0.13169310763478279,
      "num_tokens": 21285769.0,
      "step": 6370
    },
    {
      "entropy": 6.475272083282471,
      "epoch": 1.2436835430689688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004874198363500235,
      "loss": 6.4435,
      "mean_token_accuracy": 0.12616746127605438,
      "num_tokens": 21301986.0,
      "step": 6375
    },
    {
      "entropy": 6.572941017150879,
      "epoch": 1.244659057652912,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873966372830964,
      "loss": 6.5404,
      "mean_token_accuracy": 0.12172101512551307,
      "num_tokens": 21318755.0,
      "step": 6380
    },
    {
      "entropy": 6.26752986907959,
      "epoch": 1.245634572236855,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000487373417461527,
      "loss": 6.1794,
      "mean_token_accuracy": 0.15411261841654778,
      "num_tokens": 21335188.0,
      "step": 6385
    },
    {
      "entropy": 6.443607044219971,
      "epoch": 1.2466100868207979,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048735017688758406,
      "loss": 6.3817,
      "mean_token_accuracy": 0.13121942952275276,
      "num_tokens": 21351976.0,
      "step": 6390
    },
    {
      "entropy": 6.526898908615112,
      "epoch": 1.247585601404741,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004873269155635387,
      "loss": 6.4053,
      "mean_token_accuracy": 0.12783444076776504,
      "num_tokens": 21369043.0,
      "step": 6395
    },
    {
      "entropy": 6.475271272659302,
      "epoch": 1.248561115988684,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048730363349166393,
      "loss": 6.3642,
      "mean_token_accuracy": 0.13036131635308265,
      "num_tokens": 21386583.0,
      "step": 6400
    },
    {
      "entropy": 6.532704877853393,
      "epoch": 1.249536630572627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004872803306742348,
      "loss": 6.3434,
      "mean_token_accuracy": 0.13562069609761238,
      "num_tokens": 21403455.0,
      "step": 6405
    },
    {
      "entropy": 6.3780049800872805,
      "epoch": 1.2505121451565702,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004872570071135282,
      "loss": 6.408,
      "mean_token_accuracy": 0.12577296867966653,
      "num_tokens": 21420107.0,
      "step": 6410
    },
    {
      "entropy": 6.564887285232544,
      "epoch": 1.2514876597405131,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048723366281182344,
      "loss": 6.4088,
      "mean_token_accuracy": 0.12612811028957366,
      "num_tokens": 21436629.0,
      "step": 6415
    },
    {
      "entropy": 6.438782835006714,
      "epoch": 1.252463174324456,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004872102977714016,
      "loss": 6.3647,
      "mean_token_accuracy": 0.1319246031343937,
      "num_tokens": 21454364.0,
      "step": 6420
    },
    {
      "entropy": 6.418986129760742,
      "epoch": 1.2534386889083993,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048718691199454577,
      "loss": 6.301,
      "mean_token_accuracy": 0.13637648299336433,
      "num_tokens": 21472056.0,
      "step": 6425
    },
    {
      "entropy": 6.528123664855957,
      "epoch": 1.2544142034923422,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004871635054835412,
      "loss": 6.4325,
      "mean_token_accuracy": 0.12380208820104599,
      "num_tokens": 21489125.0,
      "step": 6430
    },
    {
      "entropy": 6.482168960571289,
      "epoch": 1.2553897180762852,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004871400782406749,
      "loss": 6.44,
      "mean_token_accuracy": 0.12502530887722968,
      "num_tokens": 21505772.0,
      "step": 6435
    },
    {
      "entropy": 6.498317050933838,
      "epoch": 1.2563652326602281,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871166302682363,
      "loss": 6.3325,
      "mean_token_accuracy": 0.1333508275449276,
      "num_tokens": 21522628.0,
      "step": 6440
    },
    {
      "entropy": 6.2946573257446286,
      "epoch": 1.2573407472441713,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870931615685166,
      "loss": 6.3152,
      "mean_token_accuracy": 0.13791919648647308,
      "num_tokens": 21540379.0,
      "step": 6445
    },
    {
      "entropy": 6.513365697860718,
      "epoch": 1.2583162618281143,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004870696721438091,
      "loss": 6.3748,
      "mean_token_accuracy": 0.1268819712102413,
      "num_tokens": 21556698.0,
      "step": 6450
    },
    {
      "entropy": 6.492498350143433,
      "epoch": 1.2592917764120575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048704616199640906,
      "loss": 6.305,
      "mean_token_accuracy": 0.13633760958909988,
      "num_tokens": 21573005.0,
      "step": 6455
    },
    {
      "entropy": 6.32011308670044,
      "epoch": 1.2602672909960004,
      "grad_norm": 1.0,
      "learning_rate": 0.00048702263112861377,
      "loss": 6.3074,
      "mean_token_accuracy": 0.14146097153425216,
      "num_tokens": 21591171.0,
      "step": 6460
    },
    {
      "entropy": 6.51646990776062,
      "epoch": 1.2612428055799434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004869990795427226,
      "loss": 6.3104,
      "mean_token_accuracy": 0.13623958304524422,
      "num_tokens": 21606912.0,
      "step": 6465
    },
    {
      "entropy": 6.452241039276123,
      "epoch": 1.2622183201638864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000486975507241037,
      "loss": 6.423,
      "mean_token_accuracy": 0.13142117708921433,
      "num_tokens": 21623392.0,
      "step": 6470
    },
    {
      "entropy": 6.480606174468994,
      "epoch": 1.2631938347478295,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004869519142258604,
      "loss": 6.3944,
      "mean_token_accuracy": 0.13230801597237588,
      "num_tokens": 21639706.0,
      "step": 6475
    },
    {
      "entropy": 6.313494443893433,
      "epoch": 1.2641693493317725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048692830049949803,
      "loss": 6.2291,
      "mean_token_accuracy": 0.1441313736140728,
      "num_tokens": 21656487.0,
      "step": 6480
    },
    {
      "entropy": 6.459343910217285,
      "epoch": 1.2651448639157157,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004869046660642575,
      "loss": 6.3739,
      "mean_token_accuracy": 0.12967009469866753,
      "num_tokens": 21673950.0,
      "step": 6485
    },
    {
      "entropy": 6.503636217117309,
      "epoch": 1.2661203784996586,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048688101092244826,
      "loss": 6.4683,
      "mean_token_accuracy": 0.12482526451349259,
      "num_tokens": 21692296.0,
      "step": 6490
    },
    {
      "entropy": 6.41991810798645,
      "epoch": 1.2670958930836016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004868573350763818,
      "loss": 6.3331,
      "mean_token_accuracy": 0.13989758342504502,
      "num_tokens": 21710021.0,
      "step": 6495
    },
    {
      "entropy": 6.445886564254761,
      "epoch": 1.2680714076675446,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004868336385283716,
      "loss": 6.3184,
      "mean_token_accuracy": 0.13760883435606958,
      "num_tokens": 21726660.0,
      "step": 6500
    },
    {
      "entropy": 6.289924383163452,
      "epoch": 1.2690469222514877,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048680992128073323,
      "loss": 6.1911,
      "mean_token_accuracy": 0.15483601093292237,
      "num_tokens": 21743840.0,
      "step": 6505
    },
    {
      "entropy": 6.479497766494751,
      "epoch": 1.2700224368354307,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048678618333578434,
      "loss": 6.3256,
      "mean_token_accuracy": 0.12751531451940537,
      "num_tokens": 21761000.0,
      "step": 6510
    },
    {
      "entropy": 6.458607149124146,
      "epoch": 1.2709979514193737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004867624246958443,
      "loss": 6.3969,
      "mean_token_accuracy": 0.1313244178891182,
      "num_tokens": 21778564.0,
      "step": 6515
    },
    {
      "entropy": 6.509606552124024,
      "epoch": 1.2719734660033168,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004867386453632349,
      "loss": 6.356,
      "mean_token_accuracy": 0.13098517209291458,
      "num_tokens": 21794780.0,
      "step": 6520
    },
    {
      "entropy": 6.45938663482666,
      "epoch": 1.2729489805872598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004867148453402798,
      "loss": 6.3848,
      "mean_token_accuracy": 0.12514541819691657,
      "num_tokens": 21810692.0,
      "step": 6525
    },
    {
      "entropy": 6.50916051864624,
      "epoch": 1.2739244951712028,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004866910246293045,
      "loss": 6.3348,
      "mean_token_accuracy": 0.1295309543609619,
      "num_tokens": 21826488.0,
      "step": 6530
    },
    {
      "entropy": 6.467811965942383,
      "epoch": 1.2749000097551457,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004866671832326368,
      "loss": 6.3602,
      "mean_token_accuracy": 0.12980129048228264,
      "num_tokens": 21842417.0,
      "step": 6535
    },
    {
      "entropy": 6.441570520401001,
      "epoch": 1.275875524339089,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004866433211526063,
      "loss": 6.3351,
      "mean_token_accuracy": 0.131972486525774,
      "num_tokens": 21858496.0,
      "step": 6540
    },
    {
      "entropy": 6.495133399963379,
      "epoch": 1.2768510389230319,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004866194383915448,
      "loss": 6.3905,
      "mean_token_accuracy": 0.1271805003285408,
      "num_tokens": 21874888.0,
      "step": 6545
    },
    {
      "entropy": 6.386703729629517,
      "epoch": 1.277826553506975,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000486595534951786,
      "loss": 6.249,
      "mean_token_accuracy": 0.13902777731418609,
      "num_tokens": 21891789.0,
      "step": 6550
    },
    {
      "entropy": 6.438479709625244,
      "epoch": 1.278802068090918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004865716108356656,
      "loss": 6.3782,
      "mean_token_accuracy": 0.13205621987581254,
      "num_tokens": 21907681.0,
      "step": 6555
    },
    {
      "entropy": 6.42714114189148,
      "epoch": 1.279777582674861,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004865476660455215,
      "loss": 6.3305,
      "mean_token_accuracy": 0.13146371617913247,
      "num_tokens": 21925852.0,
      "step": 6560
    },
    {
      "entropy": 6.505952596664429,
      "epoch": 1.280753097258804,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004865237005836935,
      "loss": 6.417,
      "mean_token_accuracy": 0.13148983269929887,
      "num_tokens": 21942515.0,
      "step": 6565
    },
    {
      "entropy": 6.494324207305908,
      "epoch": 1.2817286118427471,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048649971445252334,
      "loss": 6.3728,
      "mean_token_accuracy": 0.13455812707543374,
      "num_tokens": 21961183.0,
      "step": 6570
    },
    {
      "entropy": 6.4071990013122555,
      "epoch": 1.28270412642669,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004864757076543548,
      "loss": 6.3764,
      "mean_token_accuracy": 0.12977641448378563,
      "num_tokens": 21977806.0,
      "step": 6575
    },
    {
      "entropy": 6.385750675201416,
      "epoch": 1.283679641010633,
      "grad_norm": 1.0,
      "learning_rate": 0.00048645168019153384,
      "loss": 6.3099,
      "mean_token_accuracy": 0.13807467520236968,
      "num_tokens": 21994606.0,
      "step": 6580
    },
    {
      "entropy": 6.570514249801636,
      "epoch": 1.2846551555945762,
      "grad_norm": 1.0,
      "learning_rate": 0.00048642763206640843,
      "loss": 6.4778,
      "mean_token_accuracy": 0.124488964676857,
      "num_tokens": 22010386.0,
      "step": 6585
    },
    {
      "entropy": 6.34291262626648,
      "epoch": 1.2856306701785192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004864035632813282,
      "loss": 6.2119,
      "mean_token_accuracy": 0.14671706333756446,
      "num_tokens": 22027310.0,
      "step": 6590
    },
    {
      "entropy": 6.455181407928467,
      "epoch": 1.2866061847624621,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004863794738386454,
      "loss": 6.3792,
      "mean_token_accuracy": 0.12563801929354668,
      "num_tokens": 22043040.0,
      "step": 6595
    },
    {
      "entropy": 6.385380697250366,
      "epoch": 1.287581699346405,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048635536374071366,
      "loss": 6.2854,
      "mean_token_accuracy": 0.13746293485164643,
      "num_tokens": 22059521.0,
      "step": 6600
    },
    {
      "entropy": 6.510375118255615,
      "epoch": 1.2885572139303483,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048633123298988905,
      "loss": 6.3524,
      "mean_token_accuracy": 0.12806536108255387,
      "num_tokens": 22075705.0,
      "step": 6605
    },
    {
      "entropy": 6.465872526168823,
      "epoch": 1.2895327285142912,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004863070815885296,
      "loss": 6.4002,
      "mean_token_accuracy": 0.12938489839434625,
      "num_tokens": 22092366.0,
      "step": 6610
    },
    {
      "entropy": 6.474816370010376,
      "epoch": 1.2905082430982344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004862829095389952,
      "loss": 6.3718,
      "mean_token_accuracy": 0.12513460218906403,
      "num_tokens": 22108414.0,
      "step": 6615
    },
    {
      "entropy": 6.461866426467895,
      "epoch": 1.2914837576821774,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048625871684364796,
      "loss": 6.344,
      "mean_token_accuracy": 0.12851746901869773,
      "num_tokens": 22123975.0,
      "step": 6620
    },
    {
      "entropy": 6.453787517547608,
      "epoch": 1.2924592722661203,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004862345035048519,
      "loss": 6.3854,
      "mean_token_accuracy": 0.1330207757651806,
      "num_tokens": 22139529.0,
      "step": 6625
    },
    {
      "entropy": 6.447563362121582,
      "epoch": 1.2934347868500633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004862102695249729,
      "loss": 6.3945,
      "mean_token_accuracy": 0.13230193629860879,
      "num_tokens": 22155680.0,
      "step": 6630
    },
    {
      "entropy": 6.519727659225464,
      "epoch": 1.2944103014340065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048618601490637913,
      "loss": 6.4654,
      "mean_token_accuracy": 0.12100845053792,
      "num_tokens": 22172698.0,
      "step": 6635
    },
    {
      "entropy": 6.506167221069336,
      "epoch": 1.2953858160179494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048616173965144077,
      "loss": 6.3721,
      "mean_token_accuracy": 0.1288725934922695,
      "num_tokens": 22189745.0,
      "step": 6640
    },
    {
      "entropy": 6.529246282577515,
      "epoch": 1.2963613306018926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048613744376252974,
      "loss": 6.4425,
      "mean_token_accuracy": 0.1292098082602024,
      "num_tokens": 22205931.0,
      "step": 6645
    },
    {
      "entropy": 6.386481189727784,
      "epoch": 1.2973368451858356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004861131272420202,
      "loss": 6.2822,
      "mean_token_accuracy": 0.1326431654393673,
      "num_tokens": 22222232.0,
      "step": 6650
    },
    {
      "entropy": 6.492442846298218,
      "epoch": 1.2983123597697785,
      "grad_norm": 1.0,
      "learning_rate": 0.0004860887900922883,
      "loss": 6.3761,
      "mean_token_accuracy": 0.13536139726638793,
      "num_tokens": 22240082.0,
      "step": 6655
    },
    {
      "entropy": 6.327947425842285,
      "epoch": 1.2992878743537215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004860644323157122,
      "loss": 6.299,
      "mean_token_accuracy": 0.13727333769202232,
      "num_tokens": 22256105.0,
      "step": 6660
    },
    {
      "entropy": 6.387172555923462,
      "epoch": 1.3002633889376647,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004860400539146719,
      "loss": 6.2747,
      "mean_token_accuracy": 0.13837094902992247,
      "num_tokens": 22273400.0,
      "step": 6665
    },
    {
      "entropy": 6.43378586769104,
      "epoch": 1.3012389035216076,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004860156548915498,
      "loss": 6.229,
      "mean_token_accuracy": 0.14260004982352256,
      "num_tokens": 22289322.0,
      "step": 6670
    },
    {
      "entropy": 6.356115627288818,
      "epoch": 1.3022144181055506,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048599123524872996,
      "loss": 6.2825,
      "mean_token_accuracy": 0.13438496738672256,
      "num_tokens": 22306182.0,
      "step": 6675
    },
    {
      "entropy": 6.451721572875977,
      "epoch": 1.3031899326894938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004859667949885985,
      "loss": 6.3691,
      "mean_token_accuracy": 0.13371109515428542,
      "num_tokens": 22322761.0,
      "step": 6680
    },
    {
      "entropy": 6.401122665405273,
      "epoch": 1.3041654472734368,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048594233411354386,
      "loss": 6.3512,
      "mean_token_accuracy": 0.13336415514349936,
      "num_tokens": 22340224.0,
      "step": 6685
    },
    {
      "entropy": 6.387690639495849,
      "epoch": 1.3051409618573797,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004859178526259561,
      "loss": 6.2577,
      "mean_token_accuracy": 0.13855350390076637,
      "num_tokens": 22357070.0,
      "step": 6690
    },
    {
      "entropy": 6.4780275344848635,
      "epoch": 1.3061164764413227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004858933505282276,
      "loss": 6.3553,
      "mean_token_accuracy": 0.13408863693475723,
      "num_tokens": 22372723.0,
      "step": 6695
    },
    {
      "entropy": 6.411314010620117,
      "epoch": 1.3070919910252659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004858688278227524,
      "loss": 6.3143,
      "mean_token_accuracy": 0.13309935927391053,
      "num_tokens": 22388486.0,
      "step": 6700
    },
    {
      "entropy": 6.47727222442627,
      "epoch": 1.3080675056092088,
      "grad_norm": 1.0,
      "learning_rate": 0.0004858442845119269,
      "loss": 6.4466,
      "mean_token_accuracy": 0.12280274331569671,
      "num_tokens": 22405212.0,
      "step": 6705
    },
    {
      "entropy": 6.487826013565064,
      "epoch": 1.309043020193152,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004858197205981494,
      "loss": 6.3926,
      "mean_token_accuracy": 0.12588296085596085,
      "num_tokens": 22422899.0,
      "step": 6710
    },
    {
      "entropy": 6.423446083068848,
      "epoch": 1.310018534777095,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857951360838202,
      "loss": 6.3728,
      "mean_token_accuracy": 0.13206155225634575,
      "num_tokens": 22439702.0,
      "step": 6715
    },
    {
      "entropy": 6.500381565093994,
      "epoch": 1.310994049361038,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857705309713416,
      "loss": 6.4108,
      "mean_token_accuracy": 0.13082364946603775,
      "num_tokens": 22457725.0,
      "step": 6720
    },
    {
      "entropy": 6.324027252197266,
      "epoch": 1.3119695639449809,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857459052631179,
      "loss": 6.2625,
      "mean_token_accuracy": 0.1410088337957859,
      "num_tokens": 22475607.0,
      "step": 6725
    },
    {
      "entropy": 6.317308473587036,
      "epoch": 1.312945078528924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048572125896155537,
      "loss": 6.2107,
      "mean_token_accuracy": 0.14619195461273193,
      "num_tokens": 22493818.0,
      "step": 6730
    },
    {
      "entropy": 6.422386741638183,
      "epoch": 1.313920593112867,
      "grad_norm": 1.0,
      "learning_rate": 0.00048569659206906254,
      "loss": 6.4229,
      "mean_token_accuracy": 0.12803472951054573,
      "num_tokens": 22509803.0,
      "step": 6735
    },
    {
      "entropy": 6.488404703140259,
      "epoch": 1.3148961076968102,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004856719045880496,
      "loss": 6.3455,
      "mean_token_accuracy": 0.13581307902932166,
      "num_tokens": 22525335.0,
      "step": 6740
    },
    {
      "entropy": 6.431107378005981,
      "epoch": 1.3158716222807532,
      "grad_norm": 1.0,
      "learning_rate": 0.0004856471965209291,
      "loss": 6.3733,
      "mean_token_accuracy": 0.1288575917482376,
      "num_tokens": 22542726.0,
      "step": 6745
    },
    {
      "entropy": 6.38066782951355,
      "epoch": 1.3168471368646961,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004856224678701152,
      "loss": 6.2255,
      "mean_token_accuracy": 0.14016918390989302,
      "num_tokens": 22559162.0,
      "step": 6750
    },
    {
      "entropy": 6.497272968292236,
      "epoch": 1.317822651448639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004855977186380245,
      "loss": 6.3857,
      "mean_token_accuracy": 0.12673513665795327,
      "num_tokens": 22575247.0,
      "step": 6755
    },
    {
      "entropy": 6.429878997802734,
      "epoch": 1.3187981660325823,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004855729488270752,
      "loss": 6.4014,
      "mean_token_accuracy": 0.1244536004960537,
      "num_tokens": 22591604.0,
      "step": 6760
    },
    {
      "entropy": 6.386602783203125,
      "epoch": 1.3197736806165252,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004855481584396879,
      "loss": 6.3503,
      "mean_token_accuracy": 0.1333194427192211,
      "num_tokens": 22608095.0,
      "step": 6765
    },
    {
      "entropy": 6.473023414611816,
      "epoch": 1.3207491952004682,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048552334747828504,
      "loss": 6.3802,
      "mean_token_accuracy": 0.12514028325676918,
      "num_tokens": 22624090.0,
      "step": 6770
    },
    {
      "entropy": 6.5202258110046385,
      "epoch": 1.3217247097844114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048549851594529095,
      "loss": 6.4011,
      "mean_token_accuracy": 0.13002821430563927,
      "num_tokens": 22641593.0,
      "step": 6775
    },
    {
      "entropy": 6.479304075241089,
      "epoch": 1.3227002243683543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048547366384313204,
      "loss": 6.4488,
      "mean_token_accuracy": 0.12232746407389641,
      "num_tokens": 22657601.0,
      "step": 6780
    },
    {
      "entropy": 6.562661743164062,
      "epoch": 1.3236757389522973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048544879117423683,
      "loss": 6.4586,
      "mean_token_accuracy": 0.12635868638753892,
      "num_tokens": 22674665.0,
      "step": 6785
    },
    {
      "entropy": 6.428364419937134,
      "epoch": 1.3246512535362402,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004854238979410359,
      "loss": 6.2982,
      "mean_token_accuracy": 0.13467704728245736,
      "num_tokens": 22692178.0,
      "step": 6790
    },
    {
      "entropy": 6.4164238452911375,
      "epoch": 1.3256267681201834,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004853989841459616,
      "loss": 6.4013,
      "mean_token_accuracy": 0.12922243550419807,
      "num_tokens": 22709318.0,
      "step": 6795
    },
    {
      "entropy": 6.380286073684692,
      "epoch": 1.3266022827041264,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048537404979144837,
      "loss": 6.2474,
      "mean_token_accuracy": 0.14113229513168335,
      "num_tokens": 22725142.0,
      "step": 6800
    },
    {
      "entropy": 6.428120422363281,
      "epoch": 1.3275777972880696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048534909487993284,
      "loss": 6.2913,
      "mean_token_accuracy": 0.14387671872973443,
      "num_tokens": 22741494.0,
      "step": 6805
    },
    {
      "entropy": 6.398183012008667,
      "epoch": 1.3285533118720125,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048532411941385346,
      "loss": 6.2942,
      "mean_token_accuracy": 0.13323121070861815,
      "num_tokens": 22757232.0,
      "step": 6810
    },
    {
      "entropy": 6.532809638977051,
      "epoch": 1.3295288264559555,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048529912339565076,
      "loss": 6.4261,
      "mean_token_accuracy": 0.12231530249118805,
      "num_tokens": 22772846.0,
      "step": 6815
    },
    {
      "entropy": 6.41149525642395,
      "epoch": 1.3305043410398985,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004852741068277672,
      "loss": 6.353,
      "mean_token_accuracy": 0.13288354948163034,
      "num_tokens": 22789712.0,
      "step": 6820
    },
    {
      "entropy": 6.413954496383667,
      "epoch": 1.3314798556238416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048524906971264737,
      "loss": 6.3016,
      "mean_token_accuracy": 0.12552105262875557,
      "num_tokens": 22805743.0,
      "step": 6825
    },
    {
      "entropy": 6.423403358459472,
      "epoch": 1.3324553702077846,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048522401205273774,
      "loss": 6.311,
      "mean_token_accuracy": 0.13405733406543732,
      "num_tokens": 22822947.0,
      "step": 6830
    },
    {
      "entropy": 6.346534967422485,
      "epoch": 1.3334308847917276,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004851989338504869,
      "loss": 6.2252,
      "mean_token_accuracy": 0.15375058576464654,
      "num_tokens": 22839767.0,
      "step": 6835
    },
    {
      "entropy": 6.430522441864014,
      "epoch": 1.3344063993756707,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048517383510834547,
      "loss": 6.3375,
      "mean_token_accuracy": 0.13524805307388305,
      "num_tokens": 22854523.0,
      "step": 6840
    },
    {
      "entropy": 6.300710248947143,
      "epoch": 1.3353819139596137,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004851487158287659,
      "loss": 6.2804,
      "mean_token_accuracy": 0.13998101502656937,
      "num_tokens": 22871580.0,
      "step": 6845
    },
    {
      "entropy": 6.398343276977539,
      "epoch": 1.3363574285435567,
      "grad_norm": 1.0,
      "learning_rate": 0.00048512357601420287,
      "loss": 6.2293,
      "mean_token_accuracy": 0.13318464681506156,
      "num_tokens": 22887123.0,
      "step": 6850
    },
    {
      "entropy": 6.508041572570801,
      "epoch": 1.3373329431274996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850984156671128,
      "loss": 6.4055,
      "mean_token_accuracy": 0.13236077576875688,
      "num_tokens": 22905616.0,
      "step": 6855
    },
    {
      "entropy": 6.416907501220703,
      "epoch": 1.3383084577114428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004850732347899544,
      "loss": 6.2765,
      "mean_token_accuracy": 0.13641469031572342,
      "num_tokens": 22921829.0,
      "step": 6860
    },
    {
      "entropy": 6.329412031173706,
      "epoch": 1.3392839722953858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004850480333851882,
      "loss": 6.2847,
      "mean_token_accuracy": 0.14315056651830674,
      "num_tokens": 22938766.0,
      "step": 6865
    },
    {
      "entropy": 6.393110942840576,
      "epoch": 1.340259486879329,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004850228114552768,
      "loss": 6.3546,
      "mean_token_accuracy": 0.1352493979036808,
      "num_tokens": 22956428.0,
      "step": 6870
    },
    {
      "entropy": 6.548415088653565,
      "epoch": 1.341235001463272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048499756900268485,
      "loss": 6.4136,
      "mean_token_accuracy": 0.12831211537122728,
      "num_tokens": 22972538.0,
      "step": 6875
    },
    {
      "entropy": 6.520471858978271,
      "epoch": 1.3422105160472149,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004849723060298788,
      "loss": 6.4288,
      "mean_token_accuracy": 0.12977882996201515,
      "num_tokens": 22988015.0,
      "step": 6880
    },
    {
      "entropy": 6.456299924850464,
      "epoch": 1.3431860306311578,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048494702253932745,
      "loss": 6.4178,
      "mean_token_accuracy": 0.12471770271658897,
      "num_tokens": 23004486.0,
      "step": 6885
    },
    {
      "entropy": 6.4076526165008545,
      "epoch": 1.344161545215101,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048492171853350134,
      "loss": 6.2232,
      "mean_token_accuracy": 0.1409214183688164,
      "num_tokens": 23021583.0,
      "step": 6890
    },
    {
      "entropy": 6.4423035144805905,
      "epoch": 1.345137059799044,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048489639401487293,
      "loss": 6.3302,
      "mean_token_accuracy": 0.13439780846238136,
      "num_tokens": 23036920.0,
      "step": 6895
    },
    {
      "entropy": 6.360282707214355,
      "epoch": 1.3461125743829871,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048487104898591716,
      "loss": 6.2112,
      "mean_token_accuracy": 0.1423999287188053,
      "num_tokens": 23054752.0,
      "step": 6900
    },
    {
      "entropy": 6.435593843460083,
      "epoch": 1.34708808896693,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004848456834491103,
      "loss": 6.3817,
      "mean_token_accuracy": 0.12630595788359641,
      "num_tokens": 23069861.0,
      "step": 6905
    },
    {
      "entropy": 6.460933589935303,
      "epoch": 1.348063603550873,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004848202974069313,
      "loss": 6.3764,
      "mean_token_accuracy": 0.13135069683194162,
      "num_tokens": 23086349.0,
      "step": 6910
    },
    {
      "entropy": 6.405533742904663,
      "epoch": 1.349039118134816,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048479489086186055,
      "loss": 6.249,
      "mean_token_accuracy": 0.13927001729607583,
      "num_tokens": 23103116.0,
      "step": 6915
    },
    {
      "entropy": 6.405908107757568,
      "epoch": 1.3500146327187592,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004847694638163809,
      "loss": 6.2682,
      "mean_token_accuracy": 0.13608039915561676,
      "num_tokens": 23120505.0,
      "step": 6920
    },
    {
      "entropy": 6.41163215637207,
      "epoch": 1.3509901473027022,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004847440162729768,
      "loss": 6.3142,
      "mean_token_accuracy": 0.1290516123175621,
      "num_tokens": 23136751.0,
      "step": 6925
    },
    {
      "entropy": 6.320538759231567,
      "epoch": 1.3519656618866451,
      "grad_norm": 1.046875,
      "learning_rate": 0.000484718548234135,
      "loss": 6.2351,
      "mean_token_accuracy": 0.1464291125535965,
      "num_tokens": 23153581.0,
      "step": 6930
    },
    {
      "entropy": 6.357593536376953,
      "epoch": 1.3529411764705883,
      "grad_norm": 1.109375,
      "learning_rate": 0.000484693059702344,
      "loss": 6.2426,
      "mean_token_accuracy": 0.13756882920861244,
      "num_tokens": 23169650.0,
      "step": 6935
    },
    {
      "entropy": 6.507209062576294,
      "epoch": 1.3539166910545313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004846675506800947,
      "loss": 6.4547,
      "mean_token_accuracy": 0.12677338272333144,
      "num_tokens": 23186170.0,
      "step": 6940
    },
    {
      "entropy": 6.494417810440064,
      "epoch": 1.3548922056384742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048464202116987957,
      "loss": 6.3649,
      "mean_token_accuracy": 0.12990618050098418,
      "num_tokens": 23203567.0,
      "step": 6945
    },
    {
      "entropy": 6.385674524307251,
      "epoch": 1.3558677202224172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048461647117419335,
      "loss": 6.2578,
      "mean_token_accuracy": 0.14242292195558548,
      "num_tokens": 23220551.0,
      "step": 6950
    },
    {
      "entropy": 6.288053894042969,
      "epoch": 1.3568432348063604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048459090069553256,
      "loss": 6.1902,
      "mean_token_accuracy": 0.14410581216216087,
      "num_tokens": 23237390.0,
      "step": 6955
    },
    {
      "entropy": 6.415642976760864,
      "epoch": 1.3578187493903033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048456530973639604,
      "loss": 6.2594,
      "mean_token_accuracy": 0.13911432400345802,
      "num_tokens": 23254644.0,
      "step": 6960
    },
    {
      "entropy": 6.398448801040649,
      "epoch": 1.3587942639742465,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048453969829928437,
      "loss": 6.2857,
      "mean_token_accuracy": 0.13506311997771264,
      "num_tokens": 23273261.0,
      "step": 6965
    },
    {
      "entropy": 6.470621347427368,
      "epoch": 1.3597697785581895,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004845140663867001,
      "loss": 6.3866,
      "mean_token_accuracy": 0.12734778225421906,
      "num_tokens": 23290366.0,
      "step": 6970
    },
    {
      "entropy": 6.50848879814148,
      "epoch": 1.3607452931421324,
      "grad_norm": 0.921875,
      "learning_rate": 0.000484488414001148,
      "loss": 6.3904,
      "mean_token_accuracy": 0.12644991353154184,
      "num_tokens": 23306629.0,
      "step": 6975
    },
    {
      "entropy": 6.303738927841186,
      "epoch": 1.3617208077260754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004844627411451348,
      "loss": 6.2192,
      "mean_token_accuracy": 0.14309400841593742,
      "num_tokens": 23322446.0,
      "step": 6980
    },
    {
      "entropy": 6.390560960769653,
      "epoch": 1.3626963223100186,
      "grad_norm": 1.015625,
      "learning_rate": 0.000484437047821169,
      "loss": 6.2875,
      "mean_token_accuracy": 0.1382938154041767,
      "num_tokens": 23340447.0,
      "step": 6985
    },
    {
      "entropy": 6.369408226013183,
      "epoch": 1.3636718368939615,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048441133403176137,
      "loss": 6.2919,
      "mean_token_accuracy": 0.14039860367774964,
      "num_tokens": 23356571.0,
      "step": 6990
    },
    {
      "entropy": 6.490394592285156,
      "epoch": 1.3646473514779047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048438559977942447,
      "loss": 6.4203,
      "mean_token_accuracy": 0.12743518948554994,
      "num_tokens": 23374012.0,
      "step": 6995
    },
    {
      "entropy": 6.446965885162354,
      "epoch": 1.3656228660618477,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048435984506667307,
      "loss": 6.34,
      "mean_token_accuracy": 0.13338804841041565,
      "num_tokens": 23391011.0,
      "step": 7000
    },
    {
      "entropy": 6.499673461914062,
      "epoch": 1.3665983806457906,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004843340698960238,
      "loss": 6.452,
      "mean_token_accuracy": 0.12398158982396126,
      "num_tokens": 23408080.0,
      "step": 7005
    },
    {
      "entropy": 6.308070707321167,
      "epoch": 1.3675738952297336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004843082742699953,
      "loss": 6.0955,
      "mean_token_accuracy": 0.15313742607831954,
      "num_tokens": 23424060.0,
      "step": 7010
    },
    {
      "entropy": 6.356409502029419,
      "epoch": 1.3685494098136768,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004842824581911082,
      "loss": 6.2801,
      "mean_token_accuracy": 0.1314744122326374,
      "num_tokens": 23441900.0,
      "step": 7015
    },
    {
      "entropy": 6.492970943450928,
      "epoch": 1.3695249243976197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004842566216618852,
      "loss": 6.329,
      "mean_token_accuracy": 0.13227075710892677,
      "num_tokens": 23457032.0,
      "step": 7020
    },
    {
      "entropy": 6.436800622940064,
      "epoch": 1.3705004389815627,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004842307646848509,
      "loss": 6.413,
      "mean_token_accuracy": 0.12278950661420822,
      "num_tokens": 23473934.0,
      "step": 7025
    },
    {
      "entropy": 6.572264766693115,
      "epoch": 1.371475953565506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048420488726253196,
      "loss": 6.4555,
      "mean_token_accuracy": 0.12009092345833779,
      "num_tokens": 23491024.0,
      "step": 7030
    },
    {
      "entropy": 6.439138460159302,
      "epoch": 1.3724514681494488,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004841789893974571,
      "loss": 6.3772,
      "mean_token_accuracy": 0.1296961523592472,
      "num_tokens": 23506763.0,
      "step": 7035
    },
    {
      "entropy": 6.421232891082764,
      "epoch": 1.3734269827333918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004841530710921569,
      "loss": 6.3439,
      "mean_token_accuracy": 0.13281450271606446,
      "num_tokens": 23523101.0,
      "step": 7040
    },
    {
      "entropy": 6.400172758102417,
      "epoch": 1.3744024973173348,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004841271323491642,
      "loss": 6.2699,
      "mean_token_accuracy": 0.14064371436834336,
      "num_tokens": 23539787.0,
      "step": 7045
    },
    {
      "entropy": 6.363674926757812,
      "epoch": 1.375378011901278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004841011731710133,
      "loss": 6.355,
      "mean_token_accuracy": 0.13419481739401817,
      "num_tokens": 23555800.0,
      "step": 7050
    },
    {
      "entropy": 6.478248023986817,
      "epoch": 1.376353526485221,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048407519356024107,
      "loss": 6.3193,
      "mean_token_accuracy": 0.13539755791425706,
      "num_tokens": 23572234.0,
      "step": 7055
    },
    {
      "entropy": 6.431790065765381,
      "epoch": 1.377329041069164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048404919351938597,
      "loss": 6.4096,
      "mean_token_accuracy": 0.1250982478260994,
      "num_tokens": 23588982.0,
      "step": 7060
    },
    {
      "entropy": 6.474983263015747,
      "epoch": 1.378304555653107,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004840231730509889,
      "loss": 6.287,
      "mean_token_accuracy": 0.13324926123023034,
      "num_tokens": 23606921.0,
      "step": 7065
    },
    {
      "entropy": 6.397536182403565,
      "epoch": 1.37928007023705,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004839971321575922,
      "loss": 6.3946,
      "mean_token_accuracy": 0.12756569758057595,
      "num_tokens": 23625562.0,
      "step": 7070
    },
    {
      "entropy": 6.389036321640015,
      "epoch": 1.380255584820993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004839710708417407,
      "loss": 6.2183,
      "mean_token_accuracy": 0.14257784709334373,
      "num_tokens": 23642586.0,
      "step": 7075
    },
    {
      "entropy": 6.449665498733521,
      "epoch": 1.3812310994049362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004839449891059809,
      "loss": 6.2941,
      "mean_token_accuracy": 0.1341830760240555,
      "num_tokens": 23659245.0,
      "step": 7080
    },
    {
      "entropy": 6.362768173217773,
      "epoch": 1.3822066139888791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004839188869528615,
      "loss": 6.2622,
      "mean_token_accuracy": 0.14065645262598991,
      "num_tokens": 23674609.0,
      "step": 7085
    },
    {
      "entropy": 6.20830807685852,
      "epoch": 1.3831821285728223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004838927643849331,
      "loss": 6.0669,
      "mean_token_accuracy": 0.15650453865528108,
      "num_tokens": 23692778.0,
      "step": 7090
    },
    {
      "entropy": 6.427032899856568,
      "epoch": 1.3841576431567653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048386662140474814,
      "loss": 6.3617,
      "mean_token_accuracy": 0.13924025371670723,
      "num_tokens": 23708855.0,
      "step": 7095
    },
    {
      "entropy": 6.470193719863891,
      "epoch": 1.3851331577407082,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048384045801486136,
      "loss": 6.3971,
      "mean_token_accuracy": 0.12732142060995102,
      "num_tokens": 23726791.0,
      "step": 7100
    },
    {
      "entropy": 6.489234876632691,
      "epoch": 1.3861086723246512,
      "grad_norm": 1.125,
      "learning_rate": 0.0004838142742178295,
      "loss": 6.3349,
      "mean_token_accuracy": 0.12395901307463646,
      "num_tokens": 23743215.0,
      "step": 7105
    },
    {
      "entropy": 6.390693044662475,
      "epoch": 1.3870841869085944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048378807001621084,
      "loss": 6.3067,
      "mean_token_accuracy": 0.12721116319298745,
      "num_tokens": 23760814.0,
      "step": 7110
    },
    {
      "entropy": 6.460228633880615,
      "epoch": 1.3880597014925373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048376184541256616,
      "loss": 6.3663,
      "mean_token_accuracy": 0.12972285225987434,
      "num_tokens": 23776408.0,
      "step": 7115
    },
    {
      "entropy": 6.466463947296143,
      "epoch": 1.3890352160764803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000483735600409458,
      "loss": 6.3055,
      "mean_token_accuracy": 0.14098479524254798,
      "num_tokens": 23792237.0,
      "step": 7120
    },
    {
      "entropy": 6.348152780532837,
      "epoch": 1.3900107306604235,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000483709335009451,
      "loss": 6.2456,
      "mean_token_accuracy": 0.13743823543190956,
      "num_tokens": 23809006.0,
      "step": 7125
    },
    {
      "entropy": 6.479160499572754,
      "epoch": 1.3909862452443664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048368304921511153,
      "loss": 6.468,
      "mean_token_accuracy": 0.12282890975475311,
      "num_tokens": 23825964.0,
      "step": 7130
    },
    {
      "entropy": 6.4250078201293945,
      "epoch": 1.3919617598283094,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004836567430290082,
      "loss": 6.2438,
      "mean_token_accuracy": 0.1402968041598797,
      "num_tokens": 23842674.0,
      "step": 7135
    },
    {
      "entropy": 6.342310810089112,
      "epoch": 1.3929372744122523,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004836304164537117,
      "loss": 6.2331,
      "mean_token_accuracy": 0.14338141977787017,
      "num_tokens": 23859414.0,
      "step": 7140
    },
    {
      "entropy": 6.360706472396851,
      "epoch": 1.3939127889961955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004836040694917945,
      "loss": 6.2729,
      "mean_token_accuracy": 0.1426244780421257,
      "num_tokens": 23875529.0,
      "step": 7145
    },
    {
      "entropy": 6.436705446243286,
      "epoch": 1.3948883035801385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000483577702145831,
      "loss": 6.2477,
      "mean_token_accuracy": 0.1404050312936306,
      "num_tokens": 23892442.0,
      "step": 7150
    },
    {
      "entropy": 6.412185478210449,
      "epoch": 1.3958638181640817,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004835513144183979,
      "loss": 6.3655,
      "mean_token_accuracy": 0.13031112477183343,
      "num_tokens": 23908624.0,
      "step": 7155
    },
    {
      "entropy": 6.494060420989991,
      "epoch": 1.3968393327480246,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048352490631207366,
      "loss": 6.315,
      "mean_token_accuracy": 0.1393343172967434,
      "num_tokens": 23925651.0,
      "step": 7160
    },
    {
      "entropy": 6.4586510181427,
      "epoch": 1.3978148473319676,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048349847782943873,
      "loss": 6.4635,
      "mean_token_accuracy": 0.12629888728260993,
      "num_tokens": 23943114.0,
      "step": 7165
    },
    {
      "entropy": 6.4417102336883545,
      "epoch": 1.3987903619159106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004834720289730757,
      "loss": 6.3378,
      "mean_token_accuracy": 0.12870060577988623,
      "num_tokens": 23960712.0,
      "step": 7170
    },
    {
      "entropy": 6.395759963989258,
      "epoch": 1.3997658764998537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048344555974556895,
      "loss": 6.2989,
      "mean_token_accuracy": 0.13676624968647957,
      "num_tokens": 23976832.0,
      "step": 7175
    },
    {
      "entropy": 6.385701465606689,
      "epoch": 1.4007413910837967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048341907014950503,
      "loss": 6.2901,
      "mean_token_accuracy": 0.13823531270027162,
      "num_tokens": 23993323.0,
      "step": 7180
    },
    {
      "entropy": 6.350649356842041,
      "epoch": 1.4017169056677397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004833925601874724,
      "loss": 6.2256,
      "mean_token_accuracy": 0.13872112855315208,
      "num_tokens": 24010164.0,
      "step": 7185
    },
    {
      "entropy": 6.40221643447876,
      "epoch": 1.4026924202516828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004833660298620615,
      "loss": 6.2999,
      "mean_token_accuracy": 0.1347912386059761,
      "num_tokens": 24025521.0,
      "step": 7190
    },
    {
      "entropy": 6.361406946182251,
      "epoch": 1.4036679348356258,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048333947917586484,
      "loss": 6.1969,
      "mean_token_accuracy": 0.14144530296325683,
      "num_tokens": 24041674.0,
      "step": 7195
    },
    {
      "entropy": 6.393558359146118,
      "epoch": 1.4046434494195688,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048331290813147677,
      "loss": 6.319,
      "mean_token_accuracy": 0.13220444396138192,
      "num_tokens": 24057175.0,
      "step": 7200
    },
    {
      "entropy": 6.388633298873901,
      "epoch": 1.4056189640035117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048328631673149377,
      "loss": 6.313,
      "mean_token_accuracy": 0.1288256160914898,
      "num_tokens": 24072602.0,
      "step": 7205
    },
    {
      "entropy": 6.451251792907715,
      "epoch": 1.406594478587455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048325970497851425,
      "loss": 6.1938,
      "mean_token_accuracy": 0.1334480196237564,
      "num_tokens": 24089518.0,
      "step": 7210
    },
    {
      "entropy": 6.331305694580078,
      "epoch": 1.4075699931713979,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004832330728751386,
      "loss": 6.2219,
      "mean_token_accuracy": 0.14645218178629876,
      "num_tokens": 24106487.0,
      "step": 7215
    },
    {
      "entropy": 6.356255626678466,
      "epoch": 1.408545507755341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048320642042396915,
      "loss": 6.3243,
      "mean_token_accuracy": 0.13420081734657288,
      "num_tokens": 24123244.0,
      "step": 7220
    },
    {
      "entropy": 6.4249067306518555,
      "epoch": 1.409521022339284,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048317974762761046,
      "loss": 6.3163,
      "mean_token_accuracy": 0.13352451771497725,
      "num_tokens": 24140221.0,
      "step": 7225
    },
    {
      "entropy": 6.444314289093017,
      "epoch": 1.410496536923227,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048315305448866873,
      "loss": 6.3504,
      "mean_token_accuracy": 0.1284119002521038,
      "num_tokens": 24156968.0,
      "step": 7230
    },
    {
      "entropy": 6.432076692581177,
      "epoch": 1.41147205150717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048312634100975245,
      "loss": 6.4293,
      "mean_token_accuracy": 0.12714660465717315,
      "num_tokens": 24172441.0,
      "step": 7235
    },
    {
      "entropy": 6.516453218460083,
      "epoch": 1.412447566091113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004830996071934719,
      "loss": 6.4553,
      "mean_token_accuracy": 0.12107547670602799,
      "num_tokens": 24190616.0,
      "step": 7240
    },
    {
      "entropy": 6.447654438018799,
      "epoch": 1.413423080675056,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048307285304243945,
      "loss": 6.277,
      "mean_token_accuracy": 0.13554348722100257,
      "num_tokens": 24207195.0,
      "step": 7245
    },
    {
      "entropy": 6.222045803070069,
      "epoch": 1.4143985952589992,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004830460785592692,
      "loss": 6.1718,
      "mean_token_accuracy": 0.14691680669784546,
      "num_tokens": 24224160.0,
      "step": 7250
    },
    {
      "entropy": 6.509744501113891,
      "epoch": 1.4153741098429422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048301928374657783,
      "loss": 6.3991,
      "mean_token_accuracy": 0.12485866621136665,
      "num_tokens": 24239907.0,
      "step": 7255
    },
    {
      "entropy": 6.467510128021241,
      "epoch": 1.4163496244268852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048299246860698336,
      "loss": 6.4075,
      "mean_token_accuracy": 0.12531250789761544,
      "num_tokens": 24255515.0,
      "step": 7260
    },
    {
      "entropy": 6.4502698421478275,
      "epoch": 1.4173251390108281,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048296563314310616,
      "loss": 6.3748,
      "mean_token_accuracy": 0.130258459597826,
      "num_tokens": 24273002.0,
      "step": 7265
    },
    {
      "entropy": 6.282334041595459,
      "epoch": 1.4183006535947713,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004829387773575686,
      "loss": 6.1921,
      "mean_token_accuracy": 0.14881466403603555,
      "num_tokens": 24291622.0,
      "step": 7270
    },
    {
      "entropy": 6.449390316009522,
      "epoch": 1.4192761681787143,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048291190125299474,
      "loss": 6.327,
      "mean_token_accuracy": 0.12799393385648727,
      "num_tokens": 24309445.0,
      "step": 7275
    },
    {
      "entropy": 6.463464736938477,
      "epoch": 1.4202516827626572,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048288500483201083,
      "loss": 6.3235,
      "mean_token_accuracy": 0.12787167951464654,
      "num_tokens": 24324667.0,
      "step": 7280
    },
    {
      "entropy": 6.408115482330322,
      "epoch": 1.4212271973466004,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048285808809724523,
      "loss": 6.2608,
      "mean_token_accuracy": 0.14399631842970848,
      "num_tokens": 24340998.0,
      "step": 7285
    },
    {
      "entropy": 6.417619419097901,
      "epoch": 1.4222027119305434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004828311510513281,
      "loss": 6.3342,
      "mean_token_accuracy": 0.1270756185054779,
      "num_tokens": 24357416.0,
      "step": 7290
    },
    {
      "entropy": 6.344970417022705,
      "epoch": 1.4231782265144863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004828041936968916,
      "loss": 6.311,
      "mean_token_accuracy": 0.1391972079873085,
      "num_tokens": 24373837.0,
      "step": 7295
    },
    {
      "entropy": 6.4589502811431885,
      "epoch": 1.4241537410984293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048277721603656986,
      "loss": 6.3509,
      "mean_token_accuracy": 0.12704649567604065,
      "num_tokens": 24389777.0,
      "step": 7300
    },
    {
      "entropy": 6.46122784614563,
      "epoch": 1.4251292556823725,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048275021807299916,
      "loss": 6.3354,
      "mean_token_accuracy": 0.13358156830072404,
      "num_tokens": 24407815.0,
      "step": 7305
    },
    {
      "entropy": 6.448368310928345,
      "epoch": 1.4261047702663154,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048272319980881756,
      "loss": 6.3858,
      "mean_token_accuracy": 0.1291820302605629,
      "num_tokens": 24422932.0,
      "step": 7310
    },
    {
      "entropy": 6.417226648330688,
      "epoch": 1.4270802848502586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004826961612466653,
      "loss": 6.2547,
      "mean_token_accuracy": 0.1365291103720665,
      "num_tokens": 24440012.0,
      "step": 7315
    },
    {
      "entropy": 6.507063388824463,
      "epoch": 1.4280557994342016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048266910238918434,
      "loss": 6.3783,
      "mean_token_accuracy": 0.13219919204711914,
      "num_tokens": 24457056.0,
      "step": 7320
    },
    {
      "entropy": 6.421991968154908,
      "epoch": 1.4290313140181445,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004826420232390188,
      "loss": 6.2118,
      "mean_token_accuracy": 0.14021918401122094,
      "num_tokens": 24474779.0,
      "step": 7325
    },
    {
      "entropy": 6.451693820953369,
      "epoch": 1.4300068286020875,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826149237988148,
      "loss": 6.3877,
      "mean_token_accuracy": 0.13076698407530785,
      "num_tokens": 24492046.0,
      "step": 7330
    },
    {
      "entropy": 6.428009796142578,
      "epoch": 1.4309823431860307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048258780407122045,
      "loss": 6.3778,
      "mean_token_accuracy": 0.12810813784599304,
      "num_tokens": 24507972.0,
      "step": 7335
    },
    {
      "entropy": 6.406046628952026,
      "epoch": 1.4319578577699736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004825606640588858,
      "loss": 6.3003,
      "mean_token_accuracy": 0.12790659442543983,
      "num_tokens": 24526350.0,
      "step": 7340
    },
    {
      "entropy": 6.323268842697144,
      "epoch": 1.4329333723539168,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048253350376446265,
      "loss": 6.2255,
      "mean_token_accuracy": 0.14121739491820334,
      "num_tokens": 24543478.0,
      "step": 7345
    },
    {
      "entropy": 6.436959886550904,
      "epoch": 1.4339088869378598,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004825063231906053,
      "loss": 6.3139,
      "mean_token_accuracy": 0.13587325438857079,
      "num_tokens": 24562285.0,
      "step": 7350
    },
    {
      "entropy": 6.415534496307373,
      "epoch": 1.4348844015218027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004824791223399696,
      "loss": 6.3207,
      "mean_token_accuracy": 0.1319422349333763,
      "num_tokens": 24578271.0,
      "step": 7355
    },
    {
      "entropy": 6.436182737350464,
      "epoch": 1.4358599161057457,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048245190121521346,
      "loss": 6.3537,
      "mean_token_accuracy": 0.1349285714328289,
      "num_tokens": 24595107.0,
      "step": 7360
    },
    {
      "entropy": 6.44680438041687,
      "epoch": 1.4368354306896889,
      "grad_norm": 1.125,
      "learning_rate": 0.000482424659818997,
      "loss": 6.3398,
      "mean_token_accuracy": 0.13149738907814026,
      "num_tokens": 24611594.0,
      "step": 7365
    },
    {
      "entropy": 6.432389640808106,
      "epoch": 1.4378109452736318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000482397398153982,
      "loss": 6.3311,
      "mean_token_accuracy": 0.1356859177350998,
      "num_tokens": 24628284.0,
      "step": 7370
    },
    {
      "entropy": 6.389889287948608,
      "epoch": 1.4387864598575748,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048237011622283244,
      "loss": 6.3027,
      "mean_token_accuracy": 0.13516581580042838,
      "num_tokens": 24644305.0,
      "step": 7375
    },
    {
      "entropy": 6.524003744125366,
      "epoch": 1.439761974441518,
      "grad_norm": 1.25,
      "learning_rate": 0.00048234281402821416,
      "loss": 6.4155,
      "mean_token_accuracy": 0.1310324065387249,
      "num_tokens": 24661314.0,
      "step": 7380
    },
    {
      "entropy": 6.496041202545166,
      "epoch": 1.440737489025461,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048231549157279514,
      "loss": 6.3175,
      "mean_token_accuracy": 0.13314620330929755,
      "num_tokens": 24677175.0,
      "step": 7385
    },
    {
      "entropy": 6.31291093826294,
      "epoch": 1.441713003609404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004822881488592452,
      "loss": 6.2768,
      "mean_token_accuracy": 0.1356297641992569,
      "num_tokens": 24695227.0,
      "step": 7390
    },
    {
      "entropy": 6.187795257568359,
      "epoch": 1.4426885181933469,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048226078589023606,
      "loss": 6.0621,
      "mean_token_accuracy": 0.1592597134411335,
      "num_tokens": 24712186.0,
      "step": 7395
    },
    {
      "entropy": 6.413484716415406,
      "epoch": 1.44366403277729,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048223340266844166,
      "loss": 6.3136,
      "mean_token_accuracy": 0.13508392572402955,
      "num_tokens": 24728874.0,
      "step": 7400
    },
    {
      "entropy": 6.339919805526733,
      "epoch": 1.444639547361233,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822059991965377,
      "loss": 6.2824,
      "mean_token_accuracy": 0.13725995346903802,
      "num_tokens": 24745534.0,
      "step": 7405
    },
    {
      "entropy": 6.4958734035491945,
      "epoch": 1.4456150619451762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048217857547720207,
      "loss": 6.3205,
      "mean_token_accuracy": 0.1262024737894535,
      "num_tokens": 24760872.0,
      "step": 7410
    },
    {
      "entropy": 6.444066667556763,
      "epoch": 1.4465905765291192,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004821511315131144,
      "loss": 6.3431,
      "mean_token_accuracy": 0.12859414666891097,
      "num_tokens": 24776265.0,
      "step": 7415
    },
    {
      "entropy": 6.391889858245849,
      "epoch": 1.4475660911130621,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004821236673069564,
      "loss": 6.275,
      "mean_token_accuracy": 0.13680661246180534,
      "num_tokens": 24791557.0,
      "step": 7420
    },
    {
      "entropy": 6.418000793457031,
      "epoch": 1.448541605697005,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048209618286141194,
      "loss": 6.2932,
      "mean_token_accuracy": 0.13506835028529168,
      "num_tokens": 24807444.0,
      "step": 7425
    },
    {
      "entropy": 6.446009254455566,
      "epoch": 1.4495171202809483,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004820686781791665,
      "loss": 6.3605,
      "mean_token_accuracy": 0.1282960891723633,
      "num_tokens": 24823971.0,
      "step": 7430
    },
    {
      "entropy": 6.298105764389038,
      "epoch": 1.4504926348648912,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048204115326290795,
      "loss": 6.2055,
      "mean_token_accuracy": 0.1496977724134922,
      "num_tokens": 24841363.0,
      "step": 7435
    },
    {
      "entropy": 6.461188316345215,
      "epoch": 1.4514681494488342,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048201360811532567,
      "loss": 6.339,
      "mean_token_accuracy": 0.1369605392217636,
      "num_tokens": 24858593.0,
      "step": 7440
    },
    {
      "entropy": 6.471303033828735,
      "epoch": 1.4524436640327774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004819860427391115,
      "loss": 6.4293,
      "mean_token_accuracy": 0.12976977080106736,
      "num_tokens": 24876303.0,
      "step": 7445
    },
    {
      "entropy": 6.3977500915527346,
      "epoch": 1.4534191786167203,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004819584571369589,
      "loss": 6.2823,
      "mean_token_accuracy": 0.12830896079540252,
      "num_tokens": 24891455.0,
      "step": 7450
    },
    {
      "entropy": 6.5057737827301025,
      "epoch": 1.4543946932006633,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004819308513115635,
      "loss": 6.3449,
      "mean_token_accuracy": 0.13328469172120094,
      "num_tokens": 24906598.0,
      "step": 7455
    },
    {
      "entropy": 6.437407827377319,
      "epoch": 1.4553702077846065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004819032252656229,
      "loss": 6.3962,
      "mean_token_accuracy": 0.12655049264431,
      "num_tokens": 24923299.0,
      "step": 7460
    },
    {
      "entropy": 6.397623348236084,
      "epoch": 1.4563457223685494,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048187557900183643,
      "loss": 6.2592,
      "mean_token_accuracy": 0.14331335499882697,
      "num_tokens": 24940510.0,
      "step": 7465
    },
    {
      "entropy": 6.451620626449585,
      "epoch": 1.4573212369524924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004818479125229057,
      "loss": 6.3264,
      "mean_token_accuracy": 0.13229208663105965,
      "num_tokens": 24957644.0,
      "step": 7470
    },
    {
      "entropy": 6.3888345718383786,
      "epoch": 1.4582967515364356,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004818202258315343,
      "loss": 6.2763,
      "mean_token_accuracy": 0.13690287247300148,
      "num_tokens": 24974405.0,
      "step": 7475
    },
    {
      "entropy": 6.423791456222534,
      "epoch": 1.4592722661203785,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004817925189304275,
      "loss": 6.3805,
      "mean_token_accuracy": 0.1278742805123329,
      "num_tokens": 24992294.0,
      "step": 7480
    },
    {
      "entropy": 6.501798963546753,
      "epoch": 1.4602477807043215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048176479182229266,
      "loss": 6.327,
      "mean_token_accuracy": 0.12691144198179244,
      "num_tokens": 25008177.0,
      "step": 7485
    },
    {
      "entropy": 6.480355834960937,
      "epoch": 1.4612232952882644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048173704450983943,
      "loss": 6.4216,
      "mean_token_accuracy": 0.12450478151440621,
      "num_tokens": 25025321.0,
      "step": 7490
    },
    {
      "entropy": 6.456436824798584,
      "epoch": 1.4621988098722076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048170927699577894,
      "loss": 6.3578,
      "mean_token_accuracy": 0.13064106926321983,
      "num_tokens": 25041751.0,
      "step": 7495
    },
    {
      "entropy": 6.394688606262207,
      "epoch": 1.4631743244561506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004816814892828247,
      "loss": 6.2542,
      "mean_token_accuracy": 0.1398377388715744,
      "num_tokens": 25058897.0,
      "step": 7500
    },
    {
      "entropy": 6.4889732837677006,
      "epoch": 1.4641498390400938,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048165368137369186,
      "loss": 6.3837,
      "mean_token_accuracy": 0.12591828927397727,
      "num_tokens": 25074750.0,
      "step": 7505
    },
    {
      "entropy": 6.414124488830566,
      "epoch": 1.4651253536240367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048162585327109794,
      "loss": 6.3397,
      "mean_token_accuracy": 0.1316986232995987,
      "num_tokens": 25089416.0,
      "step": 7510
    },
    {
      "entropy": 6.35968132019043,
      "epoch": 1.4661008682079797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000481598004977762,
      "loss": 6.2958,
      "mean_token_accuracy": 0.13567785769701005,
      "num_tokens": 25105543.0,
      "step": 7515
    },
    {
      "entropy": 6.534859323501587,
      "epoch": 1.4670763827919227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048157013649640523,
      "loss": 6.3882,
      "mean_token_accuracy": 0.12567055374383926,
      "num_tokens": 25124761.0,
      "step": 7520
    },
    {
      "entropy": 6.4166007995605465,
      "epoch": 1.4680518973758658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048154224782975103,
      "loss": 6.3353,
      "mean_token_accuracy": 0.13088393360376357,
      "num_tokens": 25142484.0,
      "step": 7525
    },
    {
      "entropy": 6.375730705261231,
      "epoch": 1.4690274119598088,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048151433898052453,
      "loss": 6.228,
      "mean_token_accuracy": 0.1457482621073723,
      "num_tokens": 25159870.0,
      "step": 7530
    },
    {
      "entropy": 6.1643565654754635,
      "epoch": 1.4700029265437518,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004814864099514529,
      "loss": 6.0818,
      "mean_token_accuracy": 0.15548867508769035,
      "num_tokens": 25176643.0,
      "step": 7535
    },
    {
      "entropy": 6.355472421646118,
      "epoch": 1.470978441127695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004814584607452651,
      "loss": 6.3406,
      "mean_token_accuracy": 0.13133197277784348,
      "num_tokens": 25191963.0,
      "step": 7540
    },
    {
      "entropy": 6.481121349334717,
      "epoch": 1.471953955711638,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004814304913646924,
      "loss": 6.4112,
      "mean_token_accuracy": 0.13144318014383316,
      "num_tokens": 25210250.0,
      "step": 7545
    },
    {
      "entropy": 6.424749040603638,
      "epoch": 1.4729294702955809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048140250181246785,
      "loss": 6.2404,
      "mean_token_accuracy": 0.13695558905601501,
      "num_tokens": 25226951.0,
      "step": 7550
    },
    {
      "entropy": 6.2781449317932125,
      "epoch": 1.4739049848795238,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004813744920913265,
      "loss": 6.1842,
      "mean_token_accuracy": 0.1393652565777302,
      "num_tokens": 25244055.0,
      "step": 7555
    },
    {
      "entropy": 6.315723323822022,
      "epoch": 1.474880499463467,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048134646220400517,
      "loss": 6.168,
      "mean_token_accuracy": 0.15632776096463202,
      "num_tokens": 25261408.0,
      "step": 7560
    },
    {
      "entropy": 6.3712536811828615,
      "epoch": 1.47585601404741,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048131841215324307,
      "loss": 6.2189,
      "mean_token_accuracy": 0.14304035902023315,
      "num_tokens": 25277452.0,
      "step": 7565
    },
    {
      "entropy": 6.46163158416748,
      "epoch": 1.4768315286313531,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000481290341941781,
      "loss": 6.4123,
      "mean_token_accuracy": 0.12807180508971214,
      "num_tokens": 25293414.0,
      "step": 7570
    },
    {
      "entropy": 6.424300956726074,
      "epoch": 1.477807043215296,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048126225157236205,
      "loss": 6.3026,
      "mean_token_accuracy": 0.1340413823723793,
      "num_tokens": 25309684.0,
      "step": 7575
    },
    {
      "entropy": 6.475788259506226,
      "epoch": 1.478782557799239,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048123414104773093,
      "loss": 6.3083,
      "mean_token_accuracy": 0.1319167956709862,
      "num_tokens": 25324810.0,
      "step": 7580
    },
    {
      "entropy": 6.523877096176148,
      "epoch": 1.479758072383182,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048120601037063466,
      "loss": 6.4496,
      "mean_token_accuracy": 0.12885732874274253,
      "num_tokens": 25342921.0,
      "step": 7585
    },
    {
      "entropy": 6.40522575378418,
      "epoch": 1.4807335869671252,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004811778595438219,
      "loss": 6.3367,
      "mean_token_accuracy": 0.13316663652658461,
      "num_tokens": 25358770.0,
      "step": 7590
    },
    {
      "entropy": 6.360969686508179,
      "epoch": 1.4817091015510682,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004811496885700436,
      "loss": 6.1673,
      "mean_token_accuracy": 0.14153073206543923,
      "num_tokens": 25375429.0,
      "step": 7595
    },
    {
      "entropy": 6.463377332687378,
      "epoch": 1.4826846161350113,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004811214974520524,
      "loss": 6.3518,
      "mean_token_accuracy": 0.1343583919107914,
      "num_tokens": 25392262.0,
      "step": 7600
    },
    {
      "entropy": 6.375286483764649,
      "epoch": 1.4836601307189543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810932861926032,
      "loss": 6.2672,
      "mean_token_accuracy": 0.1428772673010826,
      "num_tokens": 25408663.0,
      "step": 7605
    },
    {
      "entropy": 6.43287992477417,
      "epoch": 1.4846356453028973,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004810650547944526,
      "loss": 6.3311,
      "mean_token_accuracy": 0.13645109459757804,
      "num_tokens": 25425587.0,
      "step": 7610
    },
    {
      "entropy": 6.248222351074219,
      "epoch": 1.4856111598868402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004810368032603591,
      "loss": 6.1056,
      "mean_token_accuracy": 0.15326923951506616,
      "num_tokens": 25443506.0,
      "step": 7615
    },
    {
      "entropy": 6.459558820724487,
      "epoch": 1.4865866744707834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004810085315930837,
      "loss": 6.329,
      "mean_token_accuracy": 0.13796368092298508,
      "num_tokens": 25462197.0,
      "step": 7620
    },
    {
      "entropy": 6.386405515670776,
      "epoch": 1.4875621890547264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048098023979538867,
      "loss": 6.3031,
      "mean_token_accuracy": 0.13654110655188562,
      "num_tokens": 25478732.0,
      "step": 7625
    },
    {
      "entropy": 6.4521832942962645,
      "epoch": 1.4885377036386693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048095192787003886,
      "loss": 6.3592,
      "mean_token_accuracy": 0.12688954025506974,
      "num_tokens": 25494427.0,
      "step": 7630
    },
    {
      "entropy": 6.345031118392944,
      "epoch": 1.4895132182226125,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048092359581980063,
      "loss": 6.2412,
      "mean_token_accuracy": 0.13950393721461296,
      "num_tokens": 25511617.0,
      "step": 7635
    },
    {
      "entropy": 6.404146575927735,
      "epoch": 1.4904887328065555,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004808952436474425,
      "loss": 6.3623,
      "mean_token_accuracy": 0.13065383583307266,
      "num_tokens": 25529655.0,
      "step": 7640
    },
    {
      "entropy": 6.483083057403564,
      "epoch": 1.4914642473904984,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048086687135573506,
      "loss": 6.2984,
      "mean_token_accuracy": 0.1328952915966511,
      "num_tokens": 25545615.0,
      "step": 7645
    },
    {
      "entropy": 6.220514392852783,
      "epoch": 1.4924397619744414,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048083847894745064,
      "loss": 6.1629,
      "mean_token_accuracy": 0.14714162573218345,
      "num_tokens": 25562097.0,
      "step": 7650
    },
    {
      "entropy": 6.365171480178833,
      "epoch": 1.4934152765583846,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048081006642536366,
      "loss": 6.2311,
      "mean_token_accuracy": 0.14092755317687988,
      "num_tokens": 25578013.0,
      "step": 7655
    },
    {
      "entropy": 6.366628646850586,
      "epoch": 1.4943907911423275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004807816337922506,
      "loss": 6.2762,
      "mean_token_accuracy": 0.1347263388335705,
      "num_tokens": 25594451.0,
      "step": 7660
    },
    {
      "entropy": 6.364975118637085,
      "epoch": 1.4953663057262707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004807531810508896,
      "loss": 6.2135,
      "mean_token_accuracy": 0.1374681033194065,
      "num_tokens": 25611522.0,
      "step": 7665
    },
    {
      "entropy": 6.42410945892334,
      "epoch": 1.4963418203102137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004807247082040611,
      "loss": 6.3174,
      "mean_token_accuracy": 0.13571968749165536,
      "num_tokens": 25627755.0,
      "step": 7670
    },
    {
      "entropy": 6.435871219635009,
      "epoch": 1.4973173348941566,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048069621525454733,
      "loss": 6.3704,
      "mean_token_accuracy": 0.13119870126247407,
      "num_tokens": 25645059.0,
      "step": 7675
    },
    {
      "entropy": 6.465434408187866,
      "epoch": 1.4982928494780996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004806677022051325,
      "loss": 6.364,
      "mean_token_accuracy": 0.13383298069238664,
      "num_tokens": 25662324.0,
      "step": 7680
    },
    {
      "entropy": 6.4146112442016605,
      "epoch": 1.4992683640620428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048063916905860283,
      "loss": 6.2918,
      "mean_token_accuracy": 0.13527191579341888,
      "num_tokens": 25678648.0,
      "step": 7685
    },
    {
      "entropy": 6.284770774841308,
      "epoch": 1.5002438786459857,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004806106158177465,
      "loss": 6.1931,
      "mean_token_accuracy": 0.14660634845495224,
      "num_tokens": 25695694.0,
      "step": 7690
    },
    {
      "entropy": 6.351878881454468,
      "epoch": 1.501219393229929,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004805820424853536,
      "loss": 6.2254,
      "mean_token_accuracy": 0.1423006035387516,
      "num_tokens": 25713168.0,
      "step": 7695
    },
    {
      "entropy": 6.358890008926392,
      "epoch": 1.5021949078138719,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004805534490642162,
      "loss": 6.2922,
      "mean_token_accuracy": 0.13413700014352797,
      "num_tokens": 25729763.0,
      "step": 7700
    },
    {
      "entropy": 6.422443914413452,
      "epoch": 1.5031704223978148,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048052483555712834,
      "loss": 6.2782,
      "mean_token_accuracy": 0.1360203243792057,
      "num_tokens": 25747852.0,
      "step": 7705
    },
    {
      "entropy": 6.4142485618591305,
      "epoch": 1.5041459369817578,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004804962019668861,
      "loss": 6.2656,
      "mean_token_accuracy": 0.1434503674507141,
      "num_tokens": 25764594.0,
      "step": 7710
    },
    {
      "entropy": 6.433900356292725,
      "epoch": 1.5051214515657008,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004804675482962874,
      "loss": 6.2776,
      "mean_token_accuracy": 0.1379360906779766,
      "num_tokens": 25782835.0,
      "step": 7715
    },
    {
      "entropy": 6.402728033065796,
      "epoch": 1.506096966149644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004804388745481322,
      "loss": 6.2985,
      "mean_token_accuracy": 0.1348798431456089,
      "num_tokens": 25798152.0,
      "step": 7720
    },
    {
      "entropy": 6.2837591648101805,
      "epoch": 1.5070724807335871,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048041018072522224,
      "loss": 6.2306,
      "mean_token_accuracy": 0.14413685724139214,
      "num_tokens": 25816521.0,
      "step": 7725
    },
    {
      "entropy": 6.43160572052002,
      "epoch": 1.50804799531753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048038146683036163,
      "loss": 6.2828,
      "mean_token_accuracy": 0.1277213990688324,
      "num_tokens": 25832416.0,
      "step": 7730
    },
    {
      "entropy": 6.331970977783203,
      "epoch": 1.509023509901473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004803527328663561,
      "loss": 6.2039,
      "mean_token_accuracy": 0.14752345755696297,
      "num_tokens": 25850343.0,
      "step": 7735
    },
    {
      "entropy": 6.458510017395019,
      "epoch": 1.509999024485416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004803239788360134,
      "loss": 6.4037,
      "mean_token_accuracy": 0.12729395776987076,
      "num_tokens": 25868661.0,
      "step": 7740
    },
    {
      "entropy": 6.436870050430298,
      "epoch": 1.510974539069359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048029520474214316,
      "loss": 6.3628,
      "mean_token_accuracy": 0.1293166883289814,
      "num_tokens": 25886417.0,
      "step": 7745
    },
    {
      "entropy": 6.4374744415283205,
      "epoch": 1.5119500536533022,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048026641058755725,
      "loss": 6.2428,
      "mean_token_accuracy": 0.1379207879304886,
      "num_tokens": 25904271.0,
      "step": 7750
    },
    {
      "entropy": 6.34000153541565,
      "epoch": 1.512925568237245,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048023759637506934,
      "loss": 6.2207,
      "mean_token_accuracy": 0.14182295948266982,
      "num_tokens": 25920444.0,
      "step": 7755
    },
    {
      "entropy": 6.426927709579468,
      "epoch": 1.5139010828211883,
      "grad_norm": 1.125,
      "learning_rate": 0.000480208762107495,
      "loss": 6.295,
      "mean_token_accuracy": 0.12749650329351425,
      "num_tokens": 25937655.0,
      "step": 7760
    },
    {
      "entropy": 6.371528625488281,
      "epoch": 1.5148765974051313,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004801799077876517,
      "loss": 6.2255,
      "mean_token_accuracy": 0.13998317420482637,
      "num_tokens": 25954242.0,
      "step": 7765
    },
    {
      "entropy": 6.435590696334839,
      "epoch": 1.5158521119890742,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004801510334183591,
      "loss": 6.3637,
      "mean_token_accuracy": 0.1318854011595249,
      "num_tokens": 25971343.0,
      "step": 7770
    },
    {
      "entropy": 6.258959484100342,
      "epoch": 1.5168276265730172,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048012213900243877,
      "loss": 6.2109,
      "mean_token_accuracy": 0.14184840098023416,
      "num_tokens": 25988369.0,
      "step": 7775
    },
    {
      "entropy": 6.414775466918945,
      "epoch": 1.5178031411569601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000480093224542714,
      "loss": 6.3453,
      "mean_token_accuracy": 0.13084505572915078,
      "num_tokens": 26005028.0,
      "step": 7780
    },
    {
      "entropy": 6.4657281875610355,
      "epoch": 1.5187786557409033,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004800642900420104,
      "loss": 6.382,
      "mean_token_accuracy": 0.13490205258131027,
      "num_tokens": 26021529.0,
      "step": 7785
    },
    {
      "entropy": 6.429589748382568,
      "epoch": 1.5197541703248465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048003533550315517,
      "loss": 6.2764,
      "mean_token_accuracy": 0.137289709597826,
      "num_tokens": 26036804.0,
      "step": 7790
    },
    {
      "entropy": 6.334526538848877,
      "epoch": 1.5207296849087895,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048000636092897773,
      "loss": 6.1788,
      "mean_token_accuracy": 0.14671095013618468,
      "num_tokens": 26054532.0,
      "step": 7795
    },
    {
      "entropy": 6.410811090469361,
      "epoch": 1.5217051994927324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004799773663223093,
      "loss": 6.3435,
      "mean_token_accuracy": 0.13460630774497986,
      "num_tokens": 26072600.0,
      "step": 7800
    },
    {
      "entropy": 6.405370569229126,
      "epoch": 1.5226807140766754,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004799483516859833,
      "loss": 6.2725,
      "mean_token_accuracy": 0.1357073202729225,
      "num_tokens": 26087129.0,
      "step": 7805
    },
    {
      "entropy": 6.400820779800415,
      "epoch": 1.5236562286606183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004799193170228348,
      "loss": 6.263,
      "mean_token_accuracy": 0.13634018898010253,
      "num_tokens": 26102453.0,
      "step": 7810
    },
    {
      "entropy": 6.404085636138916,
      "epoch": 1.5246317432445615,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047989026233570087,
      "loss": 6.3277,
      "mean_token_accuracy": 0.13271497786045075,
      "num_tokens": 26119591.0,
      "step": 7815
    },
    {
      "entropy": 6.394654941558838,
      "epoch": 1.5256072578285045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004798611876274208,
      "loss": 6.2659,
      "mean_token_accuracy": 0.13625818341970444,
      "num_tokens": 26137155.0,
      "step": 7820
    },
    {
      "entropy": 6.334944248199463,
      "epoch": 1.5265827724124477,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047983209290083565,
      "loss": 6.2307,
      "mean_token_accuracy": 0.142851971834898,
      "num_tokens": 26153523.0,
      "step": 7825
    },
    {
      "entropy": 6.314331007003784,
      "epoch": 1.5275582869963906,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004798029781587884,
      "loss": 6.23,
      "mean_token_accuracy": 0.1436134919524193,
      "num_tokens": 26172233.0,
      "step": 7830
    },
    {
      "entropy": 6.416027021408081,
      "epoch": 1.5285338015803336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047977384340412406,
      "loss": 6.3246,
      "mean_token_accuracy": 0.13290538638830185,
      "num_tokens": 26189702.0,
      "step": 7835
    },
    {
      "entropy": 6.506917667388916,
      "epoch": 1.5295093161642765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797446886396895,
      "loss": 6.3974,
      "mean_token_accuracy": 0.12871293872594833,
      "num_tokens": 26205986.0,
      "step": 7840
    },
    {
      "entropy": 6.448561763763427,
      "epoch": 1.5304848307482197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047971551386833376,
      "loss": 6.3444,
      "mean_token_accuracy": 0.12785482555627822,
      "num_tokens": 26223667.0,
      "step": 7845
    },
    {
      "entropy": 6.3721331596374515,
      "epoch": 1.5314603453321627,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004796863190929076,
      "loss": 6.2395,
      "mean_token_accuracy": 0.14206554293632506,
      "num_tokens": 26240269.0,
      "step": 7850
    },
    {
      "entropy": 6.484223222732544,
      "epoch": 1.5324358599161059,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047965710431626394,
      "loss": 6.3657,
      "mean_token_accuracy": 0.12661567851901054,
      "num_tokens": 26257184.0,
      "step": 7855
    },
    {
      "entropy": 6.4178954601287845,
      "epoch": 1.5334113745000488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004796278695412573,
      "loss": 6.2955,
      "mean_token_accuracy": 0.13910969123244285,
      "num_tokens": 26273002.0,
      "step": 7860
    },
    {
      "entropy": 6.535676908493042,
      "epoch": 1.5343868890839918,
      "grad_norm": 1.125,
      "learning_rate": 0.0004795986147707447,
      "loss": 6.4212,
      "mean_token_accuracy": 0.12753215432167053,
      "num_tokens": 26289734.0,
      "step": 7865
    },
    {
      "entropy": 6.373676109313965,
      "epoch": 1.5353624036679347,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795693400075846,
      "loss": 6.2689,
      "mean_token_accuracy": 0.13848392367362977,
      "num_tokens": 26307409.0,
      "step": 7870
    },
    {
      "entropy": 6.401240491867066,
      "epoch": 1.5363379182518777,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004795400452546377,
      "loss": 6.3385,
      "mean_token_accuracy": 0.13819917514920235,
      "num_tokens": 26324061.0,
      "step": 7875
    },
    {
      "entropy": 6.407931613922119,
      "epoch": 1.537313432835821,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047951073051476657,
      "loss": 6.1934,
      "mean_token_accuracy": 0.14290593191981316,
      "num_tokens": 26340990.0,
      "step": 7880
    },
    {
      "entropy": 6.377027320861816,
      "epoch": 1.538288947419764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047948139579083573,
      "loss": 6.2535,
      "mean_token_accuracy": 0.1417265333235264,
      "num_tokens": 26357389.0,
      "step": 7885
    },
    {
      "entropy": 6.503249979019165,
      "epoch": 1.539264462003707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004794520410857117,
      "loss": 6.4051,
      "mean_token_accuracy": 0.12270434722304344,
      "num_tokens": 26373325.0,
      "step": 7890
    },
    {
      "entropy": 6.456487560272217,
      "epoch": 1.54023997658765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004794226664022628,
      "loss": 6.3334,
      "mean_token_accuracy": 0.13334634378552437,
      "num_tokens": 26391098.0,
      "step": 7895
    },
    {
      "entropy": 6.428577041625976,
      "epoch": 1.541215491171593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047939327174335974,
      "loss": 6.3026,
      "mean_token_accuracy": 0.12870485484600067,
      "num_tokens": 26407806.0,
      "step": 7900
    },
    {
      "entropy": 6.514582824707031,
      "epoch": 1.542191005755536,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047936385711187443,
      "loss": 6.4159,
      "mean_token_accuracy": 0.12666127756237983,
      "num_tokens": 26423670.0,
      "step": 7905
    },
    {
      "entropy": 6.398328447341919,
      "epoch": 1.543166520339479,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047933442251068143,
      "loss": 6.2969,
      "mean_token_accuracy": 0.13170450776815415,
      "num_tokens": 26440493.0,
      "step": 7910
    },
    {
      "entropy": 6.441270112991333,
      "epoch": 1.544142034923422,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004793049679426569,
      "loss": 6.3477,
      "mean_token_accuracy": 0.13024605736136435,
      "num_tokens": 26458347.0,
      "step": 7915
    },
    {
      "entropy": 6.363266706466675,
      "epoch": 1.5451175495073652,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047927549341067916,
      "loss": 6.2375,
      "mean_token_accuracy": 0.14018218368291854,
      "num_tokens": 26474832.0,
      "step": 7920
    },
    {
      "entropy": 6.4542937755584715,
      "epoch": 1.5460930640913082,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047924599891762814,
      "loss": 6.3499,
      "mean_token_accuracy": 0.1307586796581745,
      "num_tokens": 26492343.0,
      "step": 7925
    },
    {
      "entropy": 6.432386445999145,
      "epoch": 1.5470685786752512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792164844663861,
      "loss": 6.3376,
      "mean_token_accuracy": 0.13525837883353234,
      "num_tokens": 26508677.0,
      "step": 7930
    },
    {
      "entropy": 6.385792875289917,
      "epoch": 1.5480440932591941,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047918695005983705,
      "loss": 6.347,
      "mean_token_accuracy": 0.13921325132250786,
      "num_tokens": 26524409.0,
      "step": 7935
    },
    {
      "entropy": 6.434285688400268,
      "epoch": 1.5490196078431373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004791573957008669,
      "loss": 6.3384,
      "mean_token_accuracy": 0.13250871747732162,
      "num_tokens": 26540884.0,
      "step": 7940
    },
    {
      "entropy": 6.4576334953308105,
      "epoch": 1.5499951224270803,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004791278213923638,
      "loss": 6.2617,
      "mean_token_accuracy": 0.14234519898891448,
      "num_tokens": 26556636.0,
      "step": 7945
    },
    {
      "entropy": 6.4173096179962155,
      "epoch": 1.5509706370110234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004790982271372174,
      "loss": 6.3085,
      "mean_token_accuracy": 0.13910919204354286,
      "num_tokens": 26573023.0,
      "step": 7950
    },
    {
      "entropy": 6.451989555358887,
      "epoch": 1.5519461515949664,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047906861293831965,
      "loss": 6.3465,
      "mean_token_accuracy": 0.13125205412507057,
      "num_tokens": 26590524.0,
      "step": 7955
    },
    {
      "entropy": 6.280890321731567,
      "epoch": 1.5529216661789094,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047903897879856443,
      "loss": 6.0416,
      "mean_token_accuracy": 0.15794993862509726,
      "num_tokens": 26606866.0,
      "step": 7960
    },
    {
      "entropy": 6.247773504257202,
      "epoch": 1.5538971807628523,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047900932472084745,
      "loss": 6.217,
      "mean_token_accuracy": 0.13954248502850533,
      "num_tokens": 26623723.0,
      "step": 7965
    },
    {
      "entropy": 6.47849931716919,
      "epoch": 1.5548726953467953,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004789796507080663,
      "loss": 6.2886,
      "mean_token_accuracy": 0.13119535967707635,
      "num_tokens": 26639980.0,
      "step": 7970
    },
    {
      "entropy": 6.381662893295288,
      "epoch": 1.5558482099307385,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047894995676312066,
      "loss": 6.2966,
      "mean_token_accuracy": 0.13927939757704735,
      "num_tokens": 26657731.0,
      "step": 7975
    },
    {
      "entropy": 6.356800889968872,
      "epoch": 1.5568237245146817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004789202428889122,
      "loss": 6.2912,
      "mean_token_accuracy": 0.1365792505443096,
      "num_tokens": 26675754.0,
      "step": 7980
    },
    {
      "entropy": 6.521950674057007,
      "epoch": 1.5577992390986246,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004788905090883443,
      "loss": 6.3363,
      "mean_token_accuracy": 0.1289303220808506,
      "num_tokens": 26692067.0,
      "step": 7985
    },
    {
      "entropy": 6.21132345199585,
      "epoch": 1.5587747536825676,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788607553643226,
      "loss": 6.1256,
      "mean_token_accuracy": 0.15576086342334747,
      "num_tokens": 26711026.0,
      "step": 7990
    },
    {
      "entropy": 6.315177583694458,
      "epoch": 1.5597502682665105,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047883098171975445,
      "loss": 6.2278,
      "mean_token_accuracy": 0.13475936278700829,
      "num_tokens": 26728463.0,
      "step": 7995
    },
    {
      "entropy": 6.33134880065918,
      "epoch": 1.5607257828504535,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004788011881575492,
      "loss": 6.1666,
      "mean_token_accuracy": 0.13615912646055223,
      "num_tokens": 26745160.0,
      "step": 8000
    },
    {
      "entropy": 6.486618328094482,
      "epoch": 1.5617012974343967,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047877137468061825,
      "loss": 6.3705,
      "mean_token_accuracy": 0.13017848655581474,
      "num_tokens": 26761782.0,
      "step": 8005
    },
    {
      "entropy": 6.297828960418701,
      "epoch": 1.5626768120183396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047874154129187474,
      "loss": 6.124,
      "mean_token_accuracy": 0.14389088079333306,
      "num_tokens": 26779943.0,
      "step": 8010
    },
    {
      "entropy": 6.409016036987305,
      "epoch": 1.5636523266022828,
      "grad_norm": 1.0,
      "learning_rate": 0.00047871168799423406,
      "loss": 6.3247,
      "mean_token_accuracy": 0.13029307276010513,
      "num_tokens": 26797607.0,
      "step": 8015
    },
    {
      "entropy": 6.476004314422608,
      "epoch": 1.5646278411862258,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047868181479061324,
      "loss": 6.3919,
      "mean_token_accuracy": 0.12784578055143356,
      "num_tokens": 26814668.0,
      "step": 8020
    },
    {
      "entropy": 6.417320966720581,
      "epoch": 1.5656033557701687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786519216839314,
      "loss": 6.282,
      "mean_token_accuracy": 0.13755214139819144,
      "num_tokens": 26831590.0,
      "step": 8025
    },
    {
      "entropy": 6.222051334381104,
      "epoch": 1.5665788703541117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786220086771096,
      "loss": 6.0521,
      "mean_token_accuracy": 0.1481317959725857,
      "num_tokens": 26848581.0,
      "step": 8030
    },
    {
      "entropy": 6.454157400131225,
      "epoch": 1.5675543849380549,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047859207577307083,
      "loss": 6.3495,
      "mean_token_accuracy": 0.1282666213810444,
      "num_tokens": 26867599.0,
      "step": 8035
    },
    {
      "entropy": 6.359177160263061,
      "epoch": 1.5685298995219978,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047856212297474007,
      "loss": 6.1766,
      "mean_token_accuracy": 0.14543940052390097,
      "num_tokens": 26884555.0,
      "step": 8040
    },
    {
      "entropy": 6.360111141204834,
      "epoch": 1.569505414105941,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047853215028504406,
      "loss": 6.2757,
      "mean_token_accuracy": 0.14306951835751533,
      "num_tokens": 26899988.0,
      "step": 8045
    },
    {
      "entropy": 6.434762716293335,
      "epoch": 1.570480928689884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004785021577069119,
      "loss": 6.3656,
      "mean_token_accuracy": 0.13432028368115426,
      "num_tokens": 26917588.0,
      "step": 8050
    },
    {
      "entropy": 6.404541635513306,
      "epoch": 1.571456443273827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478472145243274,
      "loss": 6.2712,
      "mean_token_accuracy": 0.13514478802680968,
      "num_tokens": 26934053.0,
      "step": 8055
    },
    {
      "entropy": 6.239260625839234,
      "epoch": 1.57243195785777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047844211289706337,
      "loss": 6.1619,
      "mean_token_accuracy": 0.14952442198991775,
      "num_tokens": 26953870.0,
      "step": 8060
    },
    {
      "entropy": 6.5316849708557125,
      "epoch": 1.5734074724417129,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047841206067121447,
      "loss": 6.4356,
      "mean_token_accuracy": 0.12824916392564772,
      "num_tokens": 26971944.0,
      "step": 8065
    },
    {
      "entropy": 6.444455957412719,
      "epoch": 1.574382987025656,
      "grad_norm": 1.15625,
      "learning_rate": 0.000478381988568664,
      "loss": 6.3209,
      "mean_token_accuracy": 0.13427698537707328,
      "num_tokens": 26988723.0,
      "step": 8070
    },
    {
      "entropy": 6.418841886520386,
      "epoch": 1.5753585016095992,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004783518965923505,
      "loss": 6.315,
      "mean_token_accuracy": 0.13255077749490737,
      "num_tokens": 27004553.0,
      "step": 8075
    },
    {
      "entropy": 6.382885313034057,
      "epoch": 1.5763340161935422,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004783217847452144,
      "loss": 6.2433,
      "mean_token_accuracy": 0.13607777059078216,
      "num_tokens": 27021241.0,
      "step": 8080
    },
    {
      "entropy": 6.3123619556427,
      "epoch": 1.5773095307774851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047829165303019816,
      "loss": 6.1788,
      "mean_token_accuracy": 0.1471312053501606,
      "num_tokens": 27038718.0,
      "step": 8085
    },
    {
      "entropy": 6.392088842391968,
      "epoch": 1.578285045361428,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004782615014502462,
      "loss": 6.3086,
      "mean_token_accuracy": 0.13244988396763802,
      "num_tokens": 27055372.0,
      "step": 8090
    },
    {
      "entropy": 6.507648706436157,
      "epoch": 1.579260559945371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047823133000830467,
      "loss": 6.3368,
      "mean_token_accuracy": 0.12963684797286987,
      "num_tokens": 27070433.0,
      "step": 8095
    },
    {
      "entropy": 6.359089803695679,
      "epoch": 1.5802360745293142,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004782011387073219,
      "loss": 6.2714,
      "mean_token_accuracy": 0.1400555416941643,
      "num_tokens": 27087467.0,
      "step": 8100
    },
    {
      "entropy": 6.426755332946778,
      "epoch": 1.5812115891132572,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004781709275502481,
      "loss": 6.3143,
      "mean_token_accuracy": 0.13048561215400695,
      "num_tokens": 27103388.0,
      "step": 8105
    },
    {
      "entropy": 6.510373163223266,
      "epoch": 1.5821871036972004,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047814069654003534,
      "loss": 6.378,
      "mean_token_accuracy": 0.1318599283695221,
      "num_tokens": 27119752.0,
      "step": 8110
    },
    {
      "entropy": 6.371843481063843,
      "epoch": 1.5831626182811434,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047811044567963774,
      "loss": 6.2879,
      "mean_token_accuracy": 0.12913826182484628,
      "num_tokens": 27136060.0,
      "step": 8115
    },
    {
      "entropy": 6.42429928779602,
      "epoch": 1.5841381328650863,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004780801749720113,
      "loss": 6.3027,
      "mean_token_accuracy": 0.1274295911192894,
      "num_tokens": 27153151.0,
      "step": 8120
    },
    {
      "entropy": 6.297478818893433,
      "epoch": 1.5851136474490293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047804988442011394,
      "loss": 6.195,
      "mean_token_accuracy": 0.1472066633403301,
      "num_tokens": 27170806.0,
      "step": 8125
    },
    {
      "entropy": 6.442647981643677,
      "epoch": 1.5860891620329722,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004780195740269056,
      "loss": 6.3391,
      "mean_token_accuracy": 0.1268402136862278,
      "num_tokens": 27187082.0,
      "step": 8130
    },
    {
      "entropy": 6.337243938446045,
      "epoch": 1.5870646766169154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000477989243795348,
      "loss": 6.2032,
      "mean_token_accuracy": 0.14415137320756913,
      "num_tokens": 27204505.0,
      "step": 8135
    },
    {
      "entropy": 6.30134687423706,
      "epoch": 1.5880401912008586,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047795889372840493,
      "loss": 6.131,
      "mean_token_accuracy": 0.1551969349384308,
      "num_tokens": 27219897.0,
      "step": 8140
    },
    {
      "entropy": 6.326290702819824,
      "epoch": 1.5890157057848016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047792852382904223,
      "loss": 6.1662,
      "mean_token_accuracy": 0.14451753497123718,
      "num_tokens": 27235295.0,
      "step": 8145
    },
    {
      "entropy": 6.2921576499938965,
      "epoch": 1.5899912203687445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004778981341002274,
      "loss": 6.1872,
      "mean_token_accuracy": 0.1341145120561123,
      "num_tokens": 27253172.0,
      "step": 8150
    },
    {
      "entropy": 6.464305257797241,
      "epoch": 1.5909667349526875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047786772454493,
      "loss": 6.4125,
      "mean_token_accuracy": 0.12653393298387527,
      "num_tokens": 27269958.0,
      "step": 8155
    },
    {
      "entropy": 6.408920383453369,
      "epoch": 1.5919422495366304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047783729516612163,
      "loss": 6.1393,
      "mean_token_accuracy": 0.14824150279164314,
      "num_tokens": 27286402.0,
      "step": 8160
    },
    {
      "entropy": 6.390683794021607,
      "epoch": 1.5929177641205736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047780684596677563,
      "loss": 6.31,
      "mean_token_accuracy": 0.1262655407190323,
      "num_tokens": 27301257.0,
      "step": 8165
    },
    {
      "entropy": 6.463958501815796,
      "epoch": 1.5938932787045166,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004777763769498676,
      "loss": 6.316,
      "mean_token_accuracy": 0.13012528494000436,
      "num_tokens": 27318951.0,
      "step": 8170
    },
    {
      "entropy": 6.454387998580932,
      "epoch": 1.5948687932884598,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047774588811837465,
      "loss": 6.3751,
      "mean_token_accuracy": 0.1347372852265835,
      "num_tokens": 27334538.0,
      "step": 8175
    },
    {
      "entropy": 6.398688840866089,
      "epoch": 1.5958443078724027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047771537947527605,
      "loss": 6.2902,
      "mean_token_accuracy": 0.13444510102272034,
      "num_tokens": 27351769.0,
      "step": 8180
    },
    {
      "entropy": 6.419303560256958,
      "epoch": 1.5968198224563457,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047768485102355314,
      "loss": 6.2903,
      "mean_token_accuracy": 0.1324678525328636,
      "num_tokens": 27369085.0,
      "step": 8185
    },
    {
      "entropy": 6.493978691101074,
      "epoch": 1.5977953370402886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047765430276618896,
      "loss": 6.4346,
      "mean_token_accuracy": 0.1298512913286686,
      "num_tokens": 27385601.0,
      "step": 8190
    },
    {
      "entropy": 6.4020094871521,
      "epoch": 1.5987708516242318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004776237347061685,
      "loss": 6.2607,
      "mean_token_accuracy": 0.13325539454817772,
      "num_tokens": 27401973.0,
      "step": 8195
    },
    {
      "entropy": 6.350370597839356,
      "epoch": 1.5997463662081748,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047759314684647894,
      "loss": 6.2103,
      "mean_token_accuracy": 0.13785403966903687,
      "num_tokens": 27418363.0,
      "step": 8200
    },
    {
      "entropy": 6.191963529586792,
      "epoch": 1.600721880792118,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047756253919010907,
      "loss": 6.1053,
      "mean_token_accuracy": 0.15070576816797257,
      "num_tokens": 27435569.0,
      "step": 8205
    },
    {
      "entropy": 6.373612356185913,
      "epoch": 1.601697395376061,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047753191174004984,
      "loss": 6.2549,
      "mean_token_accuracy": 0.13246874883770943,
      "num_tokens": 27451036.0,
      "step": 8210
    },
    {
      "entropy": 6.392071151733399,
      "epoch": 1.6026729099600039,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477501264499294,
      "loss": 6.2407,
      "mean_token_accuracy": 0.14370058476924896,
      "num_tokens": 27466425.0,
      "step": 8215
    },
    {
      "entropy": 6.3580139636993405,
      "epoch": 1.6036484245439468,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004774705974708364,
      "loss": 6.2608,
      "mean_token_accuracy": 0.13429401814937592,
      "num_tokens": 27481502.0,
      "step": 8220
    },
    {
      "entropy": 6.2868125438690186,
      "epoch": 1.6046239391278898,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047743991065767364,
      "loss": 6.2223,
      "mean_token_accuracy": 0.1369063675403595,
      "num_tokens": 27498493.0,
      "step": 8225
    },
    {
      "entropy": 6.447468662261963,
      "epoch": 1.605599453711833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047740920406280423,
      "loss": 6.3083,
      "mean_token_accuracy": 0.13104493394494057,
      "num_tokens": 27514401.0,
      "step": 8230
    },
    {
      "entropy": 6.424680185317993,
      "epoch": 1.6065749682957762,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047737847768922887,
      "loss": 6.3638,
      "mean_token_accuracy": 0.131369998306036,
      "num_tokens": 27531100.0,
      "step": 8235
    },
    {
      "entropy": 6.512340354919433,
      "epoch": 1.6075504828797191,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004773477315399499,
      "loss": 6.3976,
      "mean_token_accuracy": 0.13367923796176912,
      "num_tokens": 27549097.0,
      "step": 8240
    },
    {
      "entropy": 6.48541669845581,
      "epoch": 1.608525997463662,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047731696561797185,
      "loss": 6.3243,
      "mean_token_accuracy": 0.128982300311327,
      "num_tokens": 27565485.0,
      "step": 8245
    },
    {
      "entropy": 6.329342937469482,
      "epoch": 1.609501512047605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047728617992630096,
      "loss": 6.2325,
      "mean_token_accuracy": 0.14581531211733817,
      "num_tokens": 27581832.0,
      "step": 8250
    },
    {
      "entropy": 6.378709602355957,
      "epoch": 1.610477026631548,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047725537446794547,
      "loss": 6.1789,
      "mean_token_accuracy": 0.14422704949975013,
      "num_tokens": 27598149.0,
      "step": 8255
    },
    {
      "entropy": 6.240352487564087,
      "epoch": 1.6114525412154912,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004772245492459157,
      "loss": 6.0799,
      "mean_token_accuracy": 0.15435878932476044,
      "num_tokens": 27616920.0,
      "step": 8260
    },
    {
      "entropy": 6.41365327835083,
      "epoch": 1.6124280557994342,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047719370426322374,
      "loss": 6.3166,
      "mean_token_accuracy": 0.12869892567396163,
      "num_tokens": 27632702.0,
      "step": 8265
    },
    {
      "entropy": 6.418720674514771,
      "epoch": 1.6134035703833773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004771628395228836,
      "loss": 6.3506,
      "mean_token_accuracy": 0.12933458387851715,
      "num_tokens": 27650606.0,
      "step": 8270
    },
    {
      "entropy": 6.36143159866333,
      "epoch": 1.6143790849673203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004771319550279113,
      "loss": 6.205,
      "mean_token_accuracy": 0.1441207595169544,
      "num_tokens": 27668466.0,
      "step": 8275
    },
    {
      "entropy": 6.409032773971558,
      "epoch": 1.6153545995512633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004771010507813247,
      "loss": 6.224,
      "mean_token_accuracy": 0.13409989327192307,
      "num_tokens": 27685032.0,
      "step": 8280
    },
    {
      "entropy": 6.397589921951294,
      "epoch": 1.6163301141352062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004770701267861439,
      "loss": 6.287,
      "mean_token_accuracy": 0.1316670782864094,
      "num_tokens": 27700382.0,
      "step": 8285
    },
    {
      "entropy": 6.433884859085083,
      "epoch": 1.6173056287191494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047703918304539027,
      "loss": 6.2949,
      "mean_token_accuracy": 0.13356413319706917,
      "num_tokens": 27715603.0,
      "step": 8290
    },
    {
      "entropy": 6.359610080718994,
      "epoch": 1.6182811433030924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004770082195620879,
      "loss": 6.2215,
      "mean_token_accuracy": 0.13713194727897643,
      "num_tokens": 27732341.0,
      "step": 8295
    },
    {
      "entropy": 6.438640117645264,
      "epoch": 1.6192566578870355,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004769772363392622,
      "loss": 6.2972,
      "mean_token_accuracy": 0.13653383851051332,
      "num_tokens": 27747731.0,
      "step": 8300
    },
    {
      "entropy": 6.3701379776000975,
      "epoch": 1.6202321724709785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004769462333799409,
      "loss": 6.2734,
      "mean_token_accuracy": 0.13491251245141028,
      "num_tokens": 27764873.0,
      "step": 8305
    },
    {
      "entropy": 6.3868814468383786,
      "epoch": 1.6212076870549215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047691521068715344,
      "loss": 6.2347,
      "mean_token_accuracy": 0.13403560891747474,
      "num_tokens": 27781684.0,
      "step": 8310
    },
    {
      "entropy": 6.440019702911377,
      "epoch": 1.6221832016388644,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004768841682639312,
      "loss": 6.3728,
      "mean_token_accuracy": 0.12072104662656784,
      "num_tokens": 27799596.0,
      "step": 8315
    },
    {
      "entropy": 6.443059968948364,
      "epoch": 1.6231587162228074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004768531061133075,
      "loss": 6.3552,
      "mean_token_accuracy": 0.12633103206753732,
      "num_tokens": 27816469.0,
      "step": 8320
    },
    {
      "entropy": 6.524343538284302,
      "epoch": 1.6241342308067506,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004768220242383178,
      "loss": 6.4005,
      "mean_token_accuracy": 0.1287224218249321,
      "num_tokens": 27832794.0,
      "step": 8325
    },
    {
      "entropy": 6.329700088500976,
      "epoch": 1.6251097453906937,
      "grad_norm": 1.0,
      "learning_rate": 0.00047679092264199915,
      "loss": 6.2236,
      "mean_token_accuracy": 0.14280752018094062,
      "num_tokens": 27851019.0,
      "step": 8330
    },
    {
      "entropy": 6.464788198471069,
      "epoch": 1.6260852599746367,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047675980132739067,
      "loss": 6.3502,
      "mean_token_accuracy": 0.13230849504470826,
      "num_tokens": 27869377.0,
      "step": 8335
    },
    {
      "entropy": 6.486220741271973,
      "epoch": 1.6270607745585797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004767286602975336,
      "loss": 6.3185,
      "mean_token_accuracy": 0.13716409876942634,
      "num_tokens": 27887196.0,
      "step": 8340
    },
    {
      "entropy": 6.416134834289551,
      "epoch": 1.6280362891425226,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004766974995554707,
      "loss": 6.3151,
      "mean_token_accuracy": 0.12722501754760743,
      "num_tokens": 27903536.0,
      "step": 8345
    },
    {
      "entropy": 6.311086845397949,
      "epoch": 1.6290118037264656,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004766663191042471,
      "loss": 6.1915,
      "mean_token_accuracy": 0.14616681039333343,
      "num_tokens": 27919081.0,
      "step": 8350
    },
    {
      "entropy": 6.409001970291138,
      "epoch": 1.6299873183104088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047663511894690955,
      "loss": 6.2539,
      "mean_token_accuracy": 0.13280883133411409,
      "num_tokens": 27934432.0,
      "step": 8355
    },
    {
      "entropy": 6.409823942184448,
      "epoch": 1.6309628328943517,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004766038990865068,
      "loss": 6.3071,
      "mean_token_accuracy": 0.1350394830107689,
      "num_tokens": 27950395.0,
      "step": 8360
    },
    {
      "entropy": 6.381566047668457,
      "epoch": 1.631938347478295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047657265952608953,
      "loss": 6.3216,
      "mean_token_accuracy": 0.13965593427419662,
      "num_tokens": 27968051.0,
      "step": 8365
    },
    {
      "entropy": 6.360006427764892,
      "epoch": 1.6329138620622379,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004765414002687104,
      "loss": 6.1928,
      "mean_token_accuracy": 0.14350012689828873,
      "num_tokens": 27984942.0,
      "step": 8370
    },
    {
      "entropy": 6.468754386901855,
      "epoch": 1.6338893766461808,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004765101213174239,
      "loss": 6.366,
      "mean_token_accuracy": 0.13041064590215684,
      "num_tokens": 28001861.0,
      "step": 8375
    },
    {
      "entropy": 6.424246120452881,
      "epoch": 1.6348648912301238,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047647882267528663,
      "loss": 6.2949,
      "mean_token_accuracy": 0.13332895785570145,
      "num_tokens": 28018115.0,
      "step": 8380
    },
    {
      "entropy": 6.418905878067017,
      "epoch": 1.6358404058140668,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004764475043453569,
      "loss": 6.2472,
      "mean_token_accuracy": 0.1332973450422287,
      "num_tokens": 28033929.0,
      "step": 8385
    },
    {
      "entropy": 6.26256685256958,
      "epoch": 1.63681592039801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004764161663306949,
      "loss": 6.1147,
      "mean_token_accuracy": 0.1467380255460739,
      "num_tokens": 28050876.0,
      "step": 8390
    },
    {
      "entropy": 6.467622995376587,
      "epoch": 1.6377914349819531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004763848086343631,
      "loss": 6.3535,
      "mean_token_accuracy": 0.1269426614046097,
      "num_tokens": 28067453.0,
      "step": 8395
    },
    {
      "entropy": 6.412557125091553,
      "epoch": 1.638766949565896,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004763534312594254,
      "loss": 6.3419,
      "mean_token_accuracy": 0.13129981309175492,
      "num_tokens": 28084463.0,
      "step": 8400
    },
    {
      "entropy": 6.410643529891968,
      "epoch": 1.639742464149839,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004763220342089482,
      "loss": 6.273,
      "mean_token_accuracy": 0.14637115970253944,
      "num_tokens": 28101126.0,
      "step": 8405
    },
    {
      "entropy": 6.39175329208374,
      "epoch": 1.640717978733782,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047629061748599927,
      "loss": 6.3022,
      "mean_token_accuracy": 0.13546456322073935,
      "num_tokens": 28117000.0,
      "step": 8410
    },
    {
      "entropy": 6.421971416473388,
      "epoch": 1.641693493317725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047625918109364863,
      "loss": 6.3385,
      "mean_token_accuracy": 0.12772865816950799,
      "num_tokens": 28132290.0,
      "step": 8415
    },
    {
      "entropy": 6.39679069519043,
      "epoch": 1.6426690079016681,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047622772503496806,
      "loss": 6.211,
      "mean_token_accuracy": 0.13242074400186538,
      "num_tokens": 28149884.0,
      "step": 8420
    },
    {
      "entropy": 6.359272146224976,
      "epoch": 1.6436445224856113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047619624931303145,
      "loss": 6.2256,
      "mean_token_accuracy": 0.14724080562591552,
      "num_tokens": 28166678.0,
      "step": 8425
    },
    {
      "entropy": 6.329800415039062,
      "epoch": 1.6446200370695543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004761647539309144,
      "loss": 6.2645,
      "mean_token_accuracy": 0.14528129920363425,
      "num_tokens": 28184921.0,
      "step": 8430
    },
    {
      "entropy": 6.36128044128418,
      "epoch": 1.6455955516534972,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047613323889169457,
      "loss": 6.2571,
      "mean_token_accuracy": 0.13667764365673066,
      "num_tokens": 28203503.0,
      "step": 8435
    },
    {
      "entropy": 6.440892887115479,
      "epoch": 1.6465710662374402,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004761017041984515,
      "loss": 6.3202,
      "mean_token_accuracy": 0.13588420152664185,
      "num_tokens": 28220542.0,
      "step": 8440
    },
    {
      "entropy": 6.427672290802002,
      "epoch": 1.6475465808213832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760701498542666,
      "loss": 6.2514,
      "mean_token_accuracy": 0.13742470145225524,
      "num_tokens": 28236680.0,
      "step": 8445
    },
    {
      "entropy": 6.3082459449768065,
      "epoch": 1.6485220954053263,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047603857586222333,
      "loss": 6.1099,
      "mean_token_accuracy": 0.1449614241719246,
      "num_tokens": 28253690.0,
      "step": 8450
    },
    {
      "entropy": 6.303468847274781,
      "epoch": 1.6494976099892693,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047600698222540687,
      "loss": 6.1525,
      "mean_token_accuracy": 0.14698651432991028,
      "num_tokens": 28270528.0,
      "step": 8455
    },
    {
      "entropy": 6.378865766525268,
      "epoch": 1.6504731245732125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004759753689469046,
      "loss": 6.2513,
      "mean_token_accuracy": 0.13398033007979393,
      "num_tokens": 28287470.0,
      "step": 8460
    },
    {
      "entropy": 6.412626791000366,
      "epoch": 1.6514486391571555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004759437360298055,
      "loss": 6.2881,
      "mean_token_accuracy": 0.1377560831606388,
      "num_tokens": 28304295.0,
      "step": 8465
    },
    {
      "entropy": 6.499598550796509,
      "epoch": 1.6524241537410984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047591208347720073,
      "loss": 6.3922,
      "mean_token_accuracy": 0.1281437322497368,
      "num_tokens": 28319669.0,
      "step": 8470
    },
    {
      "entropy": 6.379131937026978,
      "epoch": 1.6533996683250414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047588041129218323,
      "loss": 6.3586,
      "mean_token_accuracy": 0.1306584119796753,
      "num_tokens": 28336251.0,
      "step": 8475
    },
    {
      "entropy": 6.456495189666748,
      "epoch": 1.6543751829089843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047584871947784787,
      "loss": 6.3748,
      "mean_token_accuracy": 0.13596502915024758,
      "num_tokens": 28352916.0,
      "step": 8480
    },
    {
      "entropy": 6.4545073986053465,
      "epoch": 1.6553506974929275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004758170080372915,
      "loss": 6.2615,
      "mean_token_accuracy": 0.1311369962990284,
      "num_tokens": 28369452.0,
      "step": 8485
    },
    {
      "entropy": 6.3804045677185055,
      "epoch": 1.6563262120768707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047578527697361277,
      "loss": 6.226,
      "mean_token_accuracy": 0.14323751628398895,
      "num_tokens": 28384932.0,
      "step": 8490
    },
    {
      "entropy": 6.419934558868408,
      "epoch": 1.6573017266608137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004757535262899124,
      "loss": 6.2606,
      "mean_token_accuracy": 0.13889929726719857,
      "num_tokens": 28400460.0,
      "step": 8495
    },
    {
      "entropy": 6.356053113937378,
      "epoch": 1.6582772412447566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000475721755989293,
      "loss": 6.3089,
      "mean_token_accuracy": 0.13253170475363732,
      "num_tokens": 28417883.0,
      "step": 8500
    },
    {
      "entropy": 6.402004337310791,
      "epoch": 1.6592527558286996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047568996607485895,
      "loss": 6.2338,
      "mean_token_accuracy": 0.13264339715242385,
      "num_tokens": 28434657.0,
      "step": 8505
    },
    {
      "entropy": 6.398932981491089,
      "epoch": 1.6602282704126425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004756581565497167,
      "loss": 6.3515,
      "mean_token_accuracy": 0.13040871620178224,
      "num_tokens": 28452767.0,
      "step": 8510
    },
    {
      "entropy": 6.532329654693603,
      "epoch": 1.6612037849965857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047562632741697454,
      "loss": 6.3889,
      "mean_token_accuracy": 0.12642064690589905,
      "num_tokens": 28469470.0,
      "step": 8515
    },
    {
      "entropy": 6.377254676818848,
      "epoch": 1.6621792995805287,
      "grad_norm": 1.0,
      "learning_rate": 0.0004755944786797427,
      "loss": 6.2603,
      "mean_token_accuracy": 0.1357783257961273,
      "num_tokens": 28486767.0,
      "step": 8520
    },
    {
      "entropy": 6.441311502456665,
      "epoch": 1.6631548141644719,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004755626103411333,
      "loss": 6.3238,
      "mean_token_accuracy": 0.1310509905219078,
      "num_tokens": 28504324.0,
      "step": 8525
    },
    {
      "entropy": 6.3576733589172365,
      "epoch": 1.6641303287484148,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004755307224042605,
      "loss": 6.154,
      "mean_token_accuracy": 0.14575303643941878,
      "num_tokens": 28520774.0,
      "step": 8530
    },
    {
      "entropy": 6.3152790546417235,
      "epoch": 1.6651058433323578,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004754988148722402,
      "loss": 6.2477,
      "mean_token_accuracy": 0.13989638090133666,
      "num_tokens": 28536754.0,
      "step": 8535
    },
    {
      "entropy": 6.355329370498657,
      "epoch": 1.6660813579163007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754668877481903,
      "loss": 6.1688,
      "mean_token_accuracy": 0.14407580345869064,
      "num_tokens": 28552866.0,
      "step": 8540
    },
    {
      "entropy": 6.437228870391846,
      "epoch": 1.667056872500244,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754349410352306,
      "loss": 6.3802,
      "mean_token_accuracy": 0.12448089197278023,
      "num_tokens": 28568802.0,
      "step": 8545
    },
    {
      "entropy": 6.40399661064148,
      "epoch": 1.6680323870841869,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004754029747364828,
      "loss": 6.224,
      "mean_token_accuracy": 0.1343241773545742,
      "num_tokens": 28585050.0,
      "step": 8550
    },
    {
      "entropy": 6.345958709716797,
      "epoch": 1.66900790166813,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047537098885507055,
      "loss": 6.2649,
      "mean_token_accuracy": 0.13776325136423112,
      "num_tokens": 28602319.0,
      "step": 8555
    },
    {
      "entropy": 6.411238956451416,
      "epoch": 1.669983416252073,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047533898339411945,
      "loss": 6.2788,
      "mean_token_accuracy": 0.14414341151714324,
      "num_tokens": 28620694.0,
      "step": 8560
    },
    {
      "entropy": 6.284890127182007,
      "epoch": 1.670958930836016,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004753069583567569,
      "loss": 6.0696,
      "mean_token_accuracy": 0.1466538690030575,
      "num_tokens": 28638309.0,
      "step": 8565
    },
    {
      "entropy": 6.3787047386169435,
      "epoch": 1.671934445419959,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004752749137461123,
      "loss": 6.2541,
      "mean_token_accuracy": 0.13430361002683638,
      "num_tokens": 28654713.0,
      "step": 8570
    },
    {
      "entropy": 6.407508087158203,
      "epoch": 1.672909960003902,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004752428495653169,
      "loss": 6.2379,
      "mean_token_accuracy": 0.13870712518692016,
      "num_tokens": 28671025.0,
      "step": 8575
    },
    {
      "entropy": 6.286981678009033,
      "epoch": 1.673885474587845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047521076581750387,
      "loss": 6.1693,
      "mean_token_accuracy": 0.14008468091487886,
      "num_tokens": 28687230.0,
      "step": 8580
    },
    {
      "entropy": 6.377354574203491,
      "epoch": 1.6748609891717883,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751786625058085,
      "loss": 6.2798,
      "mean_token_accuracy": 0.13172709941864014,
      "num_tokens": 28702595.0,
      "step": 8585
    },
    {
      "entropy": 6.336711406707764,
      "epoch": 1.6758365037557312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004751465396333676,
      "loss": 6.2414,
      "mean_token_accuracy": 0.14484235644340515,
      "num_tokens": 28719656.0,
      "step": 8590
    },
    {
      "entropy": 6.169897556304932,
      "epoch": 1.6768120183396742,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047511439720332015,
      "loss": 5.9504,
      "mean_token_accuracy": 0.16674025282263755,
      "num_tokens": 28737289.0,
      "step": 8595
    },
    {
      "entropy": 6.37326226234436,
      "epoch": 1.6777875329236172,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047508223521880704,
      "loss": 6.3164,
      "mean_token_accuracy": 0.13486414551734924,
      "num_tokens": 28752214.0,
      "step": 8600
    },
    {
      "entropy": 6.333926963806152,
      "epoch": 1.6787630475075601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750500536829711,
      "loss": 6.2151,
      "mean_token_accuracy": 0.13687286749482155,
      "num_tokens": 28769845.0,
      "step": 8605
    },
    {
      "entropy": 6.485809803009033,
      "epoch": 1.6797385620915033,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047501785259895676,
      "loss": 6.3282,
      "mean_token_accuracy": 0.1273742325603962,
      "num_tokens": 28786273.0,
      "step": 8610
    },
    {
      "entropy": 6.353741598129273,
      "epoch": 1.6807140766754463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004749856319699108,
      "loss": 6.2843,
      "mean_token_accuracy": 0.1325018547475338,
      "num_tokens": 28803171.0,
      "step": 8615
    },
    {
      "entropy": 6.376746368408203,
      "epoch": 1.6816895912593894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004749533917989817,
      "loss": 6.2407,
      "mean_token_accuracy": 0.14074417501688002,
      "num_tokens": 28819585.0,
      "step": 8620
    },
    {
      "entropy": 6.297998857498169,
      "epoch": 1.6826651058433324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047492113208931974,
      "loss": 6.1316,
      "mean_token_accuracy": 0.15067402198910712,
      "num_tokens": 28837158.0,
      "step": 8625
    },
    {
      "entropy": 6.405483675003052,
      "epoch": 1.6836406204272754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047488885284407734,
      "loss": 6.2147,
      "mean_token_accuracy": 0.13543615043163298,
      "num_tokens": 28852804.0,
      "step": 8630
    },
    {
      "entropy": 6.403007221221924,
      "epoch": 1.6846161350112183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004748565540664086,
      "loss": 6.3439,
      "mean_token_accuracy": 0.13092776611447335,
      "num_tokens": 28870336.0,
      "step": 8635
    },
    {
      "entropy": 6.407288026809693,
      "epoch": 1.6855916495951615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004748242357594697,
      "loss": 6.2115,
      "mean_token_accuracy": 0.13852638527750968,
      "num_tokens": 28886197.0,
      "step": 8640
    },
    {
      "entropy": 6.3193339824676515,
      "epoch": 1.6865671641791045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004747918979264186,
      "loss": 6.174,
      "mean_token_accuracy": 0.13869947642087938,
      "num_tokens": 28902520.0,
      "step": 8645
    },
    {
      "entropy": 6.388109302520752,
      "epoch": 1.6875426787630476,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004747595405704154,
      "loss": 6.2747,
      "mean_token_accuracy": 0.13255705535411835,
      "num_tokens": 28918725.0,
      "step": 8650
    },
    {
      "entropy": 6.436076402664185,
      "epoch": 1.6885181933469906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004747271636946219,
      "loss": 6.2992,
      "mean_token_accuracy": 0.13122606948018073,
      "num_tokens": 28936001.0,
      "step": 8655
    },
    {
      "entropy": 6.485187768936157,
      "epoch": 1.6894937079309336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746947673022018,
      "loss": 6.4034,
      "mean_token_accuracy": 0.12993389517068862,
      "num_tokens": 28953832.0,
      "step": 8660
    },
    {
      "entropy": 6.436713075637817,
      "epoch": 1.6904692225148765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004746623513963206,
      "loss": 6.2318,
      "mean_token_accuracy": 0.13691591396927832,
      "num_tokens": 28970982.0,
      "step": 8665
    },
    {
      "entropy": 6.332325649261475,
      "epoch": 1.6914447370988195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047462991598014624,
      "loss": 6.2225,
      "mean_token_accuracy": 0.13032605722546578,
      "num_tokens": 28986142.0,
      "step": 8670
    },
    {
      "entropy": 6.352346181869507,
      "epoch": 1.6924202516827627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047459746105684786,
      "loss": 6.2413,
      "mean_token_accuracy": 0.13961000591516495,
      "num_tokens": 29002244.0,
      "step": 8675
    },
    {
      "entropy": 6.326392126083374,
      "epoch": 1.6933957662667058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047456498662959694,
      "loss": 6.2201,
      "mean_token_accuracy": 0.1499812848865986,
      "num_tokens": 29018667.0,
      "step": 8680
    },
    {
      "entropy": 6.415907955169677,
      "epoch": 1.6943712808506488,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004745324927015669,
      "loss": 6.2865,
      "mean_token_accuracy": 0.13229371160268782,
      "num_tokens": 29034203.0,
      "step": 8685
    },
    {
      "entropy": 6.297095155715942,
      "epoch": 1.6953467954345918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004744999792759327,
      "loss": 6.174,
      "mean_token_accuracy": 0.14022865146398544,
      "num_tokens": 29050181.0,
      "step": 8690
    },
    {
      "entropy": 6.429923248291016,
      "epoch": 1.6963223100185347,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004744674463558717,
      "loss": 6.3231,
      "mean_token_accuracy": 0.1312961086630821,
      "num_tokens": 29066822.0,
      "step": 8695
    },
    {
      "entropy": 6.380949401855469,
      "epoch": 1.6972978246024777,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047443489394456256,
      "loss": 6.2216,
      "mean_token_accuracy": 0.14204866960644721,
      "num_tokens": 29084148.0,
      "step": 8700
    },
    {
      "entropy": 6.40456862449646,
      "epoch": 1.6982733391864209,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047440232204518654,
      "loss": 6.3319,
      "mean_token_accuracy": 0.1307839848101139,
      "num_tokens": 29099811.0,
      "step": 8705
    },
    {
      "entropy": 6.437453556060791,
      "epoch": 1.6992488537703638,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004743697306609262,
      "loss": 6.232,
      "mean_token_accuracy": 0.13557403087615966,
      "num_tokens": 29115327.0,
      "step": 8710
    },
    {
      "entropy": 6.308728313446045,
      "epoch": 1.700224368354307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004743371197949664,
      "loss": 6.2415,
      "mean_token_accuracy": 0.13114385977387427,
      "num_tokens": 29132571.0,
      "step": 8715
    },
    {
      "entropy": 6.3755100727081295,
      "epoch": 1.70119988293825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430448945049363,
      "loss": 6.2462,
      "mean_token_accuracy": 0.13620246797800065,
      "num_tokens": 29149256.0,
      "step": 8720
    },
    {
      "entropy": 6.366237592697144,
      "epoch": 1.702175397522193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004742718396306966,
      "loss": 6.1581,
      "mean_token_accuracy": 0.14344330206513406,
      "num_tokens": 29166380.0,
      "step": 8725
    },
    {
      "entropy": 6.301610040664673,
      "epoch": 1.703150912106136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004742391703387656,
      "loss": 6.2227,
      "mean_token_accuracy": 0.14369737282395362,
      "num_tokens": 29183056.0,
      "step": 8730
    },
    {
      "entropy": 6.330959129333496,
      "epoch": 1.7041264266900789,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000474206481577893,
      "loss": 6.2306,
      "mean_token_accuracy": 0.1394687719643116,
      "num_tokens": 29199631.0,
      "step": 8735
    },
    {
      "entropy": 6.388570642471313,
      "epoch": 1.705101941274022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004741737733512729,
      "loss": 6.2563,
      "mean_token_accuracy": 0.13335228487849235,
      "num_tokens": 29215074.0,
      "step": 8740
    },
    {
      "entropy": 6.364251565933228,
      "epoch": 1.7060774558579652,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047414104566210157,
      "loss": 6.1691,
      "mean_token_accuracy": 0.1413283884525299,
      "num_tokens": 29232022.0,
      "step": 8745
    },
    {
      "entropy": 6.278027200698853,
      "epoch": 1.7070529704419082,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000474108298513577,
      "loss": 6.0641,
      "mean_token_accuracy": 0.14640870988368987,
      "num_tokens": 29248040.0,
      "step": 8750
    },
    {
      "entropy": 6.421377801895142,
      "epoch": 1.7080284850258511,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047407553190889916,
      "loss": 6.2998,
      "mean_token_accuracy": 0.13976552039384843,
      "num_tokens": 29264338.0,
      "step": 8755
    },
    {
      "entropy": 6.274025869369507,
      "epoch": 1.709003999609794,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004740427458512699,
      "loss": 6.1392,
      "mean_token_accuracy": 0.14241230189800264,
      "num_tokens": 29279483.0,
      "step": 8760
    },
    {
      "entropy": 6.374361944198609,
      "epoch": 1.709979514193737,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004740099403438928,
      "loss": 6.2529,
      "mean_token_accuracy": 0.13598695024847984,
      "num_tokens": 29296582.0,
      "step": 8765
    },
    {
      "entropy": 6.426046705245971,
      "epoch": 1.7109550287776802,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047397711538997366,
      "loss": 6.3659,
      "mean_token_accuracy": 0.12885866463184356,
      "num_tokens": 29312772.0,
      "step": 8770
    },
    {
      "entropy": 6.324779605865478,
      "epoch": 1.7119305433616232,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047394427099272003,
      "loss": 6.1794,
      "mean_token_accuracy": 0.14214837327599525,
      "num_tokens": 29328962.0,
      "step": 8775
    },
    {
      "entropy": 6.435729932785034,
      "epoch": 1.7129060579455664,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047391140715534127,
      "loss": 6.2789,
      "mean_token_accuracy": 0.13365039676427842,
      "num_tokens": 29346299.0,
      "step": 8780
    },
    {
      "entropy": 6.341035032272339,
      "epoch": 1.7138815725295093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004738785238810487,
      "loss": 6.1938,
      "mean_token_accuracy": 0.14215406402945518,
      "num_tokens": 29362729.0,
      "step": 8785
    },
    {
      "entropy": 6.380382776260376,
      "epoch": 1.7148570871134523,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738456211730556,
      "loss": 6.3137,
      "mean_token_accuracy": 0.13468472957611083,
      "num_tokens": 29380807.0,
      "step": 8790
    },
    {
      "entropy": 6.330184698104858,
      "epoch": 1.7158326016973953,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004738126990345771,
      "loss": 6.1959,
      "mean_token_accuracy": 0.14188284277915955,
      "num_tokens": 29397261.0,
      "step": 8795
    },
    {
      "entropy": 6.340368843078613,
      "epoch": 1.7168081162813384,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047377975746883025,
      "loss": 6.2486,
      "mean_token_accuracy": 0.13668973222374917,
      "num_tokens": 29413710.0,
      "step": 8800
    },
    {
      "entropy": 6.387281656265259,
      "epoch": 1.7177836308652814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047374679647903397,
      "loss": 6.1965,
      "mean_token_accuracy": 0.1373312681913376,
      "num_tokens": 29430178.0,
      "step": 8805
    },
    {
      "entropy": 6.34645414352417,
      "epoch": 1.7187591454492246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004737138160684091,
      "loss": 6.2864,
      "mean_token_accuracy": 0.13263029158115386,
      "num_tokens": 29447602.0,
      "step": 8810
    },
    {
      "entropy": 6.454414749145508,
      "epoch": 1.7197346600331676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004736808162401783,
      "loss": 6.3203,
      "mean_token_accuracy": 0.13160987794399262,
      "num_tokens": 29463868.0,
      "step": 8815
    },
    {
      "entropy": 6.427574157714844,
      "epoch": 1.7207101746171105,
      "grad_norm": 1.0,
      "learning_rate": 0.00047364779699756625,
      "loss": 6.2376,
      "mean_token_accuracy": 0.13990029245615004,
      "num_tokens": 29480263.0,
      "step": 8820
    },
    {
      "entropy": 6.118601226806641,
      "epoch": 1.7216856892010535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047361475834379945,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1775626242160797,
      "num_tokens": 29498938.0,
      "step": 8825
    },
    {
      "entropy": 6.362480878829956,
      "epoch": 1.7226612037849964,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004735817002821064,
      "loss": 6.2501,
      "mean_token_accuracy": 0.1337668113410473,
      "num_tokens": 29515522.0,
      "step": 8830
    },
    {
      "entropy": 6.327173519134521,
      "epoch": 1.7236367183689396,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047354862281571736,
      "loss": 6.199,
      "mean_token_accuracy": 0.14215150848031044,
      "num_tokens": 29531050.0,
      "step": 8835
    },
    {
      "entropy": 6.297528648376465,
      "epoch": 1.7246122329528828,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004735155259478645,
      "loss": 6.1176,
      "mean_token_accuracy": 0.1464240923523903,
      "num_tokens": 29547432.0,
      "step": 8840
    },
    {
      "entropy": 6.421542978286743,
      "epoch": 1.7255877475368258,
      "grad_norm": 1.171875,
      "learning_rate": 0.000473482409681782,
      "loss": 6.227,
      "mean_token_accuracy": 0.1370633840560913,
      "num_tokens": 29564272.0,
      "step": 8845
    },
    {
      "entropy": 6.4304361820220945,
      "epoch": 1.7265632621207687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047344927402070573,
      "loss": 6.2938,
      "mean_token_accuracy": 0.12798682898283004,
      "num_tokens": 29580433.0,
      "step": 8850
    },
    {
      "entropy": 6.246017026901245,
      "epoch": 1.7275387767047117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004734161189678738,
      "loss": 6.1386,
      "mean_token_accuracy": 0.14591725841164588,
      "num_tokens": 29597602.0,
      "step": 8855
    },
    {
      "entropy": 6.3113514423370365,
      "epoch": 1.7285142912886546,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047338294452652584,
      "loss": 6.1195,
      "mean_token_accuracy": 0.1445278435945511,
      "num_tokens": 29616313.0,
      "step": 8860
    },
    {
      "entropy": 6.320663690567017,
      "epoch": 1.7294898058725978,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047334975069990363,
      "loss": 6.2097,
      "mean_token_accuracy": 0.1431247040629387,
      "num_tokens": 29633786.0,
      "step": 8865
    },
    {
      "entropy": 6.398162221908569,
      "epoch": 1.7304653204565408,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004733165374912507,
      "loss": 6.282,
      "mean_token_accuracy": 0.1356651686131954,
      "num_tokens": 29650125.0,
      "step": 8870
    },
    {
      "entropy": 6.437296056747437,
      "epoch": 1.731440835040484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047328330490381255,
      "loss": 6.2957,
      "mean_token_accuracy": 0.1291682615876198,
      "num_tokens": 29667211.0,
      "step": 8875
    },
    {
      "entropy": 6.465518236160278,
      "epoch": 1.732416349624427,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047325005294083654,
      "loss": 6.3405,
      "mean_token_accuracy": 0.12921203672885895,
      "num_tokens": 29683864.0,
      "step": 8880
    },
    {
      "entropy": 6.439604377746582,
      "epoch": 1.7333918642083699,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047321678160557187,
      "loss": 6.3617,
      "mean_token_accuracy": 0.12742238715291024,
      "num_tokens": 29699275.0,
      "step": 8885
    },
    {
      "entropy": 6.353477716445923,
      "epoch": 1.7343673787923128,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047318349090126984,
      "loss": 6.2338,
      "mean_token_accuracy": 0.13567274510860444,
      "num_tokens": 29715773.0,
      "step": 8890
    },
    {
      "entropy": 6.4024557113647464,
      "epoch": 1.735342893376256,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047315018083118334,
      "loss": 6.2053,
      "mean_token_accuracy": 0.14300538823008538,
      "num_tokens": 29732383.0,
      "step": 8895
    },
    {
      "entropy": 6.370939493179321,
      "epoch": 1.736318407960199,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004731168513985674,
      "loss": 6.2094,
      "mean_token_accuracy": 0.13631932139396669,
      "num_tokens": 29749159.0,
      "step": 8900
    },
    {
      "entropy": 6.289877605438233,
      "epoch": 1.7372939225441422,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047308350260667894,
      "loss": 6.1479,
      "mean_token_accuracy": 0.14565806165337564,
      "num_tokens": 29765357.0,
      "step": 8905
    },
    {
      "entropy": 6.320123958587646,
      "epoch": 1.7382694371280851,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004730501344587765,
      "loss": 6.2096,
      "mean_token_accuracy": 0.13831088989973067,
      "num_tokens": 29781269.0,
      "step": 8910
    },
    {
      "entropy": 6.427883338928223,
      "epoch": 1.739244951712028,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004730167469581208,
      "loss": 6.2767,
      "mean_token_accuracy": 0.13175969943404198,
      "num_tokens": 29798029.0,
      "step": 8915
    },
    {
      "entropy": 6.319770765304566,
      "epoch": 1.740220466295971,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004729833401079743,
      "loss": 6.212,
      "mean_token_accuracy": 0.13645870387554168,
      "num_tokens": 29813794.0,
      "step": 8920
    },
    {
      "entropy": 6.448766851425171,
      "epoch": 1.741195980879914,
      "grad_norm": 1.125,
      "learning_rate": 0.00047294991391160147,
      "loss": 6.3648,
      "mean_token_accuracy": 0.12680112272500993,
      "num_tokens": 29829589.0,
      "step": 8925
    },
    {
      "entropy": 6.449286460876465,
      "epoch": 1.7421714954638572,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004729164683722685,
      "loss": 6.2904,
      "mean_token_accuracy": 0.1385537125170231,
      "num_tokens": 29845861.0,
      "step": 8930
    },
    {
      "entropy": 6.430850076675415,
      "epoch": 1.7431470100478004,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004728830034932437,
      "loss": 6.299,
      "mean_token_accuracy": 0.1321383498609066,
      "num_tokens": 29861711.0,
      "step": 8935
    },
    {
      "entropy": 6.354513120651245,
      "epoch": 1.7441225246317433,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047284951927779704,
      "loss": 6.2816,
      "mean_token_accuracy": 0.12844536751508712,
      "num_tokens": 29878679.0,
      "step": 8940
    },
    {
      "entropy": 6.4125590324401855,
      "epoch": 1.7450980392156863,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047281601572920053,
      "loss": 6.2569,
      "mean_token_accuracy": 0.1351689837872982,
      "num_tokens": 29895016.0,
      "step": 8945
    },
    {
      "entropy": 6.235859680175781,
      "epoch": 1.7460735537996293,
      "grad_norm": 1.171875,
      "learning_rate": 0.000472782492850728,
      "loss": 6.1031,
      "mean_token_accuracy": 0.14930107593536376,
      "num_tokens": 29913115.0,
      "step": 8950
    },
    {
      "entropy": 6.2470814228057865,
      "epoch": 1.7470490683835722,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004727489506456551,
      "loss": 6.1306,
      "mean_token_accuracy": 0.1530684068799019,
      "num_tokens": 29928954.0,
      "step": 8955
    },
    {
      "entropy": 6.357276868820191,
      "epoch": 1.7480245829675154,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047271538911725964,
      "loss": 6.2673,
      "mean_token_accuracy": 0.13633049726486207,
      "num_tokens": 29945077.0,
      "step": 8960
    },
    {
      "entropy": 6.4214966773986815,
      "epoch": 1.7490000975514584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000472681808268821,
      "loss": 6.3342,
      "mean_token_accuracy": 0.12676829397678374,
      "num_tokens": 29962480.0,
      "step": 8965
    },
    {
      "entropy": 6.3567873477935795,
      "epoch": 1.7499756121354015,
      "grad_norm": 1.125,
      "learning_rate": 0.00047264820810362054,
      "loss": 6.1742,
      "mean_token_accuracy": 0.1451256290078163,
      "num_tokens": 29978761.0,
      "step": 8970
    },
    {
      "entropy": 6.427098035812378,
      "epoch": 1.7509511267193445,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047261458862494165,
      "loss": 6.3249,
      "mean_token_accuracy": 0.13136909604072572,
      "num_tokens": 29995836.0,
      "step": 8975
    },
    {
      "entropy": 6.168182754516602,
      "epoch": 1.7519266413032875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004725809498360694,
      "loss": 5.9366,
      "mean_token_accuracy": 0.17042837142944336,
      "num_tokens": 30014221.0,
      "step": 8980
    },
    {
      "entropy": 6.407944059371948,
      "epoch": 1.7529021558872304,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472547291740291,
      "loss": 6.3688,
      "mean_token_accuracy": 0.12758428305387498,
      "num_tokens": 30030952.0,
      "step": 8985
    },
    {
      "entropy": 6.206259536743164,
      "epoch": 1.7538776704711734,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047251361434089536,
      "loss": 6.0493,
      "mean_token_accuracy": 0.15662642121315,
      "num_tokens": 30047916.0,
      "step": 8990
    },
    {
      "entropy": 6.392978525161743,
      "epoch": 1.7548531850551166,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004724799176411732,
      "loss": 6.288,
      "mean_token_accuracy": 0.13475642278790473,
      "num_tokens": 30063238.0,
      "step": 8995
    },
    {
      "entropy": 6.404321813583374,
      "epoch": 1.7558286996390597,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724462016444174,
      "loss": 6.231,
      "mean_token_accuracy": 0.1345588356256485,
      "num_tokens": 30079079.0,
      "step": 9000
    },
    {
      "epoch": 1.7558286996390597,
      "eval_entropy": 6.285039651008602,
      "eval_loss": 6.323532581329346,
      "eval_mean_token_accuracy": 0.13831896030645524,
      "eval_num_tokens": 30079079.0,
      "eval_runtime": 15.2881,
      "eval_samples_per_second": 1084.306,
      "eval_steps_per_second": 135.595,
      "step": 9000
    },
    {
      "entropy": 6.346852159500122,
      "epoch": 1.7568042142230027,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047241246635392246,
      "loss": 6.1933,
      "mean_token_accuracy": 0.14282382354140283,
      "num_tokens": 30095165.0,
      "step": 9005
    },
    {
      "entropy": 6.3754219055175785,
      "epoch": 1.7577797288069457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004723787117729848,
      "loss": 6.2685,
      "mean_token_accuracy": 0.1358795329928398,
      "num_tokens": 30111415.0,
      "step": 9010
    },
    {
      "entropy": 6.380832242965698,
      "epoch": 1.7587552433908886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047234493790490297,
      "loss": 6.2693,
      "mean_token_accuracy": 0.1371316984295845,
      "num_tokens": 30127363.0,
      "step": 9015
    },
    {
      "entropy": 6.418465948104858,
      "epoch": 1.7597307579748316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004723111447529771,
      "loss": 6.2441,
      "mean_token_accuracy": 0.13629745915532113,
      "num_tokens": 30144139.0,
      "step": 9020
    },
    {
      "entropy": 6.177949380874634,
      "epoch": 1.7607062725587748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722773323205095,
      "loss": 6.0456,
      "mean_token_accuracy": 0.15519603043794633,
      "num_tokens": 30160522.0,
      "step": 9025
    },
    {
      "entropy": 6.297994709014892,
      "epoch": 1.761681787142718,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000472243500610804,
      "loss": 6.2256,
      "mean_token_accuracy": 0.14158434867858888,
      "num_tokens": 30178140.0,
      "step": 9030
    },
    {
      "entropy": 6.447716426849365,
      "epoch": 1.762657301726661,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047220964962716675,
      "loss": 6.3119,
      "mean_token_accuracy": 0.13332033902406693,
      "num_tokens": 30194867.0,
      "step": 9035
    },
    {
      "entropy": 6.364599180221558,
      "epoch": 1.7636328163106039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047217577937290527,
      "loss": 6.2395,
      "mean_token_accuracy": 0.13923368155956267,
      "num_tokens": 30211966.0,
      "step": 9040
    },
    {
      "entropy": 6.300136756896973,
      "epoch": 1.7646083308945468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004721418898513294,
      "loss": 6.0954,
      "mean_token_accuracy": 0.1464352510869503,
      "num_tokens": 30229230.0,
      "step": 9045
    },
    {
      "entropy": 6.355046129226684,
      "epoch": 1.7655838454784898,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004721079810657507,
      "loss": 6.1763,
      "mean_token_accuracy": 0.14177559837698936,
      "num_tokens": 30244793.0,
      "step": 9050
    },
    {
      "entropy": 6.37604603767395,
      "epoch": 1.766559360062433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047207405301948264,
      "loss": 6.2814,
      "mean_token_accuracy": 0.1320166677236557,
      "num_tokens": 30260574.0,
      "step": 9055
    },
    {
      "entropy": 6.377372455596924,
      "epoch": 1.767534874646376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047204010571584045,
      "loss": 6.1765,
      "mean_token_accuracy": 0.13665746226906778,
      "num_tokens": 30277080.0,
      "step": 9060
    },
    {
      "entropy": 6.20311017036438,
      "epoch": 1.7685103892303191,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004720061391581414,
      "loss": 6.0421,
      "mean_token_accuracy": 0.1571509212255478,
      "num_tokens": 30293282.0,
      "step": 9065
    },
    {
      "entropy": 6.279948663711548,
      "epoch": 1.769485903814262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047197215334970454,
      "loss": 6.1873,
      "mean_token_accuracy": 0.14281286671757698,
      "num_tokens": 30310268.0,
      "step": 9070
    },
    {
      "entropy": 6.359273481369018,
      "epoch": 1.770461418398205,
      "grad_norm": 1.125,
      "learning_rate": 0.0004719381482938509,
      "loss": 6.2342,
      "mean_token_accuracy": 0.13754354789853096,
      "num_tokens": 30327173.0,
      "step": 9075
    },
    {
      "entropy": 6.393432521820069,
      "epoch": 1.771436932982148,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047190412399390335,
      "loss": 6.2684,
      "mean_token_accuracy": 0.1412390388548374,
      "num_tokens": 30343798.0,
      "step": 9080
    },
    {
      "entropy": 6.295350217819214,
      "epoch": 1.772412447566091,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047187008045318653,
      "loss": 6.1407,
      "mean_token_accuracy": 0.14180871695280076,
      "num_tokens": 30359710.0,
      "step": 9085
    },
    {
      "entropy": 6.38989896774292,
      "epoch": 1.7733879621500341,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004718360176750271,
      "loss": 6.2904,
      "mean_token_accuracy": 0.13518032133579255,
      "num_tokens": 30376017.0,
      "step": 9090
    },
    {
      "entropy": 6.36014404296875,
      "epoch": 1.7743634767339773,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004718019356627536,
      "loss": 6.1585,
      "mean_token_accuracy": 0.13946156874299048,
      "num_tokens": 30392805.0,
      "step": 9095
    },
    {
      "entropy": 6.3737743377685545,
      "epoch": 1.7753389913179203,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004717678344196963,
      "loss": 6.2421,
      "mean_token_accuracy": 0.13712184578180314,
      "num_tokens": 30409165.0,
      "step": 9100
    },
    {
      "entropy": 6.415667247772217,
      "epoch": 1.7763145059018632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047173371394918764,
      "loss": 6.2832,
      "mean_token_accuracy": 0.1383858159184456,
      "num_tokens": 30425186.0,
      "step": 9105
    },
    {
      "entropy": 6.384480142593384,
      "epoch": 1.7772900204858062,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004716995742545615,
      "loss": 6.2524,
      "mean_token_accuracy": 0.1387595720589161,
      "num_tokens": 30442396.0,
      "step": 9110
    },
    {
      "entropy": 6.442374038696289,
      "epoch": 1.7782655350697492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716654153391541,
      "loss": 6.4013,
      "mean_token_accuracy": 0.13012437224388124,
      "num_tokens": 30459166.0,
      "step": 9115
    },
    {
      "entropy": 6.288897132873535,
      "epoch": 1.7792410496536923,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004716312372063032,
      "loss": 6.1112,
      "mean_token_accuracy": 0.14386690706014632,
      "num_tokens": 30475226.0,
      "step": 9120
    },
    {
      "entropy": 6.4428516864776615,
      "epoch": 1.7802165642376353,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004715970398593486,
      "loss": 6.235,
      "mean_token_accuracy": 0.1290210336446762,
      "num_tokens": 30491649.0,
      "step": 9125
    },
    {
      "entropy": 6.331375598907471,
      "epoch": 1.7811920788215785,
      "grad_norm": 1.125,
      "learning_rate": 0.0004715628233016319,
      "loss": 6.2121,
      "mean_token_accuracy": 0.14169322624802588,
      "num_tokens": 30507499.0,
      "step": 9130
    },
    {
      "entropy": 6.29231367111206,
      "epoch": 1.7821675934055214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047152858753649675,
      "loss": 6.1869,
      "mean_token_accuracy": 0.1391509160399437,
      "num_tokens": 30524962.0,
      "step": 9135
    },
    {
      "entropy": 6.307570457458496,
      "epoch": 1.7831431079894644,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004714943325672884,
      "loss": 6.1784,
      "mean_token_accuracy": 0.1402794137597084,
      "num_tokens": 30540633.0,
      "step": 9140
    },
    {
      "entropy": 6.434923982620239,
      "epoch": 1.7841186225734074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004714600583973542,
      "loss": 6.2652,
      "mean_token_accuracy": 0.13567701503634452,
      "num_tokens": 30558018.0,
      "step": 9145
    },
    {
      "entropy": 6.307761716842651,
      "epoch": 1.7850941371573505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047142576503004336,
      "loss": 6.2096,
      "mean_token_accuracy": 0.1436252474784851,
      "num_tokens": 30574870.0,
      "step": 9150
    },
    {
      "entropy": 6.465546798706055,
      "epoch": 1.7860696517412935,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047139145246870686,
      "loss": 6.3069,
      "mean_token_accuracy": 0.12781654745340348,
      "num_tokens": 30590602.0,
      "step": 9155
    },
    {
      "entropy": 6.329585552215576,
      "epoch": 1.7870451663252367,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047135712071669743,
      "loss": 6.1539,
      "mean_token_accuracy": 0.14278057515621184,
      "num_tokens": 30607616.0,
      "step": 9160
    },
    {
      "entropy": 6.423104476928711,
      "epoch": 1.7880206809091796,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047132276977737014,
      "loss": 6.27,
      "mean_token_accuracy": 0.13649666607379912,
      "num_tokens": 30624288.0,
      "step": 9165
    },
    {
      "entropy": 6.256053876876831,
      "epoch": 1.7889961954931226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047128839965408136,
      "loss": 6.0941,
      "mean_token_accuracy": 0.15076804161071777,
      "num_tokens": 30640556.0,
      "step": 9170
    },
    {
      "entropy": 6.308404731750488,
      "epoch": 1.7899717100770656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047125401035018983,
      "loss": 6.2005,
      "mean_token_accuracy": 0.14093147069215775,
      "num_tokens": 30658377.0,
      "step": 9175
    },
    {
      "entropy": 6.405527353286743,
      "epoch": 1.7909472246610085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004712196018690559,
      "loss": 6.2843,
      "mean_token_accuracy": 0.1324254274368286,
      "num_tokens": 30674424.0,
      "step": 9180
    },
    {
      "entropy": 6.430455684661865,
      "epoch": 1.7919227392449517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047118517421404173,
      "loss": 6.2882,
      "mean_token_accuracy": 0.13330901488661767,
      "num_tokens": 30690945.0,
      "step": 9185
    },
    {
      "entropy": 6.446741533279419,
      "epoch": 1.792898253828895,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047115072738851155,
      "loss": 6.2716,
      "mean_token_accuracy": 0.14572949558496476,
      "num_tokens": 30707024.0,
      "step": 9190
    },
    {
      "entropy": 6.4110229969024655,
      "epoch": 1.7938737684128379,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047111626139583146,
      "loss": 6.3685,
      "mean_token_accuracy": 0.12633953541517257,
      "num_tokens": 30723888.0,
      "step": 9195
    },
    {
      "entropy": 6.408418226242065,
      "epoch": 1.7948492829967808,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047108177623936926,
      "loss": 6.2184,
      "mean_token_accuracy": 0.1409653089940548,
      "num_tokens": 30740721.0,
      "step": 9200
    },
    {
      "entropy": 6.353799724578858,
      "epoch": 1.7958247975807238,
      "grad_norm": 1.0,
      "learning_rate": 0.0004710472719224947,
      "loss": 6.207,
      "mean_token_accuracy": 0.14333074390888215,
      "num_tokens": 30758378.0,
      "step": 9205
    },
    {
      "entropy": 6.438936853408814,
      "epoch": 1.7968003121646667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047101274844857946,
      "loss": 6.2282,
      "mean_token_accuracy": 0.1395811066031456,
      "num_tokens": 30774380.0,
      "step": 9210
    },
    {
      "entropy": 6.355694103240967,
      "epoch": 1.79777582674861,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004709782058209971,
      "loss": 6.2789,
      "mean_token_accuracy": 0.13273947983980178,
      "num_tokens": 30792051.0,
      "step": 9215
    },
    {
      "entropy": 6.35457968711853,
      "epoch": 1.7987513413325529,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004709436440431229,
      "loss": 6.2234,
      "mean_token_accuracy": 0.13932136222720146,
      "num_tokens": 30808892.0,
      "step": 9220
    },
    {
      "entropy": 6.380823612213135,
      "epoch": 1.799726855916496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047090906311833417,
      "loss": 6.2497,
      "mean_token_accuracy": 0.13484877422451974,
      "num_tokens": 30825248.0,
      "step": 9225
    },
    {
      "entropy": 6.38541693687439,
      "epoch": 1.800702370500439,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047087446305001,
      "loss": 6.213,
      "mean_token_accuracy": 0.14358601719141006,
      "num_tokens": 30842388.0,
      "step": 9230
    },
    {
      "entropy": 6.311140727996826,
      "epoch": 1.801677885084382,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004708398438415314,
      "loss": 6.1752,
      "mean_token_accuracy": 0.13990469127893448,
      "num_tokens": 30859240.0,
      "step": 9235
    },
    {
      "entropy": 6.39827790260315,
      "epoch": 1.802653399668325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004708052054962813,
      "loss": 6.2859,
      "mean_token_accuracy": 0.13488940745592118,
      "num_tokens": 30875085.0,
      "step": 9240
    },
    {
      "entropy": 6.441133499145508,
      "epoch": 1.8036289142522681,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047077054801764436,
      "loss": 6.3158,
      "mean_token_accuracy": 0.13391539305448533,
      "num_tokens": 30890913.0,
      "step": 9245
    },
    {
      "entropy": 6.454874515533447,
      "epoch": 1.804604428836211,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047073587140900714,
      "loss": 6.3239,
      "mean_token_accuracy": 0.13015394583344458,
      "num_tokens": 30906354.0,
      "step": 9250
    },
    {
      "entropy": 6.316405010223389,
      "epoch": 1.8055799434201543,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004707011756737583,
      "loss": 6.2261,
      "mean_token_accuracy": 0.13662536963820457,
      "num_tokens": 30923389.0,
      "step": 9255
    },
    {
      "entropy": 6.368812513351441,
      "epoch": 1.8065554580040972,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004706664608152879,
      "loss": 6.2004,
      "mean_token_accuracy": 0.1318556122481823,
      "num_tokens": 30940445.0,
      "step": 9260
    },
    {
      "entropy": 6.400604724884033,
      "epoch": 1.8075309725880402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047063172683698844,
      "loss": 6.1976,
      "mean_token_accuracy": 0.13614652827382087,
      "num_tokens": 30956084.0,
      "step": 9265
    },
    {
      "entropy": 6.35695595741272,
      "epoch": 1.8085064871719831,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047059697374225387,
      "loss": 6.3001,
      "mean_token_accuracy": 0.13228397965431213,
      "num_tokens": 30973546.0,
      "step": 9270
    },
    {
      "entropy": 6.3678735256195065,
      "epoch": 1.809482001755926,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004705622015344801,
      "loss": 6.2413,
      "mean_token_accuracy": 0.13858142048120498,
      "num_tokens": 30991713.0,
      "step": 9275
    },
    {
      "entropy": 6.386630201339722,
      "epoch": 1.8104575163398693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000470527410217065,
      "loss": 6.1615,
      "mean_token_accuracy": 0.14019320905208588,
      "num_tokens": 31007682.0,
      "step": 9280
    },
    {
      "entropy": 6.276372289657592,
      "epoch": 1.8114330309238125,
      "grad_norm": 1.125,
      "learning_rate": 0.0004704925997934083,
      "loss": 6.1267,
      "mean_token_accuracy": 0.1400565668940544,
      "num_tokens": 31023238.0,
      "step": 9285
    },
    {
      "entropy": 6.403650093078613,
      "epoch": 1.8124085455077554,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004704577702669114,
      "loss": 6.2133,
      "mean_token_accuracy": 0.14183202832937242,
      "num_tokens": 31038891.0,
      "step": 9290
    },
    {
      "entropy": 6.407066917419433,
      "epoch": 1.8133840600916984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047042292164097796,
      "loss": 6.3187,
      "mean_token_accuracy": 0.13228942155838014,
      "num_tokens": 31055718.0,
      "step": 9295
    },
    {
      "entropy": 6.508831024169922,
      "epoch": 1.8143595746756414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000470388053919013,
      "loss": 6.3316,
      "mean_token_accuracy": 0.13479144275188445,
      "num_tokens": 31071574.0,
      "step": 9300
    },
    {
      "entropy": 6.414669036865234,
      "epoch": 1.8153350892595843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047035316710442385,
      "loss": 6.3036,
      "mean_token_accuracy": 0.12938049286603928,
      "num_tokens": 31087391.0,
      "step": 9305
    },
    {
      "entropy": 6.293859052658081,
      "epoch": 1.8163106038435275,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047031826120061946,
      "loss": 6.1853,
      "mean_token_accuracy": 0.14075756818056107,
      "num_tokens": 31105290.0,
      "step": 9310
    },
    {
      "entropy": 6.396039438247681,
      "epoch": 1.8172861184274705,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047028333621101063,
      "loss": 6.2804,
      "mean_token_accuracy": 0.13696015104651452,
      "num_tokens": 31121229.0,
      "step": 9315
    },
    {
      "entropy": 6.28475866317749,
      "epoch": 1.8182616330114136,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047024839213901027,
      "loss": 6.1128,
      "mean_token_accuracy": 0.1487472116947174,
      "num_tokens": 31139306.0,
      "step": 9320
    },
    {
      "entropy": 6.285208749771118,
      "epoch": 1.8192371475953566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047021342898803296,
      "loss": 6.1328,
      "mean_token_accuracy": 0.14296377450227737,
      "num_tokens": 31155154.0,
      "step": 9325
    },
    {
      "entropy": 6.402400064468384,
      "epoch": 1.8202126621792996,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047017844676149496,
      "loss": 6.3007,
      "mean_token_accuracy": 0.13751576393842696,
      "num_tokens": 31172816.0,
      "step": 9330
    },
    {
      "entropy": 6.472377157211303,
      "epoch": 1.8211881767632425,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047014344546281487,
      "loss": 6.2912,
      "mean_token_accuracy": 0.12994184345006943,
      "num_tokens": 31189152.0,
      "step": 9335
    },
    {
      "entropy": 6.426492834091187,
      "epoch": 1.8221636913471855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047010842509541276,
      "loss": 6.308,
      "mean_token_accuracy": 0.136446662992239,
      "num_tokens": 31205201.0,
      "step": 9340
    },
    {
      "entropy": 6.359533166885376,
      "epoch": 1.8231392059311287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004700733856627107,
      "loss": 6.1977,
      "mean_token_accuracy": 0.13919030651450157,
      "num_tokens": 31222071.0,
      "step": 9345
    },
    {
      "entropy": 6.388296031951905,
      "epoch": 1.8241147205150718,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004700383271681328,
      "loss": 6.264,
      "mean_token_accuracy": 0.13403671085834504,
      "num_tokens": 31237717.0,
      "step": 9350
    },
    {
      "entropy": 6.336244869232178,
      "epoch": 1.8250902350990148,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047000324961510455,
      "loss": 6.2647,
      "mean_token_accuracy": 0.1383986383676529,
      "num_tokens": 31255113.0,
      "step": 9355
    },
    {
      "entropy": 6.3697772979736325,
      "epoch": 1.8260657496829578,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004699681530070538,
      "loss": 6.2024,
      "mean_token_accuracy": 0.14559535831212997,
      "num_tokens": 31271246.0,
      "step": 9360
    },
    {
      "entropy": 6.516678428649902,
      "epoch": 1.8270412642669007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046993303734740996,
      "loss": 6.3658,
      "mean_token_accuracy": 0.12384085208177567,
      "num_tokens": 31289079.0,
      "step": 9365
    },
    {
      "entropy": 6.276464033126831,
      "epoch": 1.8280167788508437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046989790263960454,
      "loss": 6.0836,
      "mean_token_accuracy": 0.15658457279205323,
      "num_tokens": 31305432.0,
      "step": 9370
    },
    {
      "entropy": 6.360106754302978,
      "epoch": 1.8289922934347869,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046986274888707066,
      "loss": 6.2917,
      "mean_token_accuracy": 0.13438301309943199,
      "num_tokens": 31321683.0,
      "step": 9375
    },
    {
      "entropy": 6.411686372756958,
      "epoch": 1.8299678080187298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004698275760932435,
      "loss": 6.2508,
      "mean_token_accuracy": 0.13539228588342667,
      "num_tokens": 31337978.0,
      "step": 9380
    },
    {
      "entropy": 6.303026866912842,
      "epoch": 1.830943322602673,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004697923842615599,
      "loss": 6.2034,
      "mean_token_accuracy": 0.13995525613427162,
      "num_tokens": 31354426.0,
      "step": 9385
    },
    {
      "entropy": 6.324839019775391,
      "epoch": 1.831918837186616,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004697571733954588,
      "loss": 6.1466,
      "mean_token_accuracy": 0.14470979422330857,
      "num_tokens": 31372295.0,
      "step": 9390
    },
    {
      "entropy": 6.276195669174195,
      "epoch": 1.832894351770559,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004697219434983809,
      "loss": 6.1439,
      "mean_token_accuracy": 0.14759137630462646,
      "num_tokens": 31388915.0,
      "step": 9395
    },
    {
      "entropy": 6.451128959655762,
      "epoch": 1.8338698663545019,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046968669457376865,
      "loss": 6.3359,
      "mean_token_accuracy": 0.13097282126545906,
      "num_tokens": 31404934.0,
      "step": 9400
    },
    {
      "entropy": 6.40408730506897,
      "epoch": 1.834845380938445,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046965142662506646,
      "loss": 6.25,
      "mean_token_accuracy": 0.13547894582152367,
      "num_tokens": 31421481.0,
      "step": 9405
    },
    {
      "entropy": 6.405406713485718,
      "epoch": 1.835820895522388,
      "grad_norm": 1.25,
      "learning_rate": 0.0004696161396557206,
      "loss": 6.1725,
      "mean_token_accuracy": 0.14359065145254135,
      "num_tokens": 31437155.0,
      "step": 9410
    },
    {
      "entropy": 6.3343878269195555,
      "epoch": 1.8367964101063312,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046958083366917925,
      "loss": 6.1391,
      "mean_token_accuracy": 0.14139294847846032,
      "num_tokens": 31453303.0,
      "step": 9415
    },
    {
      "entropy": 6.36326584815979,
      "epoch": 1.8377719246902742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046954550866889236,
      "loss": 6.254,
      "mean_token_accuracy": 0.13932962864637374,
      "num_tokens": 31468818.0,
      "step": 9420
    },
    {
      "entropy": 6.400947856903076,
      "epoch": 1.8387474392742171,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004695101646583116,
      "loss": 6.2708,
      "mean_token_accuracy": 0.12851447761058807,
      "num_tokens": 31485967.0,
      "step": 9425
    },
    {
      "entropy": 6.37727632522583,
      "epoch": 1.83972295385816,
      "grad_norm": 1.140625,
      "learning_rate": 0.000469474801640891,
      "loss": 6.1822,
      "mean_token_accuracy": 0.1372745543718338,
      "num_tokens": 31501938.0,
      "step": 9430
    },
    {
      "entropy": 6.391381692886353,
      "epoch": 1.840698468442103,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046943941962008574,
      "loss": 6.2157,
      "mean_token_accuracy": 0.13603852987289428,
      "num_tokens": 31517793.0,
      "step": 9435
    },
    {
      "entropy": 6.440922594070434,
      "epoch": 1.8416739830260462,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004694040185993534,
      "loss": 6.2911,
      "mean_token_accuracy": 0.13641856536269187,
      "num_tokens": 31534186.0,
      "step": 9440
    },
    {
      "entropy": 6.389259624481201,
      "epoch": 1.8426494976099894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004693685985821532,
      "loss": 6.2304,
      "mean_token_accuracy": 0.13665117770433427,
      "num_tokens": 31551460.0,
      "step": 9445
    },
    {
      "entropy": 6.214830923080444,
      "epoch": 1.8436250121939324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046933315957194637,
      "loss": 6.0166,
      "mean_token_accuracy": 0.15803005471825599,
      "num_tokens": 31569403.0,
      "step": 9450
    },
    {
      "entropy": 6.3573816299438475,
      "epoch": 1.8446005267778753,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004692977015721958,
      "loss": 6.2073,
      "mean_token_accuracy": 0.13621388226747513,
      "num_tokens": 31586007.0,
      "step": 9455
    },
    {
      "entropy": 6.495966863632202,
      "epoch": 1.8455760413618183,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004692622245863663,
      "loss": 6.3091,
      "mean_token_accuracy": 0.12903623953461646,
      "num_tokens": 31603421.0,
      "step": 9460
    },
    {
      "entropy": 6.3327678680419925,
      "epoch": 1.8465515559457613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046922672861792456,
      "loss": 6.2325,
      "mean_token_accuracy": 0.13628780841827393,
      "num_tokens": 31620173.0,
      "step": 9465
    },
    {
      "entropy": 6.371530961990357,
      "epoch": 1.8475270705297044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046919121367033917,
      "loss": 6.1887,
      "mean_token_accuracy": 0.14214526414871215,
      "num_tokens": 31636820.0,
      "step": 9470
    },
    {
      "entropy": 6.3811363697052,
      "epoch": 1.8485025851136474,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004691556797470805,
      "loss": 6.2538,
      "mean_token_accuracy": 0.13522107601165773,
      "num_tokens": 31655099.0,
      "step": 9475
    },
    {
      "entropy": 6.411267709732056,
      "epoch": 1.8494780996975906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046912012685162074,
      "loss": 6.2579,
      "mean_token_accuracy": 0.13759189695119858,
      "num_tokens": 31671875.0,
      "step": 9480
    },
    {
      "entropy": 6.310659408569336,
      "epoch": 1.8504536142815335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004690845549874341,
      "loss": 6.1954,
      "mean_token_accuracy": 0.13684373199939728,
      "num_tokens": 31688542.0,
      "step": 9485
    },
    {
      "entropy": 6.321006393432617,
      "epoch": 1.8514291288654765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004690489641579964,
      "loss": 6.1783,
      "mean_token_accuracy": 0.14119212925434113,
      "num_tokens": 31705139.0,
      "step": 9490
    },
    {
      "entropy": 6.470886945724487,
      "epoch": 1.8524046434494195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046901335436678554,
      "loss": 6.3411,
      "mean_token_accuracy": 0.13040006384253502,
      "num_tokens": 31723038.0,
      "step": 9495
    },
    {
      "entropy": 6.323464012145996,
      "epoch": 1.8533801580333626,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004689777256172812,
      "loss": 6.1855,
      "mean_token_accuracy": 0.14331147596240043,
      "num_tokens": 31738920.0,
      "step": 9500
    },
    {
      "entropy": 6.274996566772461,
      "epoch": 1.8543556726173056,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046894207791296484,
      "loss": 6.152,
      "mean_token_accuracy": 0.14287813007831573,
      "num_tokens": 31755093.0,
      "step": 9505
    },
    {
      "entropy": 6.321342039108276,
      "epoch": 1.8553311872012488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046890641125731987,
      "loss": 6.1455,
      "mean_token_accuracy": 0.14784156680107116,
      "num_tokens": 31775364.0,
      "step": 9510
    },
    {
      "entropy": 6.4158190250396725,
      "epoch": 1.8563067017851917,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046887072565383154,
      "loss": 6.2757,
      "mean_token_accuracy": 0.1332124277949333,
      "num_tokens": 31790462.0,
      "step": 9515
    },
    {
      "entropy": 6.463978815078735,
      "epoch": 1.8572822163691347,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004688350211059868,
      "loss": 6.2891,
      "mean_token_accuracy": 0.13250985965132714,
      "num_tokens": 31806813.0,
      "step": 9520
    },
    {
      "entropy": 6.439868402481079,
      "epoch": 1.8582577309530777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004687992976172746,
      "loss": 6.2896,
      "mean_token_accuracy": 0.13312311545014383,
      "num_tokens": 31823592.0,
      "step": 9525
    },
    {
      "entropy": 6.294678115844727,
      "epoch": 1.8592332455370206,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046876355519118586,
      "loss": 6.1036,
      "mean_token_accuracy": 0.1491539590060711,
      "num_tokens": 31840110.0,
      "step": 9530
    },
    {
      "entropy": 6.333635711669922,
      "epoch": 1.8602087601209638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000468727793831213,
      "loss": 6.1768,
      "mean_token_accuracy": 0.1426729254424572,
      "num_tokens": 31858338.0,
      "step": 9535
    },
    {
      "entropy": 6.4334649562835695,
      "epoch": 1.861184274704907,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046869201354085056,
      "loss": 6.2598,
      "mean_token_accuracy": 0.13301713466644288,
      "num_tokens": 31874793.0,
      "step": 9540
    },
    {
      "entropy": 6.340746259689331,
      "epoch": 1.86215978928885,
      "grad_norm": 1.0,
      "learning_rate": 0.00046865621432359497,
      "loss": 6.2397,
      "mean_token_accuracy": 0.13510825708508492,
      "num_tokens": 31891339.0,
      "step": 9545
    },
    {
      "entropy": 6.458423757553101,
      "epoch": 1.863135303872793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046862039618294425,
      "loss": 6.3487,
      "mean_token_accuracy": 0.12797670885920526,
      "num_tokens": 31908407.0,
      "step": 9550
    },
    {
      "entropy": 6.386204719543457,
      "epoch": 1.8641108184567359,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004685845591223985,
      "loss": 6.2097,
      "mean_token_accuracy": 0.13810239657759665,
      "num_tokens": 31923653.0,
      "step": 9555
    },
    {
      "entropy": 6.206319189071655,
      "epoch": 1.8650863330406788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046854870314545966,
      "loss": 6.0817,
      "mean_token_accuracy": 0.15029369592666625,
      "num_tokens": 31940531.0,
      "step": 9560
    },
    {
      "entropy": 6.282375431060791,
      "epoch": 1.866061847624622,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004685128282556314,
      "loss": 6.0862,
      "mean_token_accuracy": 0.1443892516195774,
      "num_tokens": 31958730.0,
      "step": 9565
    },
    {
      "entropy": 6.43106484413147,
      "epoch": 1.867037362208565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004684769344564191,
      "loss": 6.2625,
      "mean_token_accuracy": 0.135828697681427,
      "num_tokens": 31976200.0,
      "step": 9570
    },
    {
      "entropy": 6.464182138442993,
      "epoch": 1.8680128767925082,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046844102175133033,
      "loss": 6.3699,
      "mean_token_accuracy": 0.12933961302042007,
      "num_tokens": 31992791.0,
      "step": 9575
    },
    {
      "entropy": 6.383312559127807,
      "epoch": 1.8689883913764511,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046840509014387443,
      "loss": 6.167,
      "mean_token_accuracy": 0.1364608347415924,
      "num_tokens": 32008936.0,
      "step": 9580
    },
    {
      "entropy": 6.3490701675415036,
      "epoch": 1.869963905960394,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004683691396375625,
      "loss": 6.2142,
      "mean_token_accuracy": 0.13272910118103026,
      "num_tokens": 32024651.0,
      "step": 9585
    },
    {
      "entropy": 6.458617353439331,
      "epoch": 1.870939420544337,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004683331702359073,
      "loss": 6.2832,
      "mean_token_accuracy": 0.1338637463748455,
      "num_tokens": 32040879.0,
      "step": 9590
    },
    {
      "entropy": 6.385991430282592,
      "epoch": 1.8719149351282802,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046829718194242375,
      "loss": 6.2507,
      "mean_token_accuracy": 0.13922645300626754,
      "num_tokens": 32058624.0,
      "step": 9595
    },
    {
      "entropy": 6.28926157951355,
      "epoch": 1.8728904497122232,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004682611747606286,
      "loss": 6.1247,
      "mean_token_accuracy": 0.14653316289186477,
      "num_tokens": 32074594.0,
      "step": 9600
    },
    {
      "entropy": 6.4140979766845705,
      "epoch": 1.8738659642961664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004682251486940402,
      "loss": 6.2316,
      "mean_token_accuracy": 0.13104119896888733,
      "num_tokens": 32090431.0,
      "step": 9605
    },
    {
      "entropy": 6.32953109741211,
      "epoch": 1.8748414788801093,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046818910374617895,
      "loss": 6.1894,
      "mean_token_accuracy": 0.13525758311152458,
      "num_tokens": 32108353.0,
      "step": 9610
    },
    {
      "entropy": 6.351568126678467,
      "epoch": 1.8758169934640523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046815303992056703,
      "loss": 6.2337,
      "mean_token_accuracy": 0.13460186943411828,
      "num_tokens": 32125747.0,
      "step": 9615
    },
    {
      "entropy": 6.346704483032227,
      "epoch": 1.8767925080479952,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004681169572207284,
      "loss": 6.1434,
      "mean_token_accuracy": 0.1397470325231552,
      "num_tokens": 32141578.0,
      "step": 9620
    },
    {
      "entropy": 6.381634283065796,
      "epoch": 1.8777680226319382,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046808085565018905,
      "loss": 6.3142,
      "mean_token_accuracy": 0.133141028881073,
      "num_tokens": 32158939.0,
      "step": 9625
    },
    {
      "entropy": 6.403763389587402,
      "epoch": 1.8787435372158814,
      "grad_norm": 1.625,
      "learning_rate": 0.0004680447352124766,
      "loss": 6.2463,
      "mean_token_accuracy": 0.13726267218589783,
      "num_tokens": 32175355.0,
      "step": 9630
    },
    {
      "entropy": 6.356957960128784,
      "epoch": 1.8797190517998246,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046800859591112064,
      "loss": 6.2581,
      "mean_token_accuracy": 0.14053439125418662,
      "num_tokens": 32190827.0,
      "step": 9635
    },
    {
      "entropy": 6.470404434204101,
      "epoch": 1.8806945663837675,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046797243774965265,
      "loss": 6.292,
      "mean_token_accuracy": 0.13623698726296424,
      "num_tokens": 32207911.0,
      "step": 9640
    },
    {
      "entropy": 6.4029449939727785,
      "epoch": 1.8816700809677105,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004679362607316057,
      "loss": 6.231,
      "mean_token_accuracy": 0.14180153235793114,
      "num_tokens": 32225200.0,
      "step": 9645
    },
    {
      "entropy": 6.280690050125122,
      "epoch": 1.8826455955516535,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004679000648605151,
      "loss": 6.1903,
      "mean_token_accuracy": 0.14931774511933327,
      "num_tokens": 32241933.0,
      "step": 9650
    },
    {
      "entropy": 6.337716817855835,
      "epoch": 1.8836211101355964,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004678638501399176,
      "loss": 6.2202,
      "mean_token_accuracy": 0.13590962663292885,
      "num_tokens": 32257856.0,
      "step": 9655
    },
    {
      "entropy": 6.3901588916778564,
      "epoch": 1.8845966247195396,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046782761657335203,
      "loss": 6.1573,
      "mean_token_accuracy": 0.13776347935199737,
      "num_tokens": 32273973.0,
      "step": 9660
    },
    {
      "entropy": 6.385250759124756,
      "epoch": 1.8855721393034826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004677913641643591,
      "loss": 6.2816,
      "mean_token_accuracy": 0.1324045330286026,
      "num_tokens": 32289731.0,
      "step": 9665
    },
    {
      "entropy": 6.397674179077148,
      "epoch": 1.8865476538874257,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004677550929164812,
      "loss": 6.2249,
      "mean_token_accuracy": 0.13663273602724074,
      "num_tokens": 32305839.0,
      "step": 9670
    },
    {
      "entropy": 6.35627932548523,
      "epoch": 1.8875231684713687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046771880283326254,
      "loss": 6.1753,
      "mean_token_accuracy": 0.14824962764978408,
      "num_tokens": 32322393.0,
      "step": 9675
    },
    {
      "entropy": 6.284825563430786,
      "epoch": 1.8884986830553117,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004676824939182493,
      "loss": 6.1147,
      "mean_token_accuracy": 0.14091624543070794,
      "num_tokens": 32340535.0,
      "step": 9680
    },
    {
      "entropy": 6.403156280517578,
      "epoch": 1.8894741976392546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046764616617498957,
      "loss": 6.2581,
      "mean_token_accuracy": 0.13995739296078683,
      "num_tokens": 32357104.0,
      "step": 9685
    },
    {
      "entropy": 6.336461448669434,
      "epoch": 1.8904497122231976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000467609819607033,
      "loss": 6.2449,
      "mean_token_accuracy": 0.13785327076911927,
      "num_tokens": 32373969.0,
      "step": 9690
    },
    {
      "entropy": 6.427638816833496,
      "epoch": 1.8914252268071408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004675734542179313,
      "loss": 6.2688,
      "mean_token_accuracy": 0.13259026706218718,
      "num_tokens": 32391021.0,
      "step": 9695
    },
    {
      "entropy": 6.297204637527466,
      "epoch": 1.892400741391084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004675370700112382,
      "loss": 6.1666,
      "mean_token_accuracy": 0.1389855057001114,
      "num_tokens": 32407415.0,
      "step": 9700
    },
    {
      "entropy": 6.386851501464844,
      "epoch": 1.893376255975027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046750066699050865,
      "loss": 6.1727,
      "mean_token_accuracy": 0.1434981942176819,
      "num_tokens": 32424525.0,
      "step": 9705
    },
    {
      "entropy": 6.400507593154908,
      "epoch": 1.8943517705589699,
      "grad_norm": 1.125,
      "learning_rate": 0.00046746424515930004,
      "loss": 6.2526,
      "mean_token_accuracy": 0.13482319191098213,
      "num_tokens": 32440975.0,
      "step": 9710
    },
    {
      "entropy": 6.111890411376953,
      "epoch": 1.8953272851429128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004674278045211714,
      "loss": 5.9456,
      "mean_token_accuracy": 0.1616691991686821,
      "num_tokens": 32458021.0,
      "step": 9715
    },
    {
      "entropy": 6.3989557266235355,
      "epoch": 1.8963027997268558,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046739134507968347,
      "loss": 6.2666,
      "mean_token_accuracy": 0.1342753492295742,
      "num_tokens": 32475354.0,
      "step": 9720
    },
    {
      "entropy": 6.37829270362854,
      "epoch": 1.897278314310799,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046735486683839903,
      "loss": 6.2464,
      "mean_token_accuracy": 0.13810083791613578,
      "num_tokens": 32491686.0,
      "step": 9725
    },
    {
      "entropy": 6.399685335159302,
      "epoch": 1.898253828894742,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046731836980088256,
      "loss": 6.206,
      "mean_token_accuracy": 0.14393048137426376,
      "num_tokens": 32508785.0,
      "step": 9730
    },
    {
      "entropy": 6.302030801773071,
      "epoch": 1.899229343478685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046728185397070047,
      "loss": 6.1076,
      "mean_token_accuracy": 0.14788153693079947,
      "num_tokens": 32525558.0,
      "step": 9735
    },
    {
      "entropy": 6.428395175933838,
      "epoch": 1.900204858062628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046724531935142077,
      "loss": 6.2952,
      "mean_token_accuracy": 0.12745456770062447,
      "num_tokens": 32541743.0,
      "step": 9740
    },
    {
      "entropy": 6.396680212020874,
      "epoch": 1.901180372646571,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046720876594661374,
      "loss": 6.2258,
      "mean_token_accuracy": 0.13473589941859246,
      "num_tokens": 32558492.0,
      "step": 9745
    },
    {
      "entropy": 6.398730754852295,
      "epoch": 1.902155887230514,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671721937598512,
      "loss": 6.1723,
      "mean_token_accuracy": 0.14717136472463607,
      "num_tokens": 32574782.0,
      "step": 9750
    },
    {
      "entropy": 6.284476470947266,
      "epoch": 1.9031314018144572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004671356027947067,
      "loss": 6.231,
      "mean_token_accuracy": 0.139958006888628,
      "num_tokens": 32591630.0,
      "step": 9755
    },
    {
      "entropy": 6.215842819213867,
      "epoch": 1.9041069163984001,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004670989930547559,
      "loss": 6.1454,
      "mean_token_accuracy": 0.14397987499833106,
      "num_tokens": 32607388.0,
      "step": 9760
    },
    {
      "entropy": 6.460389041900635,
      "epoch": 1.9050824309823433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004670623645435762,
      "loss": 6.2757,
      "mean_token_accuracy": 0.13551186993718148,
      "num_tokens": 32624598.0,
      "step": 9765
    },
    {
      "entropy": 6.375860691070557,
      "epoch": 1.9060579455662863,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004670257172647467,
      "loss": 6.2296,
      "mean_token_accuracy": 0.13526087924838065,
      "num_tokens": 32641646.0,
      "step": 9770
    },
    {
      "entropy": 6.393792963027954,
      "epoch": 1.9070334601502292,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046698905122184855,
      "loss": 6.2903,
      "mean_token_accuracy": 0.13067851066589356,
      "num_tokens": 32659760.0,
      "step": 9775
    },
    {
      "entropy": 6.303044605255127,
      "epoch": 1.9080089747341722,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004669523664184646,
      "loss": 6.0982,
      "mean_token_accuracy": 0.15796980261802673,
      "num_tokens": 32675820.0,
      "step": 9780
    },
    {
      "entropy": 6.316472291946411,
      "epoch": 1.9089844893181152,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004669156628581794,
      "loss": 6.1721,
      "mean_token_accuracy": 0.14324838370084764,
      "num_tokens": 32694675.0,
      "step": 9785
    },
    {
      "entropy": 6.397135162353516,
      "epoch": 1.9099600039020583,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004668789405445797,
      "loss": 6.2536,
      "mean_token_accuracy": 0.13197917118668556,
      "num_tokens": 32710265.0,
      "step": 9790
    },
    {
      "entropy": 6.416831111907959,
      "epoch": 1.9109355184860015,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046684219948125397,
      "loss": 6.2312,
      "mean_token_accuracy": 0.1385997325181961,
      "num_tokens": 32726709.0,
      "step": 9795
    },
    {
      "entropy": 6.3734931468963625,
      "epoch": 1.9119110330699445,
      "grad_norm": 1.125,
      "learning_rate": 0.000466805439671792,
      "loss": 6.2224,
      "mean_token_accuracy": 0.13850449919700622,
      "num_tokens": 32742043.0,
      "step": 9800
    },
    {
      "entropy": 6.314431619644165,
      "epoch": 1.9128865476538874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046676866111978614,
      "loss": 6.1518,
      "mean_token_accuracy": 0.14299796372652054,
      "num_tokens": 32758932.0,
      "step": 9805
    },
    {
      "entropy": 6.382373237609864,
      "epoch": 1.9138620622378304,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046673186382883026,
      "loss": 6.2231,
      "mean_token_accuracy": 0.13647292852401732,
      "num_tokens": 32775012.0,
      "step": 9810
    },
    {
      "entropy": 6.332773447036743,
      "epoch": 1.9148375768217734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046669504780251985,
      "loss": 6.1925,
      "mean_token_accuracy": 0.137982015311718,
      "num_tokens": 32790312.0,
      "step": 9815
    },
    {
      "entropy": 6.35599913597107,
      "epoch": 1.9158130914057165,
      "grad_norm": 1.125,
      "learning_rate": 0.00046665821304445274,
      "loss": 6.2532,
      "mean_token_accuracy": 0.14164556041359902,
      "num_tokens": 32807926.0,
      "step": 9820
    },
    {
      "entropy": 6.434288787841797,
      "epoch": 1.9167886059896595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046662135955822804,
      "loss": 6.2354,
      "mean_token_accuracy": 0.13626089170575142,
      "num_tokens": 32825129.0,
      "step": 9825
    },
    {
      "entropy": 6.3448319911956785,
      "epoch": 1.9177641205736027,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046658448734744715,
      "loss": 6.1899,
      "mean_token_accuracy": 0.14191545769572259,
      "num_tokens": 32841688.0,
      "step": 9830
    },
    {
      "entropy": 6.390042018890381,
      "epoch": 1.9187396351575456,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004665475964157128,
      "loss": 6.2377,
      "mean_token_accuracy": 0.1458202987909317,
      "num_tokens": 32860004.0,
      "step": 9835
    },
    {
      "entropy": 6.3620946407318115,
      "epoch": 1.9197151497414886,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004665106867666301,
      "loss": 6.1692,
      "mean_token_accuracy": 0.14474282637238503,
      "num_tokens": 32877633.0,
      "step": 9840
    },
    {
      "entropy": 6.383384895324707,
      "epoch": 1.9206906643254316,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004664737584038057,
      "loss": 6.202,
      "mean_token_accuracy": 0.1403236374258995,
      "num_tokens": 32892826.0,
      "step": 9845
    },
    {
      "entropy": 6.313541650772095,
      "epoch": 1.9216661789093747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046643681133084797,
      "loss": 6.1761,
      "mean_token_accuracy": 0.14005391001701356,
      "num_tokens": 32909832.0,
      "step": 9850
    },
    {
      "entropy": 6.317526865005493,
      "epoch": 1.9226416934933177,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004663998455513673,
      "loss": 6.1817,
      "mean_token_accuracy": 0.14127086624503135,
      "num_tokens": 32926687.0,
      "step": 9855
    },
    {
      "entropy": 6.322162103652954,
      "epoch": 1.9236172080772609,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046636286106897595,
      "loss": 6.0982,
      "mean_token_accuracy": 0.15041553229093552,
      "num_tokens": 32944155.0,
      "step": 9860
    },
    {
      "entropy": 6.398593425750732,
      "epoch": 1.9245927226612038,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004663258578872879,
      "loss": 6.2145,
      "mean_token_accuracy": 0.1365725204348564,
      "num_tokens": 32961013.0,
      "step": 9865
    },
    {
      "entropy": 6.36764554977417,
      "epoch": 1.9255682372451468,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004662888360099189,
      "loss": 6.2899,
      "mean_token_accuracy": 0.13166995495557784,
      "num_tokens": 32977358.0,
      "step": 9870
    },
    {
      "entropy": 6.29514889717102,
      "epoch": 1.9265437518290898,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046625179544048653,
      "loss": 6.1055,
      "mean_token_accuracy": 0.14858441054821014,
      "num_tokens": 32994668.0,
      "step": 9875
    },
    {
      "entropy": 6.447812795639038,
      "epoch": 1.9275192664130327,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004662147361826103,
      "loss": 6.3143,
      "mean_token_accuracy": 0.12947703078389167,
      "num_tokens": 33009949.0,
      "step": 9880
    },
    {
      "entropy": 6.22262716293335,
      "epoch": 1.928494780996976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004661776582399117,
      "loss": 6.0483,
      "mean_token_accuracy": 0.1506297431886196,
      "num_tokens": 33026122.0,
      "step": 9885
    },
    {
      "entropy": 6.419974660873413,
      "epoch": 1.929470295580919,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046614056161601366,
      "loss": 6.2482,
      "mean_token_accuracy": 0.13921213895082474,
      "num_tokens": 33043406.0,
      "step": 9890
    },
    {
      "entropy": 6.302284526824951,
      "epoch": 1.930445810164862,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046610344631454114,
      "loss": 6.1294,
      "mean_token_accuracy": 0.14959441870450974,
      "num_tokens": 33058788.0,
      "step": 9895
    },
    {
      "entropy": 6.350957775115967,
      "epoch": 1.931421324748805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000466066312339121,
      "loss": 6.1851,
      "mean_token_accuracy": 0.13635668382048607,
      "num_tokens": 33076111.0,
      "step": 9900
    },
    {
      "entropy": 6.362673091888428,
      "epoch": 1.932396839332748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046602915969338175,
      "loss": 6.203,
      "mean_token_accuracy": 0.13774553090333938,
      "num_tokens": 33093531.0,
      "step": 9905
    },
    {
      "entropy": 6.328547048568725,
      "epoch": 1.933372353916691,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004659919883809538,
      "loss": 6.0969,
      "mean_token_accuracy": 0.1502293087542057,
      "num_tokens": 33110621.0,
      "step": 9910
    },
    {
      "entropy": 6.171562576293946,
      "epoch": 1.9343478685006341,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004659547984054695,
      "loss": 5.9983,
      "mean_token_accuracy": 0.1531696856021881,
      "num_tokens": 33126933.0,
      "step": 9915
    },
    {
      "entropy": 6.354141616821289,
      "epoch": 1.935323383084577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046591758977056294,
      "loss": 6.2723,
      "mean_token_accuracy": 0.13164597153663635,
      "num_tokens": 33143947.0,
      "step": 9920
    },
    {
      "entropy": 6.402196550369263,
      "epoch": 1.9362988976685203,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004658803624798699,
      "loss": 6.2841,
      "mean_token_accuracy": 0.13277989774942398,
      "num_tokens": 33160375.0,
      "step": 9925
    },
    {
      "entropy": 6.420433521270752,
      "epoch": 1.9372744122524632,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046584311653702815,
      "loss": 6.2727,
      "mean_token_accuracy": 0.1342118076980114,
      "num_tokens": 33177649.0,
      "step": 9930
    },
    {
      "entropy": 6.426245927810669,
      "epoch": 1.9382499268364062,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004658058519456772,
      "loss": 6.2757,
      "mean_token_accuracy": 0.13247256055474282,
      "num_tokens": 33194938.0,
      "step": 9935
    },
    {
      "entropy": 6.369714069366455,
      "epoch": 1.9392254414203491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004657685687094585,
      "loss": 6.1634,
      "mean_token_accuracy": 0.14209077954292298,
      "num_tokens": 33210049.0,
      "step": 9940
    },
    {
      "entropy": 6.373290014266968,
      "epoch": 1.940200956004292,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046573126683201513,
      "loss": 6.2329,
      "mean_token_accuracy": 0.1362062320113182,
      "num_tokens": 33227589.0,
      "step": 9945
    },
    {
      "entropy": 6.4011787414550785,
      "epoch": 1.9411764705882353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046569394631699215,
      "loss": 6.2077,
      "mean_token_accuracy": 0.13468170017004014,
      "num_tokens": 33245812.0,
      "step": 9950
    },
    {
      "entropy": 6.376532649993896,
      "epoch": 1.9421519851721785,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004656566071680364,
      "loss": 6.2351,
      "mean_token_accuracy": 0.13496036753058432,
      "num_tokens": 33261792.0,
      "step": 9955
    },
    {
      "entropy": 6.326299381256104,
      "epoch": 1.9431274997561214,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004656192493887965,
      "loss": 6.2173,
      "mean_token_accuracy": 0.14401627257466315,
      "num_tokens": 33279083.0,
      "step": 9960
    },
    {
      "entropy": 6.2178620338439945,
      "epoch": 1.9441030143400644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004655818729829229,
      "loss": 5.9912,
      "mean_token_accuracy": 0.15430462062358857,
      "num_tokens": 33296391.0,
      "step": 9965
    },
    {
      "entropy": 6.311451482772827,
      "epoch": 1.9450785289240073,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004655444779540679,
      "loss": 6.1097,
      "mean_token_accuracy": 0.13928199037909508,
      "num_tokens": 33313168.0,
      "step": 9970
    },
    {
      "entropy": 6.3628510475158695,
      "epoch": 1.9460540435079503,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046550706430588574,
      "loss": 6.2482,
      "mean_token_accuracy": 0.1349336735904217,
      "num_tokens": 33330156.0,
      "step": 9975
    },
    {
      "entropy": 6.333376407623291,
      "epoch": 1.9470295580918935,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046546963204203214,
      "loss": 6.1372,
      "mean_token_accuracy": 0.14500358700752258,
      "num_tokens": 33347496.0,
      "step": 9980
    },
    {
      "entropy": 6.338537216186523,
      "epoch": 1.9480050726758367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004654321811661649,
      "loss": 6.1907,
      "mean_token_accuracy": 0.13796720057725906,
      "num_tokens": 33364521.0,
      "step": 9985
    },
    {
      "entropy": 6.299841451644897,
      "epoch": 1.9489805872597796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004653947116819438,
      "loss": 6.1906,
      "mean_token_accuracy": 0.14173029959201813,
      "num_tokens": 33381902.0,
      "step": 9990
    },
    {
      "entropy": 6.394353866577148,
      "epoch": 1.9499561018437226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004653572235930299,
      "loss": 6.2842,
      "mean_token_accuracy": 0.13628909289836882,
      "num_tokens": 33397654.0,
      "step": 9995
    },
    {
      "entropy": 6.420040559768677,
      "epoch": 1.9509316164276655,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004653197169030866,
      "loss": 6.2582,
      "mean_token_accuracy": 0.13331133499741554,
      "num_tokens": 33415440.0,
      "step": 10000
    },
    {
      "entropy": 6.295147848129273,
      "epoch": 1.9519071310116085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046528219161577893,
      "loss": 6.107,
      "mean_token_accuracy": 0.14137528836727142,
      "num_tokens": 33431488.0,
      "step": 10005
    },
    {
      "entropy": 6.356692981719971,
      "epoch": 1.9528826455955517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046524464773477355,
      "loss": 6.0997,
      "mean_token_accuracy": 0.13928319811820983,
      "num_tokens": 33447750.0,
      "step": 10010
    },
    {
      "entropy": 6.2497326850891115,
      "epoch": 1.9538581601794947,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046520708526373935,
      "loss": 6.1518,
      "mean_token_accuracy": 0.14207966029644012,
      "num_tokens": 33464420.0,
      "step": 10015
    },
    {
      "entropy": 6.289768075942993,
      "epoch": 1.9548336747634378,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004651695042063467,
      "loss": 6.0429,
      "mean_token_accuracy": 0.1555169329047203,
      "num_tokens": 33482114.0,
      "step": 10020
    },
    {
      "entropy": 6.427407312393188,
      "epoch": 1.9558091893473808,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004651319045662678,
      "loss": 6.2508,
      "mean_token_accuracy": 0.13760205879807472,
      "num_tokens": 33500239.0,
      "step": 10025
    },
    {
      "entropy": 6.372778081893921,
      "epoch": 1.9567847039313238,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004650942863471768,
      "loss": 6.2543,
      "mean_token_accuracy": 0.13404427245259284,
      "num_tokens": 33516832.0,
      "step": 10030
    },
    {
      "entropy": 6.368900918960572,
      "epoch": 1.9577602185152667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046505664955274974,
      "loss": 6.1955,
      "mean_token_accuracy": 0.14626014232635498,
      "num_tokens": 33532899.0,
      "step": 10035
    },
    {
      "entropy": 6.396113014221191,
      "epoch": 1.9587357330992097,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650189941866643,
      "loss": 6.2427,
      "mean_token_accuracy": 0.13594066947698594,
      "num_tokens": 33548704.0,
      "step": 10040
    },
    {
      "entropy": 6.3285445213317875,
      "epoch": 1.9597112476831529,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004649813202525999,
      "loss": 6.153,
      "mean_token_accuracy": 0.14643149599432945,
      "num_tokens": 33566179.0,
      "step": 10045
    },
    {
      "entropy": 6.38486967086792,
      "epoch": 1.960686762267096,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004649436277542379,
      "loss": 6.2942,
      "mean_token_accuracy": 0.12989352270960808,
      "num_tokens": 33582504.0,
      "step": 10050
    },
    {
      "entropy": 6.416781234741211,
      "epoch": 1.961662276851039,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046490591669526174,
      "loss": 6.187,
      "mean_token_accuracy": 0.13434750437736512,
      "num_tokens": 33597931.0,
      "step": 10055
    },
    {
      "entropy": 6.329663562774658,
      "epoch": 1.962637791434982,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004648681870793561,
      "loss": 6.1601,
      "mean_token_accuracy": 0.14246033132076263,
      "num_tokens": 33615027.0,
      "step": 10060
    },
    {
      "entropy": 6.413044452667236,
      "epoch": 1.963613306018925,
      "grad_norm": 1.234375,
      "learning_rate": 0.000464830438910208,
      "loss": 6.2548,
      "mean_token_accuracy": 0.13771324083209038,
      "num_tokens": 33632484.0,
      "step": 10065
    },
    {
      "entropy": 6.409338760375976,
      "epoch": 1.9645888206028679,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046479267219150597,
      "loss": 6.2985,
      "mean_token_accuracy": 0.1315721906721592,
      "num_tokens": 33648889.0,
      "step": 10070
    },
    {
      "entropy": 6.397893571853638,
      "epoch": 1.965564335186811,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004647548869269404,
      "loss": 6.2074,
      "mean_token_accuracy": 0.1430997744202614,
      "num_tokens": 33666105.0,
      "step": 10075
    },
    {
      "entropy": 6.398637962341309,
      "epoch": 1.966539849770754,
      "grad_norm": 1.125,
      "learning_rate": 0.0004647170831202036,
      "loss": 6.2527,
      "mean_token_accuracy": 0.134390689432621,
      "num_tokens": 33683175.0,
      "step": 10080
    },
    {
      "entropy": 6.4589684009552,
      "epoch": 1.9675153643546972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004646792607749896,
      "loss": 6.2592,
      "mean_token_accuracy": 0.13042125329375268,
      "num_tokens": 33698202.0,
      "step": 10085
    },
    {
      "entropy": 6.334667491912842,
      "epoch": 1.9684908789386402,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046464141989499433,
      "loss": 6.1373,
      "mean_token_accuracy": 0.14383300244808198,
      "num_tokens": 33714573.0,
      "step": 10090
    },
    {
      "entropy": 6.294904375076294,
      "epoch": 1.9694663935225831,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004646035604839153,
      "loss": 6.1895,
      "mean_token_accuracy": 0.13591153845191,
      "num_tokens": 33731031.0,
      "step": 10095
    },
    {
      "entropy": 6.435849046707153,
      "epoch": 1.970441908106526,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004645656825454522,
      "loss": 6.2965,
      "mean_token_accuracy": 0.13325777053833007,
      "num_tokens": 33748228.0,
      "step": 10100
    },
    {
      "entropy": 6.340862178802491,
      "epoch": 1.9714174226904693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004645277860833062,
      "loss": 6.1487,
      "mean_token_accuracy": 0.13673630282282828,
      "num_tokens": 33765361.0,
      "step": 10105
    },
    {
      "entropy": 6.378815507888794,
      "epoch": 1.9723929372744122,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046448987110118034,
      "loss": 6.1712,
      "mean_token_accuracy": 0.14271750077605247,
      "num_tokens": 33781800.0,
      "step": 10110
    },
    {
      "entropy": 6.395051050186157,
      "epoch": 1.9733684518583554,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046445193760277976,
      "loss": 6.2959,
      "mean_token_accuracy": 0.136407832801342,
      "num_tokens": 33797280.0,
      "step": 10115
    },
    {
      "entropy": 6.373971462249756,
      "epoch": 1.9743439664422984,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046441398559181096,
      "loss": 6.1414,
      "mean_token_accuracy": 0.1384057879447937,
      "num_tokens": 33814408.0,
      "step": 10120
    },
    {
      "entropy": 6.383060598373413,
      "epoch": 1.9753194810262413,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046437601507198263,
      "loss": 6.2765,
      "mean_token_accuracy": 0.13946533277630807,
      "num_tokens": 33831499.0,
      "step": 10125
    },
    {
      "entropy": 6.44655179977417,
      "epoch": 1.9762949956101843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046433802604700506,
      "loss": 6.3225,
      "mean_token_accuracy": 0.13285728320479392,
      "num_tokens": 33848456.0,
      "step": 10130
    },
    {
      "entropy": 6.465530109405518,
      "epoch": 1.9772705101941273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004643000185205903,
      "loss": 6.3278,
      "mean_token_accuracy": 0.12818075567483903,
      "num_tokens": 33863824.0,
      "step": 10135
    },
    {
      "entropy": 6.407664346694946,
      "epoch": 1.9782460247780704,
      "grad_norm": 1.125,
      "learning_rate": 0.00046426199249645253,
      "loss": 6.2591,
      "mean_token_accuracy": 0.12772907167673112,
      "num_tokens": 33879382.0,
      "step": 10140
    },
    {
      "entropy": 6.329133367538452,
      "epoch": 1.9792215393620136,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004642239479783073,
      "loss": 6.1845,
      "mean_token_accuracy": 0.13531068488955497,
      "num_tokens": 33895930.0,
      "step": 10145
    },
    {
      "entropy": 6.171057653427124,
      "epoch": 1.9801970539459566,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046418588496987233,
      "loss": 6.0142,
      "mean_token_accuracy": 0.16210030168294906,
      "num_tokens": 33912067.0,
      "step": 10150
    },
    {
      "entropy": 6.440713787078858,
      "epoch": 1.9811725685298995,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000464147803474867,
      "loss": 6.2734,
      "mean_token_accuracy": 0.1330563798546791,
      "num_tokens": 33927481.0,
      "step": 10155
    },
    {
      "entropy": 6.375605154037475,
      "epoch": 1.9821480831138425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004641097034970123,
      "loss": 6.1413,
      "mean_token_accuracy": 0.14109975472092628,
      "num_tokens": 33945373.0,
      "step": 10160
    },
    {
      "entropy": 6.391907167434693,
      "epoch": 1.9831235976977855,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046407158504003143,
      "loss": 6.242,
      "mean_token_accuracy": 0.1398923546075821,
      "num_tokens": 33962441.0,
      "step": 10165
    },
    {
      "entropy": 6.2496216773986815,
      "epoch": 1.9840991122817286,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046403344810764933,
      "loss": 6.0956,
      "mean_token_accuracy": 0.1486932121217251,
      "num_tokens": 33979597.0,
      "step": 10170
    },
    {
      "entropy": 6.403612422943115,
      "epoch": 1.9850746268656716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046399529270359214,
      "loss": 6.247,
      "mean_token_accuracy": 0.14259378388524055,
      "num_tokens": 33996079.0,
      "step": 10175
    },
    {
      "entropy": 6.266183567047119,
      "epoch": 1.9860501414496148,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046395711883158864,
      "loss": 6.0941,
      "mean_token_accuracy": 0.14537967443466188,
      "num_tokens": 34014106.0,
      "step": 10180
    },
    {
      "entropy": 6.361168432235718,
      "epoch": 1.9870256560335577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463918926495369,
      "loss": 6.2117,
      "mean_token_accuracy": 0.13594407439231873,
      "num_tokens": 34030815.0,
      "step": 10185
    },
    {
      "entropy": 6.359433269500732,
      "epoch": 1.9880011706175007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004638807156986651,
      "loss": 6.1552,
      "mean_token_accuracy": 0.14495814144611358,
      "num_tokens": 34047453.0,
      "step": 10190
    },
    {
      "entropy": 6.382708597183227,
      "epoch": 1.9889766852014437,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004638424864452109,
      "loss": 6.1599,
      "mean_token_accuracy": 0.14070889726281166,
      "num_tokens": 34065442.0,
      "step": 10195
    },
    {
      "entropy": 6.291227149963379,
      "epoch": 1.9899521997853868,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004638042387387419,
      "loss": 6.1244,
      "mean_token_accuracy": 0.14562064334750174,
      "num_tokens": 34082073.0,
      "step": 10200
    },
    {
      "entropy": 6.456820917129517,
      "epoch": 1.9909277143693298,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004637659725829957,
      "loss": 6.2612,
      "mean_token_accuracy": 0.1299303539097309,
      "num_tokens": 34097542.0,
      "step": 10205
    },
    {
      "entropy": 6.243330335617065,
      "epoch": 1.991903228953273,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004637276879817114,
      "loss": 6.1066,
      "mean_token_accuracy": 0.14370434060692788,
      "num_tokens": 34114961.0,
      "step": 10210
    },
    {
      "entropy": 6.307617235183716,
      "epoch": 1.992878743537216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046368938493863,
      "loss": 6.1228,
      "mean_token_accuracy": 0.1487034797668457,
      "num_tokens": 34131774.0,
      "step": 10215
    },
    {
      "entropy": 6.432252550125122,
      "epoch": 1.993854258121159,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046365106345749445,
      "loss": 6.2212,
      "mean_token_accuracy": 0.1345653124153614,
      "num_tokens": 34147865.0,
      "step": 10220
    },
    {
      "entropy": 6.291583919525147,
      "epoch": 1.9948297727051019,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004636127235420493,
      "loss": 6.1609,
      "mean_token_accuracy": 0.1438751071691513,
      "num_tokens": 34164029.0,
      "step": 10225
    },
    {
      "entropy": 6.156807374954224,
      "epoch": 1.9958052872890448,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046357436519604107,
      "loss": 5.9649,
      "mean_token_accuracy": 0.16321696788072587,
      "num_tokens": 34181026.0,
      "step": 10230
    },
    {
      "entropy": 6.366192102432251,
      "epoch": 1.996780801872988,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000463535988423218,
      "loss": 6.181,
      "mean_token_accuracy": 0.1451581224799156,
      "num_tokens": 34197129.0,
      "step": 10235
    },
    {
      "entropy": 6.261172914505005,
      "epoch": 1.9977563164569312,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004634975932273301,
      "loss": 6.1363,
      "mean_token_accuracy": 0.15170709416270256,
      "num_tokens": 34214277.0,
      "step": 10240
    },
    {
      "entropy": 6.373532247543335,
      "epoch": 1.9987318310408742,
      "grad_norm": 1.125,
      "learning_rate": 0.00046345917961212915,
      "loss": 6.1881,
      "mean_token_accuracy": 0.14489124491810798,
      "num_tokens": 34231491.0,
      "step": 10245
    },
    {
      "entropy": 6.358797025680542,
      "epoch": 1.9997073456248171,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046342074758136886,
      "loss": 6.2093,
      "mean_token_accuracy": 0.1394185483455658,
      "num_tokens": 34248000.0,
      "step": 10250
    },
    {
      "entropy": 6.285558117760552,
      "epoch": 2.0005853087503658,
      "grad_norm": 1.125,
      "learning_rate": 0.0004633822971388047,
      "loss": 6.1115,
      "mean_token_accuracy": 0.1551802870300081,
      "num_tokens": 34262606.0,
      "step": 10255
    },
    {
      "entropy": 6.274199724197388,
      "epoch": 2.0015608233343087,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046334382828819385,
      "loss": 5.9489,
      "mean_token_accuracy": 0.1595419891178608,
      "num_tokens": 34280759.0,
      "step": 10260
    },
    {
      "entropy": 6.311619424819947,
      "epoch": 2.0025363379182517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004633053410332953,
      "loss": 6.118,
      "mean_token_accuracy": 0.14193400144577026,
      "num_tokens": 34296420.0,
      "step": 10265
    },
    {
      "entropy": 6.410163307189942,
      "epoch": 2.003511852502195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046326683537787006,
      "loss": 6.1856,
      "mean_token_accuracy": 0.13460189923644067,
      "num_tokens": 34313171.0,
      "step": 10270
    },
    {
      "entropy": 6.401320266723633,
      "epoch": 2.004487367086138,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046322831132568067,
      "loss": 6.0841,
      "mean_token_accuracy": 0.14148477762937545,
      "num_tokens": 34330885.0,
      "step": 10275
    },
    {
      "entropy": 6.359206247329712,
      "epoch": 2.005462881670081,
      "grad_norm": 1.125,
      "learning_rate": 0.0004631897688804914,
      "loss": 6.0337,
      "mean_token_accuracy": 0.14259042963385582,
      "num_tokens": 34346661.0,
      "step": 10280
    },
    {
      "entropy": 6.314279270172119,
      "epoch": 2.006438396254024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004631512080460688,
      "loss": 6.1301,
      "mean_token_accuracy": 0.13844745382666587,
      "num_tokens": 34362480.0,
      "step": 10285
    },
    {
      "entropy": 6.275138521194458,
      "epoch": 2.007413910837967,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046311262882618064,
      "loss": 5.9907,
      "mean_token_accuracy": 0.14758023321628572,
      "num_tokens": 34379719.0,
      "step": 10290
    },
    {
      "entropy": 6.360075807571411,
      "epoch": 2.00838942542191,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004630740312245969,
      "loss": 6.1421,
      "mean_token_accuracy": 0.13957590833306313,
      "num_tokens": 34396088.0,
      "step": 10295
    },
    {
      "entropy": 6.3457998752594,
      "epoch": 2.0093649400058533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046303541524508906,
      "loss": 6.0715,
      "mean_token_accuracy": 0.14024281427264212,
      "num_tokens": 34413658.0,
      "step": 10300
    },
    {
      "entropy": 6.468014764785766,
      "epoch": 2.0103404545897963,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004629967808914307,
      "loss": 6.2207,
      "mean_token_accuracy": 0.1385574072599411,
      "num_tokens": 34430799.0,
      "step": 10305
    },
    {
      "entropy": 6.438567590713501,
      "epoch": 2.0113159691737392,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004629581281673969,
      "loss": 6.1662,
      "mean_token_accuracy": 0.14102044701576233,
      "num_tokens": 34447177.0,
      "step": 10310
    },
    {
      "entropy": 6.393231201171875,
      "epoch": 2.012291483757682,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046291945707676473,
      "loss": 6.0825,
      "mean_token_accuracy": 0.13732732981443405,
      "num_tokens": 34463175.0,
      "step": 10315
    },
    {
      "entropy": 6.255165433883667,
      "epoch": 2.013266998341625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046288076762331294,
      "loss": 6.0329,
      "mean_token_accuracy": 0.14444167912006378,
      "num_tokens": 34482233.0,
      "step": 10320
    },
    {
      "entropy": 6.399918794631958,
      "epoch": 2.014242512925568,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004628420598108222,
      "loss": 6.181,
      "mean_token_accuracy": 0.1391908533871174,
      "num_tokens": 34498780.0,
      "step": 10325
    },
    {
      "entropy": 6.2833751201629635,
      "epoch": 2.015218027509511,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046280333364307483,
      "loss": 5.9747,
      "mean_token_accuracy": 0.1487969435751438,
      "num_tokens": 34516723.0,
      "step": 10330
    },
    {
      "entropy": 6.381993770599365,
      "epoch": 2.0161935420934545,
      "grad_norm": 1.078125,
      "learning_rate": 0.000462764589123855,
      "loss": 6.121,
      "mean_token_accuracy": 0.1380957193672657,
      "num_tokens": 34534291.0,
      "step": 10335
    },
    {
      "entropy": 6.276688861846924,
      "epoch": 2.0171690566773974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004627258262569487,
      "loss": 5.9707,
      "mean_token_accuracy": 0.14855577573180198,
      "num_tokens": 34551349.0,
      "step": 10340
    },
    {
      "entropy": 6.241369915008545,
      "epoch": 2.0181445712613404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046268704504614383,
      "loss": 5.9531,
      "mean_token_accuracy": 0.1524041011929512,
      "num_tokens": 34569014.0,
      "step": 10345
    },
    {
      "entropy": 6.412798118591309,
      "epoch": 2.0191200858452834,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004626482454952298,
      "loss": 6.2085,
      "mean_token_accuracy": 0.1333499528467655,
      "num_tokens": 34586855.0,
      "step": 10350
    },
    {
      "entropy": 6.3526294231414795,
      "epoch": 2.0200956004292263,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000462609427607998,
      "loss": 6.0031,
      "mean_token_accuracy": 0.15232734233140946,
      "num_tokens": 34602809.0,
      "step": 10355
    },
    {
      "entropy": 6.336437702178955,
      "epoch": 2.0210711150131693,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004625705913882416,
      "loss": 6.1022,
      "mean_token_accuracy": 0.14183633029460907,
      "num_tokens": 34618371.0,
      "step": 10360
    },
    {
      "entropy": 6.262168979644775,
      "epoch": 2.0220466295971127,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046253173683975557,
      "loss": 6.0317,
      "mean_token_accuracy": 0.1386827901005745,
      "num_tokens": 34635118.0,
      "step": 10365
    },
    {
      "entropy": 6.378279066085815,
      "epoch": 2.0230221441810556,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004624928639663365,
      "loss": 6.0678,
      "mean_token_accuracy": 0.1471583977341652,
      "num_tokens": 34651827.0,
      "step": 10370
    },
    {
      "entropy": 6.337476444244385,
      "epoch": 2.0239976587649986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004624539727717831,
      "loss": 6.1095,
      "mean_token_accuracy": 0.13815348595380783,
      "num_tokens": 34668698.0,
      "step": 10375
    },
    {
      "entropy": 6.4178886890411375,
      "epoch": 2.0249731733489416,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046241506325989546,
      "loss": 6.1381,
      "mean_token_accuracy": 0.14280589073896408,
      "num_tokens": 34686257.0,
      "step": 10380
    },
    {
      "entropy": 6.3597485542297365,
      "epoch": 2.0259486879328845,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004623761354344759,
      "loss": 6.0876,
      "mean_token_accuracy": 0.14185886234045028,
      "num_tokens": 34702088.0,
      "step": 10385
    },
    {
      "entropy": 6.313714790344238,
      "epoch": 2.0269242025168275,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004623371892993281,
      "loss": 6.0917,
      "mean_token_accuracy": 0.14666024446487427,
      "num_tokens": 34718940.0,
      "step": 10390
    },
    {
      "entropy": 6.340856695175171,
      "epoch": 2.0278997171007704,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000462298224858258,
      "loss": 6.107,
      "mean_token_accuracy": 0.14338432475924492,
      "num_tokens": 34735228.0,
      "step": 10395
    },
    {
      "entropy": 6.413027572631836,
      "epoch": 2.028875231684714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046225924211507275,
      "loss": 6.165,
      "mean_token_accuracy": 0.13771923407912254,
      "num_tokens": 34751741.0,
      "step": 10400
    },
    {
      "entropy": 6.288215637207031,
      "epoch": 2.029850746268657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004622202410735818,
      "loss": 6.0269,
      "mean_token_accuracy": 0.1452144704759121,
      "num_tokens": 34767152.0,
      "step": 10405
    },
    {
      "entropy": 6.354459285736084,
      "epoch": 2.0308262608525998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046218122173759626,
      "loss": 6.1041,
      "mean_token_accuracy": 0.1371276021003723,
      "num_tokens": 34784419.0,
      "step": 10410
    },
    {
      "entropy": 6.257331800460816,
      "epoch": 2.0318017754365427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046214218411092867,
      "loss": 5.9919,
      "mean_token_accuracy": 0.15269352495670319,
      "num_tokens": 34799877.0,
      "step": 10415
    },
    {
      "entropy": 6.39379096031189,
      "epoch": 2.0327772900204857,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046210312819739406,
      "loss": 6.1648,
      "mean_token_accuracy": 0.13228063359856607,
      "num_tokens": 34816332.0,
      "step": 10420
    },
    {
      "entropy": 6.434387445449829,
      "epoch": 2.0337528046044286,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004620640540008084,
      "loss": 6.1455,
      "mean_token_accuracy": 0.1407581314444542,
      "num_tokens": 34833364.0,
      "step": 10425
    },
    {
      "entropy": 6.21344223022461,
      "epoch": 2.034728319188372,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004620249615249902,
      "loss": 5.8822,
      "mean_token_accuracy": 0.1635778859257698,
      "num_tokens": 34851419.0,
      "step": 10430
    },
    {
      "entropy": 6.363228607177734,
      "epoch": 2.035703833772315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004619858507737594,
      "loss": 6.1235,
      "mean_token_accuracy": 0.14288055747747422,
      "num_tokens": 34869237.0,
      "step": 10435
    },
    {
      "entropy": 6.288411092758179,
      "epoch": 2.036679348356258,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004619467217509375,
      "loss": 6.0215,
      "mean_token_accuracy": 0.14890795797109604,
      "num_tokens": 34888254.0,
      "step": 10440
    },
    {
      "entropy": 6.389755582809448,
      "epoch": 2.037654862940201,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046190757446034844,
      "loss": 6.1316,
      "mean_token_accuracy": 0.14105125591158868,
      "num_tokens": 34905257.0,
      "step": 10445
    },
    {
      "entropy": 6.348039388656616,
      "epoch": 2.038630377524144,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046186840890581724,
      "loss": 6.0576,
      "mean_token_accuracy": 0.14253288209438325,
      "num_tokens": 34922358.0,
      "step": 10450
    },
    {
      "entropy": 6.393352079391479,
      "epoch": 2.039605892108087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046182922509117115,
      "loss": 6.0569,
      "mean_token_accuracy": 0.14792812019586563,
      "num_tokens": 34939746.0,
      "step": 10455
    },
    {
      "entropy": 6.261393880844116,
      "epoch": 2.0405814066920303,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004617900230202391,
      "loss": 6.0076,
      "mean_token_accuracy": 0.14482191279530526,
      "num_tokens": 34957188.0,
      "step": 10460
    },
    {
      "entropy": 6.292717742919922,
      "epoch": 2.041556921275973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046175080269685167,
      "loss": 6.0713,
      "mean_token_accuracy": 0.15078377947211266,
      "num_tokens": 34974859.0,
      "step": 10465
    },
    {
      "entropy": 6.285018396377564,
      "epoch": 2.042532435859916,
      "grad_norm": 1.125,
      "learning_rate": 0.00046171156412484136,
      "loss": 5.9994,
      "mean_token_accuracy": 0.1530948370695114,
      "num_tokens": 34990806.0,
      "step": 10470
    },
    {
      "entropy": 6.314354038238525,
      "epoch": 2.043507950443859,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046167230730804256,
      "loss": 6.0684,
      "mean_token_accuracy": 0.1457235835492611,
      "num_tokens": 35006644.0,
      "step": 10475
    },
    {
      "entropy": 6.395410108566284,
      "epoch": 2.044483465027802,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004616330322502911,
      "loss": 6.1314,
      "mean_token_accuracy": 0.1425391025841236,
      "num_tokens": 35023818.0,
      "step": 10480
    },
    {
      "entropy": 6.373316431045533,
      "epoch": 2.045458979611745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046159373895542506,
      "loss": 6.131,
      "mean_token_accuracy": 0.14138488620519638,
      "num_tokens": 35039447.0,
      "step": 10485
    },
    {
      "entropy": 6.322789001464844,
      "epoch": 2.046434494195688,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004615544274272837,
      "loss": 6.0647,
      "mean_token_accuracy": 0.14413523748517038,
      "num_tokens": 35054831.0,
      "step": 10490
    },
    {
      "entropy": 6.351386833190918,
      "epoch": 2.0474100087796314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004615150976697087,
      "loss": 6.0514,
      "mean_token_accuracy": 0.14682291001081466,
      "num_tokens": 35071266.0,
      "step": 10495
    },
    {
      "entropy": 6.3908223628997805,
      "epoch": 2.0483855233635744,
      "grad_norm": 1.375,
      "learning_rate": 0.000461475749686543,
      "loss": 6.1257,
      "mean_token_accuracy": 0.1402203120291233,
      "num_tokens": 35087696.0,
      "step": 10500
    },
    {
      "entropy": 6.260792064666748,
      "epoch": 2.0493610379475173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004614363834816318,
      "loss": 6.0226,
      "mean_token_accuracy": 0.14492887929081916,
      "num_tokens": 35104914.0,
      "step": 10505
    },
    {
      "entropy": 6.360859823226929,
      "epoch": 2.0503365525314603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004613969990588216,
      "loss": 6.1071,
      "mean_token_accuracy": 0.1457517370581627,
      "num_tokens": 35121777.0,
      "step": 10510
    },
    {
      "entropy": 6.285934829711914,
      "epoch": 2.0513120671154033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046135759642196094,
      "loss": 6.0747,
      "mean_token_accuracy": 0.1523919127881527,
      "num_tokens": 35141293.0,
      "step": 10515
    },
    {
      "entropy": 6.43258957862854,
      "epoch": 2.052287581699346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046131817557490016,
      "loss": 6.1658,
      "mean_token_accuracy": 0.13849455416202544,
      "num_tokens": 35157596.0,
      "step": 10520
    },
    {
      "entropy": 6.269530010223389,
      "epoch": 2.0532630962832896,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004612787365214914,
      "loss": 6.0217,
      "mean_token_accuracy": 0.15021250993013383,
      "num_tokens": 35173060.0,
      "step": 10525
    },
    {
      "entropy": 6.317304182052612,
      "epoch": 2.0542386108672326,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046123927926558836,
      "loss": 6.043,
      "mean_token_accuracy": 0.13662949949502945,
      "num_tokens": 35189846.0,
      "step": 10530
    },
    {
      "entropy": 6.356263303756714,
      "epoch": 2.0552141254511755,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004611998038110467,
      "loss": 6.1142,
      "mean_token_accuracy": 0.14656300395727156,
      "num_tokens": 35206355.0,
      "step": 10535
    },
    {
      "entropy": 6.269530248641968,
      "epoch": 2.0561896400351185,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004611603101617239,
      "loss": 5.9799,
      "mean_token_accuracy": 0.14973942637443544,
      "num_tokens": 35223535.0,
      "step": 10540
    },
    {
      "entropy": 6.220055389404297,
      "epoch": 2.0571651546190615,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046112079832147894,
      "loss": 5.8926,
      "mean_token_accuracy": 0.15822109431028367,
      "num_tokens": 35241283.0,
      "step": 10545
    },
    {
      "entropy": 6.328880548477173,
      "epoch": 2.0581406692030044,
      "grad_norm": 1.28125,
      "learning_rate": 0.000461081268294173,
      "loss": 6.0677,
      "mean_token_accuracy": 0.14267542064189911,
      "num_tokens": 35257398.0,
      "step": 10550
    },
    {
      "entropy": 6.247137832641601,
      "epoch": 2.059116183786948,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004610417200836688,
      "loss": 5.9657,
      "mean_token_accuracy": 0.15300771966576576,
      "num_tokens": 35275121.0,
      "step": 10555
    },
    {
      "entropy": 6.286081838607788,
      "epoch": 2.060091698370891,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004610021536938307,
      "loss": 6.028,
      "mean_token_accuracy": 0.13992301747202873,
      "num_tokens": 35290169.0,
      "step": 10560
    },
    {
      "entropy": 6.286533164978027,
      "epoch": 2.0610672129548337,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046096256912852505,
      "loss": 6.0104,
      "mean_token_accuracy": 0.15060097351670265,
      "num_tokens": 35305824.0,
      "step": 10565
    },
    {
      "entropy": 6.391307926177978,
      "epoch": 2.0620427275387767,
      "grad_norm": 1.125,
      "learning_rate": 0.0004609229663916199,
      "loss": 6.0963,
      "mean_token_accuracy": 0.14116146564483642,
      "num_tokens": 35322182.0,
      "step": 10570
    },
    {
      "entropy": 6.362337303161621,
      "epoch": 2.0630182421227197,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046088334548698525,
      "loss": 6.1862,
      "mean_token_accuracy": 0.14082293808460236,
      "num_tokens": 35339046.0,
      "step": 10575
    },
    {
      "entropy": 6.320966148376465,
      "epoch": 2.0639937567066626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004608437064184925,
      "loss": 6.0638,
      "mean_token_accuracy": 0.14221861734986305,
      "num_tokens": 35355448.0,
      "step": 10580
    },
    {
      "entropy": 6.301726388931274,
      "epoch": 2.0649692712906056,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046080404919001506,
      "loss": 5.9674,
      "mean_token_accuracy": 0.15147449374198912,
      "num_tokens": 35372738.0,
      "step": 10585
    },
    {
      "entropy": 6.326701211929321,
      "epoch": 2.065944785874549,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046076437380542824,
      "loss": 6.0381,
      "mean_token_accuracy": 0.14420806914567946,
      "num_tokens": 35389344.0,
      "step": 10590
    },
    {
      "entropy": 6.242225074768067,
      "epoch": 2.066920300458492,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004607246802686089,
      "loss": 5.9544,
      "mean_token_accuracy": 0.14604247510433196,
      "num_tokens": 35405897.0,
      "step": 10595
    },
    {
      "entropy": 6.2751508235931395,
      "epoch": 2.067895815042435,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606849685834356,
      "loss": 6.0551,
      "mean_token_accuracy": 0.14592415392398833,
      "num_tokens": 35423084.0,
      "step": 10600
    },
    {
      "entropy": 6.270927000045776,
      "epoch": 2.068871329626378,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004606452387537891,
      "loss": 6.0083,
      "mean_token_accuracy": 0.15448917597532272,
      "num_tokens": 35440411.0,
      "step": 10605
    },
    {
      "entropy": 6.397629070281982,
      "epoch": 2.069846844210321,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004606054907835514,
      "loss": 6.1557,
      "mean_token_accuracy": 0.14384288117289543,
      "num_tokens": 35456519.0,
      "step": 10610
    },
    {
      "entropy": 6.369201755523681,
      "epoch": 2.070822358794264,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605657246766068,
      "loss": 6.0629,
      "mean_token_accuracy": 0.14277879968285562,
      "num_tokens": 35473271.0,
      "step": 10615
    },
    {
      "entropy": 6.355735492706299,
      "epoch": 2.071797873378207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046052594043684086,
      "loss": 6.1434,
      "mean_token_accuracy": 0.14157638028264047,
      "num_tokens": 35491360.0,
      "step": 10620
    },
    {
      "entropy": 6.378295803070069,
      "epoch": 2.07277338796215,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046048613806814125,
      "loss": 6.1132,
      "mean_token_accuracy": 0.14356202334165574,
      "num_tokens": 35508481.0,
      "step": 10625
    },
    {
      "entropy": 6.3669578552246096,
      "epoch": 2.073748902546093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004604463175743974,
      "loss": 6.1007,
      "mean_token_accuracy": 0.14338903576135636,
      "num_tokens": 35524895.0,
      "step": 10630
    },
    {
      "entropy": 6.331673097610474,
      "epoch": 2.074724417130036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046040647895950016,
      "loss": 6.107,
      "mean_token_accuracy": 0.1390886977314949,
      "num_tokens": 35541644.0,
      "step": 10635
    },
    {
      "entropy": 6.337795925140381,
      "epoch": 2.075699931713979,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046036662222734276,
      "loss": 6.0593,
      "mean_token_accuracy": 0.139164786785841,
      "num_tokens": 35557211.0,
      "step": 10640
    },
    {
      "entropy": 6.338967990875244,
      "epoch": 2.076675446297922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004603267473818196,
      "loss": 6.0613,
      "mean_token_accuracy": 0.14234994053840638,
      "num_tokens": 35574366.0,
      "step": 10645
    },
    {
      "entropy": 6.185725355148316,
      "epoch": 2.077650960881865,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004602868544268272,
      "loss": 5.9239,
      "mean_token_accuracy": 0.1591850198805332,
      "num_tokens": 35592172.0,
      "step": 10650
    },
    {
      "entropy": 6.325996541976929,
      "epoch": 2.0786264754658084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004602469433662639,
      "loss": 5.9865,
      "mean_token_accuracy": 0.15413146167993547,
      "num_tokens": 35608225.0,
      "step": 10655
    },
    {
      "entropy": 6.3318033695220945,
      "epoch": 2.0796019900497513,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046020701420402936,
      "loss": 6.0804,
      "mean_token_accuracy": 0.13833392336964606,
      "num_tokens": 35624474.0,
      "step": 10660
    },
    {
      "entropy": 6.398423147201538,
      "epoch": 2.0805775046336943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046016706694402556,
      "loss": 6.1167,
      "mean_token_accuracy": 0.14075322300195695,
      "num_tokens": 35639800.0,
      "step": 10665
    },
    {
      "entropy": 6.2646585464477536,
      "epoch": 2.0815530192176372,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004601271015901559,
      "loss": 5.9854,
      "mean_token_accuracy": 0.14431332498788835,
      "num_tokens": 35656830.0,
      "step": 10670
    },
    {
      "entropy": 6.2653501510620115,
      "epoch": 2.08252853380158,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004600871181463257,
      "loss": 6.0791,
      "mean_token_accuracy": 0.14371969178318977,
      "num_tokens": 35673394.0,
      "step": 10675
    },
    {
      "entropy": 6.335498189926147,
      "epoch": 2.083504048385523,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004600471166164419,
      "loss": 6.0432,
      "mean_token_accuracy": 0.14510103240609168,
      "num_tokens": 35690135.0,
      "step": 10680
    },
    {
      "entropy": 6.396342515945435,
      "epoch": 2.0844795629694666,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004600070970044135,
      "loss": 6.1967,
      "mean_token_accuracy": 0.13703661635518075,
      "num_tokens": 35706817.0,
      "step": 10685
    },
    {
      "entropy": 6.333326768875122,
      "epoch": 2.0854550775534095,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045996705931415096,
      "loss": 6.0536,
      "mean_token_accuracy": 0.14553411677479744,
      "num_tokens": 35724929.0,
      "step": 10690
    },
    {
      "entropy": 6.1982629776000975,
      "epoch": 2.0864305921373525,
      "grad_norm": 1.125,
      "learning_rate": 0.0004599270035495665,
      "loss": 5.9276,
      "mean_token_accuracy": 0.16013435646891594,
      "num_tokens": 35741892.0,
      "step": 10695
    },
    {
      "entropy": 6.3023570537567135,
      "epoch": 2.0874061067212955,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045988692971457443,
      "loss": 5.9672,
      "mean_token_accuracy": 0.14751505702733994,
      "num_tokens": 35758486.0,
      "step": 10700
    },
    {
      "entropy": 6.303978061676025,
      "epoch": 2.0883816213052384,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004598468378130906,
      "loss": 6.0556,
      "mean_token_accuracy": 0.1466790296137333,
      "num_tokens": 35774118.0,
      "step": 10705
    },
    {
      "entropy": 6.250840139389038,
      "epoch": 2.0893571358891814,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045980672784903245,
      "loss": 5.9963,
      "mean_token_accuracy": 0.14792305529117583,
      "num_tokens": 35790442.0,
      "step": 10710
    },
    {
      "entropy": 6.2868937969207765,
      "epoch": 2.0903326504731248,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045976659982631954,
      "loss": 6.0278,
      "mean_token_accuracy": 0.14657861292362212,
      "num_tokens": 35805498.0,
      "step": 10715
    },
    {
      "entropy": 6.307354593276978,
      "epoch": 2.0913081650570677,
      "grad_norm": 1.359375,
      "learning_rate": 0.000459726453748873,
      "loss": 6.0292,
      "mean_token_accuracy": 0.14975132644176484,
      "num_tokens": 35821477.0,
      "step": 10720
    },
    {
      "entropy": 6.3264830112457275,
      "epoch": 2.0922836796410107,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045968628962061585,
      "loss": 6.0414,
      "mean_token_accuracy": 0.1459795705974102,
      "num_tokens": 35838462.0,
      "step": 10725
    },
    {
      "entropy": 6.354039859771729,
      "epoch": 2.0932591942249537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004596461074454727,
      "loss": 6.0747,
      "mean_token_accuracy": 0.1415545791387558,
      "num_tokens": 35854814.0,
      "step": 10730
    },
    {
      "entropy": 6.3709475040435795,
      "epoch": 2.0942347088088966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045960590722737,
      "loss": 6.1194,
      "mean_token_accuracy": 0.13603568226099014,
      "num_tokens": 35870117.0,
      "step": 10735
    },
    {
      "entropy": 6.401092624664306,
      "epoch": 2.0952102233928396,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000459565688970236,
      "loss": 6.0908,
      "mean_token_accuracy": 0.14193982630968094,
      "num_tokens": 35884900.0,
      "step": 10740
    },
    {
      "entropy": 6.310601758956909,
      "epoch": 2.096185737976783,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004595254526780007,
      "loss": 6.0608,
      "mean_token_accuracy": 0.15269813910126687,
      "num_tokens": 35901280.0,
      "step": 10745
    },
    {
      "entropy": 6.275589036941528,
      "epoch": 2.097161252560726,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045948519835459575,
      "loss": 6.086,
      "mean_token_accuracy": 0.15062355920672416,
      "num_tokens": 35918015.0,
      "step": 10750
    },
    {
      "entropy": 6.384981441497803,
      "epoch": 2.098136767144669,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045944492600395476,
      "loss": 6.1347,
      "mean_token_accuracy": 0.14554157108068466,
      "num_tokens": 35935367.0,
      "step": 10755
    },
    {
      "entropy": 6.34754900932312,
      "epoch": 2.099112281728612,
      "grad_norm": 1.171875,
      "learning_rate": 0.000459404635630013,
      "loss": 6.0952,
      "mean_token_accuracy": 0.14267582818865776,
      "num_tokens": 35951697.0,
      "step": 10760
    },
    {
      "entropy": 6.334804534912109,
      "epoch": 2.100087796312555,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004593643272367075,
      "loss": 6.0441,
      "mean_token_accuracy": 0.1466379888355732,
      "num_tokens": 35968321.0,
      "step": 10765
    },
    {
      "entropy": 6.35356388092041,
      "epoch": 2.101063310896498,
      "grad_norm": 1.140625,
      "learning_rate": 0.000459324000827977,
      "loss": 6.0508,
      "mean_token_accuracy": 0.14554684683680535,
      "num_tokens": 35984192.0,
      "step": 10770
    },
    {
      "entropy": 6.340356063842774,
      "epoch": 2.1020388254804407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004592836564077621,
      "loss": 6.1011,
      "mean_token_accuracy": 0.13884800523519517,
      "num_tokens": 36001208.0,
      "step": 10775
    },
    {
      "entropy": 6.388402128219605,
      "epoch": 2.103014340064384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000459243293980005,
      "loss": 6.0646,
      "mean_token_accuracy": 0.14025810360908508,
      "num_tokens": 36018253.0,
      "step": 10780
    },
    {
      "entropy": 6.237870502471924,
      "epoch": 2.103989854648327,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045920291354865,
      "loss": 5.9454,
      "mean_token_accuracy": 0.14883601292967796,
      "num_tokens": 36035041.0,
      "step": 10785
    },
    {
      "entropy": 6.367056226730346,
      "epoch": 2.10496536923227,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045916251511764265,
      "loss": 6.1531,
      "mean_token_accuracy": 0.1387026861310005,
      "num_tokens": 36051497.0,
      "step": 10790
    },
    {
      "entropy": 6.3250795841217045,
      "epoch": 2.105940883816213,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004591220986909309,
      "loss": 6.0519,
      "mean_token_accuracy": 0.1409115917980671,
      "num_tokens": 36068117.0,
      "step": 10795
    },
    {
      "entropy": 6.315334272384644,
      "epoch": 2.106916398400156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045908166427246377,
      "loss": 6.0603,
      "mean_token_accuracy": 0.14130902290344238,
      "num_tokens": 36084489.0,
      "step": 10800
    },
    {
      "entropy": 6.222200155258179,
      "epoch": 2.107891912984099,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004590412118661925,
      "loss": 5.9229,
      "mean_token_accuracy": 0.1470213435590267,
      "num_tokens": 36102633.0,
      "step": 10805
    },
    {
      "entropy": 6.262992477416992,
      "epoch": 2.1088674275680424,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004590007414760698,
      "loss": 5.97,
      "mean_token_accuracy": 0.1573075234889984,
      "num_tokens": 36121581.0,
      "step": 10810
    },
    {
      "entropy": 6.253486680984497,
      "epoch": 2.1098429421519853,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045896025310605067,
      "loss": 6.0323,
      "mean_token_accuracy": 0.15125820264220238,
      "num_tokens": 36138266.0,
      "step": 10815
    },
    {
      "entropy": 6.2048064231872555,
      "epoch": 2.1108184567359283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004589197467600911,
      "loss": 5.9678,
      "mean_token_accuracy": 0.15602525919675828,
      "num_tokens": 36154365.0,
      "step": 10820
    },
    {
      "entropy": 6.234320449829101,
      "epoch": 2.1117939713198712,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004588792224421494,
      "loss": 6.0202,
      "mean_token_accuracy": 0.15072812810540198,
      "num_tokens": 36173469.0,
      "step": 10825
    },
    {
      "entropy": 6.338155841827392,
      "epoch": 2.112769485903814,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045883868015618554,
      "loss": 6.1104,
      "mean_token_accuracy": 0.14502145051956178,
      "num_tokens": 36190032.0,
      "step": 10830
    },
    {
      "entropy": 6.315576601028442,
      "epoch": 2.113745000487757,
      "grad_norm": 1.125,
      "learning_rate": 0.000458798119906161,
      "loss": 6.0779,
      "mean_token_accuracy": 0.14178778678178788,
      "num_tokens": 36207681.0,
      "step": 10835
    },
    {
      "entropy": 6.336849927902222,
      "epoch": 2.1147205150717,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004587575416960391,
      "loss": 6.082,
      "mean_token_accuracy": 0.14655794575810432,
      "num_tokens": 36225031.0,
      "step": 10840
    },
    {
      "entropy": 6.274730968475342,
      "epoch": 2.1156960296556435,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045871694552978534,
      "loss": 5.9681,
      "mean_token_accuracy": 0.14940712377429008,
      "num_tokens": 36241579.0,
      "step": 10845
    },
    {
      "entropy": 6.259823179244995,
      "epoch": 2.1166715442395865,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045867633141136643,
      "loss": 6.0124,
      "mean_token_accuracy": 0.1521951027214527,
      "num_tokens": 36257978.0,
      "step": 10850
    },
    {
      "entropy": 6.380162906646729,
      "epoch": 2.1176470588235294,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045863569934475104,
      "loss": 6.1041,
      "mean_token_accuracy": 0.1369809813797474,
      "num_tokens": 36274068.0,
      "step": 10855
    },
    {
      "entropy": 6.133871603012085,
      "epoch": 2.1186225734074724,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045859504933390947,
      "loss": 5.847,
      "mean_token_accuracy": 0.1659851536154747,
      "num_tokens": 36291033.0,
      "step": 10860
    },
    {
      "entropy": 6.280784749984742,
      "epoch": 2.1195980879914154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004585543813828141,
      "loss": 6.0466,
      "mean_token_accuracy": 0.14922668039798737,
      "num_tokens": 36307430.0,
      "step": 10865
    },
    {
      "entropy": 6.447948217391968,
      "epoch": 2.1205736025753583,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004585136954954389,
      "loss": 6.193,
      "mean_token_accuracy": 0.13242962211370468,
      "num_tokens": 36324257.0,
      "step": 10870
    },
    {
      "entropy": 6.3962139129638675,
      "epoch": 2.1215491171593017,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004584729916757593,
      "loss": 6.111,
      "mean_token_accuracy": 0.14163657054305076,
      "num_tokens": 36339864.0,
      "step": 10875
    },
    {
      "entropy": 6.24046835899353,
      "epoch": 2.1225246317432447,
      "grad_norm": 1.125,
      "learning_rate": 0.00045843226992775294,
      "loss": 6.0096,
      "mean_token_accuracy": 0.14797204732894897,
      "num_tokens": 36356066.0,
      "step": 10880
    },
    {
      "entropy": 6.261502647399903,
      "epoch": 2.1235001463271876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045839153025539883,
      "loss": 5.9998,
      "mean_token_accuracy": 0.15151288583874703,
      "num_tokens": 36374420.0,
      "step": 10885
    },
    {
      "entropy": 6.337379217147827,
      "epoch": 2.1244756609111306,
      "grad_norm": 1.125,
      "learning_rate": 0.00045835077266267803,
      "loss": 6.1377,
      "mean_token_accuracy": 0.13508511185646058,
      "num_tokens": 36390884.0,
      "step": 10890
    },
    {
      "entropy": 6.341863489151001,
      "epoch": 2.1254511754950736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045830999715357326,
      "loss": 6.0408,
      "mean_token_accuracy": 0.14294055700302125,
      "num_tokens": 36407730.0,
      "step": 10895
    },
    {
      "entropy": 6.337892389297485,
      "epoch": 2.1264266900790165,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045826920373206884,
      "loss": 6.0816,
      "mean_token_accuracy": 0.13664693608880044,
      "num_tokens": 36423470.0,
      "step": 10900
    },
    {
      "entropy": 6.339166116714478,
      "epoch": 2.1274022046629595,
      "grad_norm": 1.15625,
      "learning_rate": 0.000458228392402151,
      "loss": 6.0648,
      "mean_token_accuracy": 0.1476524606347084,
      "num_tokens": 36439478.0,
      "step": 10905
    },
    {
      "entropy": 6.350145053863526,
      "epoch": 2.128377719246903,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004581875631678077,
      "loss": 6.1208,
      "mean_token_accuracy": 0.13855265453457832,
      "num_tokens": 36455732.0,
      "step": 10910
    },
    {
      "entropy": 6.344546508789063,
      "epoch": 2.129353233830846,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004581467160330287,
      "loss": 6.0065,
      "mean_token_accuracy": 0.14933736771345138,
      "num_tokens": 36471220.0,
      "step": 10915
    },
    {
      "entropy": 6.210495662689209,
      "epoch": 2.130328748414789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004581058510018053,
      "loss": 5.8241,
      "mean_token_accuracy": 0.1623227909207344,
      "num_tokens": 36488408.0,
      "step": 10920
    },
    {
      "entropy": 6.256653642654419,
      "epoch": 2.1313042629987318,
      "grad_norm": 1.125,
      "learning_rate": 0.0004580649680781307,
      "loss": 5.9601,
      "mean_token_accuracy": 0.15203500017523766,
      "num_tokens": 36504785.0,
      "step": 10925
    },
    {
      "entropy": 6.338117265701294,
      "epoch": 2.1322797775826747,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045802406726599997,
      "loss": 6.1712,
      "mean_token_accuracy": 0.13547390103340148,
      "num_tokens": 36521880.0,
      "step": 10930
    },
    {
      "entropy": 6.269066667556762,
      "epoch": 2.1332552921666177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045798314856940956,
      "loss": 5.9615,
      "mean_token_accuracy": 0.1471077248454094,
      "num_tokens": 36537075.0,
      "step": 10935
    },
    {
      "entropy": 6.249510097503662,
      "epoch": 2.134230806750561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045794221199235827,
      "loss": 6.009,
      "mean_token_accuracy": 0.15247500091791152,
      "num_tokens": 36554244.0,
      "step": 10940
    },
    {
      "entropy": 6.156300687789917,
      "epoch": 2.135206321334504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045790125753884575,
      "loss": 5.9071,
      "mean_token_accuracy": 0.16060149073600768,
      "num_tokens": 36571900.0,
      "step": 10945
    },
    {
      "entropy": 6.241668939590454,
      "epoch": 2.136181835918447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004578602852128744,
      "loss": 5.9735,
      "mean_token_accuracy": 0.15239520519971847,
      "num_tokens": 36588420.0,
      "step": 10950
    },
    {
      "entropy": 6.345282411575317,
      "epoch": 2.13715735050239,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045781929501844765,
      "loss": 6.1417,
      "mean_token_accuracy": 0.14691154658794403,
      "num_tokens": 36605371.0,
      "step": 10955
    },
    {
      "entropy": 6.136017751693726,
      "epoch": 2.138132865086333,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045777828695957095,
      "loss": 5.9227,
      "mean_token_accuracy": 0.16494935303926467,
      "num_tokens": 36620665.0,
      "step": 10960
    },
    {
      "entropy": 6.225688314437866,
      "epoch": 2.139108379670276,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045773726104025153,
      "loss": 6.0037,
      "mean_token_accuracy": 0.14992624446749686,
      "num_tokens": 36637067.0,
      "step": 10965
    },
    {
      "entropy": 6.3343658447265625,
      "epoch": 2.1400838942542193,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004576962172644982,
      "loss": 6.0569,
      "mean_token_accuracy": 0.1425447754561901,
      "num_tokens": 36652952.0,
      "step": 10970
    },
    {
      "entropy": 6.313946199417114,
      "epoch": 2.1410594088381623,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004576551556363217,
      "loss": 6.0153,
      "mean_token_accuracy": 0.15086103230714798,
      "num_tokens": 36668275.0,
      "step": 10975
    },
    {
      "entropy": 6.31061954498291,
      "epoch": 2.142034923422105,
      "grad_norm": 1.125,
      "learning_rate": 0.00045761407615973446,
      "loss": 6.0994,
      "mean_token_accuracy": 0.145106740295887,
      "num_tokens": 36684718.0,
      "step": 10980
    },
    {
      "entropy": 6.28707070350647,
      "epoch": 2.143010438006048,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004575729788387504,
      "loss": 6.0396,
      "mean_token_accuracy": 0.14349322468042375,
      "num_tokens": 36700259.0,
      "step": 10985
    },
    {
      "entropy": 6.318219518661499,
      "epoch": 2.143985952589991,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045753186367738564,
      "loss": 6.0668,
      "mean_token_accuracy": 0.14976088404655458,
      "num_tokens": 36716949.0,
      "step": 10990
    },
    {
      "entropy": 6.398902416229248,
      "epoch": 2.144961467173934,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004574907306796577,
      "loss": 6.102,
      "mean_token_accuracy": 0.14282070994377136,
      "num_tokens": 36732927.0,
      "step": 10995
    },
    {
      "entropy": 6.299973011016846,
      "epoch": 2.1459369817578775,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004574495798495859,
      "loss": 6.1099,
      "mean_token_accuracy": 0.13994522392749786,
      "num_tokens": 36748841.0,
      "step": 11000
    },
    {
      "entropy": 6.277076005935669,
      "epoch": 2.1469124963418205,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045740841119119147,
      "loss": 6.1056,
      "mean_token_accuracy": 0.14829643368721007,
      "num_tokens": 36766048.0,
      "step": 11005
    },
    {
      "entropy": 6.264330768585205,
      "epoch": 2.1478880109257634,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004573672247084972,
      "loss": 6.0374,
      "mean_token_accuracy": 0.15447540804743767,
      "num_tokens": 36782471.0,
      "step": 11010
    },
    {
      "entropy": 6.2465558528900145,
      "epoch": 2.1488635255097064,
      "grad_norm": 1.125,
      "learning_rate": 0.0004573260204055277,
      "loss": 5.9234,
      "mean_token_accuracy": 0.15999085158109666,
      "num_tokens": 36797789.0,
      "step": 11015
    },
    {
      "entropy": 6.145596075057983,
      "epoch": 2.1498390400936493,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004572847982863094,
      "loss": 5.9504,
      "mean_token_accuracy": 0.15779200792312623,
      "num_tokens": 36814466.0,
      "step": 11020
    },
    {
      "entropy": 6.181446886062622,
      "epoch": 2.1508145546775923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004572435583548702,
      "loss": 5.9188,
      "mean_token_accuracy": 0.15894735902547835,
      "num_tokens": 36831579.0,
      "step": 11025
    },
    {
      "entropy": 6.402044582366943,
      "epoch": 2.1517900692615353,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045720230061524,
      "loss": 6.1673,
      "mean_token_accuracy": 0.13703131079673767,
      "num_tokens": 36847260.0,
      "step": 11030
    },
    {
      "entropy": 6.383153104782105,
      "epoch": 2.1527655838454787,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004571610250714504,
      "loss": 6.1532,
      "mean_token_accuracy": 0.13719891160726547,
      "num_tokens": 36865380.0,
      "step": 11035
    },
    {
      "entropy": 6.326032066345215,
      "epoch": 2.1537410984294216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004571197317275346,
      "loss": 6.1216,
      "mean_token_accuracy": 0.14770945757627488,
      "num_tokens": 36883410.0,
      "step": 11040
    },
    {
      "entropy": 6.366364860534668,
      "epoch": 2.1547166130133646,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004570784205875278,
      "loss": 6.1313,
      "mean_token_accuracy": 0.1368573896586895,
      "num_tokens": 36900270.0,
      "step": 11045
    },
    {
      "entropy": 6.24760365486145,
      "epoch": 2.1556921275973076,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004570370916554665,
      "loss": 6.0035,
      "mean_token_accuracy": 0.14915569126605988,
      "num_tokens": 36916076.0,
      "step": 11050
    },
    {
      "entropy": 6.3359733581542965,
      "epoch": 2.1566676421812505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045699574493538966,
      "loss": 6.11,
      "mean_token_accuracy": 0.1447422742843628,
      "num_tokens": 36931966.0,
      "step": 11055
    },
    {
      "entropy": 6.28674693107605,
      "epoch": 2.1576431567651935,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045695438043133707,
      "loss": 6.044,
      "mean_token_accuracy": 0.1478276640176773,
      "num_tokens": 36946615.0,
      "step": 11060
    },
    {
      "entropy": 6.322678422927856,
      "epoch": 2.158618671349137,
      "grad_norm": 1.1875,
      "learning_rate": 0.000456912998147351,
      "loss": 6.0021,
      "mean_token_accuracy": 0.14986857771873474,
      "num_tokens": 36964830.0,
      "step": 11065
    },
    {
      "entropy": 6.363794231414795,
      "epoch": 2.15959418593308,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004568715980874751,
      "loss": 6.1282,
      "mean_token_accuracy": 0.13961051553487777,
      "num_tokens": 36982248.0,
      "step": 11070
    },
    {
      "entropy": 6.262344598770142,
      "epoch": 2.160569700517023,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004568301802557548,
      "loss": 6.0316,
      "mean_token_accuracy": 0.14778862968087197,
      "num_tokens": 36998543.0,
      "step": 11075
    },
    {
      "entropy": 6.295410871505737,
      "epoch": 2.1615452151009658,
      "grad_norm": 1.125,
      "learning_rate": 0.0004567887446562373,
      "loss": 6.031,
      "mean_token_accuracy": 0.14784975498914718,
      "num_tokens": 37014385.0,
      "step": 11080
    },
    {
      "entropy": 6.307573938369751,
      "epoch": 2.1625207296849087,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045674729129297165,
      "loss": 6.0976,
      "mean_token_accuracy": 0.13842833712697028,
      "num_tokens": 37030562.0,
      "step": 11085
    },
    {
      "entropy": 6.321400356292725,
      "epoch": 2.1634962442688517,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045670582017000847,
      "loss": 6.0161,
      "mean_token_accuracy": 0.14629049301147462,
      "num_tokens": 37045618.0,
      "step": 11090
    },
    {
      "entropy": 6.298821592330933,
      "epoch": 2.1644717588527946,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004566643312914001,
      "loss": 6.0705,
      "mean_token_accuracy": 0.1480104498565197,
      "num_tokens": 37061356.0,
      "step": 11095
    },
    {
      "entropy": 6.27530083656311,
      "epoch": 2.165447273436738,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045662282466120077,
      "loss": 5.974,
      "mean_token_accuracy": 0.15751174092292786,
      "num_tokens": 37078250.0,
      "step": 11100
    },
    {
      "entropy": 6.367358064651489,
      "epoch": 2.166422788020681,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004565813002834663,
      "loss": 6.1097,
      "mean_token_accuracy": 0.13552919626235962,
      "num_tokens": 37093774.0,
      "step": 11105
    },
    {
      "entropy": 6.330095100402832,
      "epoch": 2.167398302604624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045653975816225433,
      "loss": 6.1093,
      "mean_token_accuracy": 0.14223026856780052,
      "num_tokens": 37109720.0,
      "step": 11110
    },
    {
      "entropy": 6.123889446258545,
      "epoch": 2.168373817188567,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004564981983016242,
      "loss": 5.8101,
      "mean_token_accuracy": 0.16739074587821962,
      "num_tokens": 37126872.0,
      "step": 11115
    },
    {
      "entropy": 6.295429372787476,
      "epoch": 2.16934933177251,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004564566207056371,
      "loss": 6.0652,
      "mean_token_accuracy": 0.14344586357474326,
      "num_tokens": 37144227.0,
      "step": 11120
    },
    {
      "entropy": 6.378521490097046,
      "epoch": 2.170324846356453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045641502537835567,
      "loss": 6.1574,
      "mean_token_accuracy": 0.13739594966173171,
      "num_tokens": 37161419.0,
      "step": 11125
    },
    {
      "entropy": 6.337660932540894,
      "epoch": 2.1713003609403962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045637341232384454,
      "loss": 6.0572,
      "mean_token_accuracy": 0.1431183136999607,
      "num_tokens": 37177070.0,
      "step": 11130
    },
    {
      "entropy": 6.270116710662842,
      "epoch": 2.172275875524339,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045633178154617,
      "loss": 5.9278,
      "mean_token_accuracy": 0.15373654812574386,
      "num_tokens": 37193910.0,
      "step": 11135
    },
    {
      "entropy": 6.255342674255371,
      "epoch": 2.173251390108282,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045629013304940004,
      "loss": 6.1156,
      "mean_token_accuracy": 0.14278564602136612,
      "num_tokens": 37210636.0,
      "step": 11140
    },
    {
      "entropy": 6.314127111434937,
      "epoch": 2.174226904692225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004562484668376044,
      "loss": 6.0519,
      "mean_token_accuracy": 0.15140217691659927,
      "num_tokens": 37229194.0,
      "step": 11145
    },
    {
      "entropy": 6.4186314105987545,
      "epoch": 2.175202419276168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045620678291485454,
      "loss": 6.1354,
      "mean_token_accuracy": 0.13895186185836791,
      "num_tokens": 37245409.0,
      "step": 11150
    },
    {
      "entropy": 6.297395324707031,
      "epoch": 2.176177933860111,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045616508128522374,
      "loss": 6.1649,
      "mean_token_accuracy": 0.1371605932712555,
      "num_tokens": 37260892.0,
      "step": 11155
    },
    {
      "entropy": 6.29434723854065,
      "epoch": 2.177153448444054,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045612336195278687,
      "loss": 5.9574,
      "mean_token_accuracy": 0.1540080040693283,
      "num_tokens": 37277549.0,
      "step": 11160
    },
    {
      "entropy": 6.299536085128784,
      "epoch": 2.1781289630279974,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004560816249216206,
      "loss": 5.9769,
      "mean_token_accuracy": 0.15217790752649307,
      "num_tokens": 37295169.0,
      "step": 11165
    },
    {
      "entropy": 6.214374542236328,
      "epoch": 2.1791044776119404,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045603987019580326,
      "loss": 5.9763,
      "mean_token_accuracy": 0.15145750120282173,
      "num_tokens": 37311521.0,
      "step": 11170
    },
    {
      "entropy": 6.315257215499878,
      "epoch": 2.1800799921958833,
      "grad_norm": 1.125,
      "learning_rate": 0.0004559980977794152,
      "loss": 6.0833,
      "mean_token_accuracy": 0.14113064855337143,
      "num_tokens": 37328294.0,
      "step": 11175
    },
    {
      "entropy": 6.211073303222657,
      "epoch": 2.1810555067798263,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000455956307676538,
      "loss": 5.9439,
      "mean_token_accuracy": 0.1564269855618477,
      "num_tokens": 37344818.0,
      "step": 11180
    },
    {
      "entropy": 6.448201847076416,
      "epoch": 2.1820310213637693,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045591449989125526,
      "loss": 6.2067,
      "mean_token_accuracy": 0.1323664017021656,
      "num_tokens": 37361863.0,
      "step": 11185
    },
    {
      "entropy": 6.338916397094726,
      "epoch": 2.183006535947712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004558726744276526,
      "loss": 6.1062,
      "mean_token_accuracy": 0.14500006139278412,
      "num_tokens": 37377577.0,
      "step": 11190
    },
    {
      "entropy": 6.190906095504761,
      "epoch": 2.1839820505316556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045583083128981666,
      "loss": 6.0228,
      "mean_token_accuracy": 0.15434132218360902,
      "num_tokens": 37397204.0,
      "step": 11195
    },
    {
      "entropy": 6.368147850036621,
      "epoch": 2.1849575651155986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004557889704818365,
      "loss": 6.0995,
      "mean_token_accuracy": 0.1423078954219818,
      "num_tokens": 37414053.0,
      "step": 11200
    },
    {
      "entropy": 6.331431484222412,
      "epoch": 2.1859330796995415,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045574709200780236,
      "loss": 6.1171,
      "mean_token_accuracy": 0.13835633248090745,
      "num_tokens": 37431389.0,
      "step": 11205
    },
    {
      "entropy": 6.3285318374633786,
      "epoch": 2.1869085942834845,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045570519587180666,
      "loss": 6.0602,
      "mean_token_accuracy": 0.14592101275920868,
      "num_tokens": 37448330.0,
      "step": 11210
    },
    {
      "entropy": 6.391071605682373,
      "epoch": 2.1878841088674275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004556632820779433,
      "loss": 6.1769,
      "mean_token_accuracy": 0.13912276178598404,
      "num_tokens": 37465149.0,
      "step": 11215
    },
    {
      "entropy": 6.326007223129272,
      "epoch": 2.1888596234513704,
      "grad_norm": 1.125,
      "learning_rate": 0.00045562135063030785,
      "loss": 6.0861,
      "mean_token_accuracy": 0.14384347349405288,
      "num_tokens": 37481252.0,
      "step": 11220
    },
    {
      "entropy": 6.259375858306885,
      "epoch": 2.189835138035314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045557940153299783,
      "loss": 6.0688,
      "mean_token_accuracy": 0.1485600508749485,
      "num_tokens": 37498021.0,
      "step": 11225
    },
    {
      "entropy": 6.4022692203521725,
      "epoch": 2.190810652619257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004555374347901122,
      "loss": 6.0879,
      "mean_token_accuracy": 0.15341648310422898,
      "num_tokens": 37514715.0,
      "step": 11230
    },
    {
      "entropy": 6.331283044815064,
      "epoch": 2.1917861672031997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004554954504057519,
      "loss": 6.1303,
      "mean_token_accuracy": 0.1393424667418003,
      "num_tokens": 37531069.0,
      "step": 11235
    },
    {
      "entropy": 6.2525004863739015,
      "epoch": 2.1927616817871427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004554534483840195,
      "loss": 5.9668,
      "mean_token_accuracy": 0.15702627450227738,
      "num_tokens": 37548434.0,
      "step": 11240
    },
    {
      "entropy": 6.17685170173645,
      "epoch": 2.1937371963710857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004554114287290194,
      "loss": 5.8847,
      "mean_token_accuracy": 0.1522437497973442,
      "num_tokens": 37564233.0,
      "step": 11245
    },
    {
      "entropy": 6.366413211822509,
      "epoch": 2.1947127109550286,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045536939144485733,
      "loss": 6.1348,
      "mean_token_accuracy": 0.13927585035562515,
      "num_tokens": 37579356.0,
      "step": 11250
    },
    {
      "entropy": 6.263905763626099,
      "epoch": 2.195688225538972,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004553273365356412,
      "loss": 6.0561,
      "mean_token_accuracy": 0.137955678999424,
      "num_tokens": 37595458.0,
      "step": 11255
    },
    {
      "entropy": 6.338302850723267,
      "epoch": 2.196663740122915,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045528526400548044,
      "loss": 6.1159,
      "mean_token_accuracy": 0.14469828233122825,
      "num_tokens": 37613421.0,
      "step": 11260
    },
    {
      "entropy": 6.122881841659546,
      "epoch": 2.197639254706858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045524317385848623,
      "loss": 5.8183,
      "mean_token_accuracy": 0.166697908192873,
      "num_tokens": 37630610.0,
      "step": 11265
    },
    {
      "entropy": 6.214668464660645,
      "epoch": 2.198614769290801,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045520106609877144,
      "loss": 5.9598,
      "mean_token_accuracy": 0.14662431851029395,
      "num_tokens": 37646520.0,
      "step": 11270
    },
    {
      "entropy": 6.21582407951355,
      "epoch": 2.199590283874744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004551589407304508,
      "loss": 5.9957,
      "mean_token_accuracy": 0.14900809600949289,
      "num_tokens": 37662874.0,
      "step": 11275
    },
    {
      "entropy": 6.2054389953613285,
      "epoch": 2.200565798458687,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045511679775764053,
      "loss": 5.9579,
      "mean_token_accuracy": 0.1633349195122719,
      "num_tokens": 37679745.0,
      "step": 11280
    },
    {
      "entropy": 6.376394653320313,
      "epoch": 2.20154131304263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045507463718445876,
      "loss": 6.1104,
      "mean_token_accuracy": 0.14479258581995963,
      "num_tokens": 37697681.0,
      "step": 11285
    },
    {
      "entropy": 6.364546632766723,
      "epoch": 2.202516827626573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045503245901502517,
      "loss": 6.1551,
      "mean_token_accuracy": 0.1407507613301277,
      "num_tokens": 37714333.0,
      "step": 11290
    },
    {
      "entropy": 6.343272304534912,
      "epoch": 2.203492342210516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004549902632534614,
      "loss": 6.1059,
      "mean_token_accuracy": 0.1412137784063816,
      "num_tokens": 37730439.0,
      "step": 11295
    },
    {
      "entropy": 6.272779369354248,
      "epoch": 2.204467856794459,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045494804990389053,
      "loss": 5.9542,
      "mean_token_accuracy": 0.1501714766025543,
      "num_tokens": 37746836.0,
      "step": 11300
    },
    {
      "entropy": 6.295292901992798,
      "epoch": 2.205443371378402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045490581897043766,
      "loss": 6.011,
      "mean_token_accuracy": 0.14481884464621544,
      "num_tokens": 37763755.0,
      "step": 11305
    },
    {
      "entropy": 6.342422676086426,
      "epoch": 2.206418885962345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004548635704572293,
      "loss": 6.1539,
      "mean_token_accuracy": 0.13867631629109384,
      "num_tokens": 37780065.0,
      "step": 11310
    },
    {
      "entropy": 6.341325855255127,
      "epoch": 2.207394400546288,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004548213043683939,
      "loss": 6.0954,
      "mean_token_accuracy": 0.14121278002858162,
      "num_tokens": 37796630.0,
      "step": 11315
    },
    {
      "entropy": 6.3688348770141605,
      "epoch": 2.2083699151302314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045477902070806156,
      "loss": 6.1148,
      "mean_token_accuracy": 0.13888032138347625,
      "num_tokens": 37812257.0,
      "step": 11320
    },
    {
      "entropy": 6.251611709594727,
      "epoch": 2.2093454297141744,
      "grad_norm": 1.3125,
      "learning_rate": 0.000454736719480364,
      "loss": 6.0195,
      "mean_token_accuracy": 0.14746871888637542,
      "num_tokens": 37829702.0,
      "step": 11325
    },
    {
      "entropy": 6.313324403762818,
      "epoch": 2.2103209442981173,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004546944006894348,
      "loss": 6.1139,
      "mean_token_accuracy": 0.13852482587099074,
      "num_tokens": 37846335.0,
      "step": 11330
    },
    {
      "entropy": 6.279605054855347,
      "epoch": 2.2112964588820603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004546520643394093,
      "loss": 5.9833,
      "mean_token_accuracy": 0.14558463171124458,
      "num_tokens": 37863341.0,
      "step": 11335
    },
    {
      "entropy": 6.253409767150879,
      "epoch": 2.2122719734660032,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004546097104344242,
      "loss": 6.057,
      "mean_token_accuracy": 0.1529025763273239,
      "num_tokens": 37879277.0,
      "step": 11340
    },
    {
      "entropy": 6.289165496826172,
      "epoch": 2.213247488049946,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004545673389786185,
      "loss": 6.0005,
      "mean_token_accuracy": 0.14708464741706848,
      "num_tokens": 37895719.0,
      "step": 11345
    },
    {
      "entropy": 6.189546012878418,
      "epoch": 2.214223002633889,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045452494997613236,
      "loss": 5.9586,
      "mean_token_accuracy": 0.15204283446073533,
      "num_tokens": 37912799.0,
      "step": 11350
    },
    {
      "entropy": 6.35507607460022,
      "epoch": 2.2151985172178326,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000454482543431108,
      "loss": 6.0444,
      "mean_token_accuracy": 0.1425374425947666,
      "num_tokens": 37930875.0,
      "step": 11355
    },
    {
      "entropy": 6.210252857208252,
      "epoch": 2.2161740318017755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004544401193476891,
      "loss": 6.0182,
      "mean_token_accuracy": 0.1573004923760891,
      "num_tokens": 37948545.0,
      "step": 11360
    },
    {
      "entropy": 6.292483472824097,
      "epoch": 2.2171495463857185,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004543976777300213,
      "loss": 5.981,
      "mean_token_accuracy": 0.1534133031964302,
      "num_tokens": 37965967.0,
      "step": 11365
    },
    {
      "entropy": 6.076048707962036,
      "epoch": 2.2181250609696614,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045435521858225184,
      "loss": 5.8485,
      "mean_token_accuracy": 0.16972576677799225,
      "num_tokens": 37984285.0,
      "step": 11370
    },
    {
      "entropy": 6.3944840908050535,
      "epoch": 2.2191005755536044,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004543127419085296,
      "loss": 6.1464,
      "mean_token_accuracy": 0.13775423988699914,
      "num_tokens": 38001022.0,
      "step": 11375
    },
    {
      "entropy": 6.278937244415284,
      "epoch": 2.2200760901375474,
      "grad_norm": 1.0,
      "learning_rate": 0.00045427024771300527,
      "loss": 6.0824,
      "mean_token_accuracy": 0.1511298932135105,
      "num_tokens": 38017724.0,
      "step": 11380
    },
    {
      "entropy": 6.304081964492798,
      "epoch": 2.2210516047214908,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045422773599983137,
      "loss": 6.0853,
      "mean_token_accuracy": 0.14059553667902946,
      "num_tokens": 38033873.0,
      "step": 11385
    },
    {
      "entropy": 6.311473321914673,
      "epoch": 2.2220271193054337,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045418520677316187,
      "loss": 6.0309,
      "mean_token_accuracy": 0.147953300178051,
      "num_tokens": 38049705.0,
      "step": 11390
    },
    {
      "entropy": 6.324868965148926,
      "epoch": 2.2230026338893767,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004541426600371526,
      "loss": 6.0755,
      "mean_token_accuracy": 0.14711323156952857,
      "num_tokens": 38066756.0,
      "step": 11395
    },
    {
      "entropy": 6.421079158782959,
      "epoch": 2.2239781484733196,
      "grad_norm": 1.1875,
      "learning_rate": 0.000454100095795961,
      "loss": 6.1602,
      "mean_token_accuracy": 0.13620849549770356,
      "num_tokens": 38083103.0,
      "step": 11400
    },
    {
      "entropy": 6.255079174041748,
      "epoch": 2.2249536630572626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004540575140537464,
      "loss": 5.9884,
      "mean_token_accuracy": 0.1512216880917549,
      "num_tokens": 38100149.0,
      "step": 11405
    },
    {
      "entropy": 6.221665287017823,
      "epoch": 2.2259291776412056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045401491481466974,
      "loss": 5.9888,
      "mean_token_accuracy": 0.15392559319734572,
      "num_tokens": 38118295.0,
      "step": 11410
    },
    {
      "entropy": 6.347079849243164,
      "epoch": 2.226904692225149,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004539722980828935,
      "loss": 6.0534,
      "mean_token_accuracy": 0.1415476158261299,
      "num_tokens": 38134861.0,
      "step": 11415
    },
    {
      "entropy": 6.3304530620574955,
      "epoch": 2.227880206809092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045392966386258225,
      "loss": 6.0874,
      "mean_token_accuracy": 0.1442916102707386,
      "num_tokens": 38151070.0,
      "step": 11420
    },
    {
      "entropy": 6.276222229003906,
      "epoch": 2.228855721393035,
      "grad_norm": 1.078125,
      "learning_rate": 0.000453887012157902,
      "loss": 6.0259,
      "mean_token_accuracy": 0.15072644576430322,
      "num_tokens": 38168939.0,
      "step": 11425
    },
    {
      "entropy": 6.27154221534729,
      "epoch": 2.229831235976978,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045384434297302044,
      "loss": 6.0725,
      "mean_token_accuracy": 0.14397984594106675,
      "num_tokens": 38185779.0,
      "step": 11430
    },
    {
      "entropy": 6.40225772857666,
      "epoch": 2.230806750560921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045380165631210713,
      "loss": 6.1222,
      "mean_token_accuracy": 0.14208736270666122,
      "num_tokens": 38203415.0,
      "step": 11435
    },
    {
      "entropy": 6.368290758132934,
      "epoch": 2.2317822651448638,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004537589521793332,
      "loss": 6.0572,
      "mean_token_accuracy": 0.1402736485004425,
      "num_tokens": 38219953.0,
      "step": 11440
    },
    {
      "entropy": 6.384019279479981,
      "epoch": 2.232757779728807,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004537162305788716,
      "loss": 6.144,
      "mean_token_accuracy": 0.1385796383023262,
      "num_tokens": 38236258.0,
      "step": 11445
    },
    {
      "entropy": 6.360015535354615,
      "epoch": 2.23373329431275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045367349151489685,
      "loss": 6.1498,
      "mean_token_accuracy": 0.13953615874052047,
      "num_tokens": 38252878.0,
      "step": 11450
    },
    {
      "entropy": 6.232423782348633,
      "epoch": 2.234708808896693,
      "grad_norm": 1.125,
      "learning_rate": 0.0004536307349915853,
      "loss": 5.9587,
      "mean_token_accuracy": 0.16083403825759887,
      "num_tokens": 38270215.0,
      "step": 11455
    },
    {
      "entropy": 6.176731061935425,
      "epoch": 2.235684323480636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000453587961013115,
      "loss": 5.9929,
      "mean_token_accuracy": 0.15173392593860627,
      "num_tokens": 38287970.0,
      "step": 11460
    },
    {
      "entropy": 6.2655144214630125,
      "epoch": 2.236659838064579,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045354516958366565,
      "loss": 5.9762,
      "mean_token_accuracy": 0.1559681087732315,
      "num_tokens": 38305345.0,
      "step": 11465
    },
    {
      "entropy": 6.263467073440552,
      "epoch": 2.237635352648522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004535023607074187,
      "loss": 6.0056,
      "mean_token_accuracy": 0.15326197668910027,
      "num_tokens": 38322410.0,
      "step": 11470
    },
    {
      "entropy": 6.265497922897339,
      "epoch": 2.238610867232465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045345953438855725,
      "loss": 6.0156,
      "mean_token_accuracy": 0.14782543778419494,
      "num_tokens": 38339708.0,
      "step": 11475
    },
    {
      "entropy": 6.384361267089844,
      "epoch": 2.2395863818164083,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004534166906312661,
      "loss": 6.0891,
      "mean_token_accuracy": 0.14603593051433564,
      "num_tokens": 38356776.0,
      "step": 11480
    },
    {
      "entropy": 6.299145126342774,
      "epoch": 2.2405618964003513,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004533738294397319,
      "loss": 6.0205,
      "mean_token_accuracy": 0.14812699407339097,
      "num_tokens": 38372916.0,
      "step": 11485
    },
    {
      "entropy": 6.248420143127442,
      "epoch": 2.2415374109842943,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045333095081814275,
      "loss": 6.0474,
      "mean_token_accuracy": 0.1496947303414345,
      "num_tokens": 38389401.0,
      "step": 11490
    },
    {
      "entropy": 6.3306882858276365,
      "epoch": 2.2425129255682372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045328805477068873,
      "loss": 6.0629,
      "mean_token_accuracy": 0.14868413656949997,
      "num_tokens": 38404985.0,
      "step": 11495
    },
    {
      "entropy": 6.3090431690216064,
      "epoch": 2.24348844015218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004532451413015614,
      "loss": 6.0626,
      "mean_token_accuracy": 0.1490450918674469,
      "num_tokens": 38420748.0,
      "step": 11500
    },
    {
      "entropy": 6.34416937828064,
      "epoch": 2.244463954736123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004532022104149541,
      "loss": 6.1243,
      "mean_token_accuracy": 0.14566347301006316,
      "num_tokens": 38437043.0,
      "step": 11505
    },
    {
      "entropy": 6.331855344772339,
      "epoch": 2.2454394693200666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000453159262115062,
      "loss": 6.1323,
      "mean_token_accuracy": 0.14682493656873702,
      "num_tokens": 38453469.0,
      "step": 11510
    },
    {
      "entropy": 6.410295486450195,
      "epoch": 2.2464149839040095,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004531162964060817,
      "loss": 6.1282,
      "mean_token_accuracy": 0.13948519229888917,
      "num_tokens": 38469162.0,
      "step": 11515
    },
    {
      "entropy": 6.391000270843506,
      "epoch": 2.2473904984879525,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004530733132922118,
      "loss": 6.1509,
      "mean_token_accuracy": 0.13573283851146697,
      "num_tokens": 38485425.0,
      "step": 11520
    },
    {
      "entropy": 6.104939603805542,
      "epoch": 2.2483660130718954,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045303031277765235,
      "loss": 5.7801,
      "mean_token_accuracy": 0.17179450243711472,
      "num_tokens": 38503437.0,
      "step": 11525
    },
    {
      "entropy": 6.297632694244385,
      "epoch": 2.2493415276558384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004529872948666052,
      "loss": 6.0459,
      "mean_token_accuracy": 0.14749653935432433,
      "num_tokens": 38520206.0,
      "step": 11530
    },
    {
      "entropy": 6.352658748626709,
      "epoch": 2.2503170422397814,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004529442595632741,
      "loss": 6.1096,
      "mean_token_accuracy": 0.14250742346048356,
      "num_tokens": 38536717.0,
      "step": 11535
    },
    {
      "entropy": 6.263523530960083,
      "epoch": 2.2512925568237243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045290120687186406,
      "loss": 6.0558,
      "mean_token_accuracy": 0.14891740903258324,
      "num_tokens": 38553930.0,
      "step": 11540
    },
    {
      "entropy": 6.273693609237671,
      "epoch": 2.2522680714076677,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004528581367965821,
      "loss": 5.9274,
      "mean_token_accuracy": 0.15725503116846085,
      "num_tokens": 38569798.0,
      "step": 11545
    },
    {
      "entropy": 6.200993013381958,
      "epoch": 2.2532435859916107,
      "grad_norm": 1.25,
      "learning_rate": 0.000452815049341637,
      "loss": 5.9511,
      "mean_token_accuracy": 0.1493321903049946,
      "num_tokens": 38587878.0,
      "step": 11550
    },
    {
      "entropy": 6.21127667427063,
      "epoch": 2.2542191005755536,
      "grad_norm": 1.125,
      "learning_rate": 0.000452771944511239,
      "loss": 5.9336,
      "mean_token_accuracy": 0.14924389794468879,
      "num_tokens": 38604601.0,
      "step": 11555
    },
    {
      "entropy": 6.285223007202148,
      "epoch": 2.2551946151594966,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004527288223096001,
      "loss": 5.959,
      "mean_token_accuracy": 0.1479227811098099,
      "num_tokens": 38621086.0,
      "step": 11560
    },
    {
      "entropy": 6.124997138977051,
      "epoch": 2.2561701297434396,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004526856827409342,
      "loss": 5.8525,
      "mean_token_accuracy": 0.1632825493812561,
      "num_tokens": 38637078.0,
      "step": 11565
    },
    {
      "entropy": 6.1947074890136715,
      "epoch": 2.2571456443273825,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045264252580945664,
      "loss": 5.9585,
      "mean_token_accuracy": 0.15949101895093917,
      "num_tokens": 38653957.0,
      "step": 11570
    },
    {
      "entropy": 6.283156967163086,
      "epoch": 2.258121158911326,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045259935151938454,
      "loss": 6.0982,
      "mean_token_accuracy": 0.1430782400071621,
      "num_tokens": 38671985.0,
      "step": 11575
    },
    {
      "entropy": 6.399336481094361,
      "epoch": 2.259096673495269,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004525561598749368,
      "loss": 6.1256,
      "mean_token_accuracy": 0.14154182225465775,
      "num_tokens": 38688412.0,
      "step": 11580
    },
    {
      "entropy": 6.311310958862305,
      "epoch": 2.260072188079212,
      "grad_norm": 1.171875,
      "learning_rate": 0.000452512950880334,
      "loss": 6.0369,
      "mean_token_accuracy": 0.1457868278026581,
      "num_tokens": 38705442.0,
      "step": 11585
    },
    {
      "entropy": 6.306019926071167,
      "epoch": 2.261047702663155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004524697245397983,
      "loss": 6.1388,
      "mean_token_accuracy": 0.1405431978404522,
      "num_tokens": 38722094.0,
      "step": 11590
    },
    {
      "entropy": 6.297365331649781,
      "epoch": 2.2620232172470978,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004524264808575536,
      "loss": 6.0401,
      "mean_token_accuracy": 0.15699629187583924,
      "num_tokens": 38738287.0,
      "step": 11595
    },
    {
      "entropy": 6.353013229370117,
      "epoch": 2.2629987318310407,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045238321983782556,
      "loss": 6.0752,
      "mean_token_accuracy": 0.14521036073565483,
      "num_tokens": 38754374.0,
      "step": 11600
    },
    {
      "entropy": 6.2874367237091064,
      "epoch": 2.2639742464149837,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045233994148484154,
      "loss": 6.0226,
      "mean_token_accuracy": 0.1483229413628578,
      "num_tokens": 38770397.0,
      "step": 11605
    },
    {
      "entropy": 6.290653705596924,
      "epoch": 2.264949760998927,
      "grad_norm": 1.125,
      "learning_rate": 0.00045229664580283036,
      "loss": 6.0788,
      "mean_token_accuracy": 0.1470076248049736,
      "num_tokens": 38785657.0,
      "step": 11610
    },
    {
      "entropy": 6.335720539093018,
      "epoch": 2.26592527558287,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004522533327960229,
      "loss": 6.0761,
      "mean_token_accuracy": 0.14463562816381453,
      "num_tokens": 38801644.0,
      "step": 11615
    },
    {
      "entropy": 6.3558539867401125,
      "epoch": 2.266900790166813,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045221000246865157,
      "loss": 6.1023,
      "mean_token_accuracy": 0.13997844457626343,
      "num_tokens": 38818406.0,
      "step": 11620
    },
    {
      "entropy": 6.19734296798706,
      "epoch": 2.267876304750756,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004521666548249503,
      "loss": 6.0057,
      "mean_token_accuracy": 0.15307257398962976,
      "num_tokens": 38834805.0,
      "step": 11625
    },
    {
      "entropy": 6.434165525436401,
      "epoch": 2.268851819334699,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045212328986915496,
      "loss": 6.1608,
      "mean_token_accuracy": 0.14514754563570023,
      "num_tokens": 38850571.0,
      "step": 11630
    },
    {
      "entropy": 6.277410459518433,
      "epoch": 2.2698273339186423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045207990760550305,
      "loss": 5.9858,
      "mean_token_accuracy": 0.15320515334606172,
      "num_tokens": 38867476.0,
      "step": 11635
    },
    {
      "entropy": 6.246854972839356,
      "epoch": 2.2708028485025853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004520365080382337,
      "loss": 6.0838,
      "mean_token_accuracy": 0.1463797815144062,
      "num_tokens": 38882979.0,
      "step": 11640
    },
    {
      "entropy": 6.27082576751709,
      "epoch": 2.2717783630865283,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045199309117158773,
      "loss": 5.9542,
      "mean_token_accuracy": 0.15088210478425026,
      "num_tokens": 38899585.0,
      "step": 11645
    },
    {
      "entropy": 6.352978754043579,
      "epoch": 2.272753877670471,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004519496570098077,
      "loss": 6.1013,
      "mean_token_accuracy": 0.14001759514212608,
      "num_tokens": 38915707.0,
      "step": 11650
    },
    {
      "entropy": 6.261731672286987,
      "epoch": 2.273729392254414,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004519062055571379,
      "loss": 6.0137,
      "mean_token_accuracy": 0.1464573211967945,
      "num_tokens": 38931608.0,
      "step": 11655
    },
    {
      "entropy": 6.269374132156372,
      "epoch": 2.274704906838357,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004518627368178241,
      "loss": 6.0182,
      "mean_token_accuracy": 0.14834654852747917,
      "num_tokens": 38948675.0,
      "step": 11660
    },
    {
      "entropy": 6.2643575191497805,
      "epoch": 2.2756804214223,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004518192507961141,
      "loss": 6.0128,
      "mean_token_accuracy": 0.15395441353321077,
      "num_tokens": 38964568.0,
      "step": 11665
    },
    {
      "entropy": 6.30595293045044,
      "epoch": 2.276655936006243,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045177574749625713,
      "loss": 6.0446,
      "mean_token_accuracy": 0.14948033019900323,
      "num_tokens": 38981192.0,
      "step": 11670
    },
    {
      "entropy": 6.325563859939575,
      "epoch": 2.2776314505901865,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045173222692250405,
      "loss": 6.0054,
      "mean_token_accuracy": 0.14628474414348602,
      "num_tokens": 38997160.0,
      "step": 11675
    },
    {
      "entropy": 6.311353015899658,
      "epoch": 2.2786069651741294,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045168868907910763,
      "loss": 6.0746,
      "mean_token_accuracy": 0.1459297150373459,
      "num_tokens": 39013834.0,
      "step": 11680
    },
    {
      "entropy": 6.338512659072876,
      "epoch": 2.2795824797580724,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004516451339703224,
      "loss": 6.0789,
      "mean_token_accuracy": 0.13714150488376617,
      "num_tokens": 39029935.0,
      "step": 11685
    },
    {
      "entropy": 6.386590099334716,
      "epoch": 2.2805579943420153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045160156160040403,
      "loss": 6.0963,
      "mean_token_accuracy": 0.14036734849214555,
      "num_tokens": 39048022.0,
      "step": 11690
    },
    {
      "entropy": 6.372249269485474,
      "epoch": 2.2815335089259583,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004515579719736107,
      "loss": 6.1553,
      "mean_token_accuracy": 0.13512196242809296,
      "num_tokens": 39064036.0,
      "step": 11695
    },
    {
      "entropy": 6.242185735702515,
      "epoch": 2.2825090235099017,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004515143650942015,
      "loss": 5.9775,
      "mean_token_accuracy": 0.15831967443227768,
      "num_tokens": 39081607.0,
      "step": 11700
    },
    {
      "entropy": 6.317063570022583,
      "epoch": 2.2834845380938447,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045147074096643766,
      "loss": 5.9576,
      "mean_token_accuracy": 0.15396093726158142,
      "num_tokens": 39098101.0,
      "step": 11705
    },
    {
      "entropy": 6.2532007694244385,
      "epoch": 2.2844600526777876,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045142709959458194,
      "loss": 6.0643,
      "mean_token_accuracy": 0.14849617555737496,
      "num_tokens": 39115381.0,
      "step": 11710
    },
    {
      "entropy": 6.364624071121216,
      "epoch": 2.2854355672617306,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045138344098289893,
      "loss": 6.0937,
      "mean_token_accuracy": 0.1426037445664406,
      "num_tokens": 39131601.0,
      "step": 11715
    },
    {
      "entropy": 6.2795257568359375,
      "epoch": 2.2864110818456735,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045133976513565463,
      "loss": 5.9714,
      "mean_token_accuracy": 0.15361961126327514,
      "num_tokens": 39146551.0,
      "step": 11720
    },
    {
      "entropy": 6.23606915473938,
      "epoch": 2.2873865964296165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045129607205711696,
      "loss": 5.9481,
      "mean_token_accuracy": 0.14945898652076722,
      "num_tokens": 39163519.0,
      "step": 11725
    },
    {
      "entropy": 6.287294006347656,
      "epoch": 2.2883621110135595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004512523617515555,
      "loss": 6.0323,
      "mean_token_accuracy": 0.1447359524667263,
      "num_tokens": 39179428.0,
      "step": 11730
    },
    {
      "entropy": 6.282284116744995,
      "epoch": 2.289337625597503,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045120863422324145,
      "loss": 6.0387,
      "mean_token_accuracy": 0.1539314478635788,
      "num_tokens": 39196384.0,
      "step": 11735
    },
    {
      "entropy": 6.286980056762696,
      "epoch": 2.290313140181446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004511648894764476,
      "loss": 6.0214,
      "mean_token_accuracy": 0.14681141823530197,
      "num_tokens": 39213375.0,
      "step": 11740
    },
    {
      "entropy": 6.299848318099976,
      "epoch": 2.291288654765389,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004511211275154487,
      "loss": 6.0479,
      "mean_token_accuracy": 0.1430096685886383,
      "num_tokens": 39229697.0,
      "step": 11745
    },
    {
      "entropy": 6.245846366882324,
      "epoch": 2.2922641693493317,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045107734834452085,
      "loss": 6.0002,
      "mean_token_accuracy": 0.15243843272328378,
      "num_tokens": 39245763.0,
      "step": 11750
    },
    {
      "entropy": 6.342855548858642,
      "epoch": 2.2932396839332747,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004510335519679421,
      "loss": 6.0737,
      "mean_token_accuracy": 0.1404600240290165,
      "num_tokens": 39262298.0,
      "step": 11755
    },
    {
      "entropy": 6.3062365531921385,
      "epoch": 2.2942151985172177,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004509897383899921,
      "loss": 5.9722,
      "mean_token_accuracy": 0.1519713930785656,
      "num_tokens": 39278532.0,
      "step": 11760
    },
    {
      "entropy": 6.287449884414673,
      "epoch": 2.295190713101161,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045094590761495207,
      "loss": 6.1146,
      "mean_token_accuracy": 0.14025090485811234,
      "num_tokens": 39294542.0,
      "step": 11765
    },
    {
      "entropy": 6.295427656173706,
      "epoch": 2.296166227685104,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045090205964710495,
      "loss": 6.0411,
      "mean_token_accuracy": 0.14164887592196465,
      "num_tokens": 39311318.0,
      "step": 11770
    },
    {
      "entropy": 6.344098377227783,
      "epoch": 2.297141742269047,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004508581944907356,
      "loss": 6.0639,
      "mean_token_accuracy": 0.14884432554244995,
      "num_tokens": 39328633.0,
      "step": 11775
    },
    {
      "entropy": 6.333061647415161,
      "epoch": 2.29811725685299,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508143121501302,
      "loss": 6.1334,
      "mean_token_accuracy": 0.1416197083890438,
      "num_tokens": 39345063.0,
      "step": 11780
    },
    {
      "entropy": 6.351585483551025,
      "epoch": 2.299092771436933,
      "grad_norm": 1.125,
      "learning_rate": 0.00045077041262957685,
      "loss": 6.0886,
      "mean_token_accuracy": 0.14890801012516022,
      "num_tokens": 39360163.0,
      "step": 11785
    },
    {
      "entropy": 6.226489782333374,
      "epoch": 2.300068286020876,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004507264959333652,
      "loss": 5.9942,
      "mean_token_accuracy": 0.15249196588993072,
      "num_tokens": 39376888.0,
      "step": 11790
    },
    {
      "entropy": 6.337301540374756,
      "epoch": 2.301043800604819,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045068256206578665,
      "loss": 6.0649,
      "mean_token_accuracy": 0.14445698112249375,
      "num_tokens": 39394544.0,
      "step": 11795
    },
    {
      "entropy": 6.315564298629761,
      "epoch": 2.3020193151887622,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045063861103113433,
      "loss": 5.9808,
      "mean_token_accuracy": 0.1503424435853958,
      "num_tokens": 39410118.0,
      "step": 11800
    },
    {
      "entropy": 6.330679988861084,
      "epoch": 2.302994829772705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045059464283370295,
      "loss": 6.1235,
      "mean_token_accuracy": 0.1395067445933819,
      "num_tokens": 39427307.0,
      "step": 11805
    },
    {
      "entropy": 6.288217973709107,
      "epoch": 2.303970344356648,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505506574777888,
      "loss": 6.1006,
      "mean_token_accuracy": 0.13832788914442062,
      "num_tokens": 39443400.0,
      "step": 11810
    },
    {
      "entropy": 6.277277183532715,
      "epoch": 2.304945858940591,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045050665496769015,
      "loss": 6.0586,
      "mean_token_accuracy": 0.14200622588396072,
      "num_tokens": 39459392.0,
      "step": 11815
    },
    {
      "entropy": 6.338297510147095,
      "epoch": 2.305921373524534,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004504626353077068,
      "loss": 6.1169,
      "mean_token_accuracy": 0.13958825096487998,
      "num_tokens": 39475795.0,
      "step": 11820
    },
    {
      "entropy": 6.282676696777344,
      "epoch": 2.306896888108477,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004504185985021399,
      "loss": 6.0176,
      "mean_token_accuracy": 0.15281772911548613,
      "num_tokens": 39492883.0,
      "step": 11825
    },
    {
      "entropy": 6.29112982749939,
      "epoch": 2.3078724026924204,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045037454455529286,
      "loss": 6.0534,
      "mean_token_accuracy": 0.1457516983151436,
      "num_tokens": 39510834.0,
      "step": 11830
    },
    {
      "entropy": 6.4144855499267575,
      "epoch": 2.3088479172763634,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045033047347147034,
      "loss": 6.1268,
      "mean_token_accuracy": 0.14280546456575394,
      "num_tokens": 39527052.0,
      "step": 11835
    },
    {
      "entropy": 6.286240673065185,
      "epoch": 2.3098234318603064,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045028638525497884,
      "loss": 5.9978,
      "mean_token_accuracy": 0.1529225990176201,
      "num_tokens": 39543694.0,
      "step": 11840
    },
    {
      "entropy": 6.24630970954895,
      "epoch": 2.3107989464442493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004502422799101265,
      "loss": 6.0009,
      "mean_token_accuracy": 0.15159681290388108,
      "num_tokens": 39561223.0,
      "step": 11845
    },
    {
      "entropy": 6.334225225448608,
      "epoch": 2.3117744610281923,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045019815744122316,
      "loss": 6.0551,
      "mean_token_accuracy": 0.1465018004179001,
      "num_tokens": 39577513.0,
      "step": 11850
    },
    {
      "entropy": 6.302358198165893,
      "epoch": 2.3127499756121352,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045015401785258034,
      "loss": 5.9881,
      "mean_token_accuracy": 0.1497272416949272,
      "num_tokens": 39595172.0,
      "step": 11855
    },
    {
      "entropy": 6.2433114528656,
      "epoch": 2.313725490196078,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045010986114851103,
      "loss": 6.0437,
      "mean_token_accuracy": 0.15781751722097398,
      "num_tokens": 39611285.0,
      "step": 11860
    },
    {
      "entropy": 6.383315801620483,
      "epoch": 2.3147010047800216,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045006568733333026,
      "loss": 6.1231,
      "mean_token_accuracy": 0.13524405658245087,
      "num_tokens": 39628545.0,
      "step": 11865
    },
    {
      "entropy": 6.383442020416259,
      "epoch": 2.3156765193639646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004500214964113545,
      "loss": 6.0554,
      "mean_token_accuracy": 0.14483529329299927,
      "num_tokens": 39645868.0,
      "step": 11870
    },
    {
      "entropy": 6.267084550857544,
      "epoch": 2.3166520339479075,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004499772883869018,
      "loss": 6.0607,
      "mean_token_accuracy": 0.1463887169957161,
      "num_tokens": 39662781.0,
      "step": 11875
    },
    {
      "entropy": 6.295947027206421,
      "epoch": 2.3176275485318505,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004499330632642922,
      "loss": 6.0361,
      "mean_token_accuracy": 0.15231706872582434,
      "num_tokens": 39679675.0,
      "step": 11880
    },
    {
      "entropy": 6.329856920242309,
      "epoch": 2.3186030631157935,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044988882104784713,
      "loss": 6.0962,
      "mean_token_accuracy": 0.14306997880339622,
      "num_tokens": 39697602.0,
      "step": 11885
    },
    {
      "entropy": 6.3113641262054445,
      "epoch": 2.319578577699737,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044984456174188974,
      "loss": 6.0776,
      "mean_token_accuracy": 0.141340783983469,
      "num_tokens": 39713951.0,
      "step": 11890
    },
    {
      "entropy": 6.307555770874023,
      "epoch": 2.32055409228368,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044980028535074497,
      "loss": 6.0476,
      "mean_token_accuracy": 0.1510847568511963,
      "num_tokens": 39730648.0,
      "step": 11895
    },
    {
      "entropy": 6.245632219314575,
      "epoch": 2.3215296068676228,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044975599187873933,
      "loss": 5.9642,
      "mean_token_accuracy": 0.15377690121531487,
      "num_tokens": 39748228.0,
      "step": 11900
    },
    {
      "entropy": 6.27507758140564,
      "epoch": 2.3225051214515657,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044971168133020096,
      "loss": 6.0146,
      "mean_token_accuracy": 0.147166758030653,
      "num_tokens": 39765248.0,
      "step": 11905
    },
    {
      "entropy": 6.3325950622558596,
      "epoch": 2.3234806360355087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044966735370945973,
      "loss": 6.0232,
      "mean_token_accuracy": 0.14042915478348733,
      "num_tokens": 39781799.0,
      "step": 11910
    },
    {
      "entropy": 6.223082542419434,
      "epoch": 2.3244561506194517,
      "grad_norm": 1.125,
      "learning_rate": 0.0004496230090208473,
      "loss": 5.9669,
      "mean_token_accuracy": 0.15768114179372789,
      "num_tokens": 39798952.0,
      "step": 11915
    },
    {
      "entropy": 6.213258504867554,
      "epoch": 2.3254316652033946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004495786472686967,
      "loss": 6.0161,
      "mean_token_accuracy": 0.1536577172577381,
      "num_tokens": 39814923.0,
      "step": 11920
    },
    {
      "entropy": 6.294979476928711,
      "epoch": 2.3264071797873376,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004495342684573431,
      "loss": 6.0645,
      "mean_token_accuracy": 0.14394317716360092,
      "num_tokens": 39831629.0,
      "step": 11925
    },
    {
      "entropy": 6.261340379714966,
      "epoch": 2.327382694371281,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044948987259112264,
      "loss": 6.002,
      "mean_token_accuracy": 0.14681392312049865,
      "num_tokens": 39848103.0,
      "step": 11930
    },
    {
      "entropy": 6.297618579864502,
      "epoch": 2.328358208955224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044944545967437385,
      "loss": 6.0697,
      "mean_token_accuracy": 0.14484669864177704,
      "num_tokens": 39864969.0,
      "step": 11935
    },
    {
      "entropy": 6.320592021942138,
      "epoch": 2.329333723539167,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044940102971143637,
      "loss": 6.0764,
      "mean_token_accuracy": 0.14754758775234222,
      "num_tokens": 39880636.0,
      "step": 11940
    },
    {
      "entropy": 6.2356866836547855,
      "epoch": 2.33030923812311,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004493565827066519,
      "loss": 6.0312,
      "mean_token_accuracy": 0.14293385297060013,
      "num_tokens": 39898712.0,
      "step": 11945
    },
    {
      "entropy": 6.377697467803955,
      "epoch": 2.331284752707053,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044931211866436354,
      "loss": 6.0656,
      "mean_token_accuracy": 0.14663574621081352,
      "num_tokens": 39914601.0,
      "step": 11950
    },
    {
      "entropy": 6.238736534118653,
      "epoch": 2.3322602672909962,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004492676375889163,
      "loss": 6.0186,
      "mean_token_accuracy": 0.1483923077583313,
      "num_tokens": 39931564.0,
      "step": 11955
    },
    {
      "entropy": 6.243938255310058,
      "epoch": 2.333235781874939,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004492231394846565,
      "loss": 6.0244,
      "mean_token_accuracy": 0.1525435835123062,
      "num_tokens": 39949191.0,
      "step": 11960
    },
    {
      "entropy": 6.320635509490967,
      "epoch": 2.334211296458882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004491786243559325,
      "loss": 6.0112,
      "mean_token_accuracy": 0.14567956551909447,
      "num_tokens": 39965560.0,
      "step": 11965
    },
    {
      "entropy": 6.150753164291382,
      "epoch": 2.335186811042825,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004491340922070942,
      "loss": 5.8474,
      "mean_token_accuracy": 0.1627658486366272,
      "num_tokens": 39982113.0,
      "step": 11970
    },
    {
      "entropy": 6.207383871078491,
      "epoch": 2.336162325626768,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044908954304249287,
      "loss": 6.0337,
      "mean_token_accuracy": 0.14928779751062393,
      "num_tokens": 39998449.0,
      "step": 11975
    },
    {
      "entropy": 6.248074626922607,
      "epoch": 2.337137840210711,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044904497686648197,
      "loss": 6.0556,
      "mean_token_accuracy": 0.1490597054362297,
      "num_tokens": 40015419.0,
      "step": 11980
    },
    {
      "entropy": 6.315632486343384,
      "epoch": 2.338113354794654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044900039368341623,
      "loss": 6.0622,
      "mean_token_accuracy": 0.1445282816886902,
      "num_tokens": 40030698.0,
      "step": 11985
    },
    {
      "entropy": 6.198047494888305,
      "epoch": 2.3390888693785974,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044895579349765215,
      "loss": 5.9601,
      "mean_token_accuracy": 0.1578473225235939,
      "num_tokens": 40047385.0,
      "step": 11990
    },
    {
      "entropy": 6.35274977684021,
      "epoch": 2.3400643839625404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004489111763135479,
      "loss": 6.059,
      "mean_token_accuracy": 0.144583797454834,
      "num_tokens": 40062709.0,
      "step": 11995
    },
    {
      "entropy": 6.397759819030762,
      "epoch": 2.3410398985464833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044886654213546333,
      "loss": 6.124,
      "mean_token_accuracy": 0.1418415315449238,
      "num_tokens": 40079777.0,
      "step": 12000
    },
    {
      "epoch": 2.3410398985464833,
      "eval_entropy": 6.17707198746125,
      "eval_loss": 6.200720310211182,
      "eval_mean_token_accuracy": 0.14580881343485075,
      "eval_num_tokens": 40079777.0,
      "eval_runtime": 15.3512,
      "eval_samples_per_second": 1079.851,
      "eval_steps_per_second": 135.038,
      "step": 12000
    },
    {
      "entropy": 6.366254615783691,
      "epoch": 2.3420154131304263,
      "grad_norm": 1.125,
      "learning_rate": 0.00044882189096775993,
      "loss": 6.1225,
      "mean_token_accuracy": 0.1399042397737503,
      "num_tokens": 40096778.0,
      "step": 12005
    },
    {
      "entropy": 6.242882156372071,
      "epoch": 2.3429909277143692,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044877722281480093,
      "loss": 6.0305,
      "mean_token_accuracy": 0.1527033716440201,
      "num_tokens": 40112105.0,
      "step": 12010
    },
    {
      "entropy": 6.126497602462768,
      "epoch": 2.343966442298312,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000448732537680951,
      "loss": 5.8632,
      "mean_token_accuracy": 0.15735645294189454,
      "num_tokens": 40128611.0,
      "step": 12015
    },
    {
      "entropy": 6.400521659851075,
      "epoch": 2.3449419568822556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004486878355705767,
      "loss": 6.1451,
      "mean_token_accuracy": 0.13585125878453255,
      "num_tokens": 40145530.0,
      "step": 12020
    },
    {
      "entropy": 6.351988029479981,
      "epoch": 2.3459174714661986,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044864311648804613,
      "loss": 6.0925,
      "mean_token_accuracy": 0.1435452163219452,
      "num_tokens": 40162175.0,
      "step": 12025
    },
    {
      "entropy": 6.339210796356201,
      "epoch": 2.3468929860501415,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004485983804377291,
      "loss": 6.1184,
      "mean_token_accuracy": 0.1470786824822426,
      "num_tokens": 40178119.0,
      "step": 12030
    },
    {
      "entropy": 6.252596282958985,
      "epoch": 2.3478685006340845,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044855362742399706,
      "loss": 5.9937,
      "mean_token_accuracy": 0.15516366362571715,
      "num_tokens": 40193465.0,
      "step": 12035
    },
    {
      "entropy": 6.230870628356934,
      "epoch": 2.3488440152180274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004485088574512231,
      "loss": 6.0052,
      "mean_token_accuracy": 0.15420106798410416,
      "num_tokens": 40209673.0,
      "step": 12040
    },
    {
      "entropy": 6.26609787940979,
      "epoch": 2.3498195298019704,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044846407052378194,
      "loss": 6.0758,
      "mean_token_accuracy": 0.1480748660862446,
      "num_tokens": 40225174.0,
      "step": 12045
    },
    {
      "entropy": 6.32801661491394,
      "epoch": 2.3507950443859134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044841926664605016,
      "loss": 6.0582,
      "mean_token_accuracy": 0.15632283687591553,
      "num_tokens": 40242211.0,
      "step": 12050
    },
    {
      "entropy": 6.35921893119812,
      "epoch": 2.3517705589698568,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004483744458224056,
      "loss": 6.1774,
      "mean_token_accuracy": 0.13873545378446578,
      "num_tokens": 40258371.0,
      "step": 12055
    },
    {
      "entropy": 6.300609111785889,
      "epoch": 2.3527460735537997,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044832960805722824,
      "loss": 6.0861,
      "mean_token_accuracy": 0.14973325729370118,
      "num_tokens": 40275574.0,
      "step": 12060
    },
    {
      "entropy": 6.359575033187866,
      "epoch": 2.3537215881377427,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004482847533548992,
      "loss": 6.0228,
      "mean_token_accuracy": 0.1485180988907814,
      "num_tokens": 40292575.0,
      "step": 12065
    },
    {
      "entropy": 6.256726551055908,
      "epoch": 2.3546971027216856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044823988171980173,
      "loss": 6.1152,
      "mean_token_accuracy": 0.1442251279950142,
      "num_tokens": 40310001.0,
      "step": 12070
    },
    {
      "entropy": 6.258537769317627,
      "epoch": 2.3556726173056286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044819499315632043,
      "loss": 5.9713,
      "mean_token_accuracy": 0.14843790531158446,
      "num_tokens": 40327014.0,
      "step": 12075
    },
    {
      "entropy": 6.36438479423523,
      "epoch": 2.3566481318895716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044815008766884167,
      "loss": 6.0628,
      "mean_token_accuracy": 0.1505453035235405,
      "num_tokens": 40344628.0,
      "step": 12080
    },
    {
      "entropy": 6.389486694335938,
      "epoch": 2.357623646473515,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004481051652617535,
      "loss": 6.1694,
      "mean_token_accuracy": 0.1375954568386078,
      "num_tokens": 40361336.0,
      "step": 12085
    },
    {
      "entropy": 6.286440372467041,
      "epoch": 2.358599161057458,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004480602259394455,
      "loss": 6.0137,
      "mean_token_accuracy": 0.15383198782801627,
      "num_tokens": 40378378.0,
      "step": 12090
    },
    {
      "entropy": 6.243458032608032,
      "epoch": 2.359574675641401,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044801526970630906,
      "loss": 5.9833,
      "mean_token_accuracy": 0.15212817043066024,
      "num_tokens": 40395580.0,
      "step": 12095
    },
    {
      "entropy": 6.2438126564025875,
      "epoch": 2.360550190225344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004479702965667371,
      "loss": 6.0299,
      "mean_token_accuracy": 0.14944574236869812,
      "num_tokens": 40412395.0,
      "step": 12100
    },
    {
      "entropy": 6.240514039993286,
      "epoch": 2.361525704809287,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004479253065251242,
      "loss": 6.0439,
      "mean_token_accuracy": 0.14997108280658722,
      "num_tokens": 40428736.0,
      "step": 12105
    },
    {
      "entropy": 6.367838335037232,
      "epoch": 2.3625012193932298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004478802995858667,
      "loss": 6.0547,
      "mean_token_accuracy": 0.14342073649168013,
      "num_tokens": 40445890.0,
      "step": 12110
    },
    {
      "entropy": 6.231545877456665,
      "epoch": 2.3634767339771727,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478352757533624,
      "loss": 5.9492,
      "mean_token_accuracy": 0.15602618008852004,
      "num_tokens": 40462268.0,
      "step": 12115
    },
    {
      "entropy": 6.320216035842895,
      "epoch": 2.364452248561116,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044779023503201105,
      "loss": 6.0613,
      "mean_token_accuracy": 0.14880476668477058,
      "num_tokens": 40480111.0,
      "step": 12120
    },
    {
      "entropy": 6.298566770553589,
      "epoch": 2.365427763145059,
      "grad_norm": 1.53125,
      "learning_rate": 0.00044774517742621374,
      "loss": 5.9535,
      "mean_token_accuracy": 0.14598459750413895,
      "num_tokens": 40496868.0,
      "step": 12125
    },
    {
      "entropy": 6.222031688690185,
      "epoch": 2.366403277729002,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044770010294037333,
      "loss": 5.9811,
      "mean_token_accuracy": 0.15026386752724646,
      "num_tokens": 40512533.0,
      "step": 12130
    },
    {
      "entropy": 6.326044082641602,
      "epoch": 2.367378792312945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044765501157889446,
      "loss": 6.0864,
      "mean_token_accuracy": 0.14227018356323243,
      "num_tokens": 40530843.0,
      "step": 12135
    },
    {
      "entropy": 6.209378480911255,
      "epoch": 2.368354306896888,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044760990334618314,
      "loss": 5.9162,
      "mean_token_accuracy": 0.15214115679264067,
      "num_tokens": 40546513.0,
      "step": 12140
    },
    {
      "entropy": 6.341739702224731,
      "epoch": 2.3693298214808314,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044756477824664726,
      "loss": 6.1137,
      "mean_token_accuracy": 0.13975422456860542,
      "num_tokens": 40563763.0,
      "step": 12145
    },
    {
      "entropy": 6.275077152252197,
      "epoch": 2.3703053360647743,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475196362846963,
      "loss": 5.9518,
      "mean_token_accuracy": 0.15280189663171767,
      "num_tokens": 40580558.0,
      "step": 12150
    },
    {
      "entropy": 6.315053272247314,
      "epoch": 2.3712808506487173,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474744774647413,
      "loss": 6.0554,
      "mean_token_accuracy": 0.14046467691659928,
      "num_tokens": 40598053.0,
      "step": 12155
    },
    {
      "entropy": 6.308943128585815,
      "epoch": 2.3722563652326603,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044742930179119517,
      "loss": 6.0041,
      "mean_token_accuracy": 0.14799989834427835,
      "num_tokens": 40614435.0,
      "step": 12160
    },
    {
      "entropy": 6.283658409118653,
      "epoch": 2.373231879816603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004473841092684722,
      "loss": 6.02,
      "mean_token_accuracy": 0.14920701831579208,
      "num_tokens": 40630070.0,
      "step": 12165
    },
    {
      "entropy": 6.260366344451905,
      "epoch": 2.374207394400546,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044733889990098834,
      "loss": 6.0574,
      "mean_token_accuracy": 0.14703690856695176,
      "num_tokens": 40647777.0,
      "step": 12170
    },
    {
      "entropy": 6.350447511672973,
      "epoch": 2.375182908984489,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004472936736931615,
      "loss": 6.1384,
      "mean_token_accuracy": 0.13648258820176123,
      "num_tokens": 40665035.0,
      "step": 12175
    },
    {
      "entropy": 6.3014730453491214,
      "epoch": 2.376158423568432,
      "grad_norm": 1.125,
      "learning_rate": 0.00044724843064941096,
      "loss": 6.0655,
      "mean_token_accuracy": 0.1436515711247921,
      "num_tokens": 40681507.0,
      "step": 12180
    },
    {
      "entropy": 6.323202514648438,
      "epoch": 2.3771339381523755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004472031707741577,
      "loss": 5.9919,
      "mean_token_accuracy": 0.14767343550920486,
      "num_tokens": 40698613.0,
      "step": 12185
    },
    {
      "entropy": 6.193684673309326,
      "epoch": 2.3781094527363185,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004471578940718242,
      "loss": 5.8866,
      "mean_token_accuracy": 0.16256210803985596,
      "num_tokens": 40716127.0,
      "step": 12190
    },
    {
      "entropy": 6.188415765762329,
      "epoch": 2.3790849673202614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044711260054683497,
      "loss": 5.9039,
      "mean_token_accuracy": 0.16541376560926438,
      "num_tokens": 40734847.0,
      "step": 12195
    },
    {
      "entropy": 6.358506488800049,
      "epoch": 2.3800604819042044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004470672902036158,
      "loss": 6.1063,
      "mean_token_accuracy": 0.1433217242360115,
      "num_tokens": 40752266.0,
      "step": 12200
    },
    {
      "entropy": 6.3124714374542235,
      "epoch": 2.3810359964881473,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004470219630465943,
      "loss": 6.1362,
      "mean_token_accuracy": 0.13588107451796533,
      "num_tokens": 40769011.0,
      "step": 12205
    },
    {
      "entropy": 6.330528116226196,
      "epoch": 2.3820115110720907,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044697661908019967,
      "loss": 6.0287,
      "mean_token_accuracy": 0.14310021847486495,
      "num_tokens": 40785260.0,
      "step": 12210
    },
    {
      "entropy": 6.223949146270752,
      "epoch": 2.3829870256560337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044693125830886276,
      "loss": 5.9775,
      "mean_token_accuracy": 0.15650486648082734,
      "num_tokens": 40802651.0,
      "step": 12215
    },
    {
      "entropy": 6.223441553115845,
      "epoch": 2.3839625402399767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004468858807370161,
      "loss": 5.9594,
      "mean_token_accuracy": 0.1550566479563713,
      "num_tokens": 40820290.0,
      "step": 12220
    },
    {
      "entropy": 6.23781566619873,
      "epoch": 2.3849380548239196,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044684048636909376,
      "loss": 5.8992,
      "mean_token_accuracy": 0.1633566364645958,
      "num_tokens": 40838030.0,
      "step": 12225
    },
    {
      "entropy": 6.288909292221069,
      "epoch": 2.3859135694078626,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004467950752095316,
      "loss": 6.0171,
      "mean_token_accuracy": 0.15478880181908608,
      "num_tokens": 40854979.0,
      "step": 12230
    },
    {
      "entropy": 6.334518480300903,
      "epoch": 2.3868890839918055,
      "grad_norm": 1.0,
      "learning_rate": 0.00044674964726276685,
      "loss": 6.1316,
      "mean_token_accuracy": 0.13916602581739426,
      "num_tokens": 40872474.0,
      "step": 12235
    },
    {
      "entropy": 6.340835905075073,
      "epoch": 2.3878645985757485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044670420253323874,
      "loss": 6.0956,
      "mean_token_accuracy": 0.14093562662601472,
      "num_tokens": 40888116.0,
      "step": 12240
    },
    {
      "entropy": 6.256598281860351,
      "epoch": 2.388840113159692,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000446658741025388,
      "loss": 5.9762,
      "mean_token_accuracy": 0.1451835349202156,
      "num_tokens": 40904385.0,
      "step": 12245
    },
    {
      "entropy": 6.233569765090943,
      "epoch": 2.389815627743635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004466132627436568,
      "loss": 5.988,
      "mean_token_accuracy": 0.14833227843046187,
      "num_tokens": 40923117.0,
      "step": 12250
    },
    {
      "entropy": 6.329296016693116,
      "epoch": 2.390791142327578,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044656776769248926,
      "loss": 6.0538,
      "mean_token_accuracy": 0.14391641765832902,
      "num_tokens": 40940029.0,
      "step": 12255
    },
    {
      "entropy": 6.321358633041382,
      "epoch": 2.391766656911521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004465222558763309,
      "loss": 6.0604,
      "mean_token_accuracy": 0.1439415544271469,
      "num_tokens": 40955629.0,
      "step": 12260
    },
    {
      "entropy": 6.405422401428223,
      "epoch": 2.3927421714954638,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000446476727299629,
      "loss": 6.2109,
      "mean_token_accuracy": 0.1386207953095436,
      "num_tokens": 40974665.0,
      "step": 12265
    },
    {
      "entropy": 6.26500973701477,
      "epoch": 2.3937176860794067,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044643118196683245,
      "loss": 6.0492,
      "mean_token_accuracy": 0.1539618916809559,
      "num_tokens": 40992528.0,
      "step": 12270
    },
    {
      "entropy": 6.331904458999634,
      "epoch": 2.39469320066335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004463856198823919,
      "loss": 6.0297,
      "mean_token_accuracy": 0.14831956923007966,
      "num_tokens": 41009315.0,
      "step": 12275
    },
    {
      "entropy": 6.206641435623169,
      "epoch": 2.395668715247293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004463400410507593,
      "loss": 5.9158,
      "mean_token_accuracy": 0.15694993138313293,
      "num_tokens": 41027718.0,
      "step": 12280
    },
    {
      "entropy": 6.164503955841065,
      "epoch": 2.396644229831236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044629444547638865,
      "loss": 5.9134,
      "mean_token_accuracy": 0.16655447632074355,
      "num_tokens": 41045268.0,
      "step": 12285
    },
    {
      "entropy": 6.3829254627227785,
      "epoch": 2.397619744415179,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004462488331637352,
      "loss": 6.1228,
      "mean_token_accuracy": 0.13964200466871263,
      "num_tokens": 41062522.0,
      "step": 12290
    },
    {
      "entropy": 6.2731485843658445,
      "epoch": 2.398595258999122,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462032041172562,
      "loss": 6.0794,
      "mean_token_accuracy": 0.13710760027170182,
      "num_tokens": 41079131.0,
      "step": 12295
    },
    {
      "entropy": 6.2834532260894775,
      "epoch": 2.399570773583065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461575583414102,
      "loss": 5.9602,
      "mean_token_accuracy": 0.15173972845077516,
      "num_tokens": 41095138.0,
      "step": 12300
    },
    {
      "entropy": 6.331619787216186,
      "epoch": 2.400546288167008,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044611189584065764,
      "loss": 6.0972,
      "mean_token_accuracy": 0.14884231984615326,
      "num_tokens": 41111360.0,
      "step": 12305
    },
    {
      "entropy": 6.300934982299805,
      "epoch": 2.4015218027509513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044606621661946044,
      "loss": 6.0843,
      "mean_token_accuracy": 0.14892226308584214,
      "num_tokens": 41126017.0,
      "step": 12310
    },
    {
      "entropy": 6.285093879699707,
      "epoch": 2.4024973173348942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044602052068228227,
      "loss": 6.0635,
      "mean_token_accuracy": 0.14404755979776382,
      "num_tokens": 41142418.0,
      "step": 12315
    },
    {
      "entropy": 6.299890947341919,
      "epoch": 2.403472831918837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044597480803358836,
      "loss": 5.9932,
      "mean_token_accuracy": 0.14878451228141784,
      "num_tokens": 41158597.0,
      "step": 12320
    },
    {
      "entropy": 6.161045217514038,
      "epoch": 2.40444834650278,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044592907867784556,
      "loss": 5.8905,
      "mean_token_accuracy": 0.15810055881738663,
      "num_tokens": 41175692.0,
      "step": 12325
    },
    {
      "entropy": 6.280826711654663,
      "epoch": 2.405423861086723,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004458833326195224,
      "loss": 5.9935,
      "mean_token_accuracy": 0.14941876456141473,
      "num_tokens": 41192322.0,
      "step": 12330
    },
    {
      "entropy": 6.185526609420776,
      "epoch": 2.406399375670666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000445837569863089,
      "loss": 5.9032,
      "mean_token_accuracy": 0.15820915699005128,
      "num_tokens": 41209978.0,
      "step": 12335
    },
    {
      "entropy": 6.253163385391235,
      "epoch": 2.4073748902546095,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044579179041301717,
      "loss": 6.0279,
      "mean_token_accuracy": 0.1489587888121605,
      "num_tokens": 41228053.0,
      "step": 12340
    },
    {
      "entropy": 6.2095483303070065,
      "epoch": 2.4083504048385524,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044574599427378033,
      "loss": 5.9931,
      "mean_token_accuracy": 0.15322191566228865,
      "num_tokens": 41244871.0,
      "step": 12345
    },
    {
      "entropy": 6.3025823593139645,
      "epoch": 2.4093259194224954,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044570018144985334,
      "loss": 6.0745,
      "mean_token_accuracy": 0.14169735088944435,
      "num_tokens": 41262622.0,
      "step": 12350
    },
    {
      "entropy": 6.373507308959961,
      "epoch": 2.4103014340064384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004456543519457132,
      "loss": 6.1196,
      "mean_token_accuracy": 0.14202789962291718,
      "num_tokens": 41280045.0,
      "step": 12355
    },
    {
      "entropy": 6.25492491722107,
      "epoch": 2.4112769485903813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044560850576583785,
      "loss": 5.991,
      "mean_token_accuracy": 0.1498461939394474,
      "num_tokens": 41297223.0,
      "step": 12360
    },
    {
      "entropy": 6.220216226577759,
      "epoch": 2.4122524631743243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044556264291470746,
      "loss": 5.9518,
      "mean_token_accuracy": 0.15786455944180489,
      "num_tokens": 41313558.0,
      "step": 12365
    },
    {
      "entropy": 6.3603739738464355,
      "epoch": 2.4132279777582673,
      "grad_norm": 1.125,
      "learning_rate": 0.0004455167633968035,
      "loss": 6.1389,
      "mean_token_accuracy": 0.14063668698072435,
      "num_tokens": 41329259.0,
      "step": 12370
    },
    {
      "entropy": 6.173907327651977,
      "epoch": 2.4142034923422107,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044547086721660913,
      "loss": 5.8833,
      "mean_token_accuracy": 0.15941615253686905,
      "num_tokens": 41344725.0,
      "step": 12375
    },
    {
      "entropy": 6.2798905849456785,
      "epoch": 2.4151790069261536,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004454249543786091,
      "loss": 6.0027,
      "mean_token_accuracy": 0.15158065855503083,
      "num_tokens": 41360512.0,
      "step": 12380
    },
    {
      "entropy": 6.225098514556885,
      "epoch": 2.4161545215100966,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044537902488729,
      "loss": 5.9499,
      "mean_token_accuracy": 0.1604053184390068,
      "num_tokens": 41377915.0,
      "step": 12385
    },
    {
      "entropy": 6.327434349060058,
      "epoch": 2.4171300360940395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004453330787471399,
      "loss": 6.0183,
      "mean_token_accuracy": 0.14748126417398452,
      "num_tokens": 41394515.0,
      "step": 12390
    },
    {
      "entropy": 6.316428470611572,
      "epoch": 2.4181055506779825,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004452871159626484,
      "loss": 5.994,
      "mean_token_accuracy": 0.14275717586278916,
      "num_tokens": 41410725.0,
      "step": 12395
    },
    {
      "entropy": 6.25758466720581,
      "epoch": 2.419081065261926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044524113653830673,
      "loss": 6.0344,
      "mean_token_accuracy": 0.14752697199583054,
      "num_tokens": 41428139.0,
      "step": 12400
    },
    {
      "entropy": 6.168247413635254,
      "epoch": 2.420056579845869,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004451951404786081,
      "loss": 5.8742,
      "mean_token_accuracy": 0.16368794292211533,
      "num_tokens": 41444210.0,
      "step": 12405
    },
    {
      "entropy": 6.207807826995849,
      "epoch": 2.421032094429812,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044514912778804684,
      "loss": 5.9406,
      "mean_token_accuracy": 0.15703011006116868,
      "num_tokens": 41461107.0,
      "step": 12410
    },
    {
      "entropy": 6.183495998382568,
      "epoch": 2.422007609013755,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044510309847111925,
      "loss": 5.8975,
      "mean_token_accuracy": 0.16524085104465486,
      "num_tokens": 41476731.0,
      "step": 12415
    },
    {
      "entropy": 6.306670045852661,
      "epoch": 2.4229831235976977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044505705253232317,
      "loss": 6.0551,
      "mean_token_accuracy": 0.1464880034327507,
      "num_tokens": 41494467.0,
      "step": 12420
    },
    {
      "entropy": 6.171340036392212,
      "epoch": 2.4239586381816407,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044501098997615793,
      "loss": 5.9321,
      "mean_token_accuracy": 0.16078600138425828,
      "num_tokens": 41512094.0,
      "step": 12425
    },
    {
      "entropy": 6.288385677337646,
      "epoch": 2.4249341527655837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044496491080712477,
      "loss": 6.0341,
      "mean_token_accuracy": 0.1471555531024933,
      "num_tokens": 41529058.0,
      "step": 12430
    },
    {
      "entropy": 6.354226541519165,
      "epoch": 2.425909667349527,
      "grad_norm": 1.125,
      "learning_rate": 0.0004449188150297262,
      "loss": 6.169,
      "mean_token_accuracy": 0.14114080518484115,
      "num_tokens": 41547051.0,
      "step": 12435
    },
    {
      "entropy": 6.426115655899048,
      "epoch": 2.42688518193347,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044487270264846666,
      "loss": 6.1767,
      "mean_token_accuracy": 0.1378535136580467,
      "num_tokens": 41563847.0,
      "step": 12440
    },
    {
      "entropy": 6.275805521011352,
      "epoch": 2.427860696517413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044482657366785205,
      "loss": 5.991,
      "mean_token_accuracy": 0.1492268353700638,
      "num_tokens": 41581090.0,
      "step": 12445
    },
    {
      "entropy": 6.309862279891968,
      "epoch": 2.428836211101356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044478042809239,
      "loss": 6.0272,
      "mean_token_accuracy": 0.15168378502130508,
      "num_tokens": 41598177.0,
      "step": 12450
    },
    {
      "entropy": 6.120148849487305,
      "epoch": 2.429811725685299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044473426592658963,
      "loss": 5.7889,
      "mean_token_accuracy": 0.17274364531040193,
      "num_tokens": 41615389.0,
      "step": 12455
    },
    {
      "entropy": 6.323758840560913,
      "epoch": 2.430787240269242,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044468808717496163,
      "loss": 6.0671,
      "mean_token_accuracy": 0.14720668122172356,
      "num_tokens": 41632639.0,
      "step": 12460
    },
    {
      "entropy": 6.225213623046875,
      "epoch": 2.4317627548531853,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004446418918420186,
      "loss": 5.991,
      "mean_token_accuracy": 0.15342367067933083,
      "num_tokens": 41648523.0,
      "step": 12465
    },
    {
      "entropy": 6.332315826416016,
      "epoch": 2.4327382694371282,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004445956799322745,
      "loss": 6.0287,
      "mean_token_accuracy": 0.14218362420797348,
      "num_tokens": 41665999.0,
      "step": 12470
    },
    {
      "entropy": 6.263474798202514,
      "epoch": 2.433713784021071,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044454945145024505,
      "loss": 5.9783,
      "mean_token_accuracy": 0.15212302803993225,
      "num_tokens": 41685737.0,
      "step": 12475
    },
    {
      "entropy": 6.361418056488037,
      "epoch": 2.434689298605014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004445032064004474,
      "loss": 6.1021,
      "mean_token_accuracy": 0.14292964041233064,
      "num_tokens": 41703511.0,
      "step": 12480
    },
    {
      "entropy": 6.324593162536621,
      "epoch": 2.435664813188957,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044445694478740063,
      "loss": 6.0637,
      "mean_token_accuracy": 0.14737412855029106,
      "num_tokens": 41720775.0,
      "step": 12485
    },
    {
      "entropy": 6.2425089359283445,
      "epoch": 2.4366403277729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004444106666156251,
      "loss": 5.9883,
      "mean_token_accuracy": 0.1542596325278282,
      "num_tokens": 41738836.0,
      "step": 12490
    },
    {
      "entropy": 6.291443681716919,
      "epoch": 2.437615842356843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044436437188964304,
      "loss": 6.0856,
      "mean_token_accuracy": 0.14670897871255875,
      "num_tokens": 41755140.0,
      "step": 12495
    },
    {
      "entropy": 6.305994606018066,
      "epoch": 2.4385913569407864,
      "grad_norm": 1.125,
      "learning_rate": 0.00044431806061397816,
      "loss": 6.0976,
      "mean_token_accuracy": 0.14306779205799103,
      "num_tokens": 41770794.0,
      "step": 12500
    },
    {
      "entropy": 6.334402942657471,
      "epoch": 2.4395668715247294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044427173279315584,
      "loss": 6.0622,
      "mean_token_accuracy": 0.1465136229991913,
      "num_tokens": 41787464.0,
      "step": 12505
    },
    {
      "entropy": 6.327424955368042,
      "epoch": 2.4405423861086724,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004442253884317032,
      "loss": 6.0653,
      "mean_token_accuracy": 0.14711453318595885,
      "num_tokens": 41804851.0,
      "step": 12510
    },
    {
      "entropy": 6.2554295539855955,
      "epoch": 2.4415179006926153,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004441790275341487,
      "loss": 6.0063,
      "mean_token_accuracy": 0.15246487259864808,
      "num_tokens": 41821959.0,
      "step": 12515
    },
    {
      "entropy": 6.1962052345275875,
      "epoch": 2.4424934152765583,
      "grad_norm": 1.25,
      "learning_rate": 0.0004441326501050225,
      "loss": 5.9367,
      "mean_token_accuracy": 0.16227926462888717,
      "num_tokens": 41837581.0,
      "step": 12520
    },
    {
      "entropy": 6.244371652603149,
      "epoch": 2.4434689298605012,
      "grad_norm": 1.125,
      "learning_rate": 0.00044408625614885663,
      "loss": 5.9425,
      "mean_token_accuracy": 0.15860750675201415,
      "num_tokens": 41854497.0,
      "step": 12525
    },
    {
      "entropy": 6.337510681152343,
      "epoch": 2.4444444444444446,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004440398456701844,
      "loss": 6.1058,
      "mean_token_accuracy": 0.14168099611997603,
      "num_tokens": 41870968.0,
      "step": 12530
    },
    {
      "entropy": 6.269718837738037,
      "epoch": 2.4454199590283876,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044399341867354093,
      "loss": 5.9373,
      "mean_token_accuracy": 0.15353601574897766,
      "num_tokens": 41886734.0,
      "step": 12535
    },
    {
      "entropy": 6.254725551605224,
      "epoch": 2.4463954736123306,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044394697516346286,
      "loss": 6.035,
      "mean_token_accuracy": 0.1520478293299675,
      "num_tokens": 41903728.0,
      "step": 12540
    },
    {
      "entropy": 6.323126602172851,
      "epoch": 2.4473709881962735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044390051514448854,
      "loss": 6.0544,
      "mean_token_accuracy": 0.14168013334274293,
      "num_tokens": 41919967.0,
      "step": 12545
    },
    {
      "entropy": 6.1558679103851315,
      "epoch": 2.4483465027802165,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044385403862115785,
      "loss": 5.8317,
      "mean_token_accuracy": 0.16359133422374725,
      "num_tokens": 41937330.0,
      "step": 12550
    },
    {
      "entropy": 6.247432518005371,
      "epoch": 2.4493220173641594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044380754559801235,
      "loss": 5.9937,
      "mean_token_accuracy": 0.14394376054406166,
      "num_tokens": 41953261.0,
      "step": 12555
    },
    {
      "entropy": 6.329224920272827,
      "epoch": 2.4502975319481024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044376103607959513,
      "loss": 6.0908,
      "mean_token_accuracy": 0.13971101567149163,
      "num_tokens": 41970468.0,
      "step": 12560
    },
    {
      "entropy": 6.280830669403076,
      "epoch": 2.451273046532046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044371451007045083,
      "loss": 6.0216,
      "mean_token_accuracy": 0.14887230843305588,
      "num_tokens": 41986235.0,
      "step": 12565
    },
    {
      "entropy": 6.268201303482056,
      "epoch": 2.4522485611159888,
      "grad_norm": 1.109375,
      "learning_rate": 0.000443667967575126,
      "loss": 5.9523,
      "mean_token_accuracy": 0.1414072521030903,
      "num_tokens": 42004539.0,
      "step": 12570
    },
    {
      "entropy": 6.273130607604981,
      "epoch": 2.4532240756999317,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004436214085981686,
      "loss": 5.9812,
      "mean_token_accuracy": 0.15652548372745514,
      "num_tokens": 42022937.0,
      "step": 12575
    },
    {
      "entropy": 6.315567350387573,
      "epoch": 2.4541995902838747,
      "grad_norm": 1.140625,
      "learning_rate": 0.000443574833144128,
      "loss": 5.9617,
      "mean_token_accuracy": 0.14969747215509416,
      "num_tokens": 42039260.0,
      "step": 12580
    },
    {
      "entropy": 6.232639026641846,
      "epoch": 2.4551751048678176,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004435282412175556,
      "loss": 5.9942,
      "mean_token_accuracy": 0.14727186933159828,
      "num_tokens": 42055894.0,
      "step": 12585
    },
    {
      "entropy": 6.286185312271118,
      "epoch": 2.456150619451761,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044348163282300403,
      "loss": 6.0552,
      "mean_token_accuracy": 0.14695927649736404,
      "num_tokens": 42070748.0,
      "step": 12590
    },
    {
      "entropy": 6.363253498077393,
      "epoch": 2.457126134035704,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044343500796502783,
      "loss": 6.0504,
      "mean_token_accuracy": 0.1457233816385269,
      "num_tokens": 42088016.0,
      "step": 12595
    },
    {
      "entropy": 6.29396915435791,
      "epoch": 2.458101648619647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004433883666481829,
      "loss": 5.9526,
      "mean_token_accuracy": 0.15457767620682716,
      "num_tokens": 42105422.0,
      "step": 12600
    },
    {
      "entropy": 6.225633859634399,
      "epoch": 2.45907716320359,
      "grad_norm": 1.140625,
      "learning_rate": 0.000443341708877027,
      "loss": 6.0038,
      "mean_token_accuracy": 0.1486043192446232,
      "num_tokens": 42120534.0,
      "step": 12605
    },
    {
      "entropy": 6.285325717926026,
      "epoch": 2.460052677787533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004432950346561193,
      "loss": 5.968,
      "mean_token_accuracy": 0.15298973470926286,
      "num_tokens": 42137413.0,
      "step": 12610
    },
    {
      "entropy": 6.254862880706787,
      "epoch": 2.461028192371476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004432483439900206,
      "loss": 6.035,
      "mean_token_accuracy": 0.14902981370687485,
      "num_tokens": 42153006.0,
      "step": 12615
    },
    {
      "entropy": 6.363480520248413,
      "epoch": 2.462003706955419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004432016368832934,
      "loss": 6.1566,
      "mean_token_accuracy": 0.1371052235364914,
      "num_tokens": 42170966.0,
      "step": 12620
    },
    {
      "entropy": 6.354564809799195,
      "epoch": 2.4629792215393618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044315491334050166,
      "loss": 6.0196,
      "mean_token_accuracy": 0.14387698024511336,
      "num_tokens": 42187349.0,
      "step": 12625
    },
    {
      "entropy": 6.227053260803222,
      "epoch": 2.463954736123305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004431081733662112,
      "loss": 5.9732,
      "mean_token_accuracy": 0.15851733088493347,
      "num_tokens": 42204486.0,
      "step": 12630
    },
    {
      "entropy": 6.275281763076782,
      "epoch": 2.464930250707248,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044306141696498914,
      "loss": 6.0092,
      "mean_token_accuracy": 0.14965377300977706,
      "num_tokens": 42220316.0,
      "step": 12635
    },
    {
      "entropy": 6.171428060531616,
      "epoch": 2.465905765291191,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044301464414140437,
      "loss": 5.9198,
      "mean_token_accuracy": 0.1535326212644577,
      "num_tokens": 42237734.0,
      "step": 12640
    },
    {
      "entropy": 6.378137826919556,
      "epoch": 2.466881279875134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004429678549000274,
      "loss": 6.1148,
      "mean_token_accuracy": 0.13592008650302886,
      "num_tokens": 42253458.0,
      "step": 12645
    },
    {
      "entropy": 6.156814002990723,
      "epoch": 2.467856794459077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004429210492454303,
      "loss": 5.9153,
      "mean_token_accuracy": 0.15701181143522264,
      "num_tokens": 42270031.0,
      "step": 12650
    },
    {
      "entropy": 6.359789419174194,
      "epoch": 2.4688323090430204,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044287422718218675,
      "loss": 6.0497,
      "mean_token_accuracy": 0.14856518507003785,
      "num_tokens": 42286227.0,
      "step": 12655
    },
    {
      "entropy": 6.2775993824005125,
      "epoch": 2.4698078236269634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044282738871487205,
      "loss": 6.01,
      "mean_token_accuracy": 0.1509188622236252,
      "num_tokens": 42302361.0,
      "step": 12660
    },
    {
      "entropy": 6.314933109283447,
      "epoch": 2.4707833382109063,
      "grad_norm": 1.15625,
      "learning_rate": 0.000442780533848063,
      "loss": 6.0812,
      "mean_token_accuracy": 0.14433905780315398,
      "num_tokens": 42319462.0,
      "step": 12665
    },
    {
      "entropy": 6.306931686401367,
      "epoch": 2.4717588527948493,
      "grad_norm": 1.25,
      "learning_rate": 0.00044273366258633823,
      "loss": 6.0068,
      "mean_token_accuracy": 0.14935044646263124,
      "num_tokens": 42335565.0,
      "step": 12670
    },
    {
      "entropy": 6.376024198532105,
      "epoch": 2.4727343673787923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004426867749342778,
      "loss": 6.1244,
      "mean_token_accuracy": 0.14132535606622695,
      "num_tokens": 42352335.0,
      "step": 12675
    },
    {
      "entropy": 6.234892702102661,
      "epoch": 2.4737098819627352,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004426398708964633,
      "loss": 6.0033,
      "mean_token_accuracy": 0.15922857969999313,
      "num_tokens": 42368320.0,
      "step": 12680
    },
    {
      "entropy": 6.202628231048584,
      "epoch": 2.474685396546678,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044259295047747814,
      "loss": 5.9618,
      "mean_token_accuracy": 0.14997186958789827,
      "num_tokens": 42384224.0,
      "step": 12685
    },
    {
      "entropy": 6.283151531219483,
      "epoch": 2.4756609111306216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004425460136819071,
      "loss": 5.9737,
      "mean_token_accuracy": 0.15086906552314758,
      "num_tokens": 42399973.0,
      "step": 12690
    },
    {
      "entropy": 6.280716133117676,
      "epoch": 2.4766364257145645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004424990605143369,
      "loss": 6.0209,
      "mean_token_accuracy": 0.1463230535387993,
      "num_tokens": 42415355.0,
      "step": 12695
    },
    {
      "entropy": 6.279669761657715,
      "epoch": 2.4776119402985075,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044245209097935537,
      "loss": 6.0455,
      "mean_token_accuracy": 0.14446472153067588,
      "num_tokens": 42431176.0,
      "step": 12700
    },
    {
      "entropy": 6.27326135635376,
      "epoch": 2.4785874548824505,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004424051050815523,
      "loss": 5.9868,
      "mean_token_accuracy": 0.15247846245765687,
      "num_tokens": 42446935.0,
      "step": 12705
    },
    {
      "entropy": 6.202051830291748,
      "epoch": 2.4795629694663934,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044235810282551905,
      "loss": 5.9068,
      "mean_token_accuracy": 0.16097103729844092,
      "num_tokens": 42463473.0,
      "step": 12710
    },
    {
      "entropy": 6.277994060516358,
      "epoch": 2.4805384840503364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004423110842158484,
      "loss": 5.9753,
      "mean_token_accuracy": 0.1579202264547348,
      "num_tokens": 42479460.0,
      "step": 12715
    },
    {
      "entropy": 6.27202296257019,
      "epoch": 2.48151399863428,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000442264049257135,
      "loss": 6.0126,
      "mean_token_accuracy": 0.15057988613843917,
      "num_tokens": 42495213.0,
      "step": 12720
    },
    {
      "entropy": 6.339045143127441,
      "epoch": 2.4824895132182228,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004422169979539748,
      "loss": 6.0705,
      "mean_token_accuracy": 0.1451515145599842,
      "num_tokens": 42513104.0,
      "step": 12725
    },
    {
      "entropy": 6.285975170135498,
      "epoch": 2.4834650278021657,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044216993031096554,
      "loss": 5.9948,
      "mean_token_accuracy": 0.1499871425330639,
      "num_tokens": 42529774.0,
      "step": 12730
    },
    {
      "entropy": 6.245811891555786,
      "epoch": 2.4844405423861087,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004421228463327064,
      "loss": 6.0195,
      "mean_token_accuracy": 0.14785803630948066,
      "num_tokens": 42545570.0,
      "step": 12735
    },
    {
      "entropy": 6.289939546585083,
      "epoch": 2.4854160569700516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044207574602379856,
      "loss": 6.0144,
      "mean_token_accuracy": 0.15230693370103837,
      "num_tokens": 42563165.0,
      "step": 12740
    },
    {
      "entropy": 6.263333988189697,
      "epoch": 2.4863915715539946,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004420286293888441,
      "loss": 6.0593,
      "mean_token_accuracy": 0.14980218261480333,
      "num_tokens": 42578870.0,
      "step": 12745
    },
    {
      "entropy": 6.241397142410278,
      "epoch": 2.4873670861379376,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004419814964324474,
      "loss": 5.9624,
      "mean_token_accuracy": 0.15781439393758773,
      "num_tokens": 42596247.0,
      "step": 12750
    },
    {
      "entropy": 6.330187129974365,
      "epoch": 2.488342600721881,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441934347159214,
      "loss": 6.1605,
      "mean_token_accuracy": 0.14140565693378448,
      "num_tokens": 42613196.0,
      "step": 12755
    },
    {
      "entropy": 6.324400806427002,
      "epoch": 2.489318115305824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000441887181573751,
      "loss": 6.0047,
      "mean_token_accuracy": 0.1445665329694748,
      "num_tokens": 42628242.0,
      "step": 12760
    },
    {
      "entropy": 6.300847768783569,
      "epoch": 2.490293629889767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004418399996806676,
      "loss": 6.0778,
      "mean_token_accuracy": 0.14420289397239686,
      "num_tokens": 42645687.0,
      "step": 12765
    },
    {
      "entropy": 6.3171875953674315,
      "epoch": 2.49126914447371,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044179280148457395,
      "loss": 6.0706,
      "mean_token_accuracy": 0.1389317661523819,
      "num_tokens": 42661998.0,
      "step": 12770
    },
    {
      "entropy": 6.355466604232788,
      "epoch": 2.492244659057653,
      "grad_norm": 1.0,
      "learning_rate": 0.00044174558699008225,
      "loss": 6.1042,
      "mean_token_accuracy": 0.144660983979702,
      "num_tokens": 42678276.0,
      "step": 12775
    },
    {
      "entropy": 6.33666410446167,
      "epoch": 2.4932201736415958,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044169835620180603,
      "loss": 6.0955,
      "mean_token_accuracy": 0.14593830406665803,
      "num_tokens": 42697094.0,
      "step": 12780
    },
    {
      "entropy": 6.3249063968658445,
      "epoch": 2.494195688225539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004416511091243606,
      "loss": 6.05,
      "mean_token_accuracy": 0.1400594159960747,
      "num_tokens": 42713188.0,
      "step": 12785
    },
    {
      "entropy": 6.3046845436096195,
      "epoch": 2.495171202809482,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004416038457623628,
      "loss": 6.063,
      "mean_token_accuracy": 0.1458211213350296,
      "num_tokens": 42729210.0,
      "step": 12790
    },
    {
      "entropy": 6.248334121704102,
      "epoch": 2.496146717393425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004415565661204309,
      "loss": 6.0091,
      "mean_token_accuracy": 0.15280887186527253,
      "num_tokens": 42746543.0,
      "step": 12795
    },
    {
      "entropy": 6.281823015213012,
      "epoch": 2.497122231977368,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000441509270203185,
      "loss": 6.0085,
      "mean_token_accuracy": 0.14936689138412476,
      "num_tokens": 42762265.0,
      "step": 12800
    },
    {
      "entropy": 6.309512615203857,
      "epoch": 2.498097746561311,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044146195801524665,
      "loss": 6.0477,
      "mean_token_accuracy": 0.1506997436285019,
      "num_tokens": 42779555.0,
      "step": 12805
    },
    {
      "entropy": 6.391895818710327,
      "epoch": 2.499073261145254,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000441414629561239,
      "loss": 6.0972,
      "mean_token_accuracy": 0.14100666865706443,
      "num_tokens": 42797211.0,
      "step": 12810
    },
    {
      "entropy": 6.351943397521973,
      "epoch": 2.500048775729197,
      "grad_norm": 1.125,
      "learning_rate": 0.0004413672848457869,
      "loss": 6.0669,
      "mean_token_accuracy": 0.14325527250766754,
      "num_tokens": 42813827.0,
      "step": 12815
    },
    {
      "entropy": 6.310732364654541,
      "epoch": 2.5010242903131403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004413199238735168,
      "loss": 6.0324,
      "mean_token_accuracy": 0.1462759181857109,
      "num_tokens": 42829458.0,
      "step": 12820
    },
    {
      "entropy": 6.337245798110962,
      "epoch": 2.5019998048970833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044127254664905636,
      "loss": 6.0896,
      "mean_token_accuracy": 0.14767247810959816,
      "num_tokens": 42844651.0,
      "step": 12825
    },
    {
      "entropy": 6.242226934432983,
      "epoch": 2.5029753194810263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004412251531770353,
      "loss": 5.9123,
      "mean_token_accuracy": 0.1561625562608242,
      "num_tokens": 42860425.0,
      "step": 12830
    },
    {
      "entropy": 6.2942627429962155,
      "epoch": 2.503950834064969,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004411777434620847,
      "loss": 6.064,
      "mean_token_accuracy": 0.1389578975737095,
      "num_tokens": 42877734.0,
      "step": 12835
    },
    {
      "entropy": 6.280549049377441,
      "epoch": 2.504926348648912,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004411303175088374,
      "loss": 6.0084,
      "mean_token_accuracy": 0.1473937824368477,
      "num_tokens": 42895143.0,
      "step": 12840
    },
    {
      "entropy": 6.316490030288696,
      "epoch": 2.5059018632328556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004410828753219274,
      "loss": 6.0049,
      "mean_token_accuracy": 0.14443905353546144,
      "num_tokens": 42910777.0,
      "step": 12845
    },
    {
      "entropy": 6.365773057937622,
      "epoch": 2.5068773778167985,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004410354169059908,
      "loss": 6.0961,
      "mean_token_accuracy": 0.1406981185078621,
      "num_tokens": 42927594.0,
      "step": 12850
    },
    {
      "entropy": 6.262895393371582,
      "epoch": 2.5078528924007415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000440987942265665,
      "loss": 6.0116,
      "mean_token_accuracy": 0.14848195761442184,
      "num_tokens": 42943896.0,
      "step": 12855
    },
    {
      "entropy": 6.173103952407837,
      "epoch": 2.5088284069846845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044094045140558914,
      "loss": 5.8825,
      "mean_token_accuracy": 0.14988213926553726,
      "num_tokens": 42960762.0,
      "step": 12860
    },
    {
      "entropy": 6.234415340423584,
      "epoch": 2.5098039215686274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004408929443304037,
      "loss": 5.8762,
      "mean_token_accuracy": 0.1643492639064789,
      "num_tokens": 42976994.0,
      "step": 12865
    },
    {
      "entropy": 6.232911729812622,
      "epoch": 2.5107794361525704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044084542104475113,
      "loss": 5.9873,
      "mean_token_accuracy": 0.15119221955537795,
      "num_tokens": 42994827.0,
      "step": 12870
    },
    {
      "entropy": 6.1335869312286375,
      "epoch": 2.5117549507365133,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044079788155327503,
      "loss": 5.8589,
      "mean_token_accuracy": 0.16690931767225264,
      "num_tokens": 43011318.0,
      "step": 12875
    },
    {
      "entropy": 6.324281072616577,
      "epoch": 2.5127304653204563,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044075032586062083,
      "loss": 6.0639,
      "mean_token_accuracy": 0.14706694036722184,
      "num_tokens": 43028956.0,
      "step": 12880
    },
    {
      "entropy": 6.258916139602661,
      "epoch": 2.5137059799043997,
      "grad_norm": 1.125,
      "learning_rate": 0.00044070275397143553,
      "loss": 5.9025,
      "mean_token_accuracy": 0.14933460503816604,
      "num_tokens": 43045178.0,
      "step": 12885
    },
    {
      "entropy": 6.329501533508301,
      "epoch": 2.5146814944883427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044065516589036777,
      "loss": 6.0528,
      "mean_token_accuracy": 0.14107375144958495,
      "num_tokens": 43061887.0,
      "step": 12890
    },
    {
      "entropy": 6.315682649612427,
      "epoch": 2.5156570090722856,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044060756162206757,
      "loss": 6.0382,
      "mean_token_accuracy": 0.14633189141750336,
      "num_tokens": 43078754.0,
      "step": 12895
    },
    {
      "entropy": 6.191524267196655,
      "epoch": 2.5166325236562286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004405599411711866,
      "loss": 5.9381,
      "mean_token_accuracy": 0.15541193783283233,
      "num_tokens": 43096599.0,
      "step": 12900
    },
    {
      "entropy": 6.300765085220337,
      "epoch": 2.5176080382401715,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004405123045423783,
      "loss": 5.9964,
      "mean_token_accuracy": 0.14469641968607902,
      "num_tokens": 43111503.0,
      "step": 12905
    },
    {
      "entropy": 6.260559558868408,
      "epoch": 2.518583552824115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004404646517402975,
      "loss": 6.0108,
      "mean_token_accuracy": 0.14862313121557236,
      "num_tokens": 43129052.0,
      "step": 12910
    },
    {
      "entropy": 6.198226690292358,
      "epoch": 2.519559067408058,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004404169827696006,
      "loss": 5.9101,
      "mean_token_accuracy": 0.15698746144771575,
      "num_tokens": 43147456.0,
      "step": 12915
    },
    {
      "entropy": 6.2765508651733395,
      "epoch": 2.520534581992001,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004403692976349458,
      "loss": 5.9641,
      "mean_token_accuracy": 0.14914196729660034,
      "num_tokens": 43163535.0,
      "step": 12920
    },
    {
      "entropy": 6.295141792297363,
      "epoch": 2.521510096575944,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004403215963409926,
      "loss": 6.0521,
      "mean_token_accuracy": 0.14389070123434067,
      "num_tokens": 43181376.0,
      "step": 12925
    },
    {
      "entropy": 6.294292688369751,
      "epoch": 2.522485611159887,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044027387889240225,
      "loss": 6.0212,
      "mean_token_accuracy": 0.15052277743816375,
      "num_tokens": 43196993.0,
      "step": 12930
    },
    {
      "entropy": 6.309130001068115,
      "epoch": 2.5234611257438297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004402261452938374,
      "loss": 6.0162,
      "mean_token_accuracy": 0.14986035823822022,
      "num_tokens": 43214355.0,
      "step": 12935
    },
    {
      "entropy": 6.34047122001648,
      "epoch": 2.5244366403277727,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004401783955499626,
      "loss": 6.1313,
      "mean_token_accuracy": 0.13880303874611855,
      "num_tokens": 43231031.0,
      "step": 12940
    },
    {
      "entropy": 6.258862781524658,
      "epoch": 2.5254121549117157,
      "grad_norm": 1.125,
      "learning_rate": 0.00044013062966544365,
      "loss": 5.9712,
      "mean_token_accuracy": 0.15144584625959395,
      "num_tokens": 43248603.0,
      "step": 12945
    },
    {
      "entropy": 6.358087491989136,
      "epoch": 2.526387669495659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004400828476449481,
      "loss": 6.0518,
      "mean_token_accuracy": 0.1475965276360512,
      "num_tokens": 43264053.0,
      "step": 12950
    },
    {
      "entropy": 6.288462734222412,
      "epoch": 2.527363184079602,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044003504949314507,
      "loss": 6.0685,
      "mean_token_accuracy": 0.14573134928941728,
      "num_tokens": 43282739.0,
      "step": 12955
    },
    {
      "entropy": 6.191963338851929,
      "epoch": 2.528338698663545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004399872352147051,
      "loss": 5.8857,
      "mean_token_accuracy": 0.16222014427185058,
      "num_tokens": 43298885.0,
      "step": 12960
    },
    {
      "entropy": 6.393913698196411,
      "epoch": 2.529314213247488,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004399394048143007,
      "loss": 6.0552,
      "mean_token_accuracy": 0.14448780044913292,
      "num_tokens": 43314960.0,
      "step": 12965
    },
    {
      "entropy": 6.141121435165405,
      "epoch": 2.5302897278314314,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004398915582966054,
      "loss": 5.9408,
      "mean_token_accuracy": 0.15149340257048607,
      "num_tokens": 43331637.0,
      "step": 12970
    },
    {
      "entropy": 6.248999738693238,
      "epoch": 2.5312652424153743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004398436956662948,
      "loss": 6.0335,
      "mean_token_accuracy": 0.1448770135641098,
      "num_tokens": 43349534.0,
      "step": 12975
    },
    {
      "entropy": 6.398053789138794,
      "epoch": 2.5322407569993173,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043979581692804565,
      "loss": 6.0222,
      "mean_token_accuracy": 0.14646478816866876,
      "num_tokens": 43366385.0,
      "step": 12980
    },
    {
      "entropy": 6.223550844192505,
      "epoch": 2.5332162715832602,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043974792208653667,
      "loss": 5.9752,
      "mean_token_accuracy": 0.15133829712867736,
      "num_tokens": 43383204.0,
      "step": 12985
    },
    {
      "entropy": 6.163386821746826,
      "epoch": 2.534191786167203,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004397000111464479,
      "loss": 5.8454,
      "mean_token_accuracy": 0.16567285358905792,
      "num_tokens": 43401509.0,
      "step": 12990
    },
    {
      "entropy": 6.205859756469726,
      "epoch": 2.535167300751146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004396520841124611,
      "loss": 5.9009,
      "mean_token_accuracy": 0.15920397788286209,
      "num_tokens": 43418958.0,
      "step": 12995
    },
    {
      "entropy": 6.373366117477417,
      "epoch": 2.536142815335089,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004396041409892594,
      "loss": 6.0968,
      "mean_token_accuracy": 0.14456070140004157,
      "num_tokens": 43436032.0,
      "step": 13000
    },
    {
      "entropy": 6.363595819473266,
      "epoch": 2.537118329919032,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043955618178152773,
      "loss": 6.1209,
      "mean_token_accuracy": 0.1421290673315525,
      "num_tokens": 43453433.0,
      "step": 13005
    },
    {
      "entropy": 6.348614978790283,
      "epoch": 2.5380938445029755,
      "grad_norm": 1.125,
      "learning_rate": 0.00043950820649395244,
      "loss": 6.0678,
      "mean_token_accuracy": 0.14029428958892823,
      "num_tokens": 43469450.0,
      "step": 13010
    },
    {
      "entropy": 6.298603487014771,
      "epoch": 2.5390693590869184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004394602151312216,
      "loss": 6.0361,
      "mean_token_accuracy": 0.15245894491672515,
      "num_tokens": 43486633.0,
      "step": 13015
    },
    {
      "entropy": 6.344520425796508,
      "epoch": 2.5400448736708614,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043941220769802466,
      "loss": 6.0231,
      "mean_token_accuracy": 0.14836737066507338,
      "num_tokens": 43503056.0,
      "step": 13020
    },
    {
      "entropy": 6.19074125289917,
      "epoch": 2.5410203882548044,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043936418419905274,
      "loss": 5.9832,
      "mean_token_accuracy": 0.1502407059073448,
      "num_tokens": 43519691.0,
      "step": 13025
    },
    {
      "entropy": 6.281252861022949,
      "epoch": 2.5419959028387473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393161446389985,
      "loss": 6.0388,
      "mean_token_accuracy": 0.14935463815927505,
      "num_tokens": 43537152.0,
      "step": 13030
    },
    {
      "entropy": 6.319410896301269,
      "epoch": 2.5429714174226907,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004392680890225563,
      "loss": 6.0374,
      "mean_token_accuracy": 0.14430280029773712,
      "num_tokens": 43554352.0,
      "step": 13035
    },
    {
      "entropy": 6.278663873672485,
      "epoch": 2.5439469320066337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004392200173544219,
      "loss": 5.9859,
      "mean_token_accuracy": 0.14896645843982698,
      "num_tokens": 43570244.0,
      "step": 13040
    },
    {
      "entropy": 6.263911533355713,
      "epoch": 2.5449224465905766,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043917192963929266,
      "loss": 5.9966,
      "mean_token_accuracy": 0.14916273280978204,
      "num_tokens": 43585701.0,
      "step": 13045
    },
    {
      "entropy": 6.294520616531372,
      "epoch": 2.5458979611745196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004391238258818676,
      "loss": 6.0166,
      "mean_token_accuracy": 0.14629407823085785,
      "num_tokens": 43600966.0,
      "step": 13050
    },
    {
      "entropy": 6.199855852127075,
      "epoch": 2.5468734757584626,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004390757060868473,
      "loss": 5.9205,
      "mean_token_accuracy": 0.1552262932062149,
      "num_tokens": 43617318.0,
      "step": 13055
    },
    {
      "entropy": 6.338507556915284,
      "epoch": 2.5478489903424055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043902757025893364,
      "loss": 6.0439,
      "mean_token_accuracy": 0.1449725493788719,
      "num_tokens": 43633124.0,
      "step": 13060
    },
    {
      "entropy": 6.275015449523925,
      "epoch": 2.5488245049263485,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043897941840283043,
      "loss": 6.0349,
      "mean_token_accuracy": 0.15156045705080032,
      "num_tokens": 43649281.0,
      "step": 13065
    },
    {
      "entropy": 6.304234600067138,
      "epoch": 2.5498000195102914,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043893125052324293,
      "loss": 6.0461,
      "mean_token_accuracy": 0.14510626271367072,
      "num_tokens": 43665572.0,
      "step": 13070
    },
    {
      "entropy": 6.314881944656372,
      "epoch": 2.550775534094235,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004388830666248779,
      "loss": 6.1431,
      "mean_token_accuracy": 0.1492261089384556,
      "num_tokens": 43682447.0,
      "step": 13075
    },
    {
      "entropy": 6.329275274276734,
      "epoch": 2.551751048678178,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004388348667124437,
      "loss": 6.0257,
      "mean_token_accuracy": 0.1497623383998871,
      "num_tokens": 43698528.0,
      "step": 13080
    },
    {
      "entropy": 6.2843749046325685,
      "epoch": 2.5527265632621208,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043878665079065025,
      "loss": 5.9853,
      "mean_token_accuracy": 0.14975618124008178,
      "num_tokens": 43715403.0,
      "step": 13085
    },
    {
      "entropy": 6.249543142318726,
      "epoch": 2.5537020778460637,
      "grad_norm": 1.0,
      "learning_rate": 0.00043873841886420893,
      "loss": 5.9873,
      "mean_token_accuracy": 0.15126800239086152,
      "num_tokens": 43732029.0,
      "step": 13090
    },
    {
      "entropy": 6.2970290184021,
      "epoch": 2.5546775924300067,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043869017093783305,
      "loss": 6.0453,
      "mean_token_accuracy": 0.15285415202379227,
      "num_tokens": 43749951.0,
      "step": 13095
    },
    {
      "entropy": 6.371600532531739,
      "epoch": 2.55565310701395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004386419070162369,
      "loss": 6.1038,
      "mean_token_accuracy": 0.1500139132142067,
      "num_tokens": 43766169.0,
      "step": 13100
    },
    {
      "entropy": 6.2822216033935545,
      "epoch": 2.556628621597893,
      "grad_norm": 1.375,
      "learning_rate": 0.00043859362710413694,
      "loss": 6.012,
      "mean_token_accuracy": 0.15140908807516099,
      "num_tokens": 43782751.0,
      "step": 13105
    },
    {
      "entropy": 6.244428110122681,
      "epoch": 2.557604136181836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004385453312062508,
      "loss": 5.9225,
      "mean_token_accuracy": 0.15262957140803338,
      "num_tokens": 43799113.0,
      "step": 13110
    },
    {
      "entropy": 6.266975069046021,
      "epoch": 2.558579650765779,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004384970193272977,
      "loss": 5.9986,
      "mean_token_accuracy": 0.14822411984205247,
      "num_tokens": 43816926.0,
      "step": 13115
    },
    {
      "entropy": 6.318001747131348,
      "epoch": 2.559555165349722,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043844869147199867,
      "loss": 5.9872,
      "mean_token_accuracy": 0.1566571280360222,
      "num_tokens": 43834557.0,
      "step": 13120
    },
    {
      "entropy": 6.263475561141968,
      "epoch": 2.560530679933665,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000438400347645076,
      "loss": 6.0102,
      "mean_token_accuracy": 0.1518239349126816,
      "num_tokens": 43850527.0,
      "step": 13125
    },
    {
      "entropy": 6.240691471099853,
      "epoch": 2.561506194517608,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043835198785125367,
      "loss": 5.9875,
      "mean_token_accuracy": 0.15113913118839264,
      "num_tokens": 43867889.0,
      "step": 13130
    },
    {
      "entropy": 6.314340448379516,
      "epoch": 2.562481709101551,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004383036120952574,
      "loss": 6.1069,
      "mean_token_accuracy": 0.1403761550784111,
      "num_tokens": 43885002.0,
      "step": 13135
    },
    {
      "entropy": 6.299450159072876,
      "epoch": 2.5634572236854942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004382552203818142,
      "loss": 6.0116,
      "mean_token_accuracy": 0.1535546973347664,
      "num_tokens": 43902161.0,
      "step": 13140
    },
    {
      "entropy": 6.290432357788086,
      "epoch": 2.564432738269437,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043820681271565265,
      "loss": 5.935,
      "mean_token_accuracy": 0.1569458030164242,
      "num_tokens": 43917580.0,
      "step": 13145
    },
    {
      "entropy": 6.220638179779053,
      "epoch": 2.56540825285338,
      "grad_norm": 1.015625,
      "learning_rate": 0.000438158389101503,
      "loss": 5.9957,
      "mean_token_accuracy": 0.1475079096853733,
      "num_tokens": 43933709.0,
      "step": 13150
    },
    {
      "entropy": 6.267092895507813,
      "epoch": 2.566383767437323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381099495440971,
      "loss": 5.9738,
      "mean_token_accuracy": 0.15077186971902848,
      "num_tokens": 43950264.0,
      "step": 13155
    },
    {
      "entropy": 6.1966194152832035,
      "epoch": 2.567359282021266,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004380614940481683,
      "loss": 5.867,
      "mean_token_accuracy": 0.1595197305083275,
      "num_tokens": 43966722.0,
      "step": 13160
    },
    {
      "entropy": 6.210998487472534,
      "epoch": 2.5683347966052095,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004380130226184514,
      "loss": 5.9409,
      "mean_token_accuracy": 0.15551941841840744,
      "num_tokens": 43984780.0,
      "step": 13165
    },
    {
      "entropy": 6.275571823120117,
      "epoch": 2.5693103111891524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043796453525968304,
      "loss": 5.9957,
      "mean_token_accuracy": 0.14929342940449714,
      "num_tokens": 44000954.0,
      "step": 13170
    },
    {
      "entropy": 6.330821800231933,
      "epoch": 2.5702858257730954,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043791603197660095,
      "loss": 6.055,
      "mean_token_accuracy": 0.1486253559589386,
      "num_tokens": 44017214.0,
      "step": 13175
    },
    {
      "entropy": 6.291394996643066,
      "epoch": 2.5712613403570383,
      "grad_norm": 1.109375,
      "learning_rate": 0.000437867512773945,
      "loss": 6.0455,
      "mean_token_accuracy": 0.14661432430148125,
      "num_tokens": 44032793.0,
      "step": 13180
    },
    {
      "entropy": 6.2623084545135494,
      "epoch": 2.5722368549409813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043781897765645615,
      "loss": 5.9865,
      "mean_token_accuracy": 0.1530916780233383,
      "num_tokens": 44047831.0,
      "step": 13185
    },
    {
      "entropy": 6.320028257369995,
      "epoch": 2.5732123695249243,
      "grad_norm": 1.078125,
      "learning_rate": 0.000437770426628877,
      "loss": 6.1001,
      "mean_token_accuracy": 0.14304030314087868,
      "num_tokens": 44063861.0,
      "step": 13190
    },
    {
      "entropy": 6.3557219982147215,
      "epoch": 2.5741878841088672,
      "grad_norm": 1.046875,
      "learning_rate": 0.000437721859695952,
      "loss": 6.0881,
      "mean_token_accuracy": 0.1442723885178566,
      "num_tokens": 44080740.0,
      "step": 13195
    },
    {
      "entropy": 6.3088775157928465,
      "epoch": 2.57516339869281,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004376732768624267,
      "loss": 6.0026,
      "mean_token_accuracy": 0.14386448115110398,
      "num_tokens": 44096581.0,
      "step": 13200
    },
    {
      "entropy": 6.2245033264160154,
      "epoch": 2.5761389132767536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004376246781330487,
      "loss": 5.9236,
      "mean_token_accuracy": 0.15642841309309005,
      "num_tokens": 44113556.0,
      "step": 13205
    },
    {
      "entropy": 6.199865531921387,
      "epoch": 2.5771144278606966,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043757606351256664,
      "loss": 5.945,
      "mean_token_accuracy": 0.15962808579206467,
      "num_tokens": 44131440.0,
      "step": 13210
    },
    {
      "entropy": 6.252062273025513,
      "epoch": 2.5780899424446395,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043752743300573115,
      "loss": 6.0099,
      "mean_token_accuracy": 0.15132390409708024,
      "num_tokens": 44148982.0,
      "step": 13215
    },
    {
      "entropy": 6.3365942478179935,
      "epoch": 2.5790654570285825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004374787866172942,
      "loss": 5.9938,
      "mean_token_accuracy": 0.1507123053073883,
      "num_tokens": 44165540.0,
      "step": 13220
    },
    {
      "entropy": 6.245000505447388,
      "epoch": 2.580040971612526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004374301243520092,
      "loss": 6.0382,
      "mean_token_accuracy": 0.1442574143409729,
      "num_tokens": 44182038.0,
      "step": 13225
    },
    {
      "entropy": 6.284800243377686,
      "epoch": 2.581016486196469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043738144621463136,
      "loss": 6.0999,
      "mean_token_accuracy": 0.14790782928466797,
      "num_tokens": 44198838.0,
      "step": 13230
    },
    {
      "entropy": 6.241478681564331,
      "epoch": 2.581992000780412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004373327522099174,
      "loss": 5.909,
      "mean_token_accuracy": 0.15793198347091675,
      "num_tokens": 44215422.0,
      "step": 13235
    },
    {
      "entropy": 6.33096866607666,
      "epoch": 2.5829675153643548,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043728404234262547,
      "loss": 6.095,
      "mean_token_accuracy": 0.14112599194049835,
      "num_tokens": 44232019.0,
      "step": 13240
    },
    {
      "entropy": 6.3778797626495365,
      "epoch": 2.5839430299482977,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004372353166175153,
      "loss": 6.0233,
      "mean_token_accuracy": 0.14485273659229278,
      "num_tokens": 44246675.0,
      "step": 13245
    },
    {
      "entropy": 6.339674091339111,
      "epoch": 2.5849185445322407,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004371865750393482,
      "loss": 6.0905,
      "mean_token_accuracy": 0.1416084110736847,
      "num_tokens": 44262531.0,
      "step": 13250
    },
    {
      "entropy": 6.279878950119018,
      "epoch": 2.5858940591161836,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043713781761288705,
      "loss": 5.9959,
      "mean_token_accuracy": 0.15131533294916152,
      "num_tokens": 44279440.0,
      "step": 13255
    },
    {
      "entropy": 6.28959584236145,
      "epoch": 2.5868695737001266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004370890443428962,
      "loss": 6.05,
      "mean_token_accuracy": 0.14727651625871657,
      "num_tokens": 44296384.0,
      "step": 13260
    },
    {
      "entropy": 6.18218765258789,
      "epoch": 2.58784508828407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043704025523414163,
      "loss": 5.8727,
      "mean_token_accuracy": 0.1607759937644005,
      "num_tokens": 44313318.0,
      "step": 13265
    },
    {
      "entropy": 6.242984104156494,
      "epoch": 2.588820602868013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004369914502913909,
      "loss": 5.9169,
      "mean_token_accuracy": 0.1530502110719681,
      "num_tokens": 44329752.0,
      "step": 13270
    },
    {
      "entropy": 6.277010822296143,
      "epoch": 2.589796117451956,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004369426295194129,
      "loss": 5.9807,
      "mean_token_accuracy": 0.15146470218896865,
      "num_tokens": 44347318.0,
      "step": 13275
    },
    {
      "entropy": 6.317672920227051,
      "epoch": 2.590771632035899,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004368937929229783,
      "loss": 6.0205,
      "mean_token_accuracy": 0.14949339032173156,
      "num_tokens": 44362595.0,
      "step": 13280
    },
    {
      "entropy": 6.266662454605102,
      "epoch": 2.591747146619842,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043684494050685933,
      "loss": 6.0672,
      "mean_token_accuracy": 0.14381949454545975,
      "num_tokens": 44378611.0,
      "step": 13285
    },
    {
      "entropy": 6.328682279586792,
      "epoch": 2.5927226612037853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004367960722758296,
      "loss": 6.0329,
      "mean_token_accuracy": 0.14625910967588424,
      "num_tokens": 44394427.0,
      "step": 13290
    },
    {
      "entropy": 6.2726973533630375,
      "epoch": 2.593698175787728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043674718823466437,
      "loss": 6.0035,
      "mean_token_accuracy": 0.15127328783273697,
      "num_tokens": 44411952.0,
      "step": 13295
    },
    {
      "entropy": 6.27444167137146,
      "epoch": 2.594673690371671,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043669828838814037,
      "loss": 6.0001,
      "mean_token_accuracy": 0.1497980013489723,
      "num_tokens": 44428434.0,
      "step": 13300
    },
    {
      "entropy": 6.2243023872375485,
      "epoch": 2.595649204955614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004366493727410359,
      "loss": 5.9434,
      "mean_token_accuracy": 0.1558760941028595,
      "num_tokens": 44447072.0,
      "step": 13305
    },
    {
      "entropy": 6.302250957489013,
      "epoch": 2.596624719539557,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043660044129813084,
      "loss": 6.0258,
      "mean_token_accuracy": 0.14573199450969695,
      "num_tokens": 44464135.0,
      "step": 13310
    },
    {
      "entropy": 6.278775262832641,
      "epoch": 2.5976002341235,
      "grad_norm": 1.125,
      "learning_rate": 0.00043655149406420657,
      "loss": 6.0699,
      "mean_token_accuracy": 0.1478250056505203,
      "num_tokens": 44479366.0,
      "step": 13315
    },
    {
      "entropy": 6.363983249664306,
      "epoch": 2.598575748707443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004365025310440462,
      "loss": 6.0473,
      "mean_token_accuracy": 0.14568112343549727,
      "num_tokens": 44496957.0,
      "step": 13320
    },
    {
      "entropy": 6.377796363830567,
      "epoch": 2.599551263291386,
      "grad_norm": 1.109375,
      "learning_rate": 0.000436453552242434,
      "loss": 6.0676,
      "mean_token_accuracy": 0.1424328416585922,
      "num_tokens": 44514300.0,
      "step": 13325
    },
    {
      "entropy": 6.235426521301269,
      "epoch": 2.6005267778753294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043640455766415616,
      "loss": 6.0101,
      "mean_token_accuracy": 0.15003238320350648,
      "num_tokens": 44530354.0,
      "step": 13330
    },
    {
      "entropy": 6.320208168029785,
      "epoch": 2.6015022924592723,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043635554731400014,
      "loss": 6.0356,
      "mean_token_accuracy": 0.14917857348918914,
      "num_tokens": 44547175.0,
      "step": 13335
    },
    {
      "entropy": 6.246696043014526,
      "epoch": 2.6024778070432153,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043630652119675515,
      "loss": 6.0244,
      "mean_token_accuracy": 0.14774681404232978,
      "num_tokens": 44564461.0,
      "step": 13340
    },
    {
      "entropy": 6.244846391677856,
      "epoch": 2.6034533216271583,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043625747931721185,
      "loss": 6.0051,
      "mean_token_accuracy": 0.15596466213464738,
      "num_tokens": 44583031.0,
      "step": 13345
    },
    {
      "entropy": 6.300804948806762,
      "epoch": 2.604428836211101,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004362084216801624,
      "loss": 6.0156,
      "mean_token_accuracy": 0.1502944938838482,
      "num_tokens": 44600408.0,
      "step": 13350
    },
    {
      "entropy": 6.259979581832885,
      "epoch": 2.6054043507950446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043615934829040057,
      "loss": 5.9499,
      "mean_token_accuracy": 0.15652140080928803,
      "num_tokens": 44618361.0,
      "step": 13355
    },
    {
      "entropy": 6.171164083480835,
      "epoch": 2.6063798653789876,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004361102591527215,
      "loss": 5.7712,
      "mean_token_accuracy": 0.16572255194187163,
      "num_tokens": 44634246.0,
      "step": 13360
    },
    {
      "entropy": 6.157344770431519,
      "epoch": 2.6073553799629305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004360611542719222,
      "loss": 5.8935,
      "mean_token_accuracy": 0.16178106665611267,
      "num_tokens": 44650087.0,
      "step": 13365
    },
    {
      "entropy": 6.268585109710694,
      "epoch": 2.6083308945468735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004360120336528009,
      "loss": 6.0342,
      "mean_token_accuracy": 0.14989555329084397,
      "num_tokens": 44666540.0,
      "step": 13370
    },
    {
      "entropy": 6.268471527099609,
      "epoch": 2.6093064091308165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043596289730015754,
      "loss": 5.995,
      "mean_token_accuracy": 0.1568070963025093,
      "num_tokens": 44683856.0,
      "step": 13375
    },
    {
      "entropy": 6.281277990341186,
      "epoch": 2.6102819237147594,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043591374521879353,
      "loss": 5.9924,
      "mean_token_accuracy": 0.14958050847053528,
      "num_tokens": 44700297.0,
      "step": 13380
    },
    {
      "entropy": 6.222515010833741,
      "epoch": 2.6112574382987024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004358645774135118,
      "loss": 5.9273,
      "mean_token_accuracy": 0.15513278692960739,
      "num_tokens": 44716878.0,
      "step": 13385
    },
    {
      "entropy": 6.312325525283813,
      "epoch": 2.6122329528826453,
      "grad_norm": 1.0625,
      "learning_rate": 0.000435815393889117,
      "loss": 6.0374,
      "mean_token_accuracy": 0.1510889232158661,
      "num_tokens": 44734656.0,
      "step": 13390
    },
    {
      "entropy": 6.3178712844848635,
      "epoch": 2.6132084674665887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043576619465041496,
      "loss": 6.0011,
      "mean_token_accuracy": 0.1459091544151306,
      "num_tokens": 44750857.0,
      "step": 13395
    },
    {
      "entropy": 6.272499752044678,
      "epoch": 2.6141839820505317,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004357169797022134,
      "loss": 6.0072,
      "mean_token_accuracy": 0.14462407380342485,
      "num_tokens": 44767641.0,
      "step": 13400
    },
    {
      "entropy": 6.009925889968872,
      "epoch": 2.6151594966344747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004356677490493214,
      "loss": 5.711,
      "mean_token_accuracy": 0.17469221726059914,
      "num_tokens": 44785003.0,
      "step": 13405
    },
    {
      "entropy": 6.35559606552124,
      "epoch": 2.6161350112184176,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004356185026965496,
      "loss": 6.1087,
      "mean_token_accuracy": 0.14293037727475166,
      "num_tokens": 44800308.0,
      "step": 13410
    },
    {
      "entropy": 6.364641380310059,
      "epoch": 2.6171105258023606,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043556924064871023,
      "loss": 6.0345,
      "mean_token_accuracy": 0.1435229867696762,
      "num_tokens": 44817064.0,
      "step": 13415
    },
    {
      "entropy": 6.169976043701172,
      "epoch": 2.618086040386304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043551996291061684,
      "loss": 5.8784,
      "mean_token_accuracy": 0.15760020017623902,
      "num_tokens": 44834863.0,
      "step": 13420
    },
    {
      "entropy": 6.199122714996338,
      "epoch": 2.619061554970247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004354706694870848,
      "loss": 5.9115,
      "mean_token_accuracy": 0.15989896133542061,
      "num_tokens": 44852453.0,
      "step": 13425
    },
    {
      "entropy": 6.281630754470825,
      "epoch": 2.62003706955419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043542136038293087,
      "loss": 6.0043,
      "mean_token_accuracy": 0.14542725682258606,
      "num_tokens": 44868976.0,
      "step": 13430
    },
    {
      "entropy": 6.300322532653809,
      "epoch": 2.621012584138133,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353720356029733,
      "loss": 5.9746,
      "mean_token_accuracy": 0.14733068346977235,
      "num_tokens": 44886490.0,
      "step": 13435
    },
    {
      "entropy": 6.345875787734985,
      "epoch": 2.621988098722076,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004353226951520321,
      "loss": 6.0662,
      "mean_token_accuracy": 0.15007636547088624,
      "num_tokens": 44902572.0,
      "step": 13440
    },
    {
      "entropy": 6.230159044265747,
      "epoch": 2.622963613306019,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043527333903492846,
      "loss": 6.0057,
      "mean_token_accuracy": 0.14893409758806228,
      "num_tokens": 44918181.0,
      "step": 13445
    },
    {
      "entropy": 6.329685401916504,
      "epoch": 2.6239391278899618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004352239672564854,
      "loss": 6.1278,
      "mean_token_accuracy": 0.1409144379198551,
      "num_tokens": 44934809.0,
      "step": 13450
    },
    {
      "entropy": 6.365142107009888,
      "epoch": 2.6249146424739047,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043517457982152725,
      "loss": 6.0599,
      "mean_token_accuracy": 0.14461793825030328,
      "num_tokens": 44951626.0,
      "step": 13455
    },
    {
      "entropy": 6.334912824630737,
      "epoch": 2.625890157057848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004351251767348801,
      "loss": 6.087,
      "mean_token_accuracy": 0.1426209971308708,
      "num_tokens": 44968211.0,
      "step": 13460
    },
    {
      "entropy": 6.22229266166687,
      "epoch": 2.626865671641791,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043507575800137135,
      "loss": 5.9066,
      "mean_token_accuracy": 0.16053829193115235,
      "num_tokens": 44984605.0,
      "step": 13465
    },
    {
      "entropy": 6.326868772506714,
      "epoch": 2.627841186225734,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043502632362583005,
      "loss": 6.0297,
      "mean_token_accuracy": 0.14746719598770142,
      "num_tokens": 45000803.0,
      "step": 13470
    },
    {
      "entropy": 6.088940572738648,
      "epoch": 2.628816700809677,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043497687361308677,
      "loss": 5.8682,
      "mean_token_accuracy": 0.17036545723676683,
      "num_tokens": 45018362.0,
      "step": 13475
    },
    {
      "entropy": 6.188544511795044,
      "epoch": 2.6297922153936204,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043492740796797347,
      "loss": 5.8346,
      "mean_token_accuracy": 0.1653712123632431,
      "num_tokens": 45033160.0,
      "step": 13480
    },
    {
      "entropy": 6.299918508529663,
      "epoch": 2.6307677299775634,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004348779266953239,
      "loss": 6.0295,
      "mean_token_accuracy": 0.15312789678573607,
      "num_tokens": 45049631.0,
      "step": 13485
    },
    {
      "entropy": 6.202289772033692,
      "epoch": 2.6317432445615063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043482842979997323,
      "loss": 5.9295,
      "mean_token_accuracy": 0.15712662190198898,
      "num_tokens": 45067083.0,
      "step": 13490
    },
    {
      "entropy": 6.272989797592163,
      "epoch": 2.6327187591454493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004347789172867579,
      "loss": 6.0222,
      "mean_token_accuracy": 0.1526799313724041,
      "num_tokens": 45083576.0,
      "step": 13495
    },
    {
      "entropy": 6.24267430305481,
      "epoch": 2.6336942737293922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043472938916051633,
      "loss": 5.9536,
      "mean_token_accuracy": 0.1608738973736763,
      "num_tokens": 45100738.0,
      "step": 13500
    },
    {
      "entropy": 6.250103712081909,
      "epoch": 2.634669788313335,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004346798454260882,
      "loss": 6.0004,
      "mean_token_accuracy": 0.14364166632294656,
      "num_tokens": 45117254.0,
      "step": 13505
    },
    {
      "entropy": 6.106761360168457,
      "epoch": 2.635645302897278,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004346302860883146,
      "loss": 5.7851,
      "mean_token_accuracy": 0.17893586456775665,
      "num_tokens": 45135160.0,
      "step": 13510
    },
    {
      "entropy": 6.25085415840149,
      "epoch": 2.636620817481221,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043458071115203836,
      "loss": 5.936,
      "mean_token_accuracy": 0.15073021352291108,
      "num_tokens": 45152025.0,
      "step": 13515
    },
    {
      "entropy": 6.146157836914062,
      "epoch": 2.6375963320651645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043453112062210377,
      "loss": 5.8694,
      "mean_token_accuracy": 0.15590533316135408,
      "num_tokens": 45169452.0,
      "step": 13520
    },
    {
      "entropy": 6.215482425689697,
      "epoch": 2.6385718466491075,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043448151450335666,
      "loss": 5.9015,
      "mean_token_accuracy": 0.16144839376211167,
      "num_tokens": 45186401.0,
      "step": 13525
    },
    {
      "entropy": 6.254865312576294,
      "epoch": 2.6395473612330504,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043443189280064444,
      "loss": 5.9775,
      "mean_token_accuracy": 0.1495829179883003,
      "num_tokens": 45201713.0,
      "step": 13530
    },
    {
      "entropy": 6.306210851669311,
      "epoch": 2.6405228758169934,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043438225551881577,
      "loss": 6.041,
      "mean_token_accuracy": 0.14674002528190613,
      "num_tokens": 45216727.0,
      "step": 13535
    },
    {
      "entropy": 6.275552797317505,
      "epoch": 2.6414983904009364,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043433260266272116,
      "loss": 5.9635,
      "mean_token_accuracy": 0.14885595589876174,
      "num_tokens": 45231958.0,
      "step": 13540
    },
    {
      "entropy": 6.271491765975952,
      "epoch": 2.6424739049848798,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004342829342372125,
      "loss": 6.0314,
      "mean_token_accuracy": 0.14324500113725663,
      "num_tokens": 45248620.0,
      "step": 13545
    },
    {
      "entropy": 6.206141185760498,
      "epoch": 2.6434494195688227,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004342332502471432,
      "loss": 5.8897,
      "mean_token_accuracy": 0.1574290692806244,
      "num_tokens": 45264065.0,
      "step": 13550
    },
    {
      "entropy": 6.311268997192383,
      "epoch": 2.6444249341527657,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043418355069736834,
      "loss": 6.0383,
      "mean_token_accuracy": 0.14806754589080812,
      "num_tokens": 45279818.0,
      "step": 13555
    },
    {
      "entropy": 6.168830728530883,
      "epoch": 2.6454004487367087,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043413383559274404,
      "loss": 5.8933,
      "mean_token_accuracy": 0.1624483048915863,
      "num_tokens": 45296282.0,
      "step": 13560
    },
    {
      "entropy": 6.216638088226318,
      "epoch": 2.6463759633206516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004340841049381286,
      "loss": 5.9168,
      "mean_token_accuracy": 0.15362644866108893,
      "num_tokens": 45314345.0,
      "step": 13565
    },
    {
      "entropy": 6.239054584503174,
      "epoch": 2.6473514779045946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004340343587383814,
      "loss": 5.9646,
      "mean_token_accuracy": 0.15823800414800643,
      "num_tokens": 45329975.0,
      "step": 13570
    },
    {
      "entropy": 6.254324245452881,
      "epoch": 2.6483269924885375,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004339845969983634,
      "loss": 5.9462,
      "mean_token_accuracy": 0.15487911105155944,
      "num_tokens": 45347163.0,
      "step": 13575
    },
    {
      "entropy": 6.200194406509399,
      "epoch": 2.6493025070724805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004339348197229373,
      "loss": 5.9413,
      "mean_token_accuracy": 0.15957598984241486,
      "num_tokens": 45362966.0,
      "step": 13580
    },
    {
      "entropy": 6.271059226989746,
      "epoch": 2.650278021656424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004338850269169671,
      "loss": 6.0151,
      "mean_token_accuracy": 0.14654436260461806,
      "num_tokens": 45379628.0,
      "step": 13585
    },
    {
      "entropy": 6.250176811218262,
      "epoch": 2.651253536240367,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004338352185853183,
      "loss": 5.9465,
      "mean_token_accuracy": 0.15229950100183487,
      "num_tokens": 45397088.0,
      "step": 13590
    },
    {
      "entropy": 6.27294430732727,
      "epoch": 2.65222905082431,
      "grad_norm": 1.015625,
      "learning_rate": 0.000433785394732858,
      "loss": 5.9374,
      "mean_token_accuracy": 0.15492564886808396,
      "num_tokens": 45414134.0,
      "step": 13595
    },
    {
      "entropy": 6.222795152664185,
      "epoch": 2.653204565408253,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043373555536445487,
      "loss": 5.9178,
      "mean_token_accuracy": 0.15609867349267006,
      "num_tokens": 45431398.0,
      "step": 13600
    },
    {
      "entropy": 6.286456298828125,
      "epoch": 2.6541800799921957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043368570048497895,
      "loss": 6.0424,
      "mean_token_accuracy": 0.14936774522066115,
      "num_tokens": 45448867.0,
      "step": 13605
    },
    {
      "entropy": 6.304460334777832,
      "epoch": 2.655155594576139,
      "grad_norm": 1.125,
      "learning_rate": 0.0004336358300993021,
      "loss": 5.9967,
      "mean_token_accuracy": 0.14834614992141723,
      "num_tokens": 45466388.0,
      "step": 13610
    },
    {
      "entropy": 6.249767684936524,
      "epoch": 2.656131109160082,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043358594421229717,
      "loss": 5.9587,
      "mean_token_accuracy": 0.1496449664235115,
      "num_tokens": 45483711.0,
      "step": 13615
    },
    {
      "entropy": 6.218064069747925,
      "epoch": 2.657106623744025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000433536042828839,
      "loss": 5.9246,
      "mean_token_accuracy": 0.15566006451845169,
      "num_tokens": 45502228.0,
      "step": 13620
    },
    {
      "entropy": 6.193723487854004,
      "epoch": 2.658082138327968,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004334861259538038,
      "loss": 5.879,
      "mean_token_accuracy": 0.15702656358480455,
      "num_tokens": 45519278.0,
      "step": 13625
    },
    {
      "entropy": 6.282367610931397,
      "epoch": 2.659057652911911,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004334361935920693,
      "loss": 6.0538,
      "mean_token_accuracy": 0.14643797725439073,
      "num_tokens": 45535287.0,
      "step": 13630
    },
    {
      "entropy": 6.3444867610931395,
      "epoch": 2.660033167495854,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043338624574851454,
      "loss": 6.0194,
      "mean_token_accuracy": 0.14983067587018012,
      "num_tokens": 45551744.0,
      "step": 13635
    },
    {
      "entropy": 6.3258379936218265,
      "epoch": 2.661008682079797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004333362824280204,
      "loss": 6.0695,
      "mean_token_accuracy": 0.14857302010059356,
      "num_tokens": 45569089.0,
      "step": 13640
    },
    {
      "entropy": 6.251392459869384,
      "epoch": 2.66198419666374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043328630363546894,
      "loss": 6.0052,
      "mean_token_accuracy": 0.15395837724208833,
      "num_tokens": 45586750.0,
      "step": 13645
    },
    {
      "entropy": 6.273073768615722,
      "epoch": 2.6629597112476833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004332363093757442,
      "loss": 5.9776,
      "mean_token_accuracy": 0.15500842183828353,
      "num_tokens": 45604103.0,
      "step": 13650
    },
    {
      "entropy": 6.244315671920776,
      "epoch": 2.6639352258316262,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004331862996537312,
      "loss": 5.9954,
      "mean_token_accuracy": 0.14981036484241486,
      "num_tokens": 45620588.0,
      "step": 13655
    },
    {
      "entropy": 6.331147909164429,
      "epoch": 2.664910740415569,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004331362744743168,
      "loss": 6.0386,
      "mean_token_accuracy": 0.14957306236028672,
      "num_tokens": 45638289.0,
      "step": 13660
    },
    {
      "entropy": 6.1945549011230465,
      "epoch": 2.665886254999512,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043308623384238925,
      "loss": 5.8747,
      "mean_token_accuracy": 0.16567822992801667,
      "num_tokens": 45655904.0,
      "step": 13665
    },
    {
      "entropy": 6.257582187652588,
      "epoch": 2.666861769583455,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043303617776283834,
      "loss": 6.0026,
      "mean_token_accuracy": 0.1521809220314026,
      "num_tokens": 45673314.0,
      "step": 13670
    },
    {
      "entropy": 6.288122987747192,
      "epoch": 2.6678372841673985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004329861062405554,
      "loss": 5.9533,
      "mean_token_accuracy": 0.15800718516111373,
      "num_tokens": 45689082.0,
      "step": 13675
    },
    {
      "entropy": 6.298369121551514,
      "epoch": 2.6688127987513415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004329360192804332,
      "loss": 6.103,
      "mean_token_accuracy": 0.1483241379261017,
      "num_tokens": 45705979.0,
      "step": 13680
    },
    {
      "entropy": 6.423534774780274,
      "epoch": 2.6697883133352844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004328859168873661,
      "loss": 6.1459,
      "mean_token_accuracy": 0.1393655240535736,
      "num_tokens": 45722941.0,
      "step": 13685
    },
    {
      "entropy": 6.300044345855713,
      "epoch": 2.6707638279192274,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043283579906625003,
      "loss": 5.9978,
      "mean_token_accuracy": 0.1536937437951565,
      "num_tokens": 45739625.0,
      "step": 13690
    },
    {
      "entropy": 6.2493690013885494,
      "epoch": 2.6717393425031704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004327856658219821,
      "loss": 6.0318,
      "mean_token_accuracy": 0.15043622851371766,
      "num_tokens": 45756008.0,
      "step": 13695
    },
    {
      "entropy": 6.307055473327637,
      "epoch": 2.6727148570871133,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043273551715946126,
      "loss": 6.0045,
      "mean_token_accuracy": 0.14971668869256974,
      "num_tokens": 45773043.0,
      "step": 13700
    },
    {
      "entropy": 6.160358905792236,
      "epoch": 2.6736903716710563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004326853530835879,
      "loss": 5.8615,
      "mean_token_accuracy": 0.16165245026350022,
      "num_tokens": 45790987.0,
      "step": 13705
    },
    {
      "entropy": 6.3080991268157955,
      "epoch": 2.6746658862549992,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004326351735992638,
      "loss": 6.0173,
      "mean_token_accuracy": 0.14715195894241334,
      "num_tokens": 45807867.0,
      "step": 13710
    },
    {
      "entropy": 6.315697383880615,
      "epoch": 2.6756414008389426,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004325849787113924,
      "loss": 6.0582,
      "mean_token_accuracy": 0.1485403448343277,
      "num_tokens": 45824806.0,
      "step": 13715
    },
    {
      "entropy": 6.274390125274659,
      "epoch": 2.6766169154228856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004325347684248785,
      "loss": 5.9548,
      "mean_token_accuracy": 0.15337311923503877,
      "num_tokens": 45841013.0,
      "step": 13720
    },
    {
      "entropy": 6.360094308853149,
      "epoch": 2.6775924300068286,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004324845427446285,
      "loss": 6.0182,
      "mean_token_accuracy": 0.14698541387915612,
      "num_tokens": 45857618.0,
      "step": 13725
    },
    {
      "entropy": 6.253412914276123,
      "epoch": 2.6785679445907715,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043243430167555025,
      "loss": 5.9814,
      "mean_token_accuracy": 0.1509335681796074,
      "num_tokens": 45875807.0,
      "step": 13730
    },
    {
      "entropy": 6.299402236938477,
      "epoch": 2.679543459174715,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004323840452225532,
      "loss": 6.018,
      "mean_token_accuracy": 0.1466420531272888,
      "num_tokens": 45892580.0,
      "step": 13735
    },
    {
      "entropy": 6.2668030738830565,
      "epoch": 2.680518973758658,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004323337733905481,
      "loss": 6.0321,
      "mean_token_accuracy": 0.14921612292528152,
      "num_tokens": 45908717.0,
      "step": 13740
    },
    {
      "entropy": 6.2118017196655275,
      "epoch": 2.681494488342601,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043228348618444746,
      "loss": 5.9181,
      "mean_token_accuracy": 0.1598734214901924,
      "num_tokens": 45925200.0,
      "step": 13745
    },
    {
      "entropy": 6.307936191558838,
      "epoch": 2.682470002926544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004322331836091651,
      "loss": 6.013,
      "mean_token_accuracy": 0.14726333767175676,
      "num_tokens": 45942384.0,
      "step": 13750
    },
    {
      "entropy": 6.267459201812744,
      "epoch": 2.6834455175104868,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004321828656696165,
      "loss": 6.0426,
      "mean_token_accuracy": 0.1508985787630081,
      "num_tokens": 45959040.0,
      "step": 13755
    },
    {
      "entropy": 6.317785358428955,
      "epoch": 2.6844210320944297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043213253237071837,
      "loss": 6.015,
      "mean_token_accuracy": 0.14986944198608398,
      "num_tokens": 45975559.0,
      "step": 13760
    },
    {
      "entropy": 6.2903828620910645,
      "epoch": 2.6853965466783727,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043208218371738926,
      "loss": 6.063,
      "mean_token_accuracy": 0.14475433081388472,
      "num_tokens": 45992587.0,
      "step": 13765
    },
    {
      "entropy": 6.350510454177856,
      "epoch": 2.6863720612623156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043203181971454904,
      "loss": 6.0567,
      "mean_token_accuracy": 0.14557042419910432,
      "num_tokens": 46008612.0,
      "step": 13770
    },
    {
      "entropy": 6.36482048034668,
      "epoch": 2.687347575846259,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043198144036711903,
      "loss": 6.1272,
      "mean_token_accuracy": 0.1458436630666256,
      "num_tokens": 46025730.0,
      "step": 13775
    },
    {
      "entropy": 6.197410440444946,
      "epoch": 2.688323090430202,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043193104568002214,
      "loss": 5.9087,
      "mean_token_accuracy": 0.16397111266851425,
      "num_tokens": 46041389.0,
      "step": 13780
    },
    {
      "entropy": 6.361960363388062,
      "epoch": 2.689298605014145,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043188063565818277,
      "loss": 6.0281,
      "mean_token_accuracy": 0.15126081854104995,
      "num_tokens": 46056805.0,
      "step": 13785
    },
    {
      "entropy": 6.22874755859375,
      "epoch": 2.690274119598088,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004318302103065269,
      "loss": 5.9206,
      "mean_token_accuracy": 0.16211329400539398,
      "num_tokens": 46073219.0,
      "step": 13790
    },
    {
      "entropy": 6.324752855300903,
      "epoch": 2.691249634182031,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004317797696299818,
      "loss": 6.0439,
      "mean_token_accuracy": 0.1519034117460251,
      "num_tokens": 46089009.0,
      "step": 13795
    },
    {
      "entropy": 6.323377418518066,
      "epoch": 2.6922251487659743,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004317293136334763,
      "loss": 6.0161,
      "mean_token_accuracy": 0.15173562616109848,
      "num_tokens": 46104394.0,
      "step": 13800
    },
    {
      "entropy": 6.365319490432739,
      "epoch": 2.6932006633499173,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00043167884232194094,
      "loss": 6.0405,
      "mean_token_accuracy": 0.14761460497975348,
      "num_tokens": 46121183.0,
      "step": 13805
    },
    {
      "entropy": 6.185376262664795,
      "epoch": 2.69417617793386,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004316283557003075,
      "loss": 5.8496,
      "mean_token_accuracy": 0.17014867514371873,
      "num_tokens": 46138594.0,
      "step": 13810
    },
    {
      "entropy": 6.229904079437256,
      "epoch": 2.695151692517803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004315778537735094,
      "loss": 6.0145,
      "mean_token_accuracy": 0.15178247094154357,
      "num_tokens": 46155601.0,
      "step": 13815
    },
    {
      "entropy": 6.299903869628906,
      "epoch": 2.696127207101746,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043152733654648146,
      "loss": 5.9325,
      "mean_token_accuracy": 0.15356887727975846,
      "num_tokens": 46172282.0,
      "step": 13820
    },
    {
      "entropy": 6.255138874053955,
      "epoch": 2.697102721685689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043147680402416,
      "loss": 6.0009,
      "mean_token_accuracy": 0.1523323819041252,
      "num_tokens": 46188609.0,
      "step": 13825
    },
    {
      "entropy": 6.136853265762329,
      "epoch": 2.698078236269632,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000431426256211483,
      "loss": 5.8416,
      "mean_token_accuracy": 0.16691785156726838,
      "num_tokens": 46204624.0,
      "step": 13830
    },
    {
      "entropy": 6.183567333221435,
      "epoch": 2.699053750853575,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043137569311338977,
      "loss": 5.8647,
      "mean_token_accuracy": 0.16373968720436097,
      "num_tokens": 46221516.0,
      "step": 13835
    },
    {
      "entropy": 6.272261428833008,
      "epoch": 2.7000292654375184,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004313251147348211,
      "loss": 5.9618,
      "mean_token_accuracy": 0.15526060461997987,
      "num_tokens": 46238983.0,
      "step": 13840
    },
    {
      "entropy": 6.367432546615601,
      "epoch": 2.7010047800214614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043127452108071947,
      "loss": 6.0109,
      "mean_token_accuracy": 0.14654828011989593,
      "num_tokens": 46254158.0,
      "step": 13845
    },
    {
      "entropy": 6.311488151550293,
      "epoch": 2.7019802946054043,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004312239121560285,
      "loss": 6.0859,
      "mean_token_accuracy": 0.14225918427109718,
      "num_tokens": 46270566.0,
      "step": 13850
    },
    {
      "entropy": 6.302449369430542,
      "epoch": 2.7029558091893473,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043117328796569374,
      "loss": 6.008,
      "mean_token_accuracy": 0.14548689126968384,
      "num_tokens": 46288136.0,
      "step": 13855
    },
    {
      "entropy": 6.194075345993042,
      "epoch": 2.7039313237732903,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043112264851466177,
      "loss": 5.8855,
      "mean_token_accuracy": 0.15840536803007127,
      "num_tokens": 46305418.0,
      "step": 13860
    },
    {
      "entropy": 6.310913991928101,
      "epoch": 2.7049068383572337,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004310719938078811,
      "loss": 5.9912,
      "mean_token_accuracy": 0.15113988965749742,
      "num_tokens": 46321897.0,
      "step": 13865
    },
    {
      "entropy": 6.374405097961426,
      "epoch": 2.7058823529411766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004310213238503014,
      "loss": 6.1093,
      "mean_token_accuracy": 0.14547561407089232,
      "num_tokens": 46338936.0,
      "step": 13870
    },
    {
      "entropy": 6.266456460952758,
      "epoch": 2.7068578675251196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004309706386468741,
      "loss": 5.9381,
      "mean_token_accuracy": 0.1474335789680481,
      "num_tokens": 46354453.0,
      "step": 13875
    },
    {
      "entropy": 6.27151198387146,
      "epoch": 2.7078333821090625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043091993820255177,
      "loss": 5.9774,
      "mean_token_accuracy": 0.14929064363241196,
      "num_tokens": 46369441.0,
      "step": 13880
    },
    {
      "entropy": 6.212397766113281,
      "epoch": 2.7088088966930055,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043086922252228886,
      "loss": 5.8939,
      "mean_token_accuracy": 0.15947671085596085,
      "num_tokens": 46386273.0,
      "step": 13885
    },
    {
      "entropy": 6.166885042190552,
      "epoch": 2.7097844112769485,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043081849161104096,
      "loss": 5.9359,
      "mean_token_accuracy": 0.15587268322706221,
      "num_tokens": 46403274.0,
      "step": 13890
    },
    {
      "entropy": 6.213001680374146,
      "epoch": 2.7107599258608914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043076774547376554,
      "loss": 5.9833,
      "mean_token_accuracy": 0.15634427070617676,
      "num_tokens": 46422601.0,
      "step": 13895
    },
    {
      "entropy": 6.324032497406006,
      "epoch": 2.7117354404448344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004307169841154211,
      "loss": 6.0006,
      "mean_token_accuracy": 0.1507381796836853,
      "num_tokens": 46437861.0,
      "step": 13900
    },
    {
      "entropy": 6.329212427139282,
      "epoch": 2.712710955028778,
      "grad_norm": 1.125,
      "learning_rate": 0.00043066620754096793,
      "loss": 6.0513,
      "mean_token_accuracy": 0.14760782569646835,
      "num_tokens": 46454558.0,
      "step": 13905
    },
    {
      "entropy": 6.149903154373169,
      "epoch": 2.7136864696127208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004306154157553679,
      "loss": 5.8202,
      "mean_token_accuracy": 0.170747072994709,
      "num_tokens": 46470975.0,
      "step": 13910
    },
    {
      "entropy": 6.232824850082397,
      "epoch": 2.7146619841966637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043056460876358396,
      "loss": 5.9532,
      "mean_token_accuracy": 0.1604907989501953,
      "num_tokens": 46487791.0,
      "step": 13915
    },
    {
      "entropy": 6.158563232421875,
      "epoch": 2.7156374987806067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004305137865705809,
      "loss": 5.8161,
      "mean_token_accuracy": 0.16742174476385116,
      "num_tokens": 46504668.0,
      "step": 13920
    },
    {
      "entropy": 6.205779361724853,
      "epoch": 2.7166130133645496,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043046294918132496,
      "loss": 5.9387,
      "mean_token_accuracy": 0.15178529918193817,
      "num_tokens": 46519994.0,
      "step": 13925
    },
    {
      "entropy": 6.141224098205567,
      "epoch": 2.717588527948493,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004304120966007835,
      "loss": 5.8957,
      "mean_token_accuracy": 0.16030471697449683,
      "num_tokens": 46537407.0,
      "step": 13930
    },
    {
      "entropy": 6.274479103088379,
      "epoch": 2.718564042532436,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000430361228833926,
      "loss": 5.97,
      "mean_token_accuracy": 0.1588030993938446,
      "num_tokens": 46555518.0,
      "step": 13935
    },
    {
      "entropy": 6.3004203796386715,
      "epoch": 2.719539557116379,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043031034588572294,
      "loss": 5.9929,
      "mean_token_accuracy": 0.1486786663532257,
      "num_tokens": 46573404.0,
      "step": 13940
    },
    {
      "entropy": 6.178061532974243,
      "epoch": 2.720515071700322,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004302594477611464,
      "loss": 5.8393,
      "mean_token_accuracy": 0.16207536607980727,
      "num_tokens": 46588634.0,
      "step": 13945
    },
    {
      "entropy": 6.172136926651001,
      "epoch": 2.721490586284265,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043020853446517005,
      "loss": 5.9295,
      "mean_token_accuracy": 0.15693214684724807,
      "num_tokens": 46606105.0,
      "step": 13950
    },
    {
      "entropy": 6.226840972900391,
      "epoch": 2.722466100868208,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043015760600276877,
      "loss": 5.8527,
      "mean_token_accuracy": 0.16347473561763765,
      "num_tokens": 46623586.0,
      "step": 13955
    },
    {
      "entropy": 6.219112491607666,
      "epoch": 2.723441615452151,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043010666237891927,
      "loss": 5.9084,
      "mean_token_accuracy": 0.16023678332567215,
      "num_tokens": 46640198.0,
      "step": 13960
    },
    {
      "entropy": 6.213691234588623,
      "epoch": 2.724417130036094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004300557035985995,
      "loss": 6.0237,
      "mean_token_accuracy": 0.14815241992473602,
      "num_tokens": 46656892.0,
      "step": 13965
    },
    {
      "entropy": 6.327918195724488,
      "epoch": 2.725392644620037,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043000472966678906,
      "loss": 5.9972,
      "mean_token_accuracy": 0.15665410757064818,
      "num_tokens": 46673796.0,
      "step": 13970
    },
    {
      "entropy": 6.143670272827149,
      "epoch": 2.72636815920398,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004299537405884689,
      "loss": 5.8488,
      "mean_token_accuracy": 0.16740578263998032,
      "num_tokens": 46691399.0,
      "step": 13975
    },
    {
      "entropy": 6.140094852447509,
      "epoch": 2.727343673787923,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042990273636862137,
      "loss": 5.8515,
      "mean_token_accuracy": 0.16018384248018264,
      "num_tokens": 46708738.0,
      "step": 13980
    },
    {
      "entropy": 6.320862913131714,
      "epoch": 2.728319188371866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004298517170122307,
      "loss": 6.0091,
      "mean_token_accuracy": 0.1543658435344696,
      "num_tokens": 46725481.0,
      "step": 13985
    },
    {
      "entropy": 6.276321649551392,
      "epoch": 2.7292947029558094,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042980068252428203,
      "loss": 5.9413,
      "mean_token_accuracy": 0.1513228729367256,
      "num_tokens": 46741758.0,
      "step": 13990
    },
    {
      "entropy": 6.28612642288208,
      "epoch": 2.7302702175397524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042974963290976235,
      "loss": 6.0434,
      "mean_token_accuracy": 0.1467214584350586,
      "num_tokens": 46758448.0,
      "step": 13995
    },
    {
      "entropy": 6.2244117736816404,
      "epoch": 2.7312457321236954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004296985681736601,
      "loss": 5.9349,
      "mean_token_accuracy": 0.15568670630455017,
      "num_tokens": 46775261.0,
      "step": 14000
    },
    {
      "entropy": 6.202472972869873,
      "epoch": 2.7322212467076383,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042964748832096517,
      "loss": 5.888,
      "mean_token_accuracy": 0.16172611117362976,
      "num_tokens": 46792484.0,
      "step": 14005
    },
    {
      "entropy": 6.289637422561645,
      "epoch": 2.7331967612915813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004295963933566688,
      "loss": 6.0602,
      "mean_token_accuracy": 0.14233896434307097,
      "num_tokens": 46810078.0,
      "step": 14010
    },
    {
      "entropy": 6.237289524078369,
      "epoch": 2.7341722758755242,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004295452832857639,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1528948649764061,
      "num_tokens": 46826524.0,
      "step": 14015
    },
    {
      "entropy": 6.231190729141235,
      "epoch": 2.735147790459467,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004294941581132447,
      "loss": 5.8987,
      "mean_token_accuracy": 0.15983614325523376,
      "num_tokens": 46842939.0,
      "step": 14020
    },
    {
      "entropy": 6.271780109405517,
      "epoch": 2.73612330504341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042944301784410705,
      "loss": 6.0747,
      "mean_token_accuracy": 0.14597091153264047,
      "num_tokens": 46858060.0,
      "step": 14025
    },
    {
      "entropy": 6.285999345779419,
      "epoch": 2.7370988196273536,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004293918624833481,
      "loss": 5.9495,
      "mean_token_accuracy": 0.15466022342443467,
      "num_tokens": 46873434.0,
      "step": 14030
    },
    {
      "entropy": 6.276752138137818,
      "epoch": 2.7380743342112965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004293406920359665,
      "loss": 6.0473,
      "mean_token_accuracy": 0.15315240547060965,
      "num_tokens": 46890689.0,
      "step": 14035
    },
    {
      "entropy": 6.344392204284668,
      "epoch": 2.7390498487952395,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042928950650696274,
      "loss": 6.0427,
      "mean_token_accuracy": 0.14695354253053666,
      "num_tokens": 46906384.0,
      "step": 14040
    },
    {
      "entropy": 6.254943132400513,
      "epoch": 2.7400253633791825,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042923830590133825,
      "loss": 6.0335,
      "mean_token_accuracy": 0.1464530646800995,
      "num_tokens": 46922154.0,
      "step": 14045
    },
    {
      "entropy": 6.350614356994629,
      "epoch": 2.7410008779631254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042918709022409624,
      "loss": 6.0437,
      "mean_token_accuracy": 0.14182298481464387,
      "num_tokens": 46939213.0,
      "step": 14050
    },
    {
      "entropy": 6.250848627090454,
      "epoch": 2.741976392547069,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004291358594802412,
      "loss": 5.973,
      "mean_token_accuracy": 0.15557192414999008,
      "num_tokens": 46956207.0,
      "step": 14055
    },
    {
      "entropy": 6.29512095451355,
      "epoch": 2.742951907131012,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004290846136747795,
      "loss": 6.026,
      "mean_token_accuracy": 0.1549872487783432,
      "num_tokens": 46972228.0,
      "step": 14060
    },
    {
      "entropy": 6.350634765625,
      "epoch": 2.7439274217149547,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004290333528127184,
      "loss": 6.0352,
      "mean_token_accuracy": 0.15253800451755523,
      "num_tokens": 46988398.0,
      "step": 14065
    },
    {
      "entropy": 6.218199253082275,
      "epoch": 2.7449029362988977,
      "grad_norm": 1.0,
      "learning_rate": 0.000428982076899067,
      "loss": 5.8914,
      "mean_token_accuracy": 0.16032354682683944,
      "num_tokens": 47006178.0,
      "step": 14070
    },
    {
      "entropy": 6.295246505737305,
      "epoch": 2.7458784508828407,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004289307859388358,
      "loss": 6.0558,
      "mean_token_accuracy": 0.1530913382768631,
      "num_tokens": 47022703.0,
      "step": 14075
    },
    {
      "entropy": 6.2664632320404055,
      "epoch": 2.7468539654667836,
      "grad_norm": 1.109375,
      "learning_rate": 0.000428879479937037,
      "loss": 6.0552,
      "mean_token_accuracy": 0.1503320649266243,
      "num_tokens": 47038164.0,
      "step": 14080
    },
    {
      "entropy": 6.202656698226929,
      "epoch": 2.7478294800507266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004288281588986838,
      "loss": 5.9703,
      "mean_token_accuracy": 0.15179618671536446,
      "num_tokens": 47055080.0,
      "step": 14085
    },
    {
      "entropy": 6.376747512817383,
      "epoch": 2.7488049946346695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004287768228287912,
      "loss": 6.0011,
      "mean_token_accuracy": 0.1494814157485962,
      "num_tokens": 47071121.0,
      "step": 14090
    },
    {
      "entropy": 6.041538429260254,
      "epoch": 2.749780509218613,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004287254717323754,
      "loss": 5.6749,
      "mean_token_accuracy": 0.18591688126325606,
      "num_tokens": 47088303.0,
      "step": 14095
    },
    {
      "entropy": 6.3234474658966064,
      "epoch": 2.750756023802556,
      "grad_norm": 1.0,
      "learning_rate": 0.00042867410561445445,
      "loss": 6.0777,
      "mean_token_accuracy": 0.14542057365179062,
      "num_tokens": 47104637.0,
      "step": 14100
    },
    {
      "entropy": 6.350788307189942,
      "epoch": 2.751731538386499,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004286227244800477,
      "loss": 6.0518,
      "mean_token_accuracy": 0.15049393922090532,
      "num_tokens": 47121234.0,
      "step": 14105
    },
    {
      "entropy": 6.142922973632812,
      "epoch": 2.752707052970442,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042857132833417575,
      "loss": 5.8691,
      "mean_token_accuracy": 0.17053837031126023,
      "num_tokens": 47139264.0,
      "step": 14110
    },
    {
      "entropy": 6.201049041748047,
      "epoch": 2.753682567554385,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004285199171818609,
      "loss": 5.8038,
      "mean_token_accuracy": 0.16298413872718812,
      "num_tokens": 47155546.0,
      "step": 14115
    },
    {
      "entropy": 6.275006246566773,
      "epoch": 2.754658082138328,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042846849102812695,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1549730658531189,
      "num_tokens": 47171468.0,
      "step": 14120
    },
    {
      "entropy": 6.112595081329346,
      "epoch": 2.755633596722271,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004284170498779991,
      "loss": 5.796,
      "mean_token_accuracy": 0.16660883873701096,
      "num_tokens": 47187937.0,
      "step": 14125
    },
    {
      "entropy": 6.2292245388031,
      "epoch": 2.756609111306214,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004283655937365038,
      "loss": 5.936,
      "mean_token_accuracy": 0.16003748551011085,
      "num_tokens": 47204306.0,
      "step": 14130
    },
    {
      "entropy": 6.17051944732666,
      "epoch": 2.757584625890157,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004283141226086693,
      "loss": 5.8914,
      "mean_token_accuracy": 0.16046102643013,
      "num_tokens": 47222040.0,
      "step": 14135
    },
    {
      "entropy": 6.168228101730347,
      "epoch": 2.7585601404741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004282626364995252,
      "loss": 5.8476,
      "mean_token_accuracy": 0.15722304135560988,
      "num_tokens": 47237889.0,
      "step": 14140
    },
    {
      "entropy": 6.178996849060058,
      "epoch": 2.759535655058043,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042821113541410247,
      "loss": 5.9389,
      "mean_token_accuracy": 0.1565202608704567,
      "num_tokens": 47253752.0,
      "step": 14145
    },
    {
      "entropy": 6.218247890472412,
      "epoch": 2.760511169641986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042815961935743367,
      "loss": 5.9438,
      "mean_token_accuracy": 0.15213825702667236,
      "num_tokens": 47271469.0,
      "step": 14150
    },
    {
      "entropy": 6.24660177230835,
      "epoch": 2.761486684225929,
      "grad_norm": 1.0,
      "learning_rate": 0.00042810808833455274,
      "loss": 5.9901,
      "mean_token_accuracy": 0.15906620770692825,
      "num_tokens": 47288746.0,
      "step": 14155
    },
    {
      "entropy": 6.276077079772949,
      "epoch": 2.7624621988098723,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00042805654235049514,
      "loss": 5.9424,
      "mean_token_accuracy": 0.15555969178676604,
      "num_tokens": 47308101.0,
      "step": 14160
    },
    {
      "entropy": 6.296984434127808,
      "epoch": 2.7634377133938153,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004280049814102976,
      "loss": 6.0187,
      "mean_token_accuracy": 0.14910627007484437,
      "num_tokens": 47324441.0,
      "step": 14165
    },
    {
      "entropy": 6.31481294631958,
      "epoch": 2.7644132279777582,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004279534055189987,
      "loss": 6.0291,
      "mean_token_accuracy": 0.15069946497678757,
      "num_tokens": 47341664.0,
      "step": 14170
    },
    {
      "entropy": 6.254401302337646,
      "epoch": 2.765388742561701,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042790181468163813,
      "loss": 6.0201,
      "mean_token_accuracy": 0.153005912899971,
      "num_tokens": 47358424.0,
      "step": 14175
    },
    {
      "entropy": 6.198832893371582,
      "epoch": 2.7663642571456446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004278502089032572,
      "loss": 5.8798,
      "mean_token_accuracy": 0.16274735927581788,
      "num_tokens": 47375211.0,
      "step": 14180
    },
    {
      "entropy": 6.201909017562866,
      "epoch": 2.7673397717295876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004277985881888986,
      "loss": 5.8619,
      "mean_token_accuracy": 0.1620197579264641,
      "num_tokens": 47393947.0,
      "step": 14185
    },
    {
      "entropy": 6.336785793304443,
      "epoch": 2.7683152863135305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004277469525436065,
      "loss": 6.0149,
      "mean_token_accuracy": 0.1484317436814308,
      "num_tokens": 47410519.0,
      "step": 14190
    },
    {
      "entropy": 6.242080640792847,
      "epoch": 2.7692908008974735,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042769530197242674,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1551412433385849,
      "num_tokens": 47426672.0,
      "step": 14195
    },
    {
      "entropy": 6.341984701156616,
      "epoch": 2.7702663154814164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004276436364804061,
      "loss": 6.0861,
      "mean_token_accuracy": 0.14852617383003236,
      "num_tokens": 47443545.0,
      "step": 14200
    },
    {
      "entropy": 6.366255092620849,
      "epoch": 2.7712418300653594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042759195607259344,
      "loss": 6.0062,
      "mean_token_accuracy": 0.14768428355455399,
      "num_tokens": 47459543.0,
      "step": 14205
    },
    {
      "entropy": 6.257879829406738,
      "epoch": 2.7722173446493024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042754026075403866,
      "loss": 6.0037,
      "mean_token_accuracy": 0.14882007911801337,
      "num_tokens": 47476240.0,
      "step": 14210
    },
    {
      "entropy": 6.1007080078125,
      "epoch": 2.7731928592332453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042748855052979326,
      "loss": 5.8207,
      "mean_token_accuracy": 0.17655903100967407,
      "num_tokens": 47493855.0,
      "step": 14215
    },
    {
      "entropy": 6.197121047973633,
      "epoch": 2.7741683738171887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042743682540491017,
      "loss": 5.9052,
      "mean_token_accuracy": 0.15821613743901253,
      "num_tokens": 47510175.0,
      "step": 14220
    },
    {
      "entropy": 6.144206428527832,
      "epoch": 2.7751438884011317,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042738508538444383,
      "loss": 5.8605,
      "mean_token_accuracy": 0.16011997684836388,
      "num_tokens": 47528346.0,
      "step": 14225
    },
    {
      "entropy": 6.318954753875732,
      "epoch": 2.7761194029850746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042733333047345005,
      "loss": 5.9538,
      "mean_token_accuracy": 0.15896907299757004,
      "num_tokens": 47545069.0,
      "step": 14230
    },
    {
      "entropy": 6.150177240371704,
      "epoch": 2.7770949175690176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042728156067698607,
      "loss": 5.7888,
      "mean_token_accuracy": 0.1739795386791229,
      "num_tokens": 47562255.0,
      "step": 14235
    },
    {
      "entropy": 6.270524215698242,
      "epoch": 2.7780704321529606,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004272297760001108,
      "loss": 5.949,
      "mean_token_accuracy": 0.15622272938489914,
      "num_tokens": 47577940.0,
      "step": 14240
    },
    {
      "entropy": 6.33545298576355,
      "epoch": 2.779045946736904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271779764478844,
      "loss": 6.0156,
      "mean_token_accuracy": 0.15103597790002823,
      "num_tokens": 47593728.0,
      "step": 14245
    },
    {
      "entropy": 6.225175380706787,
      "epoch": 2.780021461320847,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042712616202536844,
      "loss": 6.0081,
      "mean_token_accuracy": 0.16153389662504197,
      "num_tokens": 47610367.0,
      "step": 14250
    },
    {
      "entropy": 6.3132408618927,
      "epoch": 2.78099697590479,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042707433273762613,
      "loss": 6.0786,
      "mean_token_accuracy": 0.1473083183169365,
      "num_tokens": 47628690.0,
      "step": 14255
    },
    {
      "entropy": 6.328036832809448,
      "epoch": 2.781972490488733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042702248858972206,
      "loss": 6.0032,
      "mean_token_accuracy": 0.15322353765368463,
      "num_tokens": 47643787.0,
      "step": 14260
    },
    {
      "entropy": 6.1490867137908936,
      "epoch": 2.782948005072676,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042697062958672215,
      "loss": 5.8657,
      "mean_token_accuracy": 0.1663847878575325,
      "num_tokens": 47660163.0,
      "step": 14265
    },
    {
      "entropy": 6.306946706771851,
      "epoch": 2.7839235196566188,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004269187557336941,
      "loss": 6.0449,
      "mean_token_accuracy": 0.15044256001710893,
      "num_tokens": 47675959.0,
      "step": 14270
    },
    {
      "entropy": 6.383178758621216,
      "epoch": 2.7848990342405617,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042686686703570655,
      "loss": 6.1132,
      "mean_token_accuracy": 0.14280450567603112,
      "num_tokens": 47693318.0,
      "step": 14275
    },
    {
      "entropy": 6.341831970214844,
      "epoch": 2.7858745488245047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004268149634978301,
      "loss": 6.0536,
      "mean_token_accuracy": 0.14487432539463044,
      "num_tokens": 47709130.0,
      "step": 14280
    },
    {
      "entropy": 6.378940439224243,
      "epoch": 2.786850063408448,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004267630451251364,
      "loss": 6.0717,
      "mean_token_accuracy": 0.14501933753490448,
      "num_tokens": 47725015.0,
      "step": 14285
    },
    {
      "entropy": 6.27600884437561,
      "epoch": 2.787825577992391,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004267111119226989,
      "loss": 6.0291,
      "mean_token_accuracy": 0.1498055562376976,
      "num_tokens": 47741322.0,
      "step": 14290
    },
    {
      "entropy": 6.2666528701782225,
      "epoch": 2.788801092576334,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042665916389559216,
      "loss": 5.9666,
      "mean_token_accuracy": 0.15242172926664352,
      "num_tokens": 47759252.0,
      "step": 14295
    },
    {
      "entropy": 6.300078964233398,
      "epoch": 2.789776607160277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042660720104889256,
      "loss": 5.9968,
      "mean_token_accuracy": 0.15305197536945342,
      "num_tokens": 47775258.0,
      "step": 14300
    },
    {
      "entropy": 6.183231592178345,
      "epoch": 2.79075212174422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042655522338767756,
      "loss": 5.9145,
      "mean_token_accuracy": 0.15822312980890274,
      "num_tokens": 47791072.0,
      "step": 14305
    },
    {
      "entropy": 6.323027801513672,
      "epoch": 2.7917276363281633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004265032309170263,
      "loss": 6.0535,
      "mean_token_accuracy": 0.15556752532720566,
      "num_tokens": 47807946.0,
      "step": 14310
    },
    {
      "entropy": 6.318481492996216,
      "epoch": 2.7927031509121063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004264512236420193,
      "loss": 6.0552,
      "mean_token_accuracy": 0.14500646740198136,
      "num_tokens": 47825295.0,
      "step": 14315
    },
    {
      "entropy": 6.336012363433838,
      "epoch": 2.7936786654960493,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004263992015677384,
      "loss": 6.06,
      "mean_token_accuracy": 0.14825536757707597,
      "num_tokens": 47841302.0,
      "step": 14320
    },
    {
      "entropy": 6.290349292755127,
      "epoch": 2.7946541800799922,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042634716469926723,
      "loss": 5.9526,
      "mean_token_accuracy": 0.15454126745462418,
      "num_tokens": 47856871.0,
      "step": 14325
    },
    {
      "entropy": 6.216133260726929,
      "epoch": 2.795629694663935,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004262951130416905,
      "loss": 5.9689,
      "mean_token_accuracy": 0.15686491876840591,
      "num_tokens": 47872842.0,
      "step": 14330
    },
    {
      "entropy": 6.27184534072876,
      "epoch": 2.796605209247878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004262430466000946,
      "loss": 5.9997,
      "mean_token_accuracy": 0.15690599903464317,
      "num_tokens": 47889093.0,
      "step": 14335
    },
    {
      "entropy": 6.380626249313354,
      "epoch": 2.797580723831821,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042619096537956715,
      "loss": 6.0596,
      "mean_token_accuracy": 0.14685216546058655,
      "num_tokens": 47904953.0,
      "step": 14340
    },
    {
      "entropy": 6.255375099182129,
      "epoch": 2.798556238415764,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004261388693851974,
      "loss": 5.9519,
      "mean_token_accuracy": 0.15236919820308686,
      "num_tokens": 47921591.0,
      "step": 14345
    },
    {
      "entropy": 6.311122941970825,
      "epoch": 2.7995317529997075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004260867586220761,
      "loss": 6.0059,
      "mean_token_accuracy": 0.14817462861537933,
      "num_tokens": 47938463.0,
      "step": 14350
    },
    {
      "entropy": 6.116691446304321,
      "epoch": 2.8005072675836504,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042603463309529515,
      "loss": 5.8089,
      "mean_token_accuracy": 0.17171406745910645,
      "num_tokens": 47956247.0,
      "step": 14355
    },
    {
      "entropy": 6.277556610107422,
      "epoch": 2.8014827821675934,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042598249280994817,
      "loss": 5.9941,
      "mean_token_accuracy": 0.1557093843817711,
      "num_tokens": 47973305.0,
      "step": 14360
    },
    {
      "entropy": 6.291027545928955,
      "epoch": 2.8024582967515363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042593033777113017,
      "loss": 5.9116,
      "mean_token_accuracy": 0.1674983024597168,
      "num_tokens": 47989674.0,
      "step": 14365
    },
    {
      "entropy": 6.233205938339234,
      "epoch": 2.8034338113354793,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004258781679839374,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1556605279445648,
      "num_tokens": 48004814.0,
      "step": 14370
    },
    {
      "entropy": 6.2570469856262205,
      "epoch": 2.8044093259194227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042582598345346787,
      "loss": 5.9758,
      "mean_token_accuracy": 0.152692811191082,
      "num_tokens": 48021174.0,
      "step": 14375
    },
    {
      "entropy": 6.25882740020752,
      "epoch": 2.8053848405033657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004257737841848207,
      "loss": 6.0342,
      "mean_token_accuracy": 0.1499677076935768,
      "num_tokens": 48038179.0,
      "step": 14380
    },
    {
      "entropy": 6.344936943054199,
      "epoch": 2.8063603550873086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042572157018309675,
      "loss": 6.0093,
      "mean_token_accuracy": 0.1538663014769554,
      "num_tokens": 48054933.0,
      "step": 14385
    },
    {
      "entropy": 6.234951114654541,
      "epoch": 2.8073358696712516,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042566934145339816,
      "loss": 5.9297,
      "mean_token_accuracy": 0.16390707790851594,
      "num_tokens": 48071908.0,
      "step": 14390
    },
    {
      "entropy": 6.269456672668457,
      "epoch": 2.8083113842551946,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042561709800082847,
      "loss": 6.0001,
      "mean_token_accuracy": 0.15069614946842194,
      "num_tokens": 48088385.0,
      "step": 14395
    },
    {
      "entropy": 6.194689178466797,
      "epoch": 2.8092868988391375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004255648398304928,
      "loss": 5.8738,
      "mean_token_accuracy": 0.16868603304028512,
      "num_tokens": 48105331.0,
      "step": 14400
    },
    {
      "entropy": 6.061970901489258,
      "epoch": 2.8102624134230805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042551256694749764,
      "loss": 5.7424,
      "mean_token_accuracy": 0.17838589251041412,
      "num_tokens": 48123092.0,
      "step": 14405
    },
    {
      "entropy": 6.236318922042846,
      "epoch": 2.8112379280070234,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004254602793569509,
      "loss": 5.9374,
      "mean_token_accuracy": 0.15803902745246887,
      "num_tokens": 48140096.0,
      "step": 14410
    },
    {
      "entropy": 6.334773588180542,
      "epoch": 2.812213442590967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004254079770639618,
      "loss": 6.0372,
      "mean_token_accuracy": 0.1477818086743355,
      "num_tokens": 48156427.0,
      "step": 14415
    },
    {
      "entropy": 6.336965894699096,
      "epoch": 2.81318895717491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004253556600736413,
      "loss": 6.0589,
      "mean_token_accuracy": 0.14795497357845305,
      "num_tokens": 48172867.0,
      "step": 14420
    },
    {
      "entropy": 6.276050186157226,
      "epoch": 2.8141644717588528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042530332839110156,
      "loss": 5.9872,
      "mean_token_accuracy": 0.15345713645219802,
      "num_tokens": 48189106.0,
      "step": 14425
    },
    {
      "entropy": 6.270062255859375,
      "epoch": 2.8151399863427957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004252509820214563,
      "loss": 5.9745,
      "mean_token_accuracy": 0.14786391854286193,
      "num_tokens": 48205400.0,
      "step": 14430
    },
    {
      "entropy": 6.292786979675293,
      "epoch": 2.816115500926739,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042519862096982054,
      "loss": 6.0121,
      "mean_token_accuracy": 0.15168064087629318,
      "num_tokens": 48221937.0,
      "step": 14435
    },
    {
      "entropy": 6.34468903541565,
      "epoch": 2.817091015510682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004251462452413109,
      "loss": 6.0074,
      "mean_token_accuracy": 0.15040035396814347,
      "num_tokens": 48238060.0,
      "step": 14440
    },
    {
      "entropy": 6.172699022293091,
      "epoch": 2.818066530094625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042509385484104526,
      "loss": 5.8805,
      "mean_token_accuracy": 0.1588948592543602,
      "num_tokens": 48255813.0,
      "step": 14445
    },
    {
      "entropy": 6.193691730499268,
      "epoch": 2.819042044678568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042504144977414303,
      "loss": 5.9165,
      "mean_token_accuracy": 0.15829538106918334,
      "num_tokens": 48272989.0,
      "step": 14450
    },
    {
      "entropy": 6.3097367763519285,
      "epoch": 2.820017559262511,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042498903004572515,
      "loss": 5.991,
      "mean_token_accuracy": 0.1512141928076744,
      "num_tokens": 48289415.0,
      "step": 14455
    },
    {
      "entropy": 6.164159202575684,
      "epoch": 2.820993073846454,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004249365956609139,
      "loss": 5.8992,
      "mean_token_accuracy": 0.16272807568311692,
      "num_tokens": 48306740.0,
      "step": 14460
    },
    {
      "entropy": 6.192884111404419,
      "epoch": 2.821968588430397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004248841466248327,
      "loss": 5.9266,
      "mean_token_accuracy": 0.15368503779172898,
      "num_tokens": 48322934.0,
      "step": 14465
    },
    {
      "entropy": 6.229930353164673,
      "epoch": 2.82294410301434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042483168294260706,
      "loss": 5.9577,
      "mean_token_accuracy": 0.15467606782913207,
      "num_tokens": 48338835.0,
      "step": 14470
    },
    {
      "entropy": 6.1721251010894775,
      "epoch": 2.8239196175982832,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004247792046193633,
      "loss": 5.8522,
      "mean_token_accuracy": 0.1557396799325943,
      "num_tokens": 48355996.0,
      "step": 14475
    },
    {
      "entropy": 6.380327892303467,
      "epoch": 2.824895132182226,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042472671166022955,
      "loss": 6.1162,
      "mean_token_accuracy": 0.146868759393692,
      "num_tokens": 48372775.0,
      "step": 14480
    },
    {
      "entropy": 6.302918434143066,
      "epoch": 2.825870646766169,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004246742040703351,
      "loss": 5.9508,
      "mean_token_accuracy": 0.15138621330261232,
      "num_tokens": 48389018.0,
      "step": 14485
    },
    {
      "entropy": 6.213649940490723,
      "epoch": 2.826846161350112,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004246216818548109,
      "loss": 5.9599,
      "mean_token_accuracy": 0.15802441462874411,
      "num_tokens": 48404298.0,
      "step": 14490
    },
    {
      "entropy": 6.264807033538818,
      "epoch": 2.827821675934055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042456914501878926,
      "loss": 5.9839,
      "mean_token_accuracy": 0.1522671952843666,
      "num_tokens": 48421477.0,
      "step": 14495
    },
    {
      "entropy": 6.291463088989258,
      "epoch": 2.8287971905179985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245165935674038,
      "loss": 5.9505,
      "mean_token_accuracy": 0.15795964524149894,
      "num_tokens": 48439357.0,
      "step": 14500
    },
    {
      "entropy": 6.202819919586181,
      "epoch": 2.8297727051019415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042446402750578974,
      "loss": 5.8879,
      "mean_token_accuracy": 0.16068982928991318,
      "num_tokens": 48454109.0,
      "step": 14505
    },
    {
      "entropy": 6.247951698303223,
      "epoch": 2.8307482196858844,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042441144683908356,
      "loss": 5.964,
      "mean_token_accuracy": 0.14951967298984528,
      "num_tokens": 48470274.0,
      "step": 14510
    },
    {
      "entropy": 6.246515893936158,
      "epoch": 2.8317237342698274,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042435885157242326,
      "loss": 5.9222,
      "mean_token_accuracy": 0.15496788024902344,
      "num_tokens": 48485190.0,
      "step": 14515
    },
    {
      "entropy": 6.284344148635864,
      "epoch": 2.8326992488537703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004243062417109484,
      "loss": 6.0075,
      "mean_token_accuracy": 0.1504894345998764,
      "num_tokens": 48501757.0,
      "step": 14520
    },
    {
      "entropy": 6.268646907806397,
      "epoch": 2.8336747634377133,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042425361725979973,
      "loss": 5.9764,
      "mean_token_accuracy": 0.14980892091989517,
      "num_tokens": 48518539.0,
      "step": 14525
    },
    {
      "entropy": 6.273557138442993,
      "epoch": 2.8346502780216563,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042420097822411945,
      "loss": 5.9699,
      "mean_token_accuracy": 0.14798097535967827,
      "num_tokens": 48535575.0,
      "step": 14530
    },
    {
      "entropy": 6.277518653869629,
      "epoch": 2.835625792605599,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042414832460905145,
      "loss": 5.9556,
      "mean_token_accuracy": 0.15443243086338043,
      "num_tokens": 48552569.0,
      "step": 14535
    },
    {
      "entropy": 6.281630134582519,
      "epoch": 2.8366013071895426,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004240956564197407,
      "loss": 6.0695,
      "mean_token_accuracy": 0.15063308626413346,
      "num_tokens": 48571151.0,
      "step": 14540
    },
    {
      "entropy": 6.221500015258789,
      "epoch": 2.8375768217734856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042404297366133376,
      "loss": 5.8743,
      "mean_token_accuracy": 0.16020936369895936,
      "num_tokens": 48588228.0,
      "step": 14545
    },
    {
      "entropy": 6.19497504234314,
      "epoch": 2.8385523363574285,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004239902763389786,
      "loss": 5.9288,
      "mean_token_accuracy": 0.15267824828624726,
      "num_tokens": 48604936.0,
      "step": 14550
    },
    {
      "entropy": 6.252235746383667,
      "epoch": 2.8395278509413715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042393756445782466,
      "loss": 5.9372,
      "mean_token_accuracy": 0.15545026063919068,
      "num_tokens": 48622064.0,
      "step": 14555
    },
    {
      "entropy": 6.213218355178833,
      "epoch": 2.8405033655253145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042388483802302285,
      "loss": 5.9003,
      "mean_token_accuracy": 0.15352746471762657,
      "num_tokens": 48638161.0,
      "step": 14560
    },
    {
      "entropy": 6.2758220672607425,
      "epoch": 2.841478880109258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042383209703972523,
      "loss": 5.9347,
      "mean_token_accuracy": 0.15938062965869904,
      "num_tokens": 48654521.0,
      "step": 14565
    },
    {
      "entropy": 6.244696426391601,
      "epoch": 2.842454394693201,
      "grad_norm": 1.125,
      "learning_rate": 0.00042377934151308555,
      "loss": 5.9833,
      "mean_token_accuracy": 0.1545104831457138,
      "num_tokens": 48671083.0,
      "step": 14570
    },
    {
      "entropy": 6.399435806274414,
      "epoch": 2.843429909277144,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042372657144825886,
      "loss": 6.0399,
      "mean_token_accuracy": 0.1509805992245674,
      "num_tokens": 48687917.0,
      "step": 14575
    },
    {
      "entropy": 6.250348615646362,
      "epoch": 2.8444054238610867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004236737868504017,
      "loss": 6.0182,
      "mean_token_accuracy": 0.15278672724962233,
      "num_tokens": 48703626.0,
      "step": 14580
    },
    {
      "entropy": 6.299494791030884,
      "epoch": 2.8453809384450297,
      "grad_norm": 1.15625,
      "learning_rate": 0.000423620987724672,
      "loss": 6.0386,
      "mean_token_accuracy": 0.14168829172849656,
      "num_tokens": 48720020.0,
      "step": 14585
    },
    {
      "entropy": 6.3286162376403805,
      "epoch": 2.8463564530289727,
      "grad_norm": 1.0,
      "learning_rate": 0.000423568174076229,
      "loss": 6.0563,
      "mean_token_accuracy": 0.14742302969098092,
      "num_tokens": 48737185.0,
      "step": 14590
    },
    {
      "entropy": 6.207276678085327,
      "epoch": 2.8473319676129156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042351534591023363,
      "loss": 5.8845,
      "mean_token_accuracy": 0.16119541376829147,
      "num_tokens": 48754106.0,
      "step": 14595
    },
    {
      "entropy": 6.216215801239014,
      "epoch": 2.8483074821968586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004234625032318479,
      "loss": 5.9172,
      "mean_token_accuracy": 0.16026968359947205,
      "num_tokens": 48770847.0,
      "step": 14600
    },
    {
      "entropy": 6.168266296386719,
      "epoch": 2.849282996780802,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042340964604623547,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1613479271531105,
      "num_tokens": 48787891.0,
      "step": 14605
    },
    {
      "entropy": 6.2511368751525875,
      "epoch": 2.850258511364745,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004233567743585613,
      "loss": 5.9844,
      "mean_token_accuracy": 0.15206416100263595,
      "num_tokens": 48804418.0,
      "step": 14610
    },
    {
      "entropy": 6.248455476760864,
      "epoch": 2.851234025948688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004233038881739921,
      "loss": 5.9245,
      "mean_token_accuracy": 0.16042139381170273,
      "num_tokens": 48821521.0,
      "step": 14615
    },
    {
      "entropy": 6.30354528427124,
      "epoch": 2.852209540532631,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042325098749769534,
      "loss": 5.9614,
      "mean_token_accuracy": 0.15344172716140747,
      "num_tokens": 48838325.0,
      "step": 14620
    },
    {
      "entropy": 6.254622507095337,
      "epoch": 2.853185055116574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004231980723348406,
      "loss": 5.9361,
      "mean_token_accuracy": 0.15501099675893784,
      "num_tokens": 48853857.0,
      "step": 14625
    },
    {
      "entropy": 6.219304418563842,
      "epoch": 2.8541605697005172,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042314514269059826,
      "loss": 5.9414,
      "mean_token_accuracy": 0.15797580778598785,
      "num_tokens": 48870782.0,
      "step": 14630
    },
    {
      "entropy": 6.22442102432251,
      "epoch": 2.85513608428446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004230921985701407,
      "loss": 5.8915,
      "mean_token_accuracy": 0.16344546675682067,
      "num_tokens": 48888068.0,
      "step": 14635
    },
    {
      "entropy": 6.185567188262939,
      "epoch": 2.856111598868403,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004230392399786413,
      "loss": 5.7743,
      "mean_token_accuracy": 0.16747819185256957,
      "num_tokens": 48904174.0,
      "step": 14640
    },
    {
      "entropy": 6.258308696746826,
      "epoch": 2.857087113452346,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004229862669212749,
      "loss": 5.9628,
      "mean_token_accuracy": 0.15003951638936996,
      "num_tokens": 48921224.0,
      "step": 14645
    },
    {
      "entropy": 6.16907000541687,
      "epoch": 2.858062628036289,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042293327940321797,
      "loss": 5.865,
      "mean_token_accuracy": 0.15985428392887116,
      "num_tokens": 48937626.0,
      "step": 14650
    },
    {
      "entropy": 6.21954607963562,
      "epoch": 2.859038142620232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042288027742964825,
      "loss": 5.8711,
      "mean_token_accuracy": 0.15928345769643784,
      "num_tokens": 48955148.0,
      "step": 14655
    },
    {
      "entropy": 6.28087010383606,
      "epoch": 2.860013657204175,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042282726100574487,
      "loss": 5.98,
      "mean_token_accuracy": 0.15226489901542664,
      "num_tokens": 48972619.0,
      "step": 14660
    },
    {
      "entropy": 6.300771427154541,
      "epoch": 2.860989171788118,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042277423013668835,
      "loss": 5.9785,
      "mean_token_accuracy": 0.1521027371287346,
      "num_tokens": 48989909.0,
      "step": 14665
    },
    {
      "entropy": 6.213870525360107,
      "epoch": 2.8619646863720614,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004227211848276608,
      "loss": 5.9041,
      "mean_token_accuracy": 0.15843137949705124,
      "num_tokens": 49005312.0,
      "step": 14670
    },
    {
      "entropy": 6.258660411834716,
      "epoch": 2.8629402009560043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042266812508384553,
      "loss": 5.9447,
      "mean_token_accuracy": 0.1616652488708496,
      "num_tokens": 49021002.0,
      "step": 14675
    },
    {
      "entropy": 6.198076248168945,
      "epoch": 2.8639157155399473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004226150509104274,
      "loss": 5.972,
      "mean_token_accuracy": 0.1567449986934662,
      "num_tokens": 49037813.0,
      "step": 14680
    },
    {
      "entropy": 6.203907060623169,
      "epoch": 2.8648912301238902,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004225619623125926,
      "loss": 5.8721,
      "mean_token_accuracy": 0.15840365439653398,
      "num_tokens": 49054477.0,
      "step": 14685
    },
    {
      "entropy": 6.34458909034729,
      "epoch": 2.8658667447078336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042250885929552876,
      "loss": 6.0094,
      "mean_token_accuracy": 0.1483165979385376,
      "num_tokens": 49070432.0,
      "step": 14690
    },
    {
      "entropy": 6.2780983448028564,
      "epoch": 2.8668422592917766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004224557418644249,
      "loss": 5.9833,
      "mean_token_accuracy": 0.15418137311935426,
      "num_tokens": 49087984.0,
      "step": 14695
    },
    {
      "entropy": 6.345433759689331,
      "epoch": 2.8678177738757196,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004224026100244715,
      "loss": 6.0363,
      "mean_token_accuracy": 0.149482923746109,
      "num_tokens": 49104899.0,
      "step": 14700
    },
    {
      "entropy": 6.235807466506958,
      "epoch": 2.8687932884596625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042234946378086043,
      "loss": 5.9693,
      "mean_token_accuracy": 0.1526598274707794,
      "num_tokens": 49121306.0,
      "step": 14705
    },
    {
      "entropy": 6.2309064865112305,
      "epoch": 2.8697688030436055,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004222963031387848,
      "loss": 5.9835,
      "mean_token_accuracy": 0.14828818291425705,
      "num_tokens": 49136793.0,
      "step": 14710
    },
    {
      "entropy": 6.298818206787109,
      "epoch": 2.8707443176275484,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222431281034395,
      "loss": 5.9582,
      "mean_token_accuracy": 0.1562913879752159,
      "num_tokens": 49154241.0,
      "step": 14715
    },
    {
      "entropy": 6.221034955978394,
      "epoch": 2.8717198322114914,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042218993868002055,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1545931190252304,
      "num_tokens": 49170978.0,
      "step": 14720
    },
    {
      "entropy": 6.3035660743713375,
      "epoch": 2.8726953467954344,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004221367348737253,
      "loss": 6.0283,
      "mean_token_accuracy": 0.14675103574991227,
      "num_tokens": 49187576.0,
      "step": 14725
    },
    {
      "entropy": 6.30269603729248,
      "epoch": 2.8736708613793778,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042208351668975274,
      "loss": 6.015,
      "mean_token_accuracy": 0.15091644525527953,
      "num_tokens": 49206404.0,
      "step": 14730
    },
    {
      "entropy": 6.292906618118286,
      "epoch": 2.8746463759633207,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004220302841333032,
      "loss": 6.0113,
      "mean_token_accuracy": 0.15467593893408776,
      "num_tokens": 49222365.0,
      "step": 14735
    },
    {
      "entropy": 6.218842649459839,
      "epoch": 2.8756218905472637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004219770372095783,
      "loss": 5.8948,
      "mean_token_accuracy": 0.1629005953669548,
      "num_tokens": 49240471.0,
      "step": 14740
    },
    {
      "entropy": 6.1816877841949465,
      "epoch": 2.8765974051312067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004219237759237811,
      "loss": 5.9505,
      "mean_token_accuracy": 0.1581816293299198,
      "num_tokens": 49255718.0,
      "step": 14745
    },
    {
      "entropy": 6.267602872848511,
      "epoch": 2.8775729197151496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004218705002811162,
      "loss": 5.957,
      "mean_token_accuracy": 0.1581806868314743,
      "num_tokens": 49272746.0,
      "step": 14750
    },
    {
      "entropy": 6.319127702713013,
      "epoch": 2.878548434299093,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004218172102867895,
      "loss": 5.9726,
      "mean_token_accuracy": 0.15299030840396882,
      "num_tokens": 49290478.0,
      "step": 14755
    },
    {
      "entropy": 6.234460306167603,
      "epoch": 2.879523948883036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004217639059460082,
      "loss": 5.8964,
      "mean_token_accuracy": 0.1581451937556267,
      "num_tokens": 49308105.0,
      "step": 14760
    },
    {
      "entropy": 6.227606534957886,
      "epoch": 2.880499463466979,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042171058726398115,
      "loss": 5.888,
      "mean_token_accuracy": 0.156497086584568,
      "num_tokens": 49323519.0,
      "step": 14765
    },
    {
      "entropy": 6.302317523956299,
      "epoch": 2.881474978050922,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004216572542459184,
      "loss": 5.9527,
      "mean_token_accuracy": 0.15427714437246323,
      "num_tokens": 49340162.0,
      "step": 14770
    },
    {
      "entropy": 6.300655651092529,
      "epoch": 2.882450492634865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042160390689703137,
      "loss": 5.9917,
      "mean_token_accuracy": 0.15171339809894563,
      "num_tokens": 49356831.0,
      "step": 14775
    },
    {
      "entropy": 6.263460302352906,
      "epoch": 2.883426007218808,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042155054522253317,
      "loss": 5.9689,
      "mean_token_accuracy": 0.15242429375648497,
      "num_tokens": 49374368.0,
      "step": 14780
    },
    {
      "entropy": 6.248424911499024,
      "epoch": 2.884401521802751,
      "grad_norm": 1.046875,
      "learning_rate": 0.000421497169227638,
      "loss": 5.9068,
      "mean_token_accuracy": 0.15938707441091537,
      "num_tokens": 49390124.0,
      "step": 14785
    },
    {
      "entropy": 6.290268325805664,
      "epoch": 2.8853770363866937,
      "grad_norm": 1.125,
      "learning_rate": 0.00042144377891756155,
      "loss": 5.9938,
      "mean_token_accuracy": 0.14825453311204911,
      "num_tokens": 49406826.0,
      "step": 14790
    },
    {
      "entropy": 6.197624397277832,
      "epoch": 2.886352550970637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042139037429752093,
      "loss": 5.8505,
      "mean_token_accuracy": 0.16434415727853774,
      "num_tokens": 49422539.0,
      "step": 14795
    },
    {
      "entropy": 6.331915473937988,
      "epoch": 2.88732806555458,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004213369553727347,
      "loss": 6.0923,
      "mean_token_accuracy": 0.14388992935419082,
      "num_tokens": 49438958.0,
      "step": 14800
    },
    {
      "entropy": 6.3133988857269285,
      "epoch": 2.888303580138523,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042128352214842275,
      "loss": 6.0038,
      "mean_token_accuracy": 0.15051837563514708,
      "num_tokens": 49454044.0,
      "step": 14805
    },
    {
      "entropy": 6.171556615829468,
      "epoch": 2.889279094722466,
      "grad_norm": 1.125,
      "learning_rate": 0.0004212300746298064,
      "loss": 5.8344,
      "mean_token_accuracy": 0.17098671346902847,
      "num_tokens": 49472198.0,
      "step": 14810
    },
    {
      "entropy": 6.210758686065674,
      "epoch": 2.890254609306409,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042117661282210827,
      "loss": 5.9411,
      "mean_token_accuracy": 0.1572197526693344,
      "num_tokens": 49489198.0,
      "step": 14815
    },
    {
      "entropy": 6.233023405075073,
      "epoch": 2.8912301238903524,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042112313673055263,
      "loss": 5.8841,
      "mean_token_accuracy": 0.16231495141983032,
      "num_tokens": 49504730.0,
      "step": 14820
    },
    {
      "entropy": 6.1392096996307375,
      "epoch": 2.8922056384742953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004210696463603647,
      "loss": 5.8501,
      "mean_token_accuracy": 0.16900878995656968,
      "num_tokens": 49521348.0,
      "step": 14825
    },
    {
      "entropy": 6.216991996765136,
      "epoch": 2.8931811530582383,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042101614171677164,
      "loss": 5.8613,
      "mean_token_accuracy": 0.15350247025489808,
      "num_tokens": 49537618.0,
      "step": 14830
    },
    {
      "entropy": 6.256294250488281,
      "epoch": 2.8941566676421813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004209626228050017,
      "loss": 6.0162,
      "mean_token_accuracy": 0.1548929825425148,
      "num_tokens": 49555898.0,
      "step": 14835
    },
    {
      "entropy": 5.989927816390991,
      "epoch": 2.8951321822261242,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042090908963028433,
      "loss": 5.6483,
      "mean_token_accuracy": 0.18598980605602264,
      "num_tokens": 49573310.0,
      "step": 14840
    },
    {
      "entropy": 6.296886920928955,
      "epoch": 2.896107696810067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042085554219785083,
      "loss": 5.9629,
      "mean_token_accuracy": 0.1600142776966095,
      "num_tokens": 49588931.0,
      "step": 14845
    },
    {
      "entropy": 6.241195201873779,
      "epoch": 2.89708321139401,
      "grad_norm": 1.0,
      "learning_rate": 0.00042080198051293354,
      "loss": 5.9407,
      "mean_token_accuracy": 0.15802027136087418,
      "num_tokens": 49605331.0,
      "step": 14850
    },
    {
      "entropy": 6.268488264083862,
      "epoch": 2.898058725977953,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042074840458076644,
      "loss": 6.0155,
      "mean_token_accuracy": 0.15243614763021468,
      "num_tokens": 49621917.0,
      "step": 14855
    },
    {
      "entropy": 6.307451915740967,
      "epoch": 2.8990342405618965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004206948144065846,
      "loss": 5.9169,
      "mean_token_accuracy": 0.1579698219895363,
      "num_tokens": 49638809.0,
      "step": 14860
    },
    {
      "entropy": 6.262693452835083,
      "epoch": 2.9000097551458395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004206412099956247,
      "loss": 5.998,
      "mean_token_accuracy": 0.15403331816196442,
      "num_tokens": 49654363.0,
      "step": 14865
    },
    {
      "entropy": 6.2110978126525875,
      "epoch": 2.9009852697297824,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042058759135312504,
      "loss": 5.8875,
      "mean_token_accuracy": 0.16389259845018386,
      "num_tokens": 49671894.0,
      "step": 14870
    },
    {
      "entropy": 6.273136234283447,
      "epoch": 2.9019607843137254,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004205339584843247,
      "loss": 6.0027,
      "mean_token_accuracy": 0.1566162109375,
      "num_tokens": 49689690.0,
      "step": 14875
    },
    {
      "entropy": 6.280284690856933,
      "epoch": 2.9029362988976684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004204803113944646,
      "loss": 5.9724,
      "mean_token_accuracy": 0.15606181249022483,
      "num_tokens": 49706308.0,
      "step": 14880
    },
    {
      "entropy": 6.193595838546753,
      "epoch": 2.9039118134816118,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004204266500887871,
      "loss": 5.8105,
      "mean_token_accuracy": 0.1668577015399933,
      "num_tokens": 49723531.0,
      "step": 14885
    },
    {
      "entropy": 6.018686437606812,
      "epoch": 2.9048873280655547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042037297457253554,
      "loss": 5.7512,
      "mean_token_accuracy": 0.17463786751031876,
      "num_tokens": 49740644.0,
      "step": 14890
    },
    {
      "entropy": 6.287517929077149,
      "epoch": 2.9058628426494977,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004203192848509551,
      "loss": 6.0578,
      "mean_token_accuracy": 0.14434486776590347,
      "num_tokens": 49758789.0,
      "step": 14895
    },
    {
      "entropy": 6.314220714569092,
      "epoch": 2.9068383572334406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042026558092929205,
      "loss": 5.9946,
      "mean_token_accuracy": 0.15495295226573944,
      "num_tokens": 49775516.0,
      "step": 14900
    },
    {
      "entropy": 6.242131567001342,
      "epoch": 2.9078138718173836,
      "grad_norm": 1.0,
      "learning_rate": 0.00042021186281279417,
      "loss": 5.9468,
      "mean_token_accuracy": 0.15357396453619004,
      "num_tokens": 49790834.0,
      "step": 14905
    },
    {
      "entropy": 6.283976697921753,
      "epoch": 2.9087893864013266,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004201581305067106,
      "loss": 5.946,
      "mean_token_accuracy": 0.1496359683573246,
      "num_tokens": 49808939.0,
      "step": 14910
    },
    {
      "entropy": 6.333368110656738,
      "epoch": 2.9097649009852695,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004201043840162919,
      "loss": 6.0139,
      "mean_token_accuracy": 0.15070301443338394,
      "num_tokens": 49825512.0,
      "step": 14915
    },
    {
      "entropy": 6.229922771453857,
      "epoch": 2.910740415569213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004200506233467899,
      "loss": 5.8775,
      "mean_token_accuracy": 0.16091395318508148,
      "num_tokens": 49841239.0,
      "step": 14920
    },
    {
      "entropy": 6.2825250148773195,
      "epoch": 2.911715930153156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000419996848503458,
      "loss": 5.9679,
      "mean_token_accuracy": 0.1535348504781723,
      "num_tokens": 49858352.0,
      "step": 14925
    },
    {
      "entropy": 6.302318000793457,
      "epoch": 2.912691444737099,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00041994305949155083,
      "loss": 6.0022,
      "mean_token_accuracy": 0.15294022634625434,
      "num_tokens": 49874862.0,
      "step": 14930
    },
    {
      "entropy": 6.231199169158936,
      "epoch": 2.913666959321042,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004198892563163244,
      "loss": 5.92,
      "mean_token_accuracy": 0.15854197591543198,
      "num_tokens": 49891575.0,
      "step": 14935
    },
    {
      "entropy": 6.246474933624268,
      "epoch": 2.9146424739049848,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041983543898303636,
      "loss": 5.9384,
      "mean_token_accuracy": 0.15324636697769164,
      "num_tokens": 49907607.0,
      "step": 14940
    },
    {
      "entropy": 6.32549147605896,
      "epoch": 2.915617988488928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004197816074969453,
      "loss": 6.0745,
      "mean_token_accuracy": 0.1457786738872528,
      "num_tokens": 49924525.0,
      "step": 14945
    },
    {
      "entropy": 6.235596084594727,
      "epoch": 2.916593503072871,
      "grad_norm": 1.0,
      "learning_rate": 0.00041972776186331164,
      "loss": 6.0108,
      "mean_token_accuracy": 0.15012683421373368,
      "num_tokens": 49942456.0,
      "step": 14950
    },
    {
      "entropy": 6.313733100891113,
      "epoch": 2.917569017656814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041967390208739693,
      "loss": 5.9767,
      "mean_token_accuracy": 0.14829600527882575,
      "num_tokens": 49961481.0,
      "step": 14955
    },
    {
      "entropy": 6.331619310379028,
      "epoch": 2.918544532240757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041962002817446406,
      "loss": 6.0054,
      "mean_token_accuracy": 0.1522788554430008,
      "num_tokens": 49978397.0,
      "step": 14960
    },
    {
      "entropy": 6.173291158676148,
      "epoch": 2.9195200468247,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041956614012977753,
      "loss": 5.8962,
      "mean_token_accuracy": 0.1643626019358635,
      "num_tokens": 49995334.0,
      "step": 14965
    },
    {
      "entropy": 6.2346038818359375,
      "epoch": 2.920495561408643,
      "grad_norm": 1.0,
      "learning_rate": 0.000419512237958603,
      "loss": 5.965,
      "mean_token_accuracy": 0.15910590291023255,
      "num_tokens": 50012342.0,
      "step": 14970
    },
    {
      "entropy": 6.285765504837036,
      "epoch": 2.921471075992586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041945832166620766,
      "loss": 6.0187,
      "mean_token_accuracy": 0.14651061594486237,
      "num_tokens": 50028679.0,
      "step": 14975
    },
    {
      "entropy": 6.264839839935303,
      "epoch": 2.922446590576529,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041940439125785994,
      "loss": 5.9356,
      "mean_token_accuracy": 0.14989166036248208,
      "num_tokens": 50044280.0,
      "step": 14980
    },
    {
      "entropy": 6.341350650787353,
      "epoch": 2.9234221051604723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004193504467388298,
      "loss": 5.9767,
      "mean_token_accuracy": 0.15000933408737183,
      "num_tokens": 50061069.0,
      "step": 14985
    },
    {
      "entropy": 6.218218469619751,
      "epoch": 2.9243976197444153,
      "grad_norm": 1.0,
      "learning_rate": 0.00041929648811438847,
      "loss": 5.9182,
      "mean_token_accuracy": 0.15436751395463943,
      "num_tokens": 50077552.0,
      "step": 14990
    },
    {
      "entropy": 6.204817819595337,
      "epoch": 2.925373134328358,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004192425153898087,
      "loss": 5.9318,
      "mean_token_accuracy": 0.15890475809574128,
      "num_tokens": 50095296.0,
      "step": 14995
    },
    {
      "entropy": 6.310628461837768,
      "epoch": 2.926348648912301,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041918852857036426,
      "loss": 5.9424,
      "mean_token_accuracy": 0.15493861883878707,
      "num_tokens": 50112472.0,
      "step": 15000
    },
    {
      "epoch": 2.926348648912301,
      "eval_entropy": 6.118985965984545,
      "eval_loss": 6.054816722869873,
      "eval_mean_token_accuracy": 0.15721502609899582,
      "eval_num_tokens": 50112472.0,
      "eval_runtime": 15.1147,
      "eval_samples_per_second": 1096.744,
      "eval_steps_per_second": 137.151,
      "step": 15000
    },
    {
      "entropy": 6.2919269561767575,
      "epoch": 2.927324163496244,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004191345276613308,
      "loss": 6.0238,
      "mean_token_accuracy": 0.14851917922496796,
      "num_tokens": 50130152.0,
      "step": 15005
    },
    {
      "entropy": 6.316591739654541,
      "epoch": 2.9282996780801875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041908051266798495,
      "loss": 6.033,
      "mean_token_accuracy": 0.1479869544506073,
      "num_tokens": 50146572.0,
      "step": 15010
    },
    {
      "entropy": 6.212274980545044,
      "epoch": 2.9292751926641305,
      "grad_norm": 1.015625,
      "learning_rate": 0.000419026483595605,
      "loss": 5.902,
      "mean_token_accuracy": 0.16266471296548843,
      "num_tokens": 50163501.0,
      "step": 15015
    },
    {
      "entropy": 6.230529642105102,
      "epoch": 2.9302507072480735,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004189724404494703,
      "loss": 5.9679,
      "mean_token_accuracy": 0.15274003148078918,
      "num_tokens": 50181417.0,
      "step": 15020
    },
    {
      "entropy": 6.372236013412476,
      "epoch": 2.9312262218320164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041891838323486187,
      "loss": 6.0975,
      "mean_token_accuracy": 0.1441318780183792,
      "num_tokens": 50198071.0,
      "step": 15025
    },
    {
      "entropy": 6.364988803863525,
      "epoch": 2.9322017364159594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000418864311957062,
      "loss": 6.0122,
      "mean_token_accuracy": 0.15416181087493896,
      "num_tokens": 50213577.0,
      "step": 15030
    },
    {
      "entropy": 6.184256839752197,
      "epoch": 2.9331772509999023,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041881022662135426,
      "loss": 5.9729,
      "mean_token_accuracy": 0.15659223943948747,
      "num_tokens": 50231023.0,
      "step": 15035
    },
    {
      "entropy": 6.183809375762939,
      "epoch": 2.9341527655838453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041875612723302376,
      "loss": 5.7937,
      "mean_token_accuracy": 0.16394634395837784,
      "num_tokens": 50247435.0,
      "step": 15040
    },
    {
      "entropy": 6.325147294998169,
      "epoch": 2.9351282801677883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041870201379735683,
      "loss": 5.9942,
      "mean_token_accuracy": 0.15148528590798377,
      "num_tokens": 50264937.0,
      "step": 15045
    },
    {
      "entropy": 6.220269680023193,
      "epoch": 2.9361037947517317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004186478863196413,
      "loss": 5.8939,
      "mean_token_accuracy": 0.1584813877940178,
      "num_tokens": 50280653.0,
      "step": 15050
    },
    {
      "entropy": 6.343685293197632,
      "epoch": 2.9370793093356746,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041859374480516625,
      "loss": 6.0749,
      "mean_token_accuracy": 0.14702347964048385,
      "num_tokens": 50297327.0,
      "step": 15055
    },
    {
      "entropy": 6.23912615776062,
      "epoch": 2.9380548239196176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004185395892592223,
      "loss": 5.9441,
      "mean_token_accuracy": 0.1529322013258934,
      "num_tokens": 50314457.0,
      "step": 15060
    },
    {
      "entropy": 6.205979919433593,
      "epoch": 2.9390303385035605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004184854196871012,
      "loss": 5.829,
      "mean_token_accuracy": 0.16631554067134857,
      "num_tokens": 50330249.0,
      "step": 15065
    },
    {
      "entropy": 6.215948295593262,
      "epoch": 2.9400058530875035,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041843123609409635,
      "loss": 5.9104,
      "mean_token_accuracy": 0.16530630439519883,
      "num_tokens": 50347494.0,
      "step": 15070
    },
    {
      "entropy": 6.306450223922729,
      "epoch": 2.940981367671447,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004183770384855022,
      "loss": 5.9561,
      "mean_token_accuracy": 0.15236890986561774,
      "num_tokens": 50364530.0,
      "step": 15075
    },
    {
      "entropy": 6.30326566696167,
      "epoch": 2.94195688225539,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000418322826866615,
      "loss": 5.9412,
      "mean_token_accuracy": 0.15731252431869508,
      "num_tokens": 50382689.0,
      "step": 15080
    },
    {
      "entropy": 6.2135217666625975,
      "epoch": 2.942932396839333,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004182686012427319,
      "loss": 5.9823,
      "mean_token_accuracy": 0.15385887175798416,
      "num_tokens": 50400091.0,
      "step": 15085
    },
    {
      "entropy": 6.239255046844482,
      "epoch": 2.943907911423276,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041821436161915165,
      "loss": 5.8904,
      "mean_token_accuracy": 0.15836466550827027,
      "num_tokens": 50417441.0,
      "step": 15090
    },
    {
      "entropy": 6.217005634307862,
      "epoch": 2.9448834260072188,
      "grad_norm": 1.0,
      "learning_rate": 0.0004181601080011745,
      "loss": 5.8987,
      "mean_token_accuracy": 0.16772947162389756,
      "num_tokens": 50434539.0,
      "step": 15095
    },
    {
      "entropy": 6.160098266601563,
      "epoch": 2.9458589405911617,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041810584039410174,
      "loss": 5.8375,
      "mean_token_accuracy": 0.16344287395477294,
      "num_tokens": 50450074.0,
      "step": 15100
    },
    {
      "entropy": 6.205565547943115,
      "epoch": 2.9468344551751047,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041805155880323644,
      "loss": 5.9225,
      "mean_token_accuracy": 0.1616169720888138,
      "num_tokens": 50467987.0,
      "step": 15105
    },
    {
      "entropy": 6.159100008010864,
      "epoch": 2.9478099697590476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004179972632338825,
      "loss": 5.8168,
      "mean_token_accuracy": 0.16329086273908616,
      "num_tokens": 50486168.0,
      "step": 15110
    },
    {
      "entropy": 6.2497090816497805,
      "epoch": 2.948785484342991,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004179429536913457,
      "loss": 5.9331,
      "mean_token_accuracy": 0.15443587005138398,
      "num_tokens": 50502556.0,
      "step": 15115
    },
    {
      "entropy": 6.216889238357544,
      "epoch": 2.949760998926934,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041788863018093293,
      "loss": 5.892,
      "mean_token_accuracy": 0.1658958375453949,
      "num_tokens": 50518347.0,
      "step": 15120
    },
    {
      "entropy": 6.171720886230469,
      "epoch": 2.950736513510877,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004178342927079525,
      "loss": 5.853,
      "mean_token_accuracy": 0.169654880464077,
      "num_tokens": 50533847.0,
      "step": 15125
    },
    {
      "entropy": 6.320952892303467,
      "epoch": 2.95171202809482,
      "grad_norm": 0.953125,
      "learning_rate": 0.000417779941277714,
      "loss": 5.9609,
      "mean_token_accuracy": 0.15152697414159774,
      "num_tokens": 50549410.0,
      "step": 15130
    },
    {
      "entropy": 6.239281558990479,
      "epoch": 2.9526875426787633,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041772557589552856,
      "loss": 5.8939,
      "mean_token_accuracy": 0.15562147200107573,
      "num_tokens": 50566554.0,
      "step": 15135
    },
    {
      "entropy": 6.167007350921631,
      "epoch": 2.9536630572627063,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041767119656670854,
      "loss": 5.9247,
      "mean_token_accuracy": 0.16479246467351913,
      "num_tokens": 50584688.0,
      "step": 15140
    },
    {
      "entropy": 6.281825256347656,
      "epoch": 2.9546385718466492,
      "grad_norm": 1.125,
      "learning_rate": 0.00041761680329656766,
      "loss": 5.9266,
      "mean_token_accuracy": 0.15441712588071824,
      "num_tokens": 50601085.0,
      "step": 15145
    },
    {
      "entropy": 6.188651275634766,
      "epoch": 2.955614086430592,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041756239609042105,
      "loss": 5.995,
      "mean_token_accuracy": 0.15077922940254213,
      "num_tokens": 50619030.0,
      "step": 15150
    },
    {
      "entropy": 6.331677055358886,
      "epoch": 2.956589601014535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004175079749535852,
      "loss": 5.9718,
      "mean_token_accuracy": 0.1541070058941841,
      "num_tokens": 50634511.0,
      "step": 15155
    },
    {
      "entropy": 6.285830497741699,
      "epoch": 2.957565115598478,
      "grad_norm": 1.0,
      "learning_rate": 0.000417453539891378,
      "loss": 5.9578,
      "mean_token_accuracy": 0.15197984129190445,
      "num_tokens": 50651824.0,
      "step": 15160
    },
    {
      "entropy": 6.170511245727539,
      "epoch": 2.958540630182421,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004173990909091186,
      "loss": 5.9078,
      "mean_token_accuracy": 0.17480936497449875,
      "num_tokens": 50668951.0,
      "step": 15165
    },
    {
      "entropy": 6.247426891326905,
      "epoch": 2.959516144766364,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041734462801212743,
      "loss": 5.8842,
      "mean_token_accuracy": 0.1561504602432251,
      "num_tokens": 50684864.0,
      "step": 15170
    },
    {
      "entropy": 6.2763991355896,
      "epoch": 2.9604916593503074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041729015120572655,
      "loss": 5.9854,
      "mean_token_accuracy": 0.15283723175525665,
      "num_tokens": 50701112.0,
      "step": 15175
    },
    {
      "entropy": 6.3197550773620605,
      "epoch": 2.9614671739342504,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004172356604952393,
      "loss": 6.005,
      "mean_token_accuracy": 0.15641386508941652,
      "num_tokens": 50718752.0,
      "step": 15180
    },
    {
      "entropy": 6.293487119674682,
      "epoch": 2.9624426885181934,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004171811558859902,
      "loss": 5.9873,
      "mean_token_accuracy": 0.14961712211370468,
      "num_tokens": 50736808.0,
      "step": 15185
    },
    {
      "entropy": 6.228486156463623,
      "epoch": 2.9634182031021363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041712663738330535,
      "loss": 5.9687,
      "mean_token_accuracy": 0.15422326177358628,
      "num_tokens": 50752389.0,
      "step": 15190
    },
    {
      "entropy": 6.244111680984497,
      "epoch": 2.9643937176860793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417072104992512,
      "loss": 5.9338,
      "mean_token_accuracy": 0.15867646932601928,
      "num_tokens": 50768586.0,
      "step": 15195
    },
    {
      "entropy": 6.277461671829224,
      "epoch": 2.9653692322700227,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041701755871893895,
      "loss": 5.9466,
      "mean_token_accuracy": 0.16013192981481553,
      "num_tokens": 50784716.0,
      "step": 15200
    },
    {
      "entropy": 6.287484502792358,
      "epoch": 2.9663447468539657,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004169629985679162,
      "loss": 6.0098,
      "mean_token_accuracy": 0.14913905113935472,
      "num_tokens": 50801684.0,
      "step": 15205
    },
    {
      "entropy": 6.156454849243164,
      "epoch": 2.9673202614379086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041690842454477516,
      "loss": 5.8795,
      "mean_token_accuracy": 0.1653762549161911,
      "num_tokens": 50817823.0,
      "step": 15210
    },
    {
      "entropy": 6.392890071868896,
      "epoch": 2.9682957760218516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041685383665484866,
      "loss": 6.1002,
      "mean_token_accuracy": 0.14346864074468613,
      "num_tokens": 50834578.0,
      "step": 15215
    },
    {
      "entropy": 6.216313600540161,
      "epoch": 2.9692712906057945,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041679923490347084,
      "loss": 5.8295,
      "mean_token_accuracy": 0.16747393161058427,
      "num_tokens": 50851596.0,
      "step": 15220
    },
    {
      "entropy": 6.080870485305786,
      "epoch": 2.9702468051897375,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004167446192959771,
      "loss": 5.8216,
      "mean_token_accuracy": 0.1722580686211586,
      "num_tokens": 50868827.0,
      "step": 15225
    },
    {
      "entropy": 6.293052434921265,
      "epoch": 2.9712223197736805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041668998983770446,
      "loss": 6.0233,
      "mean_token_accuracy": 0.15385959297418594,
      "num_tokens": 50884979.0,
      "step": 15230
    },
    {
      "entropy": 6.255758810043335,
      "epoch": 2.9721978343576234,
      "grad_norm": 1.125,
      "learning_rate": 0.00041663534653399095,
      "loss": 5.9644,
      "mean_token_accuracy": 0.1526950106024742,
      "num_tokens": 50900700.0,
      "step": 15235
    },
    {
      "entropy": 6.216403532028198,
      "epoch": 2.973173348941567,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041658068939017613,
      "loss": 5.9509,
      "mean_token_accuracy": 0.1522977113723755,
      "num_tokens": 50917508.0,
      "step": 15240
    },
    {
      "entropy": 6.274057865142822,
      "epoch": 2.97414886352551,
      "grad_norm": 1.09375,
      "learning_rate": 0.000416526018411601,
      "loss": 5.9649,
      "mean_token_accuracy": 0.1584109976887703,
      "num_tokens": 50934127.0,
      "step": 15245
    },
    {
      "entropy": 6.350365734100341,
      "epoch": 2.9751243781094527,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041647133360360765,
      "loss": 6.0457,
      "mean_token_accuracy": 0.15356752276420593,
      "num_tokens": 50951144.0,
      "step": 15250
    },
    {
      "entropy": 6.249305152893067,
      "epoch": 2.9760998926933957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041641663497153994,
      "loss": 5.8766,
      "mean_token_accuracy": 0.15714504420757294,
      "num_tokens": 50967506.0,
      "step": 15255
    },
    {
      "entropy": 6.268884611129761,
      "epoch": 2.9770754072773387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004163619225207426,
      "loss": 5.9905,
      "mean_token_accuracy": 0.15497533828020096,
      "num_tokens": 50983720.0,
      "step": 15260
    },
    {
      "entropy": 6.339649200439453,
      "epoch": 2.978050921861282,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004163071962565621,
      "loss": 6.0353,
      "mean_token_accuracy": 0.15166645050048827,
      "num_tokens": 51000613.0,
      "step": 15265
    },
    {
      "entropy": 6.117683553695679,
      "epoch": 2.979026436445225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004162524561843459,
      "loss": 5.7784,
      "mean_token_accuracy": 0.17290685772895814,
      "num_tokens": 51017345.0,
      "step": 15270
    },
    {
      "entropy": 6.21597957611084,
      "epoch": 2.980001951029168,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041619770230944317,
      "loss": 5.9026,
      "mean_token_accuracy": 0.15994881093502045,
      "num_tokens": 51034117.0,
      "step": 15275
    },
    {
      "entropy": 6.323807334899902,
      "epoch": 2.980977465613111,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041614293463720415,
      "loss": 5.9333,
      "mean_token_accuracy": 0.15898619294166566,
      "num_tokens": 51049958.0,
      "step": 15280
    },
    {
      "entropy": 6.198253440856933,
      "epoch": 2.981952980197054,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00041608815317298065,
      "loss": 5.901,
      "mean_token_accuracy": 0.15631738305091858,
      "num_tokens": 51067721.0,
      "step": 15285
    },
    {
      "entropy": 6.181815528869629,
      "epoch": 2.982928494780997,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004160333579221257,
      "loss": 5.892,
      "mean_token_accuracy": 0.16120053827762604,
      "num_tokens": 51083078.0,
      "step": 15290
    },
    {
      "entropy": 6.2078564167022705,
      "epoch": 2.98390400936494,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041597854888999363,
      "loss": 5.9003,
      "mean_token_accuracy": 0.1592087835073471,
      "num_tokens": 51099510.0,
      "step": 15295
    },
    {
      "entropy": 6.173479795455933,
      "epoch": 2.984879523948883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041592372608194024,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1629016727209091,
      "num_tokens": 51115855.0,
      "step": 15300
    },
    {
      "entropy": 6.127971220016479,
      "epoch": 2.985855038532826,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041586888950332257,
      "loss": 5.712,
      "mean_token_accuracy": 0.17545682936906815,
      "num_tokens": 51131912.0,
      "step": 15305
    },
    {
      "entropy": 6.313734102249145,
      "epoch": 2.986830553116769,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041581403915949913,
      "loss": 6.0605,
      "mean_token_accuracy": 0.14528320133686065,
      "num_tokens": 51149268.0,
      "step": 15310
    },
    {
      "entropy": 6.266527509689331,
      "epoch": 2.987806067700712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004157591750558296,
      "loss": 5.9134,
      "mean_token_accuracy": 0.15360779464244842,
      "num_tokens": 51166809.0,
      "step": 15315
    },
    {
      "entropy": 6.221965885162353,
      "epoch": 2.988781582284655,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041570429719767514,
      "loss": 5.9481,
      "mean_token_accuracy": 0.15433667302131654,
      "num_tokens": 51184155.0,
      "step": 15320
    },
    {
      "entropy": 6.178299331665039,
      "epoch": 2.989757096868598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041564940559039827,
      "loss": 5.8414,
      "mean_token_accuracy": 0.16523100435733795,
      "num_tokens": 51200158.0,
      "step": 15325
    },
    {
      "entropy": 6.234653282165527,
      "epoch": 2.9907326114525414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041559450023936277,
      "loss": 5.9326,
      "mean_token_accuracy": 0.1585438668727875,
      "num_tokens": 51216921.0,
      "step": 15330
    },
    {
      "entropy": 6.062778425216675,
      "epoch": 2.9917081260364844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004155395811499338,
      "loss": 5.7734,
      "mean_token_accuracy": 0.17575718387961387,
      "num_tokens": 51234587.0,
      "step": 15335
    },
    {
      "entropy": 6.2760536670684814,
      "epoch": 2.9926836406204274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004154846483274779,
      "loss": 5.8788,
      "mean_token_accuracy": 0.16005603075027466,
      "num_tokens": 51249379.0,
      "step": 15340
    },
    {
      "entropy": 6.088890266418457,
      "epoch": 2.9936591552043703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004154297017773627,
      "loss": 5.7417,
      "mean_token_accuracy": 0.17294821590185167,
      "num_tokens": 51267449.0,
      "step": 15345
    },
    {
      "entropy": 6.275409173965454,
      "epoch": 2.9946346697883133,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004153747415049577,
      "loss": 5.9607,
      "mean_token_accuracy": 0.15361598432064055,
      "num_tokens": 51284976.0,
      "step": 15350
    },
    {
      "entropy": 6.291387891769409,
      "epoch": 2.9956101843722562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004153197675156333,
      "loss": 6.0019,
      "mean_token_accuracy": 0.1552613765001297,
      "num_tokens": 51301534.0,
      "step": 15355
    },
    {
      "entropy": 6.154560804367065,
      "epoch": 2.996585698956199,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004152647798147612,
      "loss": 5.915,
      "mean_token_accuracy": 0.1610094428062439,
      "num_tokens": 51318932.0,
      "step": 15360
    },
    {
      "entropy": 6.188530826568604,
      "epoch": 2.997561213540142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004152097784077148,
      "loss": 5.8615,
      "mean_token_accuracy": 0.15992987155914307,
      "num_tokens": 51336724.0,
      "step": 15365
    },
    {
      "entropy": 6.275351715087891,
      "epoch": 2.9985367281240856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004151547632998685,
      "loss": 5.9792,
      "mean_token_accuracy": 0.1621492475271225,
      "num_tokens": 51353443.0,
      "step": 15370
    },
    {
      "entropy": 6.312172985076904,
      "epoch": 2.9995122427080285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004150997344965983,
      "loss": 5.9632,
      "mean_token_accuracy": 0.15882902294397355,
      "num_tokens": 51371068.0,
      "step": 15375
    },
    {
      "entropy": 6.21614212459988,
      "epoch": 3.000390205833577,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004150446920032814,
      "loss": 5.9531,
      "mean_token_accuracy": 0.14751821425226,
      "num_tokens": 51385621.0,
      "step": 15380
    },
    {
      "entropy": 6.28590955734253,
      "epoch": 3.00136572041752,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004149896358252962,
      "loss": 5.9204,
      "mean_token_accuracy": 0.15500352382659913,
      "num_tokens": 51401776.0,
      "step": 15385
    },
    {
      "entropy": 6.180731105804443,
      "epoch": 3.002341235001463,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004149345659680228,
      "loss": 5.7768,
      "mean_token_accuracy": 0.1716190904378891,
      "num_tokens": 51418966.0,
      "step": 15390
    },
    {
      "entropy": 6.286042022705078,
      "epoch": 3.0033167495854065,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041487948243684237,
      "loss": 5.8725,
      "mean_token_accuracy": 0.1575612396001816,
      "num_tokens": 51436227.0,
      "step": 15395
    },
    {
      "entropy": 6.189260816574096,
      "epoch": 3.0042922641693495,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041482438523713744,
      "loss": 5.8529,
      "mean_token_accuracy": 0.16374245285987854,
      "num_tokens": 51452772.0,
      "step": 15400
    },
    {
      "entropy": 6.3116906642913815,
      "epoch": 3.0052677787532924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004147692743742919,
      "loss": 5.8917,
      "mean_token_accuracy": 0.1548242136836052,
      "num_tokens": 51470050.0,
      "step": 15405
    },
    {
      "entropy": 6.272612190246582,
      "epoch": 3.0062432933372354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004147141498536911,
      "loss": 5.8529,
      "mean_token_accuracy": 0.15607229769229888,
      "num_tokens": 51487037.0,
      "step": 15410
    },
    {
      "entropy": 6.059240245819092,
      "epoch": 3.0072188079211783,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041465901168072144,
      "loss": 5.6611,
      "mean_token_accuracy": 0.17849406898021697,
      "num_tokens": 51505385.0,
      "step": 15415
    },
    {
      "entropy": 6.1249524593353275,
      "epoch": 3.0081943225051213,
      "grad_norm": 1.03125,
      "learning_rate": 0.000414603859860771,
      "loss": 5.6546,
      "mean_token_accuracy": 0.17987084835767747,
      "num_tokens": 51521664.0,
      "step": 15420
    },
    {
      "entropy": 6.247483968734741,
      "epoch": 3.0091698370890643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004145486943992288,
      "loss": 5.9275,
      "mean_token_accuracy": 0.1555858761072159,
      "num_tokens": 51537765.0,
      "step": 15425
    },
    {
      "entropy": 6.251210308074951,
      "epoch": 3.0101453516730077,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041449351530148566,
      "loss": 5.8686,
      "mean_token_accuracy": 0.16099113821983338,
      "num_tokens": 51553698.0,
      "step": 15430
    },
    {
      "entropy": 6.219232368469238,
      "epoch": 3.0111208662569506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004144383225729333,
      "loss": 5.8006,
      "mean_token_accuracy": 0.17068903595209123,
      "num_tokens": 51570234.0,
      "step": 15435
    },
    {
      "entropy": 6.158806896209716,
      "epoch": 3.0120963808408936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041438311621896505,
      "loss": 5.7794,
      "mean_token_accuracy": 0.17299237102270126,
      "num_tokens": 51586692.0,
      "step": 15440
    },
    {
      "entropy": 6.276414012908935,
      "epoch": 3.0130718954248366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041432789624497546,
      "loss": 5.8775,
      "mean_token_accuracy": 0.15775725841522217,
      "num_tokens": 51603521.0,
      "step": 15445
    },
    {
      "entropy": 6.202864027023315,
      "epoch": 3.0140474100087795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041427266265636043,
      "loss": 5.7618,
      "mean_token_accuracy": 0.17343517541885375,
      "num_tokens": 51618055.0,
      "step": 15450
    },
    {
      "entropy": 6.236621952056884,
      "epoch": 3.0150229245927225,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004142174154585172,
      "loss": 5.8364,
      "mean_token_accuracy": 0.16151363253593445,
      "num_tokens": 51633309.0,
      "step": 15455
    },
    {
      "entropy": 6.14202184677124,
      "epoch": 3.015998439176666,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004141621546568443,
      "loss": 5.7669,
      "mean_token_accuracy": 0.1704660937190056,
      "num_tokens": 51650327.0,
      "step": 15460
    },
    {
      "entropy": 6.204433345794678,
      "epoch": 3.016973953760609,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041410688025674154,
      "loss": 5.8394,
      "mean_token_accuracy": 0.1602709025144577,
      "num_tokens": 51666216.0,
      "step": 15465
    },
    {
      "entropy": 6.278918170928955,
      "epoch": 3.017949468344552,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004140515922636103,
      "loss": 5.8921,
      "mean_token_accuracy": 0.16242118626832963,
      "num_tokens": 51682361.0,
      "step": 15470
    },
    {
      "entropy": 6.30773811340332,
      "epoch": 3.0189249829284948,
      "grad_norm": 1.0,
      "learning_rate": 0.00041399629068285303,
      "loss": 5.8746,
      "mean_token_accuracy": 0.1537291243672371,
      "num_tokens": 51698826.0,
      "step": 15475
    },
    {
      "entropy": 6.23756628036499,
      "epoch": 3.0199004975124377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041394097551987366,
      "loss": 5.839,
      "mean_token_accuracy": 0.15667078047990798,
      "num_tokens": 51714838.0,
      "step": 15480
    },
    {
      "entropy": 6.228147888183594,
      "epoch": 3.0208760120963807,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041388564678007726,
      "loss": 5.8293,
      "mean_token_accuracy": 0.1597999721765518,
      "num_tokens": 51731185.0,
      "step": 15485
    },
    {
      "entropy": 6.330277729034424,
      "epoch": 3.021851526680324,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041383030446887054,
      "loss": 6.0105,
      "mean_token_accuracy": 0.15695403516292572,
      "num_tokens": 51749007.0,
      "step": 15490
    },
    {
      "entropy": 6.193541765213013,
      "epoch": 3.022827041264267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041377494859166133,
      "loss": 5.7702,
      "mean_token_accuracy": 0.1713410884141922,
      "num_tokens": 51767088.0,
      "step": 15495
    },
    {
      "entropy": 6.19540638923645,
      "epoch": 3.02380255584821,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041371957915385866,
      "loss": 5.7636,
      "mean_token_accuracy": 0.1675364300608635,
      "num_tokens": 51783361.0,
      "step": 15500
    },
    {
      "entropy": 6.2012840747833256,
      "epoch": 3.024778070432153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041366419616087314,
      "loss": 5.7827,
      "mean_token_accuracy": 0.16362987458705902,
      "num_tokens": 51799555.0,
      "step": 15505
    },
    {
      "entropy": 6.203964424133301,
      "epoch": 3.025753585016096,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004136087996181166,
      "loss": 5.7501,
      "mean_token_accuracy": 0.17340597063302993,
      "num_tokens": 51815973.0,
      "step": 15510
    },
    {
      "entropy": 6.205616283416748,
      "epoch": 3.026729099600039,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004135533895310021,
      "loss": 5.7748,
      "mean_token_accuracy": 0.16305476427078247,
      "num_tokens": 51832337.0,
      "step": 15515
    },
    {
      "entropy": 6.179690837860107,
      "epoch": 3.027704614183982,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004134979659049443,
      "loss": 5.8076,
      "mean_token_accuracy": 0.1601847380399704,
      "num_tokens": 51848465.0,
      "step": 15520
    },
    {
      "entropy": 6.236193466186523,
      "epoch": 3.0286801287679253,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041344252874535883,
      "loss": 5.8523,
      "mean_token_accuracy": 0.158984936773777,
      "num_tokens": 51865477.0,
      "step": 15525
    },
    {
      "entropy": 6.252911567687988,
      "epoch": 3.029655643351868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004133870780576628,
      "loss": 5.8915,
      "mean_token_accuracy": 0.16164444088935853,
      "num_tokens": 51881444.0,
      "step": 15530
    },
    {
      "entropy": 6.3043030261993405,
      "epoch": 3.030631157935811,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004133316138472748,
      "loss": 5.9352,
      "mean_token_accuracy": 0.1589430332183838,
      "num_tokens": 51897564.0,
      "step": 15535
    },
    {
      "entropy": 6.194598770141601,
      "epoch": 3.031606672519754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004132761361196145,
      "loss": 5.8182,
      "mean_token_accuracy": 0.16249465495347976,
      "num_tokens": 51916964.0,
      "step": 15540
    },
    {
      "entropy": 6.250267601013183,
      "epoch": 3.032582187103697,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004132206448801031,
      "loss": 5.821,
      "mean_token_accuracy": 0.16538207083940507,
      "num_tokens": 51934530.0,
      "step": 15545
    },
    {
      "entropy": 6.176346111297607,
      "epoch": 3.03355770168764,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041316514013416274,
      "loss": 5.741,
      "mean_token_accuracy": 0.1711185798048973,
      "num_tokens": 51951355.0,
      "step": 15550
    },
    {
      "entropy": 6.186114835739136,
      "epoch": 3.0345332162715835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041310962188721744,
      "loss": 5.7887,
      "mean_token_accuracy": 0.1680769681930542,
      "num_tokens": 51969406.0,
      "step": 15555
    },
    {
      "entropy": 6.220592880249024,
      "epoch": 3.0355087308555264,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041305409014469205,
      "loss": 5.8605,
      "mean_token_accuracy": 0.1596178725361824,
      "num_tokens": 51986093.0,
      "step": 15560
    },
    {
      "entropy": 6.268423128128052,
      "epoch": 3.0364842454394694,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000412998544912013,
      "loss": 5.8838,
      "mean_token_accuracy": 0.15344384163618088,
      "num_tokens": 52004360.0,
      "step": 15565
    },
    {
      "entropy": 6.236105680465698,
      "epoch": 3.0374597600234123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041294298619460804,
      "loss": 5.8221,
      "mean_token_accuracy": 0.16185527443885803,
      "num_tokens": 52018991.0,
      "step": 15570
    },
    {
      "entropy": 6.288387489318848,
      "epoch": 3.0384352746073553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000412887413997906,
      "loss": 5.8842,
      "mean_token_accuracy": 0.15377508103847504,
      "num_tokens": 52035724.0,
      "step": 15575
    },
    {
      "entropy": 6.208971548080444,
      "epoch": 3.0394107891912983,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004128318283273374,
      "loss": 5.7579,
      "mean_token_accuracy": 0.17067483365535735,
      "num_tokens": 52052369.0,
      "step": 15580
    },
    {
      "entropy": 6.276555871963501,
      "epoch": 3.040386303775241,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041277622918833374,
      "loss": 5.9483,
      "mean_token_accuracy": 0.15715281367301942,
      "num_tokens": 52069161.0,
      "step": 15585
    },
    {
      "entropy": 6.180603885650635,
      "epoch": 3.0413618183591846,
      "grad_norm": 1.0625,
      "learning_rate": 0.000412720616586328,
      "loss": 5.749,
      "mean_token_accuracy": 0.18240474611520768,
      "num_tokens": 52084980.0,
      "step": 15590
    },
    {
      "entropy": 6.184870624542237,
      "epoch": 3.0423373329431276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004126649905267545,
      "loss": 5.693,
      "mean_token_accuracy": 0.17052701711654664,
      "num_tokens": 52103566.0,
      "step": 15595
    },
    {
      "entropy": 6.236437034606934,
      "epoch": 3.0433128475270705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041260935101504877,
      "loss": 5.9104,
      "mean_token_accuracy": 0.15712258517742156,
      "num_tokens": 52119535.0,
      "step": 15600
    },
    {
      "entropy": 6.290128183364868,
      "epoch": 3.0442883621110135,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041255369805664766,
      "loss": 5.8943,
      "mean_token_accuracy": 0.15817553251981736,
      "num_tokens": 52135222.0,
      "step": 15605
    },
    {
      "entropy": 6.246860313415527,
      "epoch": 3.0452638766949565,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041249803165698944,
      "loss": 5.8934,
      "mean_token_accuracy": 0.16286389380693436,
      "num_tokens": 52151242.0,
      "step": 15610
    },
    {
      "entropy": 6.142171573638916,
      "epoch": 3.0462393912788994,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004124423518215136,
      "loss": 5.7864,
      "mean_token_accuracy": 0.16618895083665847,
      "num_tokens": 52168366.0,
      "step": 15615
    },
    {
      "entropy": 6.129471397399902,
      "epoch": 3.047214905862843,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041238665855566106,
      "loss": 5.7188,
      "mean_token_accuracy": 0.174710613489151,
      "num_tokens": 52184327.0,
      "step": 15620
    },
    {
      "entropy": 6.302042818069458,
      "epoch": 3.048190420446786,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041233095186487385,
      "loss": 5.8761,
      "mean_token_accuracy": 0.16129359155893325,
      "num_tokens": 52200974.0,
      "step": 15625
    },
    {
      "entropy": 6.247194242477417,
      "epoch": 3.0491659350307287,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004122752317545955,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16440994441509246,
      "num_tokens": 52217245.0,
      "step": 15630
    },
    {
      "entropy": 6.139833927154541,
      "epoch": 3.0501414496146717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004122194982302707,
      "loss": 5.7906,
      "mean_token_accuracy": 0.1669991508126259,
      "num_tokens": 52234924.0,
      "step": 15635
    },
    {
      "entropy": 6.2707297801971436,
      "epoch": 3.0511169641986147,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041216375129734563,
      "loss": 5.9246,
      "mean_token_accuracy": 0.15929774194955826,
      "num_tokens": 52252375.0,
      "step": 15640
    },
    {
      "entropy": 6.2664021968841555,
      "epoch": 3.0520924787825576,
      "grad_norm": 1.0,
      "learning_rate": 0.0004121079909612677,
      "loss": 5.8533,
      "mean_token_accuracy": 0.16780735403299332,
      "num_tokens": 52269751.0,
      "step": 15645
    },
    {
      "entropy": 6.06834044456482,
      "epoch": 3.053067993366501,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004120522172274855,
      "loss": 5.6246,
      "mean_token_accuracy": 0.1735197901725769,
      "num_tokens": 52286967.0,
      "step": 15650
    },
    {
      "entropy": 6.244331645965576,
      "epoch": 3.054043507950444,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041199643010144906,
      "loss": 5.8237,
      "mean_token_accuracy": 0.16442621797323226,
      "num_tokens": 52303712.0,
      "step": 15655
    },
    {
      "entropy": 6.142187213897705,
      "epoch": 3.055019022534387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004119406295886098,
      "loss": 5.6904,
      "mean_token_accuracy": 0.17513761073350906,
      "num_tokens": 52320706.0,
      "step": 15660
    },
    {
      "entropy": 6.207463359832763,
      "epoch": 3.05599453711833,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041188481569442027,
      "loss": 5.803,
      "mean_token_accuracy": 0.16695754677057267,
      "num_tokens": 52335681.0,
      "step": 15665
    },
    {
      "entropy": 6.183241176605224,
      "epoch": 3.056970051702273,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004118289884243344,
      "loss": 5.7819,
      "mean_token_accuracy": 0.1622559204697609,
      "num_tokens": 52354088.0,
      "step": 15670
    },
    {
      "entropy": 6.2671020984649655,
      "epoch": 3.057945566286216,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004117731477838074,
      "loss": 5.9028,
      "mean_token_accuracy": 0.1537437230348587,
      "num_tokens": 52372086.0,
      "step": 15675
    },
    {
      "entropy": 6.245706701278687,
      "epoch": 3.058921080870159,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000411717293778296,
      "loss": 5.876,
      "mean_token_accuracy": 0.1552209421992302,
      "num_tokens": 52389460.0,
      "step": 15680
    },
    {
      "entropy": 6.156208419799805,
      "epoch": 3.059896595454102,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004116614264132578,
      "loss": 5.7511,
      "mean_token_accuracy": 0.16637495160102844,
      "num_tokens": 52404589.0,
      "step": 15685
    },
    {
      "entropy": 6.117969703674317,
      "epoch": 3.060872110038045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004116055456941521,
      "loss": 5.7121,
      "mean_token_accuracy": 0.1747534230351448,
      "num_tokens": 52423225.0,
      "step": 15690
    },
    {
      "entropy": 6.230490779876709,
      "epoch": 3.061847624621988,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041154965162643937,
      "loss": 5.8232,
      "mean_token_accuracy": 0.1574159935116768,
      "num_tokens": 52438289.0,
      "step": 15695
    },
    {
      "entropy": 6.205915069580078,
      "epoch": 3.062823139205931,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004114937442155814,
      "loss": 5.7889,
      "mean_token_accuracy": 0.16770548969507218,
      "num_tokens": 52456418.0,
      "step": 15700
    },
    {
      "entropy": 6.251110649108886,
      "epoch": 3.063798653789874,
      "grad_norm": 1.296875,
      "learning_rate": 0.00041143782346704115,
      "loss": 5.8594,
      "mean_token_accuracy": 0.1584403172135353,
      "num_tokens": 52472327.0,
      "step": 15705
    },
    {
      "entropy": 6.247626304626465,
      "epoch": 3.064774168373817,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041138188938628315,
      "loss": 5.919,
      "mean_token_accuracy": 0.15917345732450486,
      "num_tokens": 52488530.0,
      "step": 15710
    },
    {
      "entropy": 6.272049427032471,
      "epoch": 3.0657496829577604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004113259419787728,
      "loss": 5.8887,
      "mean_token_accuracy": 0.1639714241027832,
      "num_tokens": 52505576.0,
      "step": 15715
    },
    {
      "entropy": 6.251947116851807,
      "epoch": 3.0667251975417034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041126998124997745,
      "loss": 5.7933,
      "mean_token_accuracy": 0.16930869370698928,
      "num_tokens": 52521984.0,
      "step": 15720
    },
    {
      "entropy": 6.225034284591675,
      "epoch": 3.0677007121256463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004112140072053652,
      "loss": 5.7864,
      "mean_token_accuracy": 0.16860363781452178,
      "num_tokens": 52539139.0,
      "step": 15725
    },
    {
      "entropy": 6.15387659072876,
      "epoch": 3.0686762267095893,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004111580198504055,
      "loss": 5.8072,
      "mean_token_accuracy": 0.16266367956995964,
      "num_tokens": 52555652.0,
      "step": 15730
    },
    {
      "entropy": 6.224470615386963,
      "epoch": 3.0696517412935322,
      "grad_norm": 1.125,
      "learning_rate": 0.00041110201919056954,
      "loss": 5.8257,
      "mean_token_accuracy": 0.16698003560304642,
      "num_tokens": 52572013.0,
      "step": 15735
    },
    {
      "entropy": 6.294884061813354,
      "epoch": 3.070627255877475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041104600523132924,
      "loss": 5.9032,
      "mean_token_accuracy": 0.15160734653472902,
      "num_tokens": 52587915.0,
      "step": 15740
    },
    {
      "entropy": 6.1369095802307125,
      "epoch": 3.0716027704614186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004109899779781583,
      "loss": 5.7479,
      "mean_token_accuracy": 0.17229584902524947,
      "num_tokens": 52605406.0,
      "step": 15745
    },
    {
      "entropy": 6.174232149124146,
      "epoch": 3.0725782850453616,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004109339374365312,
      "loss": 5.723,
      "mean_token_accuracy": 0.17424971163272857,
      "num_tokens": 52623246.0,
      "step": 15750
    },
    {
      "entropy": 6.150001811981201,
      "epoch": 3.0735537996293045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004108778836119244,
      "loss": 5.7324,
      "mean_token_accuracy": 0.1678655058145523,
      "num_tokens": 52640528.0,
      "step": 15755
    },
    {
      "entropy": 6.154436159133911,
      "epoch": 3.0745293142132475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000410821816509815,
      "loss": 5.7535,
      "mean_token_accuracy": 0.17023841440677642,
      "num_tokens": 52657579.0,
      "step": 15760
    },
    {
      "entropy": 6.26321759223938,
      "epoch": 3.0755048287971904,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041076573613568176,
      "loss": 5.8848,
      "mean_token_accuracy": 0.16412207931280137,
      "num_tokens": 52673250.0,
      "step": 15765
    },
    {
      "entropy": 6.13038067817688,
      "epoch": 3.0764803433811334,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004107096424950047,
      "loss": 5.7693,
      "mean_token_accuracy": 0.17106588929891586,
      "num_tokens": 52689727.0,
      "step": 15770
    },
    {
      "entropy": 6.167050886154175,
      "epoch": 3.0774558579650764,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004106535355932651,
      "loss": 5.7952,
      "mean_token_accuracy": 0.16932253912091255,
      "num_tokens": 52706724.0,
      "step": 15775
    },
    {
      "entropy": 6.216355991363526,
      "epoch": 3.0784313725490198,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041059741543594537,
      "loss": 5.7988,
      "mean_token_accuracy": 0.16183100640773773,
      "num_tokens": 52723540.0,
      "step": 15780
    },
    {
      "entropy": 6.151831483840942,
      "epoch": 3.0794068871329627,
      "grad_norm": 1.125,
      "learning_rate": 0.00041054128202852954,
      "loss": 5.7929,
      "mean_token_accuracy": 0.17093122005462646,
      "num_tokens": 52740276.0,
      "step": 15785
    },
    {
      "entropy": 6.286447954177857,
      "epoch": 3.0803824017169057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004104851353765027,
      "loss": 5.8751,
      "mean_token_accuracy": 0.16286818981170653,
      "num_tokens": 52756952.0,
      "step": 15790
    },
    {
      "entropy": 6.224555778503418,
      "epoch": 3.0813579163008487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004104289754853513,
      "loss": 5.8542,
      "mean_token_accuracy": 0.1665990561246872,
      "num_tokens": 52773088.0,
      "step": 15795
    },
    {
      "entropy": 6.034260606765747,
      "epoch": 3.0823334308847916,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004103728023605631,
      "loss": 5.6561,
      "mean_token_accuracy": 0.18083238899707793,
      "num_tokens": 52790918.0,
      "step": 15800
    },
    {
      "entropy": 6.272315597534179,
      "epoch": 3.0833089454687346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041031661600762705,
      "loss": 5.87,
      "mean_token_accuracy": 0.16242607682943344,
      "num_tokens": 52807000.0,
      "step": 15805
    },
    {
      "entropy": 6.2181402206420895,
      "epoch": 3.084284460052678,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041026041643203373,
      "loss": 5.8347,
      "mean_token_accuracy": 0.1649252399802208,
      "num_tokens": 52822994.0,
      "step": 15810
    },
    {
      "entropy": 6.224099493026733,
      "epoch": 3.085259974636621,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004102042036392744,
      "loss": 5.8076,
      "mean_token_accuracy": 0.16005063951015472,
      "num_tokens": 52838268.0,
      "step": 15815
    },
    {
      "entropy": 6.2812189102172855,
      "epoch": 3.086235489220564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004101479776348423,
      "loss": 5.9133,
      "mean_token_accuracy": 0.15770626068115234,
      "num_tokens": 52855683.0,
      "step": 15820
    },
    {
      "entropy": 6.239777135848999,
      "epoch": 3.087211003804507,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004100917384242315,
      "loss": 5.8503,
      "mean_token_accuracy": 0.16720605343580247,
      "num_tokens": 52872141.0,
      "step": 15825
    },
    {
      "entropy": 6.265920400619507,
      "epoch": 3.08818651838845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041003548601293747,
      "loss": 5.8372,
      "mean_token_accuracy": 0.16442590802907944,
      "num_tokens": 52888919.0,
      "step": 15830
    },
    {
      "entropy": 6.216282939910888,
      "epoch": 3.089162032972393,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040997922040645696,
      "loss": 5.7946,
      "mean_token_accuracy": 0.161602945625782,
      "num_tokens": 52905903.0,
      "step": 15835
    },
    {
      "entropy": 6.201101684570313,
      "epoch": 3.0901375475563357,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004099229416102881,
      "loss": 5.8169,
      "mean_token_accuracy": 0.16546008139848709,
      "num_tokens": 52924224.0,
      "step": 15840
    },
    {
      "entropy": 6.227807903289795,
      "epoch": 3.091113062140279,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004098666496299304,
      "loss": 5.8522,
      "mean_token_accuracy": 0.1619046673178673,
      "num_tokens": 52940510.0,
      "step": 15845
    },
    {
      "entropy": 6.046582269668579,
      "epoch": 3.092088576724222,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040981034447088425,
      "loss": 5.6806,
      "mean_token_accuracy": 0.1842719256877899,
      "num_tokens": 52958458.0,
      "step": 15850
    },
    {
      "entropy": 6.283346939086914,
      "epoch": 3.093064091308165,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004097540261386518,
      "loss": 5.8477,
      "mean_token_accuracy": 0.16206797659397126,
      "num_tokens": 52973709.0,
      "step": 15855
    },
    {
      "entropy": 6.2765540599823,
      "epoch": 3.094039605892108,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040969769463873617,
      "loss": 5.8178,
      "mean_token_accuracy": 0.1612459197640419,
      "num_tokens": 52989203.0,
      "step": 15860
    },
    {
      "entropy": 6.18087067604065,
      "epoch": 3.095015120476051,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040964134997664183,
      "loss": 5.7578,
      "mean_token_accuracy": 0.16331739574670792,
      "num_tokens": 53005833.0,
      "step": 15865
    },
    {
      "entropy": 6.279252767562866,
      "epoch": 3.095990635059994,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040958499215787463,
      "loss": 5.9386,
      "mean_token_accuracy": 0.15845808386802673,
      "num_tokens": 53022616.0,
      "step": 15870
    },
    {
      "entropy": 6.164662265777588,
      "epoch": 3.0969661496439373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004095286211879418,
      "loss": 5.7803,
      "mean_token_accuracy": 0.16866704374551772,
      "num_tokens": 53040473.0,
      "step": 15875
    },
    {
      "entropy": 6.137791395187378,
      "epoch": 3.0979416642278803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040947223707235147,
      "loss": 5.7111,
      "mean_token_accuracy": 0.1821012169122696,
      "num_tokens": 53057142.0,
      "step": 15880
    },
    {
      "entropy": 6.07467737197876,
      "epoch": 3.0989171788118233,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004094158398166134,
      "loss": 5.6909,
      "mean_token_accuracy": 0.17446031123399736,
      "num_tokens": 53073909.0,
      "step": 15885
    },
    {
      "entropy": 6.273946237564087,
      "epoch": 3.0998926933957662,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004093594294262386,
      "loss": 5.8719,
      "mean_token_accuracy": 0.15555986016988754,
      "num_tokens": 53091327.0,
      "step": 15890
    },
    {
      "entropy": 6.2101637840271,
      "epoch": 3.100868207979709,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040930300590673915,
      "loss": 5.8364,
      "mean_token_accuracy": 0.15907673686742782,
      "num_tokens": 53107589.0,
      "step": 15895
    },
    {
      "entropy": 6.156439638137817,
      "epoch": 3.101843722563652,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040924656926362863,
      "loss": 5.7032,
      "mean_token_accuracy": 0.17921730130910873,
      "num_tokens": 53125165.0,
      "step": 15900
    },
    {
      "entropy": 6.256997537612915,
      "epoch": 3.1028192371475956,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004091901195024218,
      "loss": 5.793,
      "mean_token_accuracy": 0.16626544594764708,
      "num_tokens": 53141704.0,
      "step": 15905
    },
    {
      "entropy": 6.223965072631836,
      "epoch": 3.1037947517315385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004091336566286348,
      "loss": 5.8264,
      "mean_token_accuracy": 0.16708132475614548,
      "num_tokens": 53158220.0,
      "step": 15910
    },
    {
      "entropy": 6.221233224868774,
      "epoch": 3.1047702663154815,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004090771806477849,
      "loss": 5.8586,
      "mean_token_accuracy": 0.16655220091342926,
      "num_tokens": 53173947.0,
      "step": 15915
    },
    {
      "entropy": 6.231815528869629,
      "epoch": 3.1057457808994244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004090206915653908,
      "loss": 5.8239,
      "mean_token_accuracy": 0.171123006939888,
      "num_tokens": 53190885.0,
      "step": 15920
    },
    {
      "entropy": 6.266730165481567,
      "epoch": 3.1067212954833674,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040896418938697246,
      "loss": 5.9107,
      "mean_token_accuracy": 0.15785178244113923,
      "num_tokens": 53209555.0,
      "step": 15925
    },
    {
      "entropy": 6.339381408691406,
      "epoch": 3.1076968100673104,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040890767411805084,
      "loss": 5.9003,
      "mean_token_accuracy": 0.15762402415275573,
      "num_tokens": 53225794.0,
      "step": 15930
    },
    {
      "entropy": 6.258233499526978,
      "epoch": 3.1086723246512538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004088511457641487,
      "loss": 5.8437,
      "mean_token_accuracy": 0.15972869247198104,
      "num_tokens": 53243070.0,
      "step": 15935
    },
    {
      "entropy": 6.110585165023804,
      "epoch": 3.1096478392351967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040879460433078954,
      "loss": 5.7547,
      "mean_token_accuracy": 0.1688948690891266,
      "num_tokens": 53260734.0,
      "step": 15940
    },
    {
      "entropy": 6.27774453163147,
      "epoch": 3.1106233538191397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004087380498234985,
      "loss": 5.9277,
      "mean_token_accuracy": 0.15289975255727767,
      "num_tokens": 53278462.0,
      "step": 15945
    },
    {
      "entropy": 6.295208501815796,
      "epoch": 3.1115988684030826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004086814822478019,
      "loss": 5.8068,
      "mean_token_accuracy": 0.16519416719675065,
      "num_tokens": 53294010.0,
      "step": 15950
    },
    {
      "entropy": 6.265837860107422,
      "epoch": 3.1125743829870256,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040862490160922734,
      "loss": 5.8917,
      "mean_token_accuracy": 0.1550367906689644,
      "num_tokens": 53310546.0,
      "step": 15955
    },
    {
      "entropy": 6.130472850799561,
      "epoch": 3.1135498975709686,
      "grad_norm": 1.25,
      "learning_rate": 0.0004085683079133036,
      "loss": 5.7626,
      "mean_token_accuracy": 0.17229465395212173,
      "num_tokens": 53328646.0,
      "step": 15960
    },
    {
      "entropy": 6.0057343482971195,
      "epoch": 3.1145254121549115,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040851170116556085,
      "loss": 5.5543,
      "mean_token_accuracy": 0.19516785591840743,
      "num_tokens": 53346116.0,
      "step": 15965
    },
    {
      "entropy": 6.16615662574768,
      "epoch": 3.115500926738855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004084550813715306,
      "loss": 5.7995,
      "mean_token_accuracy": 0.16141244024038315,
      "num_tokens": 53363067.0,
      "step": 15970
    },
    {
      "entropy": 6.246896028518677,
      "epoch": 3.116476441322798,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040839844853674537,
      "loss": 5.7981,
      "mean_token_accuracy": 0.1653435230255127,
      "num_tokens": 53378992.0,
      "step": 15975
    },
    {
      "entropy": 6.174412775039673,
      "epoch": 3.117451955906741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040834180266673917,
      "loss": 5.8069,
      "mean_token_accuracy": 0.16678315550088882,
      "num_tokens": 53395761.0,
      "step": 15980
    },
    {
      "entropy": 6.254252290725708,
      "epoch": 3.118427470490684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040828514376704724,
      "loss": 5.9143,
      "mean_token_accuracy": 0.15514134764671325,
      "num_tokens": 53412717.0,
      "step": 15985
    },
    {
      "entropy": 6.287068128585815,
      "epoch": 3.1194029850746268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040822847184320617,
      "loss": 5.8075,
      "mean_token_accuracy": 0.16678355187177657,
      "num_tokens": 53429525.0,
      "step": 15990
    },
    {
      "entropy": 6.230864334106445,
      "epoch": 3.1203784996585697,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040817178690075363,
      "loss": 5.8396,
      "mean_token_accuracy": 0.1627047747373581,
      "num_tokens": 53447195.0,
      "step": 15995
    },
    {
      "entropy": 6.216898393630982,
      "epoch": 3.121354014242513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004081150889452287,
      "loss": 5.7866,
      "mean_token_accuracy": 0.1642625406384468,
      "num_tokens": 53463671.0,
      "step": 16000
    },
    {
      "entropy": 6.217847537994385,
      "epoch": 3.122329528826456,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004080583779821717,
      "loss": 5.823,
      "mean_token_accuracy": 0.16079695224761964,
      "num_tokens": 53480721.0,
      "step": 16005
    },
    {
      "entropy": 6.190394020080566,
      "epoch": 3.123305043410399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040800165401712427,
      "loss": 5.8003,
      "mean_token_accuracy": 0.1643790617585182,
      "num_tokens": 53498292.0,
      "step": 16010
    },
    {
      "entropy": 6.220792245864868,
      "epoch": 3.124280557994342,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040794491705562917,
      "loss": 5.7932,
      "mean_token_accuracy": 0.1724562719464302,
      "num_tokens": 53515747.0,
      "step": 16015
    },
    {
      "entropy": 6.107896184921264,
      "epoch": 3.125256072578285,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040788816710323063,
      "loss": 5.7057,
      "mean_token_accuracy": 0.17410804182291031,
      "num_tokens": 53532895.0,
      "step": 16020
    },
    {
      "entropy": 6.115035581588745,
      "epoch": 3.126231587162228,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004078314041654741,
      "loss": 5.7263,
      "mean_token_accuracy": 0.17649343162775039,
      "num_tokens": 53549241.0,
      "step": 16025
    },
    {
      "entropy": 6.256380367279053,
      "epoch": 3.127207101746171,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004077746282479061,
      "loss": 5.8562,
      "mean_token_accuracy": 0.16379924565553666,
      "num_tokens": 53565882.0,
      "step": 16030
    },
    {
      "entropy": 6.274718618392944,
      "epoch": 3.1281826163301143,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004077178393560746,
      "loss": 5.8991,
      "mean_token_accuracy": 0.15876066237688063,
      "num_tokens": 53582770.0,
      "step": 16035
    },
    {
      "entropy": 6.169937038421631,
      "epoch": 3.1291581309140573,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004076610374955289,
      "loss": 5.7756,
      "mean_token_accuracy": 0.17033956497907637,
      "num_tokens": 53600649.0,
      "step": 16040
    },
    {
      "entropy": 6.18187575340271,
      "epoch": 3.130133645498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004076042226718195,
      "loss": 5.8097,
      "mean_token_accuracy": 0.1714318037033081,
      "num_tokens": 53616568.0,
      "step": 16045
    },
    {
      "entropy": 6.276512670516968,
      "epoch": 3.131109160081943,
      "grad_norm": 0.953125,
      "learning_rate": 0.000407547394890498,
      "loss": 5.8604,
      "mean_token_accuracy": 0.16256125420331954,
      "num_tokens": 53634717.0,
      "step": 16050
    },
    {
      "entropy": 6.210611391067505,
      "epoch": 3.132084674665886,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040749055415711743,
      "loss": 5.7901,
      "mean_token_accuracy": 0.1599969521164894,
      "num_tokens": 53650877.0,
      "step": 16055
    },
    {
      "entropy": 6.235774135589599,
      "epoch": 3.133060189249829,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004074337004772322,
      "loss": 5.8034,
      "mean_token_accuracy": 0.16631621718406678,
      "num_tokens": 53667906.0,
      "step": 16060
    },
    {
      "entropy": 6.0996490001678465,
      "epoch": 3.1340357038337725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040737683385639776,
      "loss": 5.758,
      "mean_token_accuracy": 0.16607977002859114,
      "num_tokens": 53682942.0,
      "step": 16065
    },
    {
      "entropy": 6.165964031219483,
      "epoch": 3.1350112184177155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040731995430017085,
      "loss": 5.8351,
      "mean_token_accuracy": 0.16445836871862413,
      "num_tokens": 53699906.0,
      "step": 16070
    },
    {
      "entropy": 6.293869161605835,
      "epoch": 3.1359867330016584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004072630618141096,
      "loss": 5.8535,
      "mean_token_accuracy": 0.16018864661455154,
      "num_tokens": 53716278.0,
      "step": 16075
    },
    {
      "entropy": 6.157254600524903,
      "epoch": 3.1369622475856014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040720615640377347,
      "loss": 5.7054,
      "mean_token_accuracy": 0.1818431168794632,
      "num_tokens": 53733317.0,
      "step": 16080
    },
    {
      "entropy": 6.2067403316497805,
      "epoch": 3.1379377621695443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040714923807472274,
      "loss": 5.7974,
      "mean_token_accuracy": 0.16177025735378264,
      "num_tokens": 53750477.0,
      "step": 16085
    },
    {
      "entropy": 6.262089633941651,
      "epoch": 3.1389132767534873,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004070923068325196,
      "loss": 5.8693,
      "mean_token_accuracy": 0.1629518210887909,
      "num_tokens": 53765755.0,
      "step": 16090
    },
    {
      "entropy": 6.180956554412842,
      "epoch": 3.1398887913374303,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040703536268272693,
      "loss": 5.8297,
      "mean_token_accuracy": 0.17208331674337388,
      "num_tokens": 53782173.0,
      "step": 16095
    },
    {
      "entropy": 6.091863012313842,
      "epoch": 3.1408643059213737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004069784056309093,
      "loss": 5.6888,
      "mean_token_accuracy": 0.17549254447221757,
      "num_tokens": 53799407.0,
      "step": 16100
    },
    {
      "entropy": 6.142131519317627,
      "epoch": 3.1418398205053166,
      "grad_norm": 1.0,
      "learning_rate": 0.0004069214356826322,
      "loss": 5.7433,
      "mean_token_accuracy": 0.17709824740886687,
      "num_tokens": 53816363.0,
      "step": 16105
    },
    {
      "entropy": 6.118453788757324,
      "epoch": 3.1428153350892596,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040686445284346254,
      "loss": 5.7124,
      "mean_token_accuracy": 0.17353785932064056,
      "num_tokens": 53833022.0,
      "step": 16110
    },
    {
      "entropy": 6.205348682403565,
      "epoch": 3.1437908496732025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004068074571189685,
      "loss": 5.7813,
      "mean_token_accuracy": 0.17576292008161545,
      "num_tokens": 53849924.0,
      "step": 16115
    },
    {
      "entropy": 6.203200054168701,
      "epoch": 3.1447663642571455,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040675044851471966,
      "loss": 5.8462,
      "mean_token_accuracy": 0.1701057657599449,
      "num_tokens": 53866291.0,
      "step": 16120
    },
    {
      "entropy": 6.284155082702637,
      "epoch": 3.1457418788410885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040669342703628645,
      "loss": 5.8826,
      "mean_token_accuracy": 0.16285464614629747,
      "num_tokens": 53882510.0,
      "step": 16125
    },
    {
      "entropy": 6.207709836959839,
      "epoch": 3.146717393425032,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004066363926892409,
      "loss": 5.7888,
      "mean_token_accuracy": 0.16349004209041595,
      "num_tokens": 53898833.0,
      "step": 16130
    },
    {
      "entropy": 6.260624742507934,
      "epoch": 3.147692908008975,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040657934547915625,
      "loss": 5.8578,
      "mean_token_accuracy": 0.16358547657728195,
      "num_tokens": 53916945.0,
      "step": 16135
    },
    {
      "entropy": 6.26761827468872,
      "epoch": 3.148668422592918,
      "grad_norm": 1.015625,
      "learning_rate": 0.000406522285411607,
      "loss": 5.8869,
      "mean_token_accuracy": 0.16068013310432433,
      "num_tokens": 53933853.0,
      "step": 16140
    },
    {
      "entropy": 6.240430212020874,
      "epoch": 3.1496439371768608,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004064652124921686,
      "loss": 5.8441,
      "mean_token_accuracy": 0.15799386650323868,
      "num_tokens": 53950098.0,
      "step": 16145
    },
    {
      "entropy": 6.220660877227783,
      "epoch": 3.1506194517608037,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040640812672641834,
      "loss": 5.7651,
      "mean_token_accuracy": 0.16783173829317094,
      "num_tokens": 53966070.0,
      "step": 16150
    },
    {
      "entropy": 6.120322799682617,
      "epoch": 3.1515949663447467,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040635102811993417,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1716659188270569,
      "num_tokens": 53981715.0,
      "step": 16155
    },
    {
      "entropy": 6.135096168518066,
      "epoch": 3.15257048092869,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004062939166782958,
      "loss": 5.727,
      "mean_token_accuracy": 0.1723276972770691,
      "num_tokens": 53999179.0,
      "step": 16160
    },
    {
      "entropy": 5.9937573909759525,
      "epoch": 3.153545995512633,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004062367924070837,
      "loss": 5.5775,
      "mean_token_accuracy": 0.18260839730501174,
      "num_tokens": 54016024.0,
      "step": 16165
    },
    {
      "entropy": 6.211577892303467,
      "epoch": 3.154521510096576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040617965531188007,
      "loss": 5.864,
      "mean_token_accuracy": 0.16289284378290175,
      "num_tokens": 54033378.0,
      "step": 16170
    },
    {
      "entropy": 6.227110433578491,
      "epoch": 3.155497024680519,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004061225053982681,
      "loss": 5.8338,
      "mean_token_accuracy": 0.1638586401939392,
      "num_tokens": 54049432.0,
      "step": 16175
    },
    {
      "entropy": 6.1753058433532715,
      "epoch": 3.156472539264462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004060653426718321,
      "loss": 5.717,
      "mean_token_accuracy": 0.171714386343956,
      "num_tokens": 54064688.0,
      "step": 16180
    },
    {
      "entropy": 6.07286434173584,
      "epoch": 3.157448053848405,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000406008167138158,
      "loss": 5.7226,
      "mean_token_accuracy": 0.1755938559770584,
      "num_tokens": 54080195.0,
      "step": 16185
    },
    {
      "entropy": 6.255313777923584,
      "epoch": 3.1584235684323483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040595097880283276,
      "loss": 5.8709,
      "mean_token_accuracy": 0.16279981136322022,
      "num_tokens": 54096138.0,
      "step": 16190
    },
    {
      "entropy": 6.178056383132935,
      "epoch": 3.1593990830162912,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040589377767144455,
      "loss": 5.7752,
      "mean_token_accuracy": 0.1773795410990715,
      "num_tokens": 54112411.0,
      "step": 16195
    },
    {
      "entropy": 6.237961387634277,
      "epoch": 3.160374597600234,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040583656374958285,
      "loss": 5.8089,
      "mean_token_accuracy": 0.16634130477905273,
      "num_tokens": 54128207.0,
      "step": 16200
    },
    {
      "entropy": 6.212740564346314,
      "epoch": 3.161350112184177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040577933704283856,
      "loss": 5.8705,
      "mean_token_accuracy": 0.15824130475521087,
      "num_tokens": 54144997.0,
      "step": 16205
    },
    {
      "entropy": 6.139139699935913,
      "epoch": 3.16232562676812,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004057220975568035,
      "loss": 5.8067,
      "mean_token_accuracy": 0.16710774302482606,
      "num_tokens": 54162733.0,
      "step": 16210
    },
    {
      "entropy": 6.222628498077393,
      "epoch": 3.163301141352063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004056648452970709,
      "loss": 5.8722,
      "mean_token_accuracy": 0.15591213554143907,
      "num_tokens": 54180712.0,
      "step": 16215
    },
    {
      "entropy": 6.190336275100708,
      "epoch": 3.164276655936006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004056075802692354,
      "loss": 5.7341,
      "mean_token_accuracy": 0.16957340836524964,
      "num_tokens": 54196797.0,
      "step": 16220
    },
    {
      "entropy": 6.142151594161987,
      "epoch": 3.1652521705199494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004055503024788927,
      "loss": 5.7738,
      "mean_token_accuracy": 0.16792965531349183,
      "num_tokens": 54214911.0,
      "step": 16225
    },
    {
      "entropy": 6.246589469909668,
      "epoch": 3.1662276851038924,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004054930119316397,
      "loss": 5.9063,
      "mean_token_accuracy": 0.16129881590604783,
      "num_tokens": 54230049.0,
      "step": 16230
    },
    {
      "entropy": 6.152826833724975,
      "epoch": 3.1672031996878354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004054357086330745,
      "loss": 5.8063,
      "mean_token_accuracy": 0.17263448387384414,
      "num_tokens": 54247490.0,
      "step": 16235
    },
    {
      "entropy": 6.332635164260864,
      "epoch": 3.1681787142717783,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040537839258879694,
      "loss": 5.8733,
      "mean_token_accuracy": 0.16470015347003936,
      "num_tokens": 54263731.0,
      "step": 16240
    },
    {
      "entropy": 6.276911878585816,
      "epoch": 3.1691542288557213,
      "grad_norm": 1.0,
      "learning_rate": 0.0004053210638044075,
      "loss": 5.8668,
      "mean_token_accuracy": 0.1595069333910942,
      "num_tokens": 54279820.0,
      "step": 16245
    },
    {
      "entropy": 6.2404632568359375,
      "epoch": 3.1701297434396642,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040526372228550804,
      "loss": 5.8851,
      "mean_token_accuracy": 0.15932225286960602,
      "num_tokens": 54296728.0,
      "step": 16250
    },
    {
      "entropy": 6.146991729736328,
      "epoch": 3.1711052580236077,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040520636803770206,
      "loss": 5.6713,
      "mean_token_accuracy": 0.18026439994573593,
      "num_tokens": 54314526.0,
      "step": 16255
    },
    {
      "entropy": 6.2244819641113285,
      "epoch": 3.1720807726075506,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004051490010665938,
      "loss": 5.8063,
      "mean_token_accuracy": 0.15908506214618684,
      "num_tokens": 54330750.0,
      "step": 16260
    },
    {
      "entropy": 6.18062424659729,
      "epoch": 3.1730562871914936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000405091621377789,
      "loss": 5.7783,
      "mean_token_accuracy": 0.17058975994586945,
      "num_tokens": 54349380.0,
      "step": 16265
    },
    {
      "entropy": 6.285933971405029,
      "epoch": 3.1740318017754365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004050342289768946,
      "loss": 5.8918,
      "mean_token_accuracy": 0.1593440815806389,
      "num_tokens": 54366034.0,
      "step": 16270
    },
    {
      "entropy": 6.186677169799805,
      "epoch": 3.1750073163593795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040497682386951886,
      "loss": 5.852,
      "mean_token_accuracy": 0.15959134250879287,
      "num_tokens": 54382293.0,
      "step": 16275
    },
    {
      "entropy": 6.2267717838287355,
      "epoch": 3.1759828309433225,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040491940606127115,
      "loss": 5.8754,
      "mean_token_accuracy": 0.16566607505083084,
      "num_tokens": 54399183.0,
      "step": 16280
    },
    {
      "entropy": 6.278177499771118,
      "epoch": 3.1769583455272654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040486197555776197,
      "loss": 5.9004,
      "mean_token_accuracy": 0.1580891951918602,
      "num_tokens": 54414611.0,
      "step": 16285
    },
    {
      "entropy": 6.19642071723938,
      "epoch": 3.177933860111209,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040480453236460355,
      "loss": 5.7476,
      "mean_token_accuracy": 0.1661606550216675,
      "num_tokens": 54431025.0,
      "step": 16290
    },
    {
      "entropy": 6.186483144760132,
      "epoch": 3.178909374695152,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040474707648740875,
      "loss": 5.7622,
      "mean_token_accuracy": 0.17054146230220796,
      "num_tokens": 54447604.0,
      "step": 16295
    },
    {
      "entropy": 6.18745698928833,
      "epoch": 3.1798848892790947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004046896079317921,
      "loss": 5.7983,
      "mean_token_accuracy": 0.16434222161769868,
      "num_tokens": 54463519.0,
      "step": 16300
    },
    {
      "entropy": 6.150858592987061,
      "epoch": 3.1808604038630377,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040463212670336924,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16986935436725617,
      "num_tokens": 54479303.0,
      "step": 16305
    },
    {
      "entropy": 5.9953882694244385,
      "epoch": 3.1818359184469807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040457463280775687,
      "loss": 5.5336,
      "mean_token_accuracy": 0.18457736670970917,
      "num_tokens": 54495462.0,
      "step": 16310
    },
    {
      "entropy": 6.209109544754028,
      "epoch": 3.1828114330309236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004045171262505732,
      "loss": 5.8549,
      "mean_token_accuracy": 0.156624037027359,
      "num_tokens": 54512849.0,
      "step": 16315
    },
    {
      "entropy": 6.236200523376465,
      "epoch": 3.183786947614867,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004044596070374376,
      "loss": 5.8533,
      "mean_token_accuracy": 0.1627502992749214,
      "num_tokens": 54529678.0,
      "step": 16320
    },
    {
      "entropy": 6.255303192138672,
      "epoch": 3.18476246219881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004044020751739706,
      "loss": 5.8382,
      "mean_token_accuracy": 0.16084229052066804,
      "num_tokens": 54546317.0,
      "step": 16325
    },
    {
      "entropy": 6.186604690551758,
      "epoch": 3.185737976782753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004043445306657941,
      "loss": 5.8264,
      "mean_token_accuracy": 0.1650228202342987,
      "num_tokens": 54562983.0,
      "step": 16330
    },
    {
      "entropy": 6.3230408191680905,
      "epoch": 3.186713491366696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040428697351853094,
      "loss": 5.9689,
      "mean_token_accuracy": 0.15704874694347382,
      "num_tokens": 54578708.0,
      "step": 16335
    },
    {
      "entropy": 6.069735908508301,
      "epoch": 3.187689005950639,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040422940373780554,
      "loss": 5.7211,
      "mean_token_accuracy": 0.1815740242600441,
      "num_tokens": 54594555.0,
      "step": 16340
    },
    {
      "entropy": 6.249438285827637,
      "epoch": 3.188664520534582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004041718213292434,
      "loss": 5.83,
      "mean_token_accuracy": 0.1648094356060028,
      "num_tokens": 54611430.0,
      "step": 16345
    },
    {
      "entropy": 6.175962734222412,
      "epoch": 3.1896400351185252,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040411422629847124,
      "loss": 5.7555,
      "mean_token_accuracy": 0.168538436293602,
      "num_tokens": 54629467.0,
      "step": 16350
    },
    {
      "entropy": 6.278278875350952,
      "epoch": 3.190615549702468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000404056618651117,
      "loss": 5.8881,
      "mean_token_accuracy": 0.15618254095315934,
      "num_tokens": 54644732.0,
      "step": 16355
    },
    {
      "entropy": 6.1532604694366455,
      "epoch": 3.191591064286411,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040399899839281013,
      "loss": 5.7948,
      "mean_token_accuracy": 0.17012899219989777,
      "num_tokens": 54661639.0,
      "step": 16360
    },
    {
      "entropy": 6.144733953475952,
      "epoch": 3.192566578870354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004039413655291807,
      "loss": 5.7562,
      "mean_token_accuracy": 0.17060239762067794,
      "num_tokens": 54678153.0,
      "step": 16365
    },
    {
      "entropy": 6.216518783569336,
      "epoch": 3.193542093454297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004038837200658608,
      "loss": 5.776,
      "mean_token_accuracy": 0.16694238185882568,
      "num_tokens": 54694215.0,
      "step": 16370
    },
    {
      "entropy": 6.157992553710938,
      "epoch": 3.19451760803824,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040382606200848285,
      "loss": 5.7934,
      "mean_token_accuracy": 0.1711350053548813,
      "num_tokens": 54710637.0,
      "step": 16375
    },
    {
      "entropy": 6.158230590820312,
      "epoch": 3.1954931226221834,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040376839136268145,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1632266715168953,
      "num_tokens": 54726638.0,
      "step": 16380
    },
    {
      "entropy": 6.16383113861084,
      "epoch": 3.1964686372061264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004037107081340918,
      "loss": 5.7558,
      "mean_token_accuracy": 0.16720911711454392,
      "num_tokens": 54744581.0,
      "step": 16385
    },
    {
      "entropy": 6.1665709018707275,
      "epoch": 3.1974441517900694,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004036530123283504,
      "loss": 5.762,
      "mean_token_accuracy": 0.17346725910902022,
      "num_tokens": 54761561.0,
      "step": 16390
    },
    {
      "entropy": 6.219540309906006,
      "epoch": 3.1984196663740123,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004035953039510952,
      "loss": 5.8446,
      "mean_token_accuracy": 0.16642138510942459,
      "num_tokens": 54778457.0,
      "step": 16395
    },
    {
      "entropy": 6.207294702529907,
      "epoch": 3.1993951809579553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040353758300796524,
      "loss": 5.7629,
      "mean_token_accuracy": 0.16747866123914718,
      "num_tokens": 54794992.0,
      "step": 16400
    },
    {
      "entropy": 6.273727035522461,
      "epoch": 3.2003706955418982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040347984950460077,
      "loss": 5.8676,
      "mean_token_accuracy": 0.16194114685058594,
      "num_tokens": 54812263.0,
      "step": 16405
    },
    {
      "entropy": 6.219223880767823,
      "epoch": 3.201346210125841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040342210344664325,
      "loss": 5.8815,
      "mean_token_accuracy": 0.1599980801343918,
      "num_tokens": 54828059.0,
      "step": 16410
    },
    {
      "entropy": 6.277783918380737,
      "epoch": 3.2023217247097846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004033643448397356,
      "loss": 5.9515,
      "mean_token_accuracy": 0.1570083335042,
      "num_tokens": 54844615.0,
      "step": 16415
    },
    {
      "entropy": 6.212590551376342,
      "epoch": 3.2032972392937276,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004033065736895215,
      "loss": 5.7715,
      "mean_token_accuracy": 0.16367797404527665,
      "num_tokens": 54861760.0,
      "step": 16420
    },
    {
      "entropy": 6.290134859085083,
      "epoch": 3.2042727538776705,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040324879000164644,
      "loss": 5.9163,
      "mean_token_accuracy": 0.15867924839258193,
      "num_tokens": 54878426.0,
      "step": 16425
    },
    {
      "entropy": 6.214656257629395,
      "epoch": 3.2052482684616135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031909937817567,
      "loss": 5.8222,
      "mean_token_accuracy": 0.1646653816103935,
      "num_tokens": 54895173.0,
      "step": 16430
    },
    {
      "entropy": 6.310809230804443,
      "epoch": 3.2062237830455564,
      "grad_norm": 1.0,
      "learning_rate": 0.00040313318503549996,
      "loss": 5.9213,
      "mean_token_accuracy": 0.157949598133564,
      "num_tokens": 54911054.0,
      "step": 16435
    },
    {
      "entropy": 6.256660509109497,
      "epoch": 3.2071992976294994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040307536376852504,
      "loss": 5.8324,
      "mean_token_accuracy": 0.1637166514992714,
      "num_tokens": 54926250.0,
      "step": 16440
    },
    {
      "entropy": 6.189906787872315,
      "epoch": 3.208174812213443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040301752998648193,
      "loss": 5.7979,
      "mean_token_accuracy": 0.17346610873937607,
      "num_tokens": 54941810.0,
      "step": 16445
    },
    {
      "entropy": 6.235698795318603,
      "epoch": 3.2091503267973858,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004029596836950222,
      "loss": 5.8862,
      "mean_token_accuracy": 0.16191161125898362,
      "num_tokens": 54957628.0,
      "step": 16450
    },
    {
      "entropy": 6.192990827560425,
      "epoch": 3.2101258413813287,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040290182489979815,
      "loss": 5.8092,
      "mean_token_accuracy": 0.16441211998462676,
      "num_tokens": 54974349.0,
      "step": 16455
    },
    {
      "entropy": 6.139688158035279,
      "epoch": 3.2111013559652717,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040284395360646366,
      "loss": 5.7453,
      "mean_token_accuracy": 0.1746838241815567,
      "num_tokens": 54991936.0,
      "step": 16460
    },
    {
      "entropy": 6.274082326889038,
      "epoch": 3.2120768705492146,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040278606982067363,
      "loss": 5.9153,
      "mean_token_accuracy": 0.1604343593120575,
      "num_tokens": 55009444.0,
      "step": 16465
    },
    {
      "entropy": 6.2613318920135494,
      "epoch": 3.2130523851331576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040272817354808425,
      "loss": 5.8482,
      "mean_token_accuracy": 0.16183757781982422,
      "num_tokens": 55025183.0,
      "step": 16470
    },
    {
      "entropy": 6.131117630004883,
      "epoch": 3.2140278997171006,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004026702647943531,
      "loss": 5.7391,
      "mean_token_accuracy": 0.1667181059718132,
      "num_tokens": 55040996.0,
      "step": 16475
    },
    {
      "entropy": 6.110746383666992,
      "epoch": 3.215003414301044,
      "grad_norm": 1.125,
      "learning_rate": 0.00040261234356513863,
      "loss": 5.7388,
      "mean_token_accuracy": 0.1693742960691452,
      "num_tokens": 55056979.0,
      "step": 16480
    },
    {
      "entropy": 6.08370008468628,
      "epoch": 3.215978928884987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004025544098661008,
      "loss": 5.6818,
      "mean_token_accuracy": 0.17895012050867082,
      "num_tokens": 55075276.0,
      "step": 16485
    },
    {
      "entropy": 6.12376594543457,
      "epoch": 3.21695444346893,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040249646370290064,
      "loss": 5.6895,
      "mean_token_accuracy": 0.17899810075759887,
      "num_tokens": 55091564.0,
      "step": 16490
    },
    {
      "entropy": 6.114178276062011,
      "epoch": 3.217929958052873,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004024385050812004,
      "loss": 5.7709,
      "mean_token_accuracy": 0.173391255736351,
      "num_tokens": 55109801.0,
      "step": 16495
    },
    {
      "entropy": 6.191365098953247,
      "epoch": 3.218905472636816,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040238053400666374,
      "loss": 5.781,
      "mean_token_accuracy": 0.16869866400957106,
      "num_tokens": 55127625.0,
      "step": 16500
    },
    {
      "entropy": 6.18901801109314,
      "epoch": 3.2198809872207588,
      "grad_norm": 1.0,
      "learning_rate": 0.0004023225504849552,
      "loss": 5.8206,
      "mean_token_accuracy": 0.16985686123371124,
      "num_tokens": 55146895.0,
      "step": 16505
    },
    {
      "entropy": 6.240416479110718,
      "epoch": 3.220856501804702,
      "grad_norm": 1.0,
      "learning_rate": 0.00040226455452174086,
      "loss": 5.8338,
      "mean_token_accuracy": 0.16268148720264436,
      "num_tokens": 55163540.0,
      "step": 16510
    },
    {
      "entropy": 6.241417646408081,
      "epoch": 3.221832016388645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004022065461226879,
      "loss": 5.8214,
      "mean_token_accuracy": 0.1628086194396019,
      "num_tokens": 55180938.0,
      "step": 16515
    },
    {
      "entropy": 6.089584875106811,
      "epoch": 3.222807530972588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040214852529346445,
      "loss": 5.6952,
      "mean_token_accuracy": 0.18189727067947387,
      "num_tokens": 55197661.0,
      "step": 16520
    },
    {
      "entropy": 6.152491617202759,
      "epoch": 3.223783045556531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040209049203974035,
      "loss": 5.7156,
      "mean_token_accuracy": 0.1700295999646187,
      "num_tokens": 55214013.0,
      "step": 16525
    },
    {
      "entropy": 6.080373287200928,
      "epoch": 3.224758560140474,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004020324463671864,
      "loss": 5.6699,
      "mean_token_accuracy": 0.18817343711853027,
      "num_tokens": 55231443.0,
      "step": 16530
    },
    {
      "entropy": 6.131638145446777,
      "epoch": 3.225734074724417,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004019743882814745,
      "loss": 5.721,
      "mean_token_accuracy": 0.1785511329770088,
      "num_tokens": 55249367.0,
      "step": 16535
    },
    {
      "entropy": 6.3049060821533205,
      "epoch": 3.22670958930836,
      "grad_norm": 1.0,
      "learning_rate": 0.0004019163177882779,
      "loss": 5.9465,
      "mean_token_accuracy": 0.16234225928783416,
      "num_tokens": 55266440.0,
      "step": 16540
    },
    {
      "entropy": 6.251637506484985,
      "epoch": 3.2276851038923033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000401858234893271,
      "loss": 5.8368,
      "mean_token_accuracy": 0.16408730894327164,
      "num_tokens": 55282708.0,
      "step": 16545
    },
    {
      "entropy": 6.000202894210815,
      "epoch": 3.2286606184762463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004018001396021296,
      "loss": 5.6324,
      "mean_token_accuracy": 0.1835852324962616,
      "num_tokens": 55299604.0,
      "step": 16550
    },
    {
      "entropy": 6.21516523361206,
      "epoch": 3.2296361330601893,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040174203192053047,
      "loss": 5.7583,
      "mean_token_accuracy": 0.17036513090133668,
      "num_tokens": 55316865.0,
      "step": 16555
    },
    {
      "entropy": 6.164998245239258,
      "epoch": 3.2306116476441322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004016839118541517,
      "loss": 5.6633,
      "mean_token_accuracy": 0.1762847736477852,
      "num_tokens": 55333855.0,
      "step": 16560
    },
    {
      "entropy": 6.167322158813477,
      "epoch": 3.231587162228075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004016257794086726,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16849977821111678,
      "num_tokens": 55350793.0,
      "step": 16565
    },
    {
      "entropy": 6.235674333572388,
      "epoch": 3.232562676812018,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040156763458977366,
      "loss": 5.7982,
      "mean_token_accuracy": 0.16675218045711518,
      "num_tokens": 55367367.0,
      "step": 16570
    },
    {
      "entropy": 6.270968008041382,
      "epoch": 3.2335381913959615,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004015094774031365,
      "loss": 5.9241,
      "mean_token_accuracy": 0.15426121205091475,
      "num_tokens": 55385931.0,
      "step": 16575
    },
    {
      "entropy": 6.224885606765747,
      "epoch": 3.2345137059799045,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004014513078544441,
      "loss": 5.8971,
      "mean_token_accuracy": 0.1646696299314499,
      "num_tokens": 55402540.0,
      "step": 16580
    },
    {
      "entropy": 6.095988130569458,
      "epoch": 3.2354892205638475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040139312594938053,
      "loss": 5.6739,
      "mean_token_accuracy": 0.1786152109503746,
      "num_tokens": 55420237.0,
      "step": 16585
    },
    {
      "entropy": 6.186000156402588,
      "epoch": 3.2364647351477904,
      "grad_norm": 1.0,
      "learning_rate": 0.00040133493169363125,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1656525433063507,
      "num_tokens": 55437274.0,
      "step": 16590
    },
    {
      "entropy": 6.279809474945068,
      "epoch": 3.2374402497317334,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040127672509288265,
      "loss": 5.9187,
      "mean_token_accuracy": 0.16225582361221313,
      "num_tokens": 55455129.0,
      "step": 16595
    },
    {
      "entropy": 6.148156785964966,
      "epoch": 3.2384157643156763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004012185061528226,
      "loss": 5.7959,
      "mean_token_accuracy": 0.1681055322289467,
      "num_tokens": 55471962.0,
      "step": 16600
    },
    {
      "entropy": 6.227354049682617,
      "epoch": 3.2393912788996198,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004011602748791399,
      "loss": 5.7805,
      "mean_token_accuracy": 0.16889325082302092,
      "num_tokens": 55490566.0,
      "step": 16605
    },
    {
      "entropy": 6.096278810501099,
      "epoch": 3.2403667934835627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040110203127752486,
      "loss": 5.755,
      "mean_token_accuracy": 0.17009971290826797,
      "num_tokens": 55507578.0,
      "step": 16610
    },
    {
      "entropy": 6.227713251113892,
      "epoch": 3.2413423080675057,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004010437753536687,
      "loss": 5.8195,
      "mean_token_accuracy": 0.1658927857875824,
      "num_tokens": 55523097.0,
      "step": 16615
    },
    {
      "entropy": 6.216927766799927,
      "epoch": 3.2423178226514486,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004009855071132641,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1647816702723503,
      "num_tokens": 55541304.0,
      "step": 16620
    },
    {
      "entropy": 6.265952777862549,
      "epoch": 3.2432933372353916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004009272265620047,
      "loss": 5.8782,
      "mean_token_accuracy": 0.1553978815674782,
      "num_tokens": 55557767.0,
      "step": 16625
    },
    {
      "entropy": 6.288684701919555,
      "epoch": 3.2442688518193346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040086893370558553,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1588510438799858,
      "num_tokens": 55573172.0,
      "step": 16630
    },
    {
      "entropy": 6.238116025924683,
      "epoch": 3.245244366403278,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040081062854970277,
      "loss": 5.8543,
      "mean_token_accuracy": 0.16024504899978637,
      "num_tokens": 55589921.0,
      "step": 16635
    },
    {
      "entropy": 6.240134191513062,
      "epoch": 3.246219880987221,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040075231110005385,
      "loss": 5.921,
      "mean_token_accuracy": 0.15424976646900176,
      "num_tokens": 55605647.0,
      "step": 16640
    },
    {
      "entropy": 6.137513208389282,
      "epoch": 3.247195395571164,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040069398136233724,
      "loss": 5.6969,
      "mean_token_accuracy": 0.17573120594024658,
      "num_tokens": 55623308.0,
      "step": 16645
    },
    {
      "entropy": 6.138785982131958,
      "epoch": 3.248170910155107,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004006356393422528,
      "loss": 5.6906,
      "mean_token_accuracy": 0.18201616853475572,
      "num_tokens": 55640245.0,
      "step": 16650
    },
    {
      "entropy": 6.2000689029693605,
      "epoch": 3.24914642473905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040057728504550136,
      "loss": 5.8071,
      "mean_token_accuracy": 0.16861167252063752,
      "num_tokens": 55657842.0,
      "step": 16655
    },
    {
      "entropy": 6.241994857788086,
      "epoch": 3.2501219393229928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004005189184777852,
      "loss": 5.7809,
      "mean_token_accuracy": 0.17051647305488588,
      "num_tokens": 55672580.0,
      "step": 16660
    },
    {
      "entropy": 6.215320491790772,
      "epoch": 3.2510974539069357,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004004605396448078,
      "loss": 5.7984,
      "mean_token_accuracy": 0.16376446932554245,
      "num_tokens": 55688806.0,
      "step": 16665
    },
    {
      "entropy": 6.241527795791626,
      "epoch": 3.252072968490879,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040040214855227346,
      "loss": 5.7942,
      "mean_token_accuracy": 0.17205523401498796,
      "num_tokens": 55704989.0,
      "step": 16670
    },
    {
      "entropy": 6.015704154968262,
      "epoch": 3.253048483074822,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004003437452058882,
      "loss": 5.6983,
      "mean_token_accuracy": 0.17922573238611222,
      "num_tokens": 55722448.0,
      "step": 16675
    },
    {
      "entropy": 6.275079202651978,
      "epoch": 3.254023997658765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004002853296113589,
      "loss": 5.8913,
      "mean_token_accuracy": 0.16363176256418227,
      "num_tokens": 55738862.0,
      "step": 16680
    },
    {
      "entropy": 6.136256122589112,
      "epoch": 3.254999512242708,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004002269017743937,
      "loss": 5.6822,
      "mean_token_accuracy": 0.17195298820734023,
      "num_tokens": 55754593.0,
      "step": 16685
    },
    {
      "entropy": 6.215645837783813,
      "epoch": 3.255975026826651,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040016846170070193,
      "loss": 5.8418,
      "mean_token_accuracy": 0.16621104180812835,
      "num_tokens": 55772039.0,
      "step": 16690
    },
    {
      "entropy": 6.27204647064209,
      "epoch": 3.256950541410594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004001100093959943,
      "loss": 5.8456,
      "mean_token_accuracy": 0.1666400194168091,
      "num_tokens": 55788977.0,
      "step": 16695
    },
    {
      "entropy": 6.2223358154296875,
      "epoch": 3.2579260559945373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004000515448659823,
      "loss": 5.8583,
      "mean_token_accuracy": 0.16171907633543015,
      "num_tokens": 55805997.0,
      "step": 16700
    },
    {
      "entropy": 6.160766458511352,
      "epoch": 3.2589015705784803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039999306811637915,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16496407240629196,
      "num_tokens": 55822108.0,
      "step": 16705
    },
    {
      "entropy": 6.301667833328247,
      "epoch": 3.2598770851624232,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003999345791528988,
      "loss": 5.9167,
      "mean_token_accuracy": 0.1572996959090233,
      "num_tokens": 55838840.0,
      "step": 16710
    },
    {
      "entropy": 6.2123517990112305,
      "epoch": 3.260852599746366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039987607798125667,
      "loss": 5.8258,
      "mean_token_accuracy": 0.16571723967790603,
      "num_tokens": 55855023.0,
      "step": 16715
    },
    {
      "entropy": 6.1946347713470455,
      "epoch": 3.261828114330309,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003998175646071693,
      "loss": 5.8232,
      "mean_token_accuracy": 0.16450294256210327,
      "num_tokens": 55870994.0,
      "step": 16720
    },
    {
      "entropy": 6.200986671447754,
      "epoch": 3.262803628914252,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039975903903635435,
      "loss": 5.7844,
      "mean_token_accuracy": 0.17239202111959456,
      "num_tokens": 55887816.0,
      "step": 16725
    },
    {
      "entropy": 6.1420385360717775,
      "epoch": 3.263779143498195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003997005012745307,
      "loss": 5.7186,
      "mean_token_accuracy": 0.1684591144323349,
      "num_tokens": 55905075.0,
      "step": 16730
    },
    {
      "entropy": 6.231030416488648,
      "epoch": 3.2647546580821385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039964195132741854,
      "loss": 5.8494,
      "mean_token_accuracy": 0.15886534303426741,
      "num_tokens": 55923240.0,
      "step": 16735
    },
    {
      "entropy": 6.198131132125854,
      "epoch": 3.2657301726660815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039958338920073914,
      "loss": 5.783,
      "mean_token_accuracy": 0.16971637606620787,
      "num_tokens": 55938854.0,
      "step": 16740
    },
    {
      "entropy": 6.186221122741699,
      "epoch": 3.2667056872500244,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039952481490021497,
      "loss": 5.8446,
      "mean_token_accuracy": 0.16709199100732802,
      "num_tokens": 55956748.0,
      "step": 16745
    },
    {
      "entropy": 6.210334205627442,
      "epoch": 3.2676812018339674,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039946622843156973,
      "loss": 5.7925,
      "mean_token_accuracy": 0.16852199137210847,
      "num_tokens": 55971773.0,
      "step": 16750
    },
    {
      "entropy": 6.133023881912232,
      "epoch": 3.2686567164179103,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003994076298005282,
      "loss": 5.7062,
      "mean_token_accuracy": 0.18014868795871736,
      "num_tokens": 55990160.0,
      "step": 16755
    },
    {
      "entropy": 6.192638540267945,
      "epoch": 3.2696322310018533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003993490190128165,
      "loss": 5.7517,
      "mean_token_accuracy": 0.17177090644836426,
      "num_tokens": 56007343.0,
      "step": 16760
    },
    {
      "entropy": 6.169049406051636,
      "epoch": 3.2706077455857967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003992903960741619,
      "loss": 5.7109,
      "mean_token_accuracy": 0.1786699652671814,
      "num_tokens": 56022882.0,
      "step": 16765
    },
    {
      "entropy": 6.180188083648682,
      "epoch": 3.2715832601697397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003992317609902927,
      "loss": 5.7689,
      "mean_token_accuracy": 0.16632729768753052,
      "num_tokens": 56039020.0,
      "step": 16770
    },
    {
      "entropy": 6.232550239562988,
      "epoch": 3.2725587747536826,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003991731137669386,
      "loss": 5.7829,
      "mean_token_accuracy": 0.1667168542742729,
      "num_tokens": 56054990.0,
      "step": 16775
    },
    {
      "entropy": 6.111131858825684,
      "epoch": 3.2735342893376256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003991144544098304,
      "loss": 5.728,
      "mean_token_accuracy": 0.16731897592544556,
      "num_tokens": 56071107.0,
      "step": 16780
    },
    {
      "entropy": 6.267436599731445,
      "epoch": 3.2745098039215685,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003990557829247,
      "loss": 5.8795,
      "mean_token_accuracy": 0.16164483577013017,
      "num_tokens": 56086275.0,
      "step": 16785
    },
    {
      "entropy": 6.09709358215332,
      "epoch": 3.2754853185055115,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003989970993172808,
      "loss": 5.7113,
      "mean_token_accuracy": 0.17741267681121825,
      "num_tokens": 56104010.0,
      "step": 16790
    },
    {
      "entropy": 6.198166561126709,
      "epoch": 3.2764608330894545,
      "grad_norm": 1.0,
      "learning_rate": 0.0003989384035933068,
      "loss": 5.8083,
      "mean_token_accuracy": 0.16752442717552185,
      "num_tokens": 56121145.0,
      "step": 16795
    },
    {
      "entropy": 6.0406153202056885,
      "epoch": 3.277436347673398,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003988796957585139,
      "loss": 5.622,
      "mean_token_accuracy": 0.18714051991701125,
      "num_tokens": 56138927.0,
      "step": 16800
    },
    {
      "entropy": 6.1508019924163815,
      "epoch": 3.278411862257341,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003988209758186386,
      "loss": 5.8081,
      "mean_token_accuracy": 0.16854393631219863,
      "num_tokens": 56157550.0,
      "step": 16805
    },
    {
      "entropy": 6.19583568572998,
      "epoch": 3.279387376841284,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003987622437794188,
      "loss": 5.8046,
      "mean_token_accuracy": 0.1639383465051651,
      "num_tokens": 56174011.0,
      "step": 16810
    },
    {
      "entropy": 6.1762806415557865,
      "epoch": 3.2803628914252267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003987034996465936,
      "loss": 5.7874,
      "mean_token_accuracy": 0.1663076788187027,
      "num_tokens": 56190630.0,
      "step": 16815
    },
    {
      "entropy": 6.30146050453186,
      "epoch": 3.2813384060091697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039864474342590333,
      "loss": 5.864,
      "mean_token_accuracy": 0.16946633756160737,
      "num_tokens": 56206397.0,
      "step": 16820
    },
    {
      "entropy": 6.080964279174805,
      "epoch": 3.282313920593113,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039858597512308945,
      "loss": 5.6033,
      "mean_token_accuracy": 0.18085575550794603,
      "num_tokens": 56223112.0,
      "step": 16825
    },
    {
      "entropy": 6.242026376724243,
      "epoch": 3.283289435177056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039852719474389454,
      "loss": 5.8598,
      "mean_token_accuracy": 0.1599734291434288,
      "num_tokens": 56238520.0,
      "step": 16830
    },
    {
      "entropy": 6.2013993740081785,
      "epoch": 3.284264949760999,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003984684022940624,
      "loss": 5.8244,
      "mean_token_accuracy": 0.16407504081726074,
      "num_tokens": 56256255.0,
      "step": 16835
    },
    {
      "entropy": 6.1731353282928465,
      "epoch": 3.285240464344942,
      "grad_norm": 1.0625,
      "learning_rate": 0.000398409597779338,
      "loss": 5.7657,
      "mean_token_accuracy": 0.16626926958560945,
      "num_tokens": 56271725.0,
      "step": 16840
    },
    {
      "entropy": 6.140598297119141,
      "epoch": 3.286215978928885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039835078120546756,
      "loss": 5.7843,
      "mean_token_accuracy": 0.16786590069532395,
      "num_tokens": 56287421.0,
      "step": 16845
    },
    {
      "entropy": 6.25272650718689,
      "epoch": 3.287191493512828,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003982919525781985,
      "loss": 5.873,
      "mean_token_accuracy": 0.16022708714008332,
      "num_tokens": 56304557.0,
      "step": 16850
    },
    {
      "entropy": 6.22704005241394,
      "epoch": 3.288167008096771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039823311190327913,
      "loss": 5.9087,
      "mean_token_accuracy": 0.16887876391410828,
      "num_tokens": 56323026.0,
      "step": 16855
    },
    {
      "entropy": 6.217720174789429,
      "epoch": 3.289142522680714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003981742591864594,
      "loss": 5.8004,
      "mean_token_accuracy": 0.16076219230890273,
      "num_tokens": 56340231.0,
      "step": 16860
    },
    {
      "entropy": 6.1592480659484865,
      "epoch": 3.2901180372646572,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039811539443349,
      "loss": 5.8098,
      "mean_token_accuracy": 0.16645825058221816,
      "num_tokens": 56357987.0,
      "step": 16865
    },
    {
      "entropy": 6.3083977699279785,
      "epoch": 3.2910935518486,
      "grad_norm": 1.03125,
      "learning_rate": 0.000398056517650123,
      "loss": 5.9393,
      "mean_token_accuracy": 0.16241712868213654,
      "num_tokens": 56374854.0,
      "step": 16870
    },
    {
      "entropy": 6.165575933456421,
      "epoch": 3.292069066432543,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003979976288421117,
      "loss": 5.7405,
      "mean_token_accuracy": 0.17377142906188964,
      "num_tokens": 56391620.0,
      "step": 16875
    },
    {
      "entropy": 6.113065385818482,
      "epoch": 3.293044581016486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003979387280152106,
      "loss": 5.7751,
      "mean_token_accuracy": 0.1682015061378479,
      "num_tokens": 56407892.0,
      "step": 16880
    },
    {
      "entropy": 6.275309896469116,
      "epoch": 3.294020095600429,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039787981517517505,
      "loss": 5.833,
      "mean_token_accuracy": 0.16175294518470765,
      "num_tokens": 56424005.0,
      "step": 16885
    },
    {
      "entropy": 6.235040998458862,
      "epoch": 3.2949956101843725,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003978208903277619,
      "loss": 5.7809,
      "mean_token_accuracy": 0.16370525062084199,
      "num_tokens": 56439936.0,
      "step": 16890
    },
    {
      "entropy": 6.1306956768035885,
      "epoch": 3.2959711247683154,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003977619534787292,
      "loss": 5.7564,
      "mean_token_accuracy": 0.16468770802021027,
      "num_tokens": 56456332.0,
      "step": 16895
    },
    {
      "entropy": 6.229488229751587,
      "epoch": 3.2969466393522584,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039770300463383575,
      "loss": 5.7934,
      "mean_token_accuracy": 0.16771951019763948,
      "num_tokens": 56472294.0,
      "step": 16900
    },
    {
      "entropy": 6.150830411911011,
      "epoch": 3.2979221539362014,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039764404379884216,
      "loss": 5.7275,
      "mean_token_accuracy": 0.17154242098331451,
      "num_tokens": 56489414.0,
      "step": 16905
    },
    {
      "entropy": 6.239172124862671,
      "epoch": 3.2988976685201443,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039758507097950965,
      "loss": 5.8998,
      "mean_token_accuracy": 0.15794903188943862,
      "num_tokens": 56506834.0,
      "step": 16910
    },
    {
      "entropy": 6.073769378662109,
      "epoch": 3.2998731831040873,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003975260861816009,
      "loss": 5.7061,
      "mean_token_accuracy": 0.17927711009979247,
      "num_tokens": 56523958.0,
      "step": 16915
    },
    {
      "entropy": 6.282330799102783,
      "epoch": 3.3008486976880302,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003974670894108798,
      "loss": 5.8213,
      "mean_token_accuracy": 0.16652171909809113,
      "num_tokens": 56539889.0,
      "step": 16920
    },
    {
      "entropy": 6.249163579940796,
      "epoch": 3.3018242122719736,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039740808067311107,
      "loss": 5.8779,
      "mean_token_accuracy": 0.16516658663749695,
      "num_tokens": 56556879.0,
      "step": 16925
    },
    {
      "entropy": 6.260849571228027,
      "epoch": 3.3027997268559166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000397349059974061,
      "loss": 5.9285,
      "mean_token_accuracy": 0.16372158229351044,
      "num_tokens": 56573077.0,
      "step": 16930
    },
    {
      "entropy": 6.282155513763428,
      "epoch": 3.3037752414398596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039729002731949694,
      "loss": 5.8425,
      "mean_token_accuracy": 0.1647841900587082,
      "num_tokens": 56589183.0,
      "step": 16935
    },
    {
      "entropy": 6.133101177215576,
      "epoch": 3.3047507560238025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039723098271518716,
      "loss": 5.7405,
      "mean_token_accuracy": 0.17355390042066574,
      "num_tokens": 56607060.0,
      "step": 16940
    },
    {
      "entropy": 6.203525352478027,
      "epoch": 3.3057262706077455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003971719261669014,
      "loss": 5.8498,
      "mean_token_accuracy": 0.16620658338069916,
      "num_tokens": 56623503.0,
      "step": 16945
    },
    {
      "entropy": 6.136212015151978,
      "epoch": 3.3067017851916884,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039711285768041045,
      "loss": 5.7174,
      "mean_token_accuracy": 0.17288971990346907,
      "num_tokens": 56639442.0,
      "step": 16950
    },
    {
      "entropy": 6.188206529617309,
      "epoch": 3.307677299775632,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003970537772614863,
      "loss": 5.8139,
      "mean_token_accuracy": 0.16283346563577653,
      "num_tokens": 56658451.0,
      "step": 16955
    },
    {
      "entropy": 6.22841215133667,
      "epoch": 3.308652814359575,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039699468491590206,
      "loss": 5.826,
      "mean_token_accuracy": 0.16128423511981965,
      "num_tokens": 56674826.0,
      "step": 16960
    },
    {
      "entropy": 6.166077280044556,
      "epoch": 3.3096283289435178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003969355806494319,
      "loss": 5.6707,
      "mean_token_accuracy": 0.17879000455141067,
      "num_tokens": 56690617.0,
      "step": 16965
    },
    {
      "entropy": 6.1155290603637695,
      "epoch": 3.3106038435274607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003968764644678515,
      "loss": 5.7586,
      "mean_token_accuracy": 0.16306644231081008,
      "num_tokens": 56707169.0,
      "step": 16970
    },
    {
      "entropy": 6.180872821807862,
      "epoch": 3.3115793581114037,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003968173363769373,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1698200598359108,
      "num_tokens": 56723448.0,
      "step": 16975
    },
    {
      "entropy": 6.167807340621948,
      "epoch": 3.3125548726953467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003967581963824672,
      "loss": 5.7379,
      "mean_token_accuracy": 0.17422424107789994,
      "num_tokens": 56740535.0,
      "step": 16980
    },
    {
      "entropy": 6.110744047164917,
      "epoch": 3.3135303872792896,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003966990444902201,
      "loss": 5.8012,
      "mean_token_accuracy": 0.1716802880167961,
      "num_tokens": 56758375.0,
      "step": 16985
    },
    {
      "entropy": 6.041568231582642,
      "epoch": 3.314505901863233,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039663988070597604,
      "loss": 5.5764,
      "mean_token_accuracy": 0.1806977093219757,
      "num_tokens": 56776090.0,
      "step": 16990
    },
    {
      "entropy": 6.326121807098389,
      "epoch": 3.315481416447176,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039658070503551646,
      "loss": 6.0127,
      "mean_token_accuracy": 0.15157385617494584,
      "num_tokens": 56792627.0,
      "step": 16995
    },
    {
      "entropy": 6.3314961910247805,
      "epoch": 3.316456931031119,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039652151748462367,
      "loss": 5.8704,
      "mean_token_accuracy": 0.15737409740686417,
      "num_tokens": 56809025.0,
      "step": 17000
    },
    {
      "entropy": 6.0861961364746096,
      "epoch": 3.317432445615062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003964623180590813,
      "loss": 5.6538,
      "mean_token_accuracy": 0.17308831810951233,
      "num_tokens": 56826043.0,
      "step": 17005
    },
    {
      "entropy": 6.328200578689575,
      "epoch": 3.318407960199005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003964031067646741,
      "loss": 5.892,
      "mean_token_accuracy": 0.15947714745998381,
      "num_tokens": 56841974.0,
      "step": 17010
    },
    {
      "entropy": 6.214485311508179,
      "epoch": 3.319383474782948,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039634388360718797,
      "loss": 5.8803,
      "mean_token_accuracy": 0.15619587600231172,
      "num_tokens": 56859404.0,
      "step": 17015
    },
    {
      "entropy": 6.247118997573852,
      "epoch": 3.3203589893668912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003962846485924101,
      "loss": 5.792,
      "mean_token_accuracy": 0.16704636216163635,
      "num_tokens": 56875727.0,
      "step": 17020
    },
    {
      "entropy": 6.165727043151856,
      "epoch": 3.321334503950834,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003962254017261286,
      "loss": 5.7439,
      "mean_token_accuracy": 0.17000497579574586,
      "num_tokens": 56891212.0,
      "step": 17025
    },
    {
      "entropy": 6.110712146759033,
      "epoch": 3.322310018534777,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039616614301413286,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1786120504140854,
      "num_tokens": 56908005.0,
      "step": 17030
    },
    {
      "entropy": 6.2068041324615475,
      "epoch": 3.32328553311872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003961068724622135,
      "loss": 5.73,
      "mean_token_accuracy": 0.17866628468036652,
      "num_tokens": 56924993.0,
      "step": 17035
    },
    {
      "entropy": 6.194933080673218,
      "epoch": 3.324261047702663,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003960475900761622,
      "loss": 5.856,
      "mean_token_accuracy": 0.16530465483665466,
      "num_tokens": 56941354.0,
      "step": 17040
    },
    {
      "entropy": 6.111437225341797,
      "epoch": 3.325236562286606,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039598829586177177,
      "loss": 5.6926,
      "mean_token_accuracy": 0.17617877572774887,
      "num_tokens": 56957908.0,
      "step": 17045
    },
    {
      "entropy": 6.275740909576416,
      "epoch": 3.326212076870549,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003959289898248363,
      "loss": 5.8516,
      "mean_token_accuracy": 0.1632503405213356,
      "num_tokens": 56973342.0,
      "step": 17050
    },
    {
      "entropy": 6.205459976196289,
      "epoch": 3.3271875914544924,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039586967197115095,
      "loss": 5.7752,
      "mean_token_accuracy": 0.16840512454509735,
      "num_tokens": 56990688.0,
      "step": 17055
    },
    {
      "entropy": 6.116314220428467,
      "epoch": 3.3281631060384353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003958103423065121,
      "loss": 5.7836,
      "mean_token_accuracy": 0.1730041027069092,
      "num_tokens": 57007516.0,
      "step": 17060
    },
    {
      "entropy": 6.103476858139038,
      "epoch": 3.3291386206223783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003957510008367171,
      "loss": 5.7065,
      "mean_token_accuracy": 0.17103757411241532,
      "num_tokens": 57023771.0,
      "step": 17065
    },
    {
      "entropy": 6.263871574401856,
      "epoch": 3.3301141352063213,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003956916475675647,
      "loss": 5.8306,
      "mean_token_accuracy": 0.16709998100996018,
      "num_tokens": 57040018.0,
      "step": 17070
    },
    {
      "entropy": 6.097398471832276,
      "epoch": 3.3310896497902642,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003956322825048546,
      "loss": 5.7333,
      "mean_token_accuracy": 0.17480166107416154,
      "num_tokens": 57055385.0,
      "step": 17075
    },
    {
      "entropy": 6.268301725387573,
      "epoch": 3.3320651643742076,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039557290565438787,
      "loss": 5.8302,
      "mean_token_accuracy": 0.16375699937343596,
      "num_tokens": 57072172.0,
      "step": 17080
    },
    {
      "entropy": 6.190821886062622,
      "epoch": 3.3330406789581506,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003955135170219665,
      "loss": 5.8069,
      "mean_token_accuracy": 0.1701032117009163,
      "num_tokens": 57088944.0,
      "step": 17085
    },
    {
      "entropy": 6.244196987152099,
      "epoch": 3.3340161935420936,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003954541166133937,
      "loss": 5.8604,
      "mean_token_accuracy": 0.1612774044275284,
      "num_tokens": 57106284.0,
      "step": 17090
    },
    {
      "entropy": 6.15723819732666,
      "epoch": 3.3349917081260365,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000395394704434474,
      "loss": 5.727,
      "mean_token_accuracy": 0.17432261854410172,
      "num_tokens": 57123507.0,
      "step": 17095
    },
    {
      "entropy": 6.167015409469604,
      "epoch": 3.3359672227099795,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003953352804910129,
      "loss": 5.7755,
      "mean_token_accuracy": 0.17442060708999635,
      "num_tokens": 57140944.0,
      "step": 17100
    },
    {
      "entropy": 6.221920776367187,
      "epoch": 3.3369427372939224,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000395275844788817,
      "loss": 5.8238,
      "mean_token_accuracy": 0.16936539560556413,
      "num_tokens": 57156582.0,
      "step": 17105
    },
    {
      "entropy": 6.181091547012329,
      "epoch": 3.3379182518778654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003952163973336943,
      "loss": 5.7899,
      "mean_token_accuracy": 0.16649939715862275,
      "num_tokens": 57172230.0,
      "step": 17110
    },
    {
      "entropy": 6.2473376274108885,
      "epoch": 3.3388937664618084,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003951569381314537,
      "loss": 5.8607,
      "mean_token_accuracy": 0.16160990446805953,
      "num_tokens": 57189176.0,
      "step": 17115
    },
    {
      "entropy": 6.156323671340942,
      "epoch": 3.3398692810457518,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003950974671879053,
      "loss": 5.6434,
      "mean_token_accuracy": 0.18203024566173553,
      "num_tokens": 57205802.0,
      "step": 17120
    },
    {
      "entropy": 6.1265300750732425,
      "epoch": 3.3408447956296947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003950379845088605,
      "loss": 5.6822,
      "mean_token_accuracy": 0.17076052129268646,
      "num_tokens": 57222224.0,
      "step": 17125
    },
    {
      "entropy": 6.102639865875244,
      "epoch": 3.3418203102136377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003949784901001315,
      "loss": 5.7478,
      "mean_token_accuracy": 0.16781802773475646,
      "num_tokens": 57236920.0,
      "step": 17130
    },
    {
      "entropy": 6.158773756027221,
      "epoch": 3.3427958247975806,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039491898396753225,
      "loss": 5.7506,
      "mean_token_accuracy": 0.17748329043388367,
      "num_tokens": 57254076.0,
      "step": 17135
    },
    {
      "entropy": 6.040859603881836,
      "epoch": 3.3437713393815236,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039485946611687725,
      "loss": 5.6267,
      "mean_token_accuracy": 0.18260992765426637,
      "num_tokens": 57269932.0,
      "step": 17140
    },
    {
      "entropy": 6.232811403274536,
      "epoch": 3.344746853965467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039479993655398236,
      "loss": 5.9062,
      "mean_token_accuracy": 0.1543854646384716,
      "num_tokens": 57286571.0,
      "step": 17145
    },
    {
      "entropy": 6.240063905715942,
      "epoch": 3.34572236854941,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003947403952846647,
      "loss": 5.7483,
      "mean_token_accuracy": 0.17909324467182158,
      "num_tokens": 57304076.0,
      "step": 17150
    },
    {
      "entropy": 6.224098491668701,
      "epoch": 3.346697883133353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003946808423147423,
      "loss": 5.8058,
      "mean_token_accuracy": 0.16894888281822204,
      "num_tokens": 57322032.0,
      "step": 17155
    },
    {
      "entropy": 6.2198995590209964,
      "epoch": 3.347673397717296,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003946212776500346,
      "loss": 5.8854,
      "mean_token_accuracy": 0.16874572336673738,
      "num_tokens": 57338101.0,
      "step": 17160
    },
    {
      "entropy": 6.2917992115020756,
      "epoch": 3.348648912301239,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000394561701296362,
      "loss": 5.8868,
      "mean_token_accuracy": 0.163589408993721,
      "num_tokens": 57355964.0,
      "step": 17165
    },
    {
      "entropy": 6.202717399597168,
      "epoch": 3.349624426885182,
      "grad_norm": 1.15625,
      "learning_rate": 0.000394502113259546,
      "loss": 5.8109,
      "mean_token_accuracy": 0.16663880944252013,
      "num_tokens": 57372645.0,
      "step": 17170
    },
    {
      "entropy": 6.098342514038086,
      "epoch": 3.3505999414691248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003944425135454095,
      "loss": 5.652,
      "mean_token_accuracy": 0.17964135855436325,
      "num_tokens": 57388415.0,
      "step": 17175
    },
    {
      "entropy": 6.202218294143677,
      "epoch": 3.351575456053068,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003943829021597762,
      "loss": 5.8427,
      "mean_token_accuracy": 0.16425636112689973,
      "num_tokens": 57404027.0,
      "step": 17180
    },
    {
      "entropy": 6.211448097229004,
      "epoch": 3.352550970637011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003943232791084712,
      "loss": 5.8457,
      "mean_token_accuracy": 0.16818261742591858,
      "num_tokens": 57421229.0,
      "step": 17185
    },
    {
      "entropy": 6.214021921157837,
      "epoch": 3.353526485220954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003942636443973207,
      "loss": 5.8203,
      "mean_token_accuracy": 0.16104209423065186,
      "num_tokens": 57436151.0,
      "step": 17190
    },
    {
      "entropy": 6.162757396697998,
      "epoch": 3.354501999804897,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003942039980321519,
      "loss": 5.7442,
      "mean_token_accuracy": 0.18094334304332732,
      "num_tokens": 57452832.0,
      "step": 17195
    },
    {
      "entropy": 6.033816289901734,
      "epoch": 3.35547751438884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039414434001879326,
      "loss": 5.6126,
      "mean_token_accuracy": 0.18607808351516725,
      "num_tokens": 57470926.0,
      "step": 17200
    },
    {
      "entropy": 6.191651678085327,
      "epoch": 3.356453028972783,
      "grad_norm": 1.0,
      "learning_rate": 0.00039408467036307435,
      "loss": 5.7946,
      "mean_token_accuracy": 0.16495394855737686,
      "num_tokens": 57489120.0,
      "step": 17205
    },
    {
      "entropy": 6.225435447692871,
      "epoch": 3.3574285435567264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039402498907082587,
      "loss": 5.7991,
      "mean_token_accuracy": 0.1720167264342308,
      "num_tokens": 57505553.0,
      "step": 17210
    },
    {
      "entropy": 6.170841312408447,
      "epoch": 3.3584040581406693,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039396529614787963,
      "loss": 5.7686,
      "mean_token_accuracy": 0.16867985278367997,
      "num_tokens": 57521827.0,
      "step": 17215
    },
    {
      "entropy": 6.157406949996949,
      "epoch": 3.3593795727246123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039390559160006874,
      "loss": 5.7784,
      "mean_token_accuracy": 0.17256633192300797,
      "num_tokens": 57537221.0,
      "step": 17220
    },
    {
      "entropy": 6.119461536407471,
      "epoch": 3.3603550873085553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003938458754332272,
      "loss": 5.7203,
      "mean_token_accuracy": 0.17937441319227218,
      "num_tokens": 57555468.0,
      "step": 17225
    },
    {
      "entropy": 6.175655746459961,
      "epoch": 3.361330601892498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003937861476531902,
      "loss": 5.7825,
      "mean_token_accuracy": 0.16885608285665513,
      "num_tokens": 57573616.0,
      "step": 17230
    },
    {
      "entropy": 6.254832983016968,
      "epoch": 3.362306116476441,
      "grad_norm": 1.0,
      "learning_rate": 0.00039372640826579433,
      "loss": 5.7948,
      "mean_token_accuracy": 0.16996069699525834,
      "num_tokens": 57591393.0,
      "step": 17235
    },
    {
      "entropy": 6.220024633407593,
      "epoch": 3.363281631060384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039366665727687695,
      "loss": 5.8269,
      "mean_token_accuracy": 0.16245007067918776,
      "num_tokens": 57607149.0,
      "step": 17240
    },
    {
      "entropy": 6.208581924438477,
      "epoch": 3.3642571456443275,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003936068946922767,
      "loss": 5.8474,
      "mean_token_accuracy": 0.16453957110643386,
      "num_tokens": 57623344.0,
      "step": 17245
    },
    {
      "entropy": 6.160965871810913,
      "epoch": 3.3652326602282705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039354712051783346,
      "loss": 5.7847,
      "mean_token_accuracy": 0.16970707178115846,
      "num_tokens": 57640190.0,
      "step": 17250
    },
    {
      "entropy": 6.20426459312439,
      "epoch": 3.3662081748122135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003934873347593882,
      "loss": 5.7668,
      "mean_token_accuracy": 0.16719861328601837,
      "num_tokens": 57657205.0,
      "step": 17255
    },
    {
      "entropy": 6.192591619491577,
      "epoch": 3.3671836893961564,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003934275374227828,
      "loss": 5.8055,
      "mean_token_accuracy": 0.16623726189136506,
      "num_tokens": 57673241.0,
      "step": 17260
    },
    {
      "entropy": 6.069849491119385,
      "epoch": 3.3681592039800994,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003933677285138605,
      "loss": 5.6625,
      "mean_token_accuracy": 0.18858612179756165,
      "num_tokens": 57691316.0,
      "step": 17265
    },
    {
      "entropy": 6.192935800552368,
      "epoch": 3.3691347185640423,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039330790803846573,
      "loss": 5.8353,
      "mean_token_accuracy": 0.16353553980588914,
      "num_tokens": 57708216.0,
      "step": 17270
    },
    {
      "entropy": 6.21717677116394,
      "epoch": 3.3701102331479857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003932480760024437,
      "loss": 5.8231,
      "mean_token_accuracy": 0.16143232733011245,
      "num_tokens": 57725567.0,
      "step": 17275
    },
    {
      "entropy": 6.199190378189087,
      "epoch": 3.3710857477319287,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039318823241164125,
      "loss": 5.7396,
      "mean_token_accuracy": 0.17333099544048308,
      "num_tokens": 57741984.0,
      "step": 17280
    },
    {
      "entropy": 6.223867893218994,
      "epoch": 3.3720612623158717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003931283772719059,
      "loss": 5.8473,
      "mean_token_accuracy": 0.16648814976215362,
      "num_tokens": 57758532.0,
      "step": 17285
    },
    {
      "entropy": 6.135135221481323,
      "epoch": 3.3730367768998146,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003930685105890866,
      "loss": 5.8114,
      "mean_token_accuracy": 0.17125010192394258,
      "num_tokens": 57776602.0,
      "step": 17290
    },
    {
      "entropy": 6.128791475296021,
      "epoch": 3.3740122914837576,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00039300863236903314,
      "loss": 5.6887,
      "mean_token_accuracy": 0.17285167574882507,
      "num_tokens": 57795269.0,
      "step": 17295
    },
    {
      "entropy": 6.101983594894409,
      "epoch": 3.3749878060677005,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003929487426175968,
      "loss": 5.7074,
      "mean_token_accuracy": 0.17895589172840118,
      "num_tokens": 57813301.0,
      "step": 17300
    },
    {
      "entropy": 6.147219467163086,
      "epoch": 3.3759633206516435,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039288884134062974,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1694385826587677,
      "num_tokens": 57830314.0,
      "step": 17305
    },
    {
      "entropy": 6.240018320083618,
      "epoch": 3.376938835235587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003928289285439852,
      "loss": 5.778,
      "mean_token_accuracy": 0.17194385081529617,
      "num_tokens": 57847309.0,
      "step": 17310
    },
    {
      "entropy": 6.2558189868927006,
      "epoch": 3.37791434981953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003927690042335178,
      "loss": 5.8503,
      "mean_token_accuracy": 0.162647108733654,
      "num_tokens": 57863050.0,
      "step": 17315
    },
    {
      "entropy": 6.182820224761963,
      "epoch": 3.378889864403473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000392709068415083,
      "loss": 5.8636,
      "mean_token_accuracy": 0.16323252171278,
      "num_tokens": 57880469.0,
      "step": 17320
    },
    {
      "entropy": 6.2592000484466555,
      "epoch": 3.379865378987416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039264912109453746,
      "loss": 5.8531,
      "mean_token_accuracy": 0.16349887400865554,
      "num_tokens": 57898090.0,
      "step": 17325
    },
    {
      "entropy": 6.2235370635986325,
      "epoch": 3.3808408935713588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003925891622777392,
      "loss": 5.758,
      "mean_token_accuracy": 0.17109051793813707,
      "num_tokens": 57914343.0,
      "step": 17330
    },
    {
      "entropy": 6.253592920303345,
      "epoch": 3.381816408155302,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003925291919705472,
      "loss": 5.853,
      "mean_token_accuracy": 0.15941972583532332,
      "num_tokens": 57930471.0,
      "step": 17335
    },
    {
      "entropy": 5.988532638549804,
      "epoch": 3.382791922739245,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039246921017882123,
      "loss": 5.5998,
      "mean_token_accuracy": 0.18521798402071,
      "num_tokens": 57947766.0,
      "step": 17340
    },
    {
      "entropy": 6.285721254348755,
      "epoch": 3.383767437323188,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003924092169084228,
      "loss": 5.9244,
      "mean_token_accuracy": 0.16200951337814332,
      "num_tokens": 57964205.0,
      "step": 17345
    },
    {
      "entropy": 6.232707262039185,
      "epoch": 3.384742951907131,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039234921216521417,
      "loss": 5.8628,
      "mean_token_accuracy": 0.16674195677042009,
      "num_tokens": 57982307.0,
      "step": 17350
    },
    {
      "entropy": 6.1864869594573975,
      "epoch": 3.385718466491074,
      "grad_norm": 1.0,
      "learning_rate": 0.00039228919595505875,
      "loss": 5.771,
      "mean_token_accuracy": 0.17079897075891495,
      "num_tokens": 57999302.0,
      "step": 17355
    },
    {
      "entropy": 6.190733432769775,
      "epoch": 3.386693981075017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003922291682838211,
      "loss": 5.7522,
      "mean_token_accuracy": 0.17222905158996582,
      "num_tokens": 58016578.0,
      "step": 17360
    },
    {
      "entropy": 6.1763170719146725,
      "epoch": 3.38766949565896,
      "grad_norm": 1.015625,
      "learning_rate": 0.000392169129157367,
      "loss": 5.7602,
      "mean_token_accuracy": 0.16745971739292145,
      "num_tokens": 58033770.0,
      "step": 17365
    },
    {
      "entropy": 6.087692928314209,
      "epoch": 3.3886450102429033,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003921090785815632,
      "loss": 5.67,
      "mean_token_accuracy": 0.1727004587650299,
      "num_tokens": 58050075.0,
      "step": 17370
    },
    {
      "entropy": 6.250836515426636,
      "epoch": 3.3896205248268463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003920490165622775,
      "loss": 5.8578,
      "mean_token_accuracy": 0.16365945488214492,
      "num_tokens": 58067020.0,
      "step": 17375
    },
    {
      "entropy": 6.225775957107544,
      "epoch": 3.3905960394107892,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039198894310537923,
      "loss": 5.7898,
      "mean_token_accuracy": 0.16803980022668838,
      "num_tokens": 58081922.0,
      "step": 17380
    },
    {
      "entropy": 6.187716722488403,
      "epoch": 3.391571553994732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003919288582167383,
      "loss": 5.7771,
      "mean_token_accuracy": 0.16972967386245727,
      "num_tokens": 58097929.0,
      "step": 17385
    },
    {
      "entropy": 6.32011308670044,
      "epoch": 3.392547068578675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039186876190222607,
      "loss": 5.8634,
      "mean_token_accuracy": 0.16467693895101548,
      "num_tokens": 58115299.0,
      "step": 17390
    },
    {
      "entropy": 6.206123399734497,
      "epoch": 3.393522583162618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039180865416771506,
      "loss": 5.7486,
      "mean_token_accuracy": 0.17502067238092422,
      "num_tokens": 58131333.0,
      "step": 17395
    },
    {
      "entropy": 6.106006193161011,
      "epoch": 3.3944980977465615,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003917485350190786,
      "loss": 5.7455,
      "mean_token_accuracy": 0.17270562797784805,
      "num_tokens": 58147556.0,
      "step": 17400
    },
    {
      "entropy": 6.197161769866943,
      "epoch": 3.3954736123305045,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039168840446219145,
      "loss": 5.8576,
      "mean_token_accuracy": 0.16344637870788575,
      "num_tokens": 58163731.0,
      "step": 17405
    },
    {
      "entropy": 6.2439727783203125,
      "epoch": 3.3964491269144474,
      "grad_norm": 1.0,
      "learning_rate": 0.00039162826250292926,
      "loss": 5.8328,
      "mean_token_accuracy": 0.16448771208524704,
      "num_tokens": 58180842.0,
      "step": 17410
    },
    {
      "entropy": 6.1581751823425295,
      "epoch": 3.3974246414983904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039156810914716897,
      "loss": 5.7904,
      "mean_token_accuracy": 0.1697973355650902,
      "num_tokens": 58198064.0,
      "step": 17415
    },
    {
      "entropy": 6.227036714553833,
      "epoch": 3.3984001560823334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039150794440078845,
      "loss": 5.7813,
      "mean_token_accuracy": 0.16811074912548066,
      "num_tokens": 58214233.0,
      "step": 17420
    },
    {
      "entropy": 6.187510204315186,
      "epoch": 3.3993756706662763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003914477682696668,
      "loss": 5.7645,
      "mean_token_accuracy": 0.1729908064007759,
      "num_tokens": 58231940.0,
      "step": 17425
    },
    {
      "entropy": 6.220187711715698,
      "epoch": 3.4003511852502193,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003913875807596844,
      "loss": 5.8044,
      "mean_token_accuracy": 0.16844303756952286,
      "num_tokens": 58248076.0,
      "step": 17430
    },
    {
      "entropy": 6.189469242095948,
      "epoch": 3.4013266998341627,
      "grad_norm": 1.0,
      "learning_rate": 0.0003913273818767223,
      "loss": 5.837,
      "mean_token_accuracy": 0.16269300878047943,
      "num_tokens": 58264548.0,
      "step": 17435
    },
    {
      "entropy": 6.178514671325684,
      "epoch": 3.4023022144181057,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039126717162666307,
      "loss": 5.7689,
      "mean_token_accuracy": 0.16825851649045945,
      "num_tokens": 58280177.0,
      "step": 17440
    },
    {
      "entropy": 6.197010564804077,
      "epoch": 3.4032777290020486,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039120695001539026,
      "loss": 5.7849,
      "mean_token_accuracy": 0.17214158922433853,
      "num_tokens": 58297083.0,
      "step": 17445
    },
    {
      "entropy": 6.235075044631958,
      "epoch": 3.4042532435859916,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039114671704878844,
      "loss": 5.839,
      "mean_token_accuracy": 0.16803935766220093,
      "num_tokens": 58313098.0,
      "step": 17450
    },
    {
      "entropy": 6.11886773109436,
      "epoch": 3.4052287581699345,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003910864727327434,
      "loss": 5.7244,
      "mean_token_accuracy": 0.17666105329990386,
      "num_tokens": 58329618.0,
      "step": 17455
    },
    {
      "entropy": 6.192620134353637,
      "epoch": 3.4062042727538775,
      "grad_norm": 1.0,
      "learning_rate": 0.00039102621707314195,
      "loss": 5.7076,
      "mean_token_accuracy": 0.17375415861606597,
      "num_tokens": 58347133.0,
      "step": 17460
    },
    {
      "entropy": 6.076377820968628,
      "epoch": 3.407179787337821,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003909659500758721,
      "loss": 5.741,
      "mean_token_accuracy": 0.17114297598600386,
      "num_tokens": 58363550.0,
      "step": 17465
    },
    {
      "entropy": 6.164313268661499,
      "epoch": 3.408155301921764,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000390905671746823,
      "loss": 5.712,
      "mean_token_accuracy": 0.1768069326877594,
      "num_tokens": 58379437.0,
      "step": 17470
    },
    {
      "entropy": 6.211909770965576,
      "epoch": 3.409130816505707,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003908453820918846,
      "loss": 5.8506,
      "mean_token_accuracy": 0.1661149486899376,
      "num_tokens": 58396459.0,
      "step": 17475
    },
    {
      "entropy": 6.1542637825012205,
      "epoch": 3.41010633108965,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039078508111694844,
      "loss": 5.7304,
      "mean_token_accuracy": 0.17524790465831758,
      "num_tokens": 58411926.0,
      "step": 17480
    },
    {
      "entropy": 6.177354431152343,
      "epoch": 3.4110818456735927,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003907247688279068,
      "loss": 5.8428,
      "mean_token_accuracy": 0.16375600248575212,
      "num_tokens": 58428024.0,
      "step": 17485
    },
    {
      "entropy": 6.162237882614136,
      "epoch": 3.4120573602575357,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003906644452306532,
      "loss": 5.782,
      "mean_token_accuracy": 0.17345110923051835,
      "num_tokens": 58446112.0,
      "step": 17490
    },
    {
      "entropy": 6.196967506408692,
      "epoch": 3.4130328748414787,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039060411033108223,
      "loss": 5.7798,
      "mean_token_accuracy": 0.1740495577454567,
      "num_tokens": 58462903.0,
      "step": 17495
    },
    {
      "entropy": 6.168573522567749,
      "epoch": 3.414008389425422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003905437641350897,
      "loss": 5.7203,
      "mean_token_accuracy": 0.17068171948194505,
      "num_tokens": 58478257.0,
      "step": 17500
    },
    {
      "entropy": 6.196128988265992,
      "epoch": 3.414983904009365,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039048340664857225,
      "loss": 5.7538,
      "mean_token_accuracy": 0.1770758017897606,
      "num_tokens": 58493742.0,
      "step": 17505
    },
    {
      "entropy": 5.8940449237823485,
      "epoch": 3.415959418593308,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003904230378774279,
      "loss": 5.5542,
      "mean_token_accuracy": 0.18784513026475907,
      "num_tokens": 58512808.0,
      "step": 17510
    },
    {
      "entropy": 6.264538383483886,
      "epoch": 3.416934933177251,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003903626578275557,
      "loss": 5.8281,
      "mean_token_accuracy": 0.1682042211294174,
      "num_tokens": 58530001.0,
      "step": 17515
    },
    {
      "entropy": 6.177198028564453,
      "epoch": 3.417910447761194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039030226650485576,
      "loss": 5.7304,
      "mean_token_accuracy": 0.17866278290748597,
      "num_tokens": 58546710.0,
      "step": 17520
    },
    {
      "entropy": 6.081159257888794,
      "epoch": 3.4188859623451373,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003902418639152293,
      "loss": 5.7213,
      "mean_token_accuracy": 0.17834618985652922,
      "num_tokens": 58563931.0,
      "step": 17525
    },
    {
      "entropy": 6.221897554397583,
      "epoch": 3.4198614769290803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003901814500645785,
      "loss": 5.8062,
      "mean_token_accuracy": 0.16645080149173735,
      "num_tokens": 58579228.0,
      "step": 17530
    },
    {
      "entropy": 6.214130210876465,
      "epoch": 3.4208369915130232,
      "grad_norm": 1.0625,
      "learning_rate": 0.000390121024958807,
      "loss": 5.8399,
      "mean_token_accuracy": 0.16121381968259813,
      "num_tokens": 58595177.0,
      "step": 17535
    },
    {
      "entropy": 6.093438529968262,
      "epoch": 3.421812506096966,
      "grad_norm": 1.125,
      "learning_rate": 0.0003900605886038192,
      "loss": 5.6728,
      "mean_token_accuracy": 0.1834900677204132,
      "num_tokens": 58612465.0,
      "step": 17540
    },
    {
      "entropy": 6.120640563964844,
      "epoch": 3.422788020680909,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039000014100552084,
      "loss": 5.7362,
      "mean_token_accuracy": 0.17900001108646393,
      "num_tokens": 58628696.0,
      "step": 17545
    },
    {
      "entropy": 6.1867430210113525,
      "epoch": 3.423763535264852,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038993968216981847,
      "loss": 5.766,
      "mean_token_accuracy": 0.17094120234251023,
      "num_tokens": 58645972.0,
      "step": 17550
    },
    {
      "entropy": 6.1800856590271,
      "epoch": 3.424739049848795,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038987921210262007,
      "loss": 5.748,
      "mean_token_accuracy": 0.17098635137081147,
      "num_tokens": 58662453.0,
      "step": 17555
    },
    {
      "entropy": 5.9814245223999025,
      "epoch": 3.425714564432738,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003898187308098345,
      "loss": 5.5254,
      "mean_token_accuracy": 0.19754383862018585,
      "num_tokens": 58680046.0,
      "step": 17560
    },
    {
      "entropy": 6.27193660736084,
      "epoch": 3.4266900790166814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003897582382973717,
      "loss": 5.8884,
      "mean_token_accuracy": 0.16085583716630936,
      "num_tokens": 58696851.0,
      "step": 17565
    },
    {
      "entropy": 6.112611865997314,
      "epoch": 3.4276655936006244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038969773457114295,
      "loss": 5.7604,
      "mean_token_accuracy": 0.16985290199518205,
      "num_tokens": 58713095.0,
      "step": 17570
    },
    {
      "entropy": 6.079631662368774,
      "epoch": 3.4286411081845674,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038963721963706035,
      "loss": 5.6773,
      "mean_token_accuracy": 0.18100662231445314,
      "num_tokens": 58729997.0,
      "step": 17575
    },
    {
      "entropy": 6.145488452911377,
      "epoch": 3.4296166227685103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003895766935010372,
      "loss": 5.7915,
      "mean_token_accuracy": 0.1751919150352478,
      "num_tokens": 58749063.0,
      "step": 17580
    },
    {
      "entropy": 6.307219648361206,
      "epoch": 3.4305921373524533,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038951615616898797,
      "loss": 5.8344,
      "mean_token_accuracy": 0.16172501891851426,
      "num_tokens": 58765158.0,
      "step": 17585
    },
    {
      "entropy": 6.165906810760498,
      "epoch": 3.4315676519363967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038945560764682807,
      "loss": 5.7688,
      "mean_token_accuracy": 0.17110703587532045,
      "num_tokens": 58781523.0,
      "step": 17590
    },
    {
      "entropy": 6.13886194229126,
      "epoch": 3.4325431665203396,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038939504794047414,
      "loss": 5.6833,
      "mean_token_accuracy": 0.1773821160197258,
      "num_tokens": 58798436.0,
      "step": 17595
    },
    {
      "entropy": 6.258548927307129,
      "epoch": 3.4335186811042826,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003893344770558438,
      "loss": 5.8733,
      "mean_token_accuracy": 0.16334510147571563,
      "num_tokens": 58814839.0,
      "step": 17600
    },
    {
      "entropy": 6.250882005691528,
      "epoch": 3.4344941956882256,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038927389499885583,
      "loss": 5.858,
      "mean_token_accuracy": 0.16432573199272155,
      "num_tokens": 58831752.0,
      "step": 17605
    },
    {
      "entropy": 6.185084104537964,
      "epoch": 3.4354697102721685,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003892133017754302,
      "loss": 5.7725,
      "mean_token_accuracy": 0.1782112494111061,
      "num_tokens": 58848975.0,
      "step": 17610
    },
    {
      "entropy": 6.20664963722229,
      "epoch": 3.4364452248561115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038915269739148774,
      "loss": 5.8343,
      "mean_token_accuracy": 0.16140740513801574,
      "num_tokens": 58864413.0,
      "step": 17615
    },
    {
      "entropy": 6.052206325531006,
      "epoch": 3.4374207394400544,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003890920818529506,
      "loss": 5.5824,
      "mean_token_accuracy": 0.1906673640012741,
      "num_tokens": 58881151.0,
      "step": 17620
    },
    {
      "entropy": 6.188439464569091,
      "epoch": 3.438396254023998,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003890314551657417,
      "loss": 5.7102,
      "mean_token_accuracy": 0.18009480983018875,
      "num_tokens": 58899699.0,
      "step": 17625
    },
    {
      "entropy": 6.148987245559693,
      "epoch": 3.439371768607941,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003889708173357856,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1729016661643982,
      "num_tokens": 58915193.0,
      "step": 17630
    },
    {
      "entropy": 6.087255382537842,
      "epoch": 3.4403472831918838,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003889101683690073,
      "loss": 5.6569,
      "mean_token_accuracy": 0.18466433584690095,
      "num_tokens": 58931385.0,
      "step": 17635
    },
    {
      "entropy": 6.199092245101928,
      "epoch": 3.4413227977758267,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038884950827133336,
      "loss": 5.843,
      "mean_token_accuracy": 0.1676139861345291,
      "num_tokens": 58946498.0,
      "step": 17640
    },
    {
      "entropy": 6.255775022506714,
      "epoch": 3.4422983123597697,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038878883704869126,
      "loss": 5.8007,
      "mean_token_accuracy": 0.17553744167089463,
      "num_tokens": 58964885.0,
      "step": 17645
    },
    {
      "entropy": 6.214127922058106,
      "epoch": 3.4432738269437126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003887281547070095,
      "loss": 5.7922,
      "mean_token_accuracy": 0.17280245423316956,
      "num_tokens": 58981866.0,
      "step": 17650
    },
    {
      "entropy": 6.107967853546143,
      "epoch": 3.444249341527656,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003886674612522179,
      "loss": 5.7519,
      "mean_token_accuracy": 0.17905376404523848,
      "num_tokens": 58999688.0,
      "step": 17655
    },
    {
      "entropy": 6.138291883468628,
      "epoch": 3.445224856111599,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038860675669024703,
      "loss": 5.6736,
      "mean_token_accuracy": 0.18582030683755874,
      "num_tokens": 59016385.0,
      "step": 17660
    },
    {
      "entropy": 6.214352130889893,
      "epoch": 3.446200370695542,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003885460410270288,
      "loss": 5.8161,
      "mean_token_accuracy": 0.1682658776640892,
      "num_tokens": 59031825.0,
      "step": 17665
    },
    {
      "entropy": 6.228054141998291,
      "epoch": 3.447175885279485,
      "grad_norm": 1.125,
      "learning_rate": 0.0003884853142684961,
      "loss": 5.8567,
      "mean_token_accuracy": 0.1597047731280327,
      "num_tokens": 59048310.0,
      "step": 17670
    },
    {
      "entropy": 6.128956413269043,
      "epoch": 3.448151399863428,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038842457642058304,
      "loss": 5.7049,
      "mean_token_accuracy": 0.17633059471845627,
      "num_tokens": 59066440.0,
      "step": 17675
    },
    {
      "entropy": 6.1378871440887455,
      "epoch": 3.449126914447371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038836382748922457,
      "loss": 5.6989,
      "mean_token_accuracy": 0.18186635673046112,
      "num_tokens": 59083904.0,
      "step": 17680
    },
    {
      "entropy": 6.123879051208496,
      "epoch": 3.450102429031314,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038830306748035686,
      "loss": 5.6808,
      "mean_token_accuracy": 0.1814774051308632,
      "num_tokens": 59099991.0,
      "step": 17685
    },
    {
      "entropy": 6.164350271224976,
      "epoch": 3.451077943615257,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003882422963999173,
      "loss": 5.7975,
      "mean_token_accuracy": 0.16798274517059325,
      "num_tokens": 59116770.0,
      "step": 17690
    },
    {
      "entropy": 6.259416341781616,
      "epoch": 3.4520534581992,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003881815142538442,
      "loss": 5.8299,
      "mean_token_accuracy": 0.17037690430879593,
      "num_tokens": 59134722.0,
      "step": 17695
    },
    {
      "entropy": 6.177161121368409,
      "epoch": 3.453028972783143,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038812072104807676,
      "loss": 5.7551,
      "mean_token_accuracy": 0.17158991396427153,
      "num_tokens": 59151345.0,
      "step": 17700
    },
    {
      "entropy": 6.154886865615845,
      "epoch": 3.454004487367086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003880599167885557,
      "loss": 5.7208,
      "mean_token_accuracy": 0.17662499248981475,
      "num_tokens": 59166492.0,
      "step": 17705
    },
    {
      "entropy": 6.25265941619873,
      "epoch": 3.454980001951029,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038799910148122255,
      "loss": 5.8478,
      "mean_token_accuracy": 0.17202709764242172,
      "num_tokens": 59183570.0,
      "step": 17710
    },
    {
      "entropy": 6.283731985092163,
      "epoch": 3.455955516534972,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038793827513201987,
      "loss": 5.9044,
      "mean_token_accuracy": 0.15755837708711623,
      "num_tokens": 59200175.0,
      "step": 17715
    },
    {
      "entropy": 6.250362777709961,
      "epoch": 3.4569310311189154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003878774377468915,
      "loss": 5.8118,
      "mean_token_accuracy": 0.16769860982894896,
      "num_tokens": 59216275.0,
      "step": 17720
    },
    {
      "entropy": 6.095877695083618,
      "epoch": 3.4579065457028584,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038781658933178224,
      "loss": 5.6747,
      "mean_token_accuracy": 0.18359975665807723,
      "num_tokens": 59233341.0,
      "step": 17725
    },
    {
      "entropy": 6.082904100418091,
      "epoch": 3.4588820602868013,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003877557298926379,
      "loss": 5.6946,
      "mean_token_accuracy": 0.1799486458301544,
      "num_tokens": 59248631.0,
      "step": 17730
    },
    {
      "entropy": 6.253630924224853,
      "epoch": 3.4598575748707443,
      "grad_norm": 1.0,
      "learning_rate": 0.0003876948594354056,
      "loss": 5.8208,
      "mean_token_accuracy": 0.17181308418512345,
      "num_tokens": 59264645.0,
      "step": 17735
    },
    {
      "entropy": 6.25147967338562,
      "epoch": 3.4608330894546873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003876339779660332,
      "loss": 5.7647,
      "mean_token_accuracy": 0.16577113270759583,
      "num_tokens": 59279830.0,
      "step": 17740
    },
    {
      "entropy": 6.177078485488892,
      "epoch": 3.46180860403863,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038757308549047,
      "loss": 5.8556,
      "mean_token_accuracy": 0.16683875620365143,
      "num_tokens": 59297958.0,
      "step": 17745
    },
    {
      "entropy": 6.204007482528686,
      "epoch": 3.462784118622573,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038751218201466605,
      "loss": 5.7604,
      "mean_token_accuracy": 0.17017522305250168,
      "num_tokens": 59314119.0,
      "step": 17750
    },
    {
      "entropy": 6.1932902336120605,
      "epoch": 3.4637596332065166,
      "grad_norm": 1.0,
      "learning_rate": 0.0003874512675445726,
      "loss": 5.7929,
      "mean_token_accuracy": 0.17145213037729262,
      "num_tokens": 59330391.0,
      "step": 17755
    },
    {
      "entropy": 6.244610118865967,
      "epoch": 3.4647351477904595,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003873903420861422,
      "loss": 5.8304,
      "mean_token_accuracy": 0.17437862306833268,
      "num_tokens": 59346449.0,
      "step": 17760
    },
    {
      "entropy": 6.282543420791626,
      "epoch": 3.4657106623744025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038732940564532804,
      "loss": 5.8792,
      "mean_token_accuracy": 0.1660800024867058,
      "num_tokens": 59363077.0,
      "step": 17765
    },
    {
      "entropy": 6.132140111923218,
      "epoch": 3.4666861769583455,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003872684582280847,
      "loss": 5.682,
      "mean_token_accuracy": 0.18322596848011016,
      "num_tokens": 59381542.0,
      "step": 17770
    },
    {
      "entropy": 6.188733243942261,
      "epoch": 3.4676616915422884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003872074998403678,
      "loss": 5.731,
      "mean_token_accuracy": 0.17578685730695726,
      "num_tokens": 59398045.0,
      "step": 17775
    },
    {
      "entropy": 6.206124544143677,
      "epoch": 3.468637206126232,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038714653048813394,
      "loss": 5.9028,
      "mean_token_accuracy": 0.16514235734939575,
      "num_tokens": 59414958.0,
      "step": 17780
    },
    {
      "entropy": 6.261943244934082,
      "epoch": 3.469612720710175,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038708555017734076,
      "loss": 5.8758,
      "mean_token_accuracy": 0.16759155094623565,
      "num_tokens": 59431585.0,
      "step": 17785
    },
    {
      "entropy": 6.228524398803711,
      "epoch": 3.4705882352941178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038702455891394716,
      "loss": 5.7272,
      "mean_token_accuracy": 0.16757213771343232,
      "num_tokens": 59447724.0,
      "step": 17790
    },
    {
      "entropy": 6.031094980239868,
      "epoch": 3.4715637498780607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038696355670391294,
      "loss": 5.6181,
      "mean_token_accuracy": 0.18891649395227433,
      "num_tokens": 59465116.0,
      "step": 17795
    },
    {
      "entropy": 6.2362758159637455,
      "epoch": 3.4725392644620037,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038690254355319893,
      "loss": 5.8457,
      "mean_token_accuracy": 0.16489522904157639,
      "num_tokens": 59481541.0,
      "step": 17800
    },
    {
      "entropy": 6.213523006439209,
      "epoch": 3.4735147790459466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038684151946776727,
      "loss": 5.7963,
      "mean_token_accuracy": 0.17233069986104965,
      "num_tokens": 59498300.0,
      "step": 17805
    },
    {
      "entropy": 6.1905396461486815,
      "epoch": 3.4744902936298896,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003867804844535809,
      "loss": 5.8153,
      "mean_token_accuracy": 0.16457686871290206,
      "num_tokens": 59514844.0,
      "step": 17810
    },
    {
      "entropy": 6.212172937393189,
      "epoch": 3.4754658082138326,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003867194385166039,
      "loss": 5.7317,
      "mean_token_accuracy": 0.1748227134346962,
      "num_tokens": 59531820.0,
      "step": 17815
    },
    {
      "entropy": 6.195404100418091,
      "epoch": 3.476441322797776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038665838166280163,
      "loss": 5.7681,
      "mean_token_accuracy": 0.1686202645301819,
      "num_tokens": 59549744.0,
      "step": 17820
    },
    {
      "entropy": 6.2421740055084225,
      "epoch": 3.477416837381719,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003865973138981403,
      "loss": 5.8727,
      "mean_token_accuracy": 0.16628369092941284,
      "num_tokens": 59567598.0,
      "step": 17825
    },
    {
      "entropy": 6.23821268081665,
      "epoch": 3.478392351965662,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038653623522858714,
      "loss": 5.788,
      "mean_token_accuracy": 0.1693824052810669,
      "num_tokens": 59583332.0,
      "step": 17830
    },
    {
      "entropy": 6.218443393707275,
      "epoch": 3.479367866549605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003864751456601106,
      "loss": 5.8032,
      "mean_token_accuracy": 0.16701114773750306,
      "num_tokens": 59600056.0,
      "step": 17835
    },
    {
      "entropy": 6.245884561538697,
      "epoch": 3.480343381133548,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038641404519868016,
      "loss": 5.8514,
      "mean_token_accuracy": 0.17037612497806548,
      "num_tokens": 59616065.0,
      "step": 17840
    },
    {
      "entropy": 6.177815008163452,
      "epoch": 3.481318895717491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003863529338502663,
      "loss": 5.7615,
      "mean_token_accuracy": 0.17610031366348267,
      "num_tokens": 59633533.0,
      "step": 17845
    },
    {
      "entropy": 6.207184839248657,
      "epoch": 3.482294410301434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038629181162084055,
      "loss": 5.7485,
      "mean_token_accuracy": 0.17688652127981186,
      "num_tokens": 59649011.0,
      "step": 17850
    },
    {
      "entropy": 6.195564413070679,
      "epoch": 3.483269924885377,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038623067851637573,
      "loss": 5.77,
      "mean_token_accuracy": 0.16310310512781143,
      "num_tokens": 59664638.0,
      "step": 17855
    },
    {
      "entropy": 6.257577991485595,
      "epoch": 3.48424543946932,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003861695345428453,
      "loss": 5.9018,
      "mean_token_accuracy": 0.15630542039871215,
      "num_tokens": 59680173.0,
      "step": 17860
    },
    {
      "entropy": 6.277031421661377,
      "epoch": 3.485220954053263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038610837970622433,
      "loss": 5.9006,
      "mean_token_accuracy": 0.16520483642816544,
      "num_tokens": 59696156.0,
      "step": 17865
    },
    {
      "entropy": 6.278731775283814,
      "epoch": 3.486196468637206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003860472140124884,
      "loss": 5.8373,
      "mean_token_accuracy": 0.17093375325202942,
      "num_tokens": 59712449.0,
      "step": 17870
    },
    {
      "entropy": 6.071975946426392,
      "epoch": 3.487171983221149,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038598603746761446,
      "loss": 5.7156,
      "mean_token_accuracy": 0.17915693521499634,
      "num_tokens": 59730896.0,
      "step": 17875
    },
    {
      "entropy": 6.142519235610962,
      "epoch": 3.4881474978050924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003859248500775805,
      "loss": 5.7237,
      "mean_token_accuracy": 0.175522381067276,
      "num_tokens": 59747040.0,
      "step": 17880
    },
    {
      "entropy": 6.305240249633789,
      "epoch": 3.4891230123890353,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038586365184836556,
      "loss": 5.8363,
      "mean_token_accuracy": 0.1653756782412529,
      "num_tokens": 59763931.0,
      "step": 17885
    },
    {
      "entropy": 6.179248809814453,
      "epoch": 3.4900985269729783,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038580244278594976,
      "loss": 5.7266,
      "mean_token_accuracy": 0.16875726878643035,
      "num_tokens": 59779581.0,
      "step": 17890
    },
    {
      "entropy": 6.2069206714630125,
      "epoch": 3.4910740415569212,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003857412228963141,
      "loss": 5.844,
      "mean_token_accuracy": 0.16791892796754837,
      "num_tokens": 59796388.0,
      "step": 17895
    },
    {
      "entropy": 6.125154876708985,
      "epoch": 3.492049556140864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003856799921854408,
      "loss": 5.7319,
      "mean_token_accuracy": 0.1793094426393509,
      "num_tokens": 59814398.0,
      "step": 17900
    },
    {
      "entropy": 6.1658368587493895,
      "epoch": 3.493025070724807,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003856187506593132,
      "loss": 5.7917,
      "mean_token_accuracy": 0.17405546605587005,
      "num_tokens": 59831772.0,
      "step": 17905
    },
    {
      "entropy": 5.898790168762207,
      "epoch": 3.4940005853087506,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003855574983239155,
      "loss": 5.4659,
      "mean_token_accuracy": 0.2025705546140671,
      "num_tokens": 59849621.0,
      "step": 17910
    },
    {
      "entropy": 6.230285882949829,
      "epoch": 3.4949760998926935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003854962351852331,
      "loss": 5.7475,
      "mean_token_accuracy": 0.17065997719764708,
      "num_tokens": 59867231.0,
      "step": 17915
    },
    {
      "entropy": 6.185370016098022,
      "epoch": 3.4959516144766365,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038543496124925243,
      "loss": 5.7608,
      "mean_token_accuracy": 0.17369667291641236,
      "num_tokens": 59884040.0,
      "step": 17920
    },
    {
      "entropy": 6.121036338806152,
      "epoch": 3.4969271290605795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003853736765219609,
      "loss": 5.685,
      "mean_token_accuracy": 0.18036148250102996,
      "num_tokens": 59900593.0,
      "step": 17925
    },
    {
      "entropy": 6.296364450454712,
      "epoch": 3.4979026436445224,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038531238100934713,
      "loss": 5.8605,
      "mean_token_accuracy": 0.16429421454668044,
      "num_tokens": 59916389.0,
      "step": 17930
    },
    {
      "entropy": 6.128342342376709,
      "epoch": 3.4988781582284654,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038525107471740064,
      "loss": 5.8016,
      "mean_token_accuracy": 0.17112081944942475,
      "num_tokens": 59933536.0,
      "step": 17935
    },
    {
      "entropy": 6.11711106300354,
      "epoch": 3.4998536728124083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003851897576521121,
      "loss": 5.6955,
      "mean_token_accuracy": 0.17576702535152436,
      "num_tokens": 59949756.0,
      "step": 17940
    },
    {
      "entropy": 6.245225811004639,
      "epoch": 3.5008291873963517,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038512842981947297,
      "loss": 5.7509,
      "mean_token_accuracy": 0.17032110095024108,
      "num_tokens": 59965791.0,
      "step": 17945
    },
    {
      "entropy": 6.116296863555908,
      "epoch": 3.5018047019802947,
      "grad_norm": 1.125,
      "learning_rate": 0.00038506709122547645,
      "loss": 5.675,
      "mean_token_accuracy": 0.18399551063776015,
      "num_tokens": 59981471.0,
      "step": 17950
    },
    {
      "entropy": 6.1950078964233395,
      "epoch": 3.5027802165642377,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038500574187611594,
      "loss": 5.7183,
      "mean_token_accuracy": 0.17098646014928817,
      "num_tokens": 59996896.0,
      "step": 17955
    },
    {
      "entropy": 6.174486351013184,
      "epoch": 3.5037557311481806,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003849443817773863,
      "loss": 5.8085,
      "mean_token_accuracy": 0.16825642436742783,
      "num_tokens": 60013110.0,
      "step": 17960
    },
    {
      "entropy": 6.198900365829468,
      "epoch": 3.5047312457321236,
      "grad_norm": 1.0,
      "learning_rate": 0.0003848830109352837,
      "loss": 5.7383,
      "mean_token_accuracy": 0.17242272794246674,
      "num_tokens": 60029763.0,
      "step": 17965
    },
    {
      "entropy": 6.158895015716553,
      "epoch": 3.505706760316067,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003848216293558048,
      "loss": 5.7375,
      "mean_token_accuracy": 0.1740889757871628,
      "num_tokens": 60046078.0,
      "step": 17970
    },
    {
      "entropy": 6.2523033618927,
      "epoch": 3.50668227490001,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038476023704494774,
      "loss": 5.8585,
      "mean_token_accuracy": 0.161104479432106,
      "num_tokens": 60062469.0,
      "step": 17975
    },
    {
      "entropy": 6.07324709892273,
      "epoch": 3.507657789483953,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003846988340087114,
      "loss": 5.6425,
      "mean_token_accuracy": 0.18482144623994828,
      "num_tokens": 60078183.0,
      "step": 17980
    },
    {
      "entropy": 6.040505838394165,
      "epoch": 3.508633304067896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003846374202530961,
      "loss": 5.6494,
      "mean_token_accuracy": 0.18096137791872025,
      "num_tokens": 60094811.0,
      "step": 17985
    },
    {
      "entropy": 6.2143150806427006,
      "epoch": 3.509608818651839,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038457599578410286,
      "loss": 5.753,
      "mean_token_accuracy": 0.17044518142938614,
      "num_tokens": 60112478.0,
      "step": 17990
    },
    {
      "entropy": 6.186296463012695,
      "epoch": 3.510584333235782,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038451456060773376,
      "loss": 5.7745,
      "mean_token_accuracy": 0.17154462784528732,
      "num_tokens": 60128486.0,
      "step": 17995
    },
    {
      "entropy": 6.134571695327759,
      "epoch": 3.5115598478197247,
      "grad_norm": 1.0,
      "learning_rate": 0.0003844531147299921,
      "loss": 5.7382,
      "mean_token_accuracy": 0.1734660431742668,
      "num_tokens": 60144303.0,
      "step": 18000
    },
    {
      "epoch": 3.5115598478197247,
      "eval_entropy": 5.979544270170868,
      "eval_loss": 5.921469688415527,
      "eval_mean_token_accuracy": 0.17202759359447034,
      "eval_num_tokens": 60144303.0,
      "eval_runtime": 15.413,
      "eval_samples_per_second": 1075.518,
      "eval_steps_per_second": 134.496,
      "step": 18000
    },
    {
      "entropy": 6.046319484710693,
      "epoch": 3.5125353624036677,
      "grad_norm": 0.859375,
      "learning_rate": 0.00038439165815688233,
      "loss": 5.6559,
      "mean_token_accuracy": 0.1805383160710335,
      "num_tokens": 60161750.0,
      "step": 18005
    },
    {
      "entropy": 6.238372039794922,
      "epoch": 3.513510876987611,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003843301908944094,
      "loss": 5.7975,
      "mean_token_accuracy": 0.16977430284023284,
      "num_tokens": 60178144.0,
      "step": 18010
    },
    {
      "entropy": 6.05722713470459,
      "epoch": 3.514486391571554,
      "grad_norm": 1.125,
      "learning_rate": 0.00038426871294858,
      "loss": 5.5706,
      "mean_token_accuracy": 0.19161845594644547,
      "num_tokens": 60196152.0,
      "step": 18015
    },
    {
      "entropy": 6.133829069137573,
      "epoch": 3.515461906155497,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038420722432540137,
      "loss": 5.7038,
      "mean_token_accuracy": 0.17330305874347687,
      "num_tokens": 60212049.0,
      "step": 18020
    },
    {
      "entropy": 6.134670448303223,
      "epoch": 3.51643742073944,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000384145725030882,
      "loss": 5.7437,
      "mean_token_accuracy": 0.17497252076864242,
      "num_tokens": 60230028.0,
      "step": 18025
    },
    {
      "entropy": 6.26535816192627,
      "epoch": 3.517412935323383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038408421507103144,
      "loss": 5.8942,
      "mean_token_accuracy": 0.16276237219572068,
      "num_tokens": 60247265.0,
      "step": 18030
    },
    {
      "entropy": 6.126549863815308,
      "epoch": 3.5183884499073264,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003840226944518601,
      "loss": 5.7234,
      "mean_token_accuracy": 0.17729570120573043,
      "num_tokens": 60263685.0,
      "step": 18035
    },
    {
      "entropy": 6.2592024326324465,
      "epoch": 3.5193639644912693,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003839611631793796,
      "loss": 5.8258,
      "mean_token_accuracy": 0.1690717101097107,
      "num_tokens": 60279574.0,
      "step": 18040
    },
    {
      "entropy": 6.159539890289307,
      "epoch": 3.5203394790752123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003838996212596027,
      "loss": 5.8131,
      "mean_token_accuracy": 0.17265928983688356,
      "num_tokens": 60296302.0,
      "step": 18045
    },
    {
      "entropy": 6.157758045196533,
      "epoch": 3.5213149936591552,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003838380686985428,
      "loss": 5.8096,
      "mean_token_accuracy": 0.17433899492025376,
      "num_tokens": 60313595.0,
      "step": 18050
    },
    {
      "entropy": 6.077081823348999,
      "epoch": 3.522290508243098,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038377650550221475,
      "loss": 5.6384,
      "mean_token_accuracy": 0.18251245319843293,
      "num_tokens": 60331286.0,
      "step": 18055
    },
    {
      "entropy": 6.219256162643433,
      "epoch": 3.523266022827041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038371493167663434,
      "loss": 5.7655,
      "mean_token_accuracy": 0.17276675701141359,
      "num_tokens": 60347704.0,
      "step": 18060
    },
    {
      "entropy": 6.220047187805176,
      "epoch": 3.524241537410984,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003836533472278183,
      "loss": 5.8146,
      "mean_token_accuracy": 0.17160067111253738,
      "num_tokens": 60364702.0,
      "step": 18065
    },
    {
      "entropy": 6.2044751167297365,
      "epoch": 3.525217051994927,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038359175216178427,
      "loss": 5.7545,
      "mean_token_accuracy": 0.16896059215068818,
      "num_tokens": 60381052.0,
      "step": 18070
    },
    {
      "entropy": 6.165033197402954,
      "epoch": 3.5261925665788705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038353014648455136,
      "loss": 5.8174,
      "mean_token_accuracy": 0.1688805416226387,
      "num_tokens": 60396772.0,
      "step": 18075
    },
    {
      "entropy": 6.1541704654693605,
      "epoch": 3.5271680811628134,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003834685302021393,
      "loss": 5.7143,
      "mean_token_accuracy": 0.17336861789226532,
      "num_tokens": 60412951.0,
      "step": 18080
    },
    {
      "entropy": 6.155292654037476,
      "epoch": 3.5281435957467564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038340690332056904,
      "loss": 5.7344,
      "mean_token_accuracy": 0.17679644525051116,
      "num_tokens": 60429560.0,
      "step": 18085
    },
    {
      "entropy": 6.226267528533936,
      "epoch": 3.5291191103306994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003833452658458626,
      "loss": 5.7839,
      "mean_token_accuracy": 0.1687949776649475,
      "num_tokens": 60444962.0,
      "step": 18090
    },
    {
      "entropy": 6.077334451675415,
      "epoch": 3.5300946249146423,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038328361778404294,
      "loss": 5.6931,
      "mean_token_accuracy": 0.17805678248405457,
      "num_tokens": 60462796.0,
      "step": 18095
    },
    {
      "entropy": 6.218808650970459,
      "epoch": 3.5310701394985857,
      "grad_norm": 1.0,
      "learning_rate": 0.0003832219591411341,
      "loss": 5.7759,
      "mean_token_accuracy": 0.16748720705509185,
      "num_tokens": 60478639.0,
      "step": 18100
    },
    {
      "entropy": 6.049761533737183,
      "epoch": 3.5320456540825287,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038316028992316106,
      "loss": 5.6322,
      "mean_token_accuracy": 0.18533750325441362,
      "num_tokens": 60493114.0,
      "step": 18105
    },
    {
      "entropy": 6.211369419097901,
      "epoch": 3.5330211686664716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038309861013614995,
      "loss": 5.724,
      "mean_token_accuracy": 0.17981152832508088,
      "num_tokens": 60507726.0,
      "step": 18110
    },
    {
      "entropy": 6.211700201034546,
      "epoch": 3.5339966832504146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038303691978612797,
      "loss": 5.8492,
      "mean_token_accuracy": 0.16457293182611465,
      "num_tokens": 60525835.0,
      "step": 18115
    },
    {
      "entropy": 6.126596403121948,
      "epoch": 3.5349721978343576,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003829752188791233,
      "loss": 5.7202,
      "mean_token_accuracy": 0.1791197508573532,
      "num_tokens": 60543333.0,
      "step": 18120
    },
    {
      "entropy": 6.112605237960816,
      "epoch": 3.5359477124183005,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000382913507421165,
      "loss": 5.6819,
      "mean_token_accuracy": 0.17352758198976517,
      "num_tokens": 60559369.0,
      "step": 18125
    },
    {
      "entropy": 5.981074857711792,
      "epoch": 3.5369232270022435,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003828517854182834,
      "loss": 5.5171,
      "mean_token_accuracy": 0.19558584839105606,
      "num_tokens": 60574575.0,
      "step": 18130
    },
    {
      "entropy": 6.114460897445679,
      "epoch": 3.5378987415861864,
      "grad_norm": 1.0,
      "learning_rate": 0.0003827900528765097,
      "loss": 5.7108,
      "mean_token_accuracy": 0.17666537910699845,
      "num_tokens": 60591185.0,
      "step": 18135
    },
    {
      "entropy": 6.121013116836548,
      "epoch": 3.53887425617013,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003827283098018762,
      "loss": 5.7313,
      "mean_token_accuracy": 0.17508092224597932,
      "num_tokens": 60607942.0,
      "step": 18140
    },
    {
      "entropy": 6.1462188243865965,
      "epoch": 3.539849770754073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003826665562004162,
      "loss": 5.7485,
      "mean_token_accuracy": 0.1788178101181984,
      "num_tokens": 60625127.0,
      "step": 18145
    },
    {
      "entropy": 6.273778104782105,
      "epoch": 3.5408252853380158,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003826047920781641,
      "loss": 5.8532,
      "mean_token_accuracy": 0.16791571229696273,
      "num_tokens": 60641980.0,
      "step": 18150
    },
    {
      "entropy": 6.155543375015259,
      "epoch": 3.5418007999219587,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003825430174411552,
      "loss": 5.7188,
      "mean_token_accuracy": 0.17656655013561248,
      "num_tokens": 60659171.0,
      "step": 18155
    },
    {
      "entropy": 6.19038233757019,
      "epoch": 3.542776314505902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003824812322954259,
      "loss": 5.8124,
      "mean_token_accuracy": 0.17543772161006926,
      "num_tokens": 60675320.0,
      "step": 18160
    },
    {
      "entropy": 6.230347490310669,
      "epoch": 3.543751829089845,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038241943664701374,
      "loss": 5.918,
      "mean_token_accuracy": 0.16489268690347672,
      "num_tokens": 60692876.0,
      "step": 18165
    },
    {
      "entropy": 6.221252202987671,
      "epoch": 3.544727343673788,
      "grad_norm": 0.96875,
      "learning_rate": 0.000382357630501957,
      "loss": 5.8489,
      "mean_token_accuracy": 0.16630346029996873,
      "num_tokens": 60708442.0,
      "step": 18170
    },
    {
      "entropy": 6.189753293991089,
      "epoch": 3.545702858257731,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003822958138662953,
      "loss": 5.6795,
      "mean_token_accuracy": 0.17741169333457946,
      "num_tokens": 60726103.0,
      "step": 18175
    },
    {
      "entropy": 6.11042160987854,
      "epoch": 3.546678372841674,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000382233986746069,
      "loss": 5.6919,
      "mean_token_accuracy": 0.18053998798131943,
      "num_tokens": 60741990.0,
      "step": 18180
    },
    {
      "entropy": 6.213805294036865,
      "epoch": 3.547653887425617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003821721491473198,
      "loss": 5.7991,
      "mean_token_accuracy": 0.17401611506938935,
      "num_tokens": 60758877.0,
      "step": 18185
    },
    {
      "entropy": 6.178156518936158,
      "epoch": 3.54862940200956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038211030107609,
      "loss": 5.7355,
      "mean_token_accuracy": 0.17499111145734786,
      "num_tokens": 60774762.0,
      "step": 18190
    },
    {
      "entropy": 6.200686073303222,
      "epoch": 3.549604916593503,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038204844253842345,
      "loss": 5.7878,
      "mean_token_accuracy": 0.17462475299835206,
      "num_tokens": 60791196.0,
      "step": 18195
    },
    {
      "entropy": 6.166643714904785,
      "epoch": 3.5505804311774463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038198657354036457,
      "loss": 5.7862,
      "mean_token_accuracy": 0.17061716467142105,
      "num_tokens": 60807688.0,
      "step": 18200
    },
    {
      "entropy": 6.223697662353516,
      "epoch": 3.551555945761389,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038192469408795905,
      "loss": 5.7869,
      "mean_token_accuracy": 0.17313445806503297,
      "num_tokens": 60823713.0,
      "step": 18205
    },
    {
      "entropy": 6.187939310073853,
      "epoch": 3.552531460345332,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038186280418725353,
      "loss": 5.8256,
      "mean_token_accuracy": 0.17040076702833176,
      "num_tokens": 60840433.0,
      "step": 18210
    },
    {
      "entropy": 6.245127105712891,
      "epoch": 3.553506974929275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038180090384429564,
      "loss": 5.8417,
      "mean_token_accuracy": 0.17270833551883696,
      "num_tokens": 60854953.0,
      "step": 18215
    },
    {
      "entropy": 6.232597923278808,
      "epoch": 3.554482489513218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038173899306513404,
      "loss": 5.8805,
      "mean_token_accuracy": 0.1687017112970352,
      "num_tokens": 60871973.0,
      "step": 18220
    },
    {
      "entropy": 6.215385293960571,
      "epoch": 3.5554580040971615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003816770718558185,
      "loss": 5.7965,
      "mean_token_accuracy": 0.1641494318842888,
      "num_tokens": 60888135.0,
      "step": 18225
    },
    {
      "entropy": 6.056157064437866,
      "epoch": 3.5564335186811045,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038161514022239974,
      "loss": 5.6004,
      "mean_token_accuracy": 0.18909138739109038,
      "num_tokens": 60905831.0,
      "step": 18230
    },
    {
      "entropy": 6.156909990310669,
      "epoch": 3.5574090332650474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003815531981709294,
      "loss": 5.7468,
      "mean_token_accuracy": 0.16797012835741043,
      "num_tokens": 60923670.0,
      "step": 18235
    },
    {
      "entropy": 6.096674394607544,
      "epoch": 3.5583845478489904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038149124570746037,
      "loss": 5.7279,
      "mean_token_accuracy": 0.17513610124588014,
      "num_tokens": 60939007.0,
      "step": 18240
    },
    {
      "entropy": 6.174310970306396,
      "epoch": 3.5593600624329333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003814292828380463,
      "loss": 5.7307,
      "mean_token_accuracy": 0.16783622950315474,
      "num_tokens": 60956421.0,
      "step": 18245
    },
    {
      "entropy": 6.0206505298614506,
      "epoch": 3.5603355770168763,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000381367309568742,
      "loss": 5.5041,
      "mean_token_accuracy": 0.19503472447395326,
      "num_tokens": 60975047.0,
      "step": 18250
    },
    {
      "entropy": 6.1401787281036375,
      "epoch": 3.5613110916008193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003813053259056033,
      "loss": 5.6876,
      "mean_token_accuracy": 0.17909881770610808,
      "num_tokens": 60991482.0,
      "step": 18255
    },
    {
      "entropy": 6.078264427185059,
      "epoch": 3.5622866061847622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038124333185468714,
      "loss": 5.6629,
      "mean_token_accuracy": 0.17988295555114747,
      "num_tokens": 61009162.0,
      "step": 18260
    },
    {
      "entropy": 6.167684316635132,
      "epoch": 3.5632621207687056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003811813274220511,
      "loss": 5.7349,
      "mean_token_accuracy": 0.1740291476249695,
      "num_tokens": 61026654.0,
      "step": 18265
    },
    {
      "entropy": 6.184818363189697,
      "epoch": 3.5642376353526486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038111931261375424,
      "loss": 5.7668,
      "mean_token_accuracy": 0.17223773151636124,
      "num_tokens": 61042808.0,
      "step": 18270
    },
    {
      "entropy": 6.1821530818939205,
      "epoch": 3.5652131499365916,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038105728743585644,
      "loss": 5.8219,
      "mean_token_accuracy": 0.15777841806411744,
      "num_tokens": 61059349.0,
      "step": 18275
    },
    {
      "entropy": 6.227806615829468,
      "epoch": 3.5661886645205345,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003809952518944185,
      "loss": 5.8473,
      "mean_token_accuracy": 0.17231039255857467,
      "num_tokens": 61077830.0,
      "step": 18280
    },
    {
      "entropy": 6.124451923370361,
      "epoch": 3.5671641791044775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003809332059955022,
      "loss": 5.7142,
      "mean_token_accuracy": 0.18271051347255707,
      "num_tokens": 61094812.0,
      "step": 18285
    },
    {
      "entropy": 6.140315771102905,
      "epoch": 3.568139693688421,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003808711497451706,
      "loss": 5.7169,
      "mean_token_accuracy": 0.18110805749893188,
      "num_tokens": 61113756.0,
      "step": 18290
    },
    {
      "entropy": 6.180631923675537,
      "epoch": 3.569115208272364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038080908314948764,
      "loss": 5.7466,
      "mean_token_accuracy": 0.18088702559471131,
      "num_tokens": 61128395.0,
      "step": 18295
    },
    {
      "entropy": 6.037953472137451,
      "epoch": 3.570090722856307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003807470062145181,
      "loss": 5.5799,
      "mean_token_accuracy": 0.17811317890882492,
      "num_tokens": 61144823.0,
      "step": 18300
    },
    {
      "entropy": 6.2101037979125975,
      "epoch": 3.5710662374402498,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038068491894632805,
      "loss": 5.7933,
      "mean_token_accuracy": 0.1683559700846672,
      "num_tokens": 61161693.0,
      "step": 18305
    },
    {
      "entropy": 6.234171628952026,
      "epoch": 3.5720417520241927,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003806228213509844,
      "loss": 5.8917,
      "mean_token_accuracy": 0.1618264600634575,
      "num_tokens": 61178807.0,
      "step": 18310
    },
    {
      "entropy": 6.0633618354797365,
      "epoch": 3.5730172666081357,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038056071343455514,
      "loss": 5.585,
      "mean_token_accuracy": 0.1893756553530693,
      "num_tokens": 61195154.0,
      "step": 18315
    },
    {
      "entropy": 6.190413141250611,
      "epoch": 3.5739927811920786,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003804985952031091,
      "loss": 5.8331,
      "mean_token_accuracy": 0.1686033308506012,
      "num_tokens": 61211825.0,
      "step": 18320
    },
    {
      "entropy": 6.244935846328735,
      "epoch": 3.5749682957760216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003804364666627164,
      "loss": 5.8609,
      "mean_token_accuracy": 0.16781862825155258,
      "num_tokens": 61228548.0,
      "step": 18325
    },
    {
      "entropy": 6.276086521148682,
      "epoch": 3.575943810359965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038037432781944797,
      "loss": 5.8268,
      "mean_token_accuracy": 0.17072461545467377,
      "num_tokens": 61245490.0,
      "step": 18330
    },
    {
      "entropy": 6.121934509277343,
      "epoch": 3.576919324943908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038031217867937566,
      "loss": 5.6359,
      "mean_token_accuracy": 0.1793843299150467,
      "num_tokens": 61261604.0,
      "step": 18335
    },
    {
      "entropy": 6.137858247756958,
      "epoch": 3.577894839527851,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038025001924857273,
      "loss": 5.7207,
      "mean_token_accuracy": 0.1809381976723671,
      "num_tokens": 61279100.0,
      "step": 18340
    },
    {
      "entropy": 6.188151502609253,
      "epoch": 3.578870354111794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000380187849533113,
      "loss": 5.7404,
      "mean_token_accuracy": 0.17274908125400543,
      "num_tokens": 61295184.0,
      "step": 18345
    },
    {
      "entropy": 6.162976026535034,
      "epoch": 3.579845868695737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038012566953907147,
      "loss": 5.7453,
      "mean_token_accuracy": 0.1821703165769577,
      "num_tokens": 61313070.0,
      "step": 18350
    },
    {
      "entropy": 6.2407519817352295,
      "epoch": 3.5808213832796802,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003800634792725242,
      "loss": 5.8345,
      "mean_token_accuracy": 0.1691788151860237,
      "num_tokens": 61330199.0,
      "step": 18355
    },
    {
      "entropy": 6.302834177017212,
      "epoch": 3.581796897863623,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038000127873954824,
      "loss": 5.8803,
      "mean_token_accuracy": 0.15879196375608445,
      "num_tokens": 61345525.0,
      "step": 18360
    },
    {
      "entropy": 6.223489570617676,
      "epoch": 3.582772412447566,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037993906794622144,
      "loss": 5.8563,
      "mean_token_accuracy": 0.17167430818080903,
      "num_tokens": 61363064.0,
      "step": 18365
    },
    {
      "entropy": 6.19888219833374,
      "epoch": 3.583747927031509,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037987684689862314,
      "loss": 5.7836,
      "mean_token_accuracy": 0.1713288277387619,
      "num_tokens": 61381092.0,
      "step": 18370
    },
    {
      "entropy": 6.178164005279541,
      "epoch": 3.584723441615452,
      "grad_norm": 0.953125,
      "learning_rate": 0.000379814615602833,
      "loss": 5.7502,
      "mean_token_accuracy": 0.17218416631221772,
      "num_tokens": 61397505.0,
      "step": 18375
    },
    {
      "entropy": 6.137833976745606,
      "epoch": 3.585698956199395,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00037975237406493215,
      "loss": 5.7003,
      "mean_token_accuracy": 0.17689979076385498,
      "num_tokens": 61414986.0,
      "step": 18380
    },
    {
      "entropy": 5.987028169631958,
      "epoch": 3.586674470783338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037969012229100276,
      "loss": 5.5228,
      "mean_token_accuracy": 0.19605806916952134,
      "num_tokens": 61431878.0,
      "step": 18385
    },
    {
      "entropy": 6.229620885848999,
      "epoch": 3.587649985367281,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037962786028712767,
      "loss": 5.761,
      "mean_token_accuracy": 0.17265518009662628,
      "num_tokens": 61447872.0,
      "step": 18390
    },
    {
      "entropy": 6.164532327651978,
      "epoch": 3.5886254999512244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037956558805939095,
      "loss": 5.707,
      "mean_token_accuracy": 0.17717968970537185,
      "num_tokens": 61465005.0,
      "step": 18395
    },
    {
      "entropy": 6.092661714553833,
      "epoch": 3.5896010145351673,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003795033056138777,
      "loss": 5.6905,
      "mean_token_accuracy": 0.17041276395320892,
      "num_tokens": 61481482.0,
      "step": 18400
    },
    {
      "entropy": 6.1962890625,
      "epoch": 3.5905765291191103,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003794410129566738,
      "loss": 5.7898,
      "mean_token_accuracy": 0.16641943752765656,
      "num_tokens": 61496611.0,
      "step": 18405
    },
    {
      "entropy": 6.145768165588379,
      "epoch": 3.5915520437030533,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003793787100938665,
      "loss": 5.7106,
      "mean_token_accuracy": 0.17813935726881028,
      "num_tokens": 61515309.0,
      "step": 18410
    },
    {
      "entropy": 6.113019704818726,
      "epoch": 3.5925275582869967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003793163970315435,
      "loss": 5.6903,
      "mean_token_accuracy": 0.1772035166621208,
      "num_tokens": 61531375.0,
      "step": 18415
    },
    {
      "entropy": 6.254645252227784,
      "epoch": 3.5935030728709396,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037925407377579405,
      "loss": 5.807,
      "mean_token_accuracy": 0.16895021945238115,
      "num_tokens": 61548626.0,
      "step": 18420
    },
    {
      "entropy": 6.148106050491333,
      "epoch": 3.5944785874548826,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003791917403327081,
      "loss": 5.7156,
      "mean_token_accuracy": 0.1789318636059761,
      "num_tokens": 61566514.0,
      "step": 18425
    },
    {
      "entropy": 6.176311492919922,
      "epoch": 3.5954541020388255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003791293967083766,
      "loss": 5.7837,
      "mean_token_accuracy": 0.16909078508615494,
      "num_tokens": 61582558.0,
      "step": 18430
    },
    {
      "entropy": 6.2440564155578615,
      "epoch": 3.5964296166227685,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003790670429088915,
      "loss": 5.8346,
      "mean_token_accuracy": 0.17110563069581985,
      "num_tokens": 61599045.0,
      "step": 18435
    },
    {
      "entropy": 6.052045726776123,
      "epoch": 3.5974051312067115,
      "grad_norm": 1.0,
      "learning_rate": 0.00037900467894034606,
      "loss": 5.65,
      "mean_token_accuracy": 0.17712507098913194,
      "num_tokens": 61613661.0,
      "step": 18440
    },
    {
      "entropy": 6.179902076721191,
      "epoch": 3.5983806457906544,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003789423048088339,
      "loss": 5.6978,
      "mean_token_accuracy": 0.17722102701663972,
      "num_tokens": 61628579.0,
      "step": 18445
    },
    {
      "entropy": 6.090974044799805,
      "epoch": 3.5993561603745974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003788799205204502,
      "loss": 5.639,
      "mean_token_accuracy": 0.18063572496175767,
      "num_tokens": 61644334.0,
      "step": 18450
    },
    {
      "entropy": 6.222491359710693,
      "epoch": 3.600331674958541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000378817526081291,
      "loss": 5.7562,
      "mean_token_accuracy": 0.1666577935218811,
      "num_tokens": 61661279.0,
      "step": 18455
    },
    {
      "entropy": 6.202485227584839,
      "epoch": 3.6013071895424837,
      "grad_norm": 1.0,
      "learning_rate": 0.0003787551214974531,
      "loss": 5.7155,
      "mean_token_accuracy": 0.16976553350687026,
      "num_tokens": 61677745.0,
      "step": 18460
    },
    {
      "entropy": 6.154550933837891,
      "epoch": 3.6022827041264267,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003786927067750345,
      "loss": 5.7666,
      "mean_token_accuracy": 0.17782085686922072,
      "num_tokens": 61693875.0,
      "step": 18465
    },
    {
      "entropy": 6.196810579299926,
      "epoch": 3.6032582187103697,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037863028192013424,
      "loss": 5.8266,
      "mean_token_accuracy": 0.1672483876347542,
      "num_tokens": 61711047.0,
      "step": 18470
    },
    {
      "entropy": 6.041038227081299,
      "epoch": 3.6042337332943126,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003785678469388521,
      "loss": 5.5745,
      "mean_token_accuracy": 0.1898118481040001,
      "num_tokens": 61728260.0,
      "step": 18475
    },
    {
      "entropy": 6.206927347183227,
      "epoch": 3.605209247878256,
      "grad_norm": 1.0,
      "learning_rate": 0.0003785054018372891,
      "loss": 5.7812,
      "mean_token_accuracy": 0.1708504870533943,
      "num_tokens": 61744056.0,
      "step": 18480
    },
    {
      "entropy": 6.047855472564697,
      "epoch": 3.606184762462199,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003784429466215471,
      "loss": 5.5992,
      "mean_token_accuracy": 0.19052072763442993,
      "num_tokens": 61760692.0,
      "step": 18485
    },
    {
      "entropy": 5.992300748825073,
      "epoch": 3.607160277046142,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037838048129772904,
      "loss": 5.5692,
      "mean_token_accuracy": 0.1997926115989685,
      "num_tokens": 61777537.0,
      "step": 18490
    },
    {
      "entropy": 6.158553552627564,
      "epoch": 3.608135791630085,
      "grad_norm": 1.0,
      "learning_rate": 0.00037831800587193883,
      "loss": 5.754,
      "mean_token_accuracy": 0.17530040591955184,
      "num_tokens": 61792535.0,
      "step": 18495
    },
    {
      "entropy": 6.126518964767456,
      "epoch": 3.609111306214028,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00037825552035028134,
      "loss": 5.7667,
      "mean_token_accuracy": 0.1695755362510681,
      "num_tokens": 61808979.0,
      "step": 18500
    },
    {
      "entropy": 6.277272129058838,
      "epoch": 3.610086820797971,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037819302473886236,
      "loss": 5.7518,
      "mean_token_accuracy": 0.17441413253545762,
      "num_tokens": 61824524.0,
      "step": 18505
    },
    {
      "entropy": 6.129389429092408,
      "epoch": 3.611062335381914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003781305190437888,
      "loss": 5.7158,
      "mean_token_accuracy": 0.17414656281471252,
      "num_tokens": 61840196.0,
      "step": 18510
    },
    {
      "entropy": 6.248560094833374,
      "epoch": 3.6120378499658568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003780680032711685,
      "loss": 5.898,
      "mean_token_accuracy": 0.16296985745429993,
      "num_tokens": 61856708.0,
      "step": 18515
    },
    {
      "entropy": 6.18530011177063,
      "epoch": 3.6130133645498,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003780054774271102,
      "loss": 5.7476,
      "mean_token_accuracy": 0.16841731518507003,
      "num_tokens": 61873405.0,
      "step": 18520
    },
    {
      "entropy": 6.223161888122559,
      "epoch": 3.613988879133743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037794294151772385,
      "loss": 5.7981,
      "mean_token_accuracy": 0.17110195904970169,
      "num_tokens": 61891066.0,
      "step": 18525
    },
    {
      "entropy": 6.225844621658325,
      "epoch": 3.614964393717686,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037788039554912013,
      "loss": 5.8452,
      "mean_token_accuracy": 0.16929514855146408,
      "num_tokens": 61906858.0,
      "step": 18530
    },
    {
      "entropy": 6.220838212966919,
      "epoch": 3.615939908301629,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003778178395274108,
      "loss": 5.8141,
      "mean_token_accuracy": 0.16686542630195617,
      "num_tokens": 61922871.0,
      "step": 18535
    },
    {
      "entropy": 6.176358461380005,
      "epoch": 3.616915422885572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003777552734587087,
      "loss": 5.7464,
      "mean_token_accuracy": 0.1796169638633728,
      "num_tokens": 61938041.0,
      "step": 18540
    },
    {
      "entropy": 6.181814193725586,
      "epoch": 3.6178909374695154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037769269734912745,
      "loss": 5.8582,
      "mean_token_accuracy": 0.16307311356067658,
      "num_tokens": 61954027.0,
      "step": 18545
    },
    {
      "entropy": 6.17427625656128,
      "epoch": 3.6188664520534584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003776301112047818,
      "loss": 5.7002,
      "mean_token_accuracy": 0.1746800273656845,
      "num_tokens": 61968666.0,
      "step": 18550
    },
    {
      "entropy": 6.165698671340943,
      "epoch": 3.6198419666374013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037756751503178755,
      "loss": 5.7629,
      "mean_token_accuracy": 0.17371155172586442,
      "num_tokens": 61985320.0,
      "step": 18555
    },
    {
      "entropy": 6.125310039520263,
      "epoch": 3.6208174812213443,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003775049088362612,
      "loss": 5.6792,
      "mean_token_accuracy": 0.17962249368429184,
      "num_tokens": 62001905.0,
      "step": 18560
    },
    {
      "entropy": 6.172470235824585,
      "epoch": 3.6217929958052872,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003774422926243206,
      "loss": 5.7184,
      "mean_token_accuracy": 0.17690739631652833,
      "num_tokens": 62017280.0,
      "step": 18565
    },
    {
      "entropy": 6.104892444610596,
      "epoch": 3.62276851038923,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037737966640208413,
      "loss": 5.6226,
      "mean_token_accuracy": 0.17962018251419068,
      "num_tokens": 62034919.0,
      "step": 18570
    },
    {
      "entropy": 6.131844568252563,
      "epoch": 3.623744024973173,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003773170301756716,
      "loss": 5.7311,
      "mean_token_accuracy": 0.17454282939434052,
      "num_tokens": 62051575.0,
      "step": 18575
    },
    {
      "entropy": 6.12718243598938,
      "epoch": 3.624719539557116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037725438395120356,
      "loss": 5.7556,
      "mean_token_accuracy": 0.1772538900375366,
      "num_tokens": 62067684.0,
      "step": 18580
    },
    {
      "entropy": 6.192548418045044,
      "epoch": 3.6256950541410595,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003771917277348016,
      "loss": 5.7625,
      "mean_token_accuracy": 0.17476568520069122,
      "num_tokens": 62085030.0,
      "step": 18585
    },
    {
      "entropy": 6.258602428436279,
      "epoch": 3.6266705687250025,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037712906153258813,
      "loss": 5.8813,
      "mean_token_accuracy": 0.1673385515809059,
      "num_tokens": 62101662.0,
      "step": 18590
    },
    {
      "entropy": 6.082759284973145,
      "epoch": 3.6276460833089454,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037706638535068686,
      "loss": 5.664,
      "mean_token_accuracy": 0.17614962607622148,
      "num_tokens": 62118879.0,
      "step": 18595
    },
    {
      "entropy": 6.189345502853394,
      "epoch": 3.6286215978928884,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003770036991952221,
      "loss": 5.7578,
      "mean_token_accuracy": 0.17635428607463838,
      "num_tokens": 62134641.0,
      "step": 18600
    },
    {
      "entropy": 6.275901651382446,
      "epoch": 3.6295971124768314,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037694100307231947,
      "loss": 5.8675,
      "mean_token_accuracy": 0.1631275713443756,
      "num_tokens": 62151616.0,
      "step": 18605
    },
    {
      "entropy": 6.224777555465698,
      "epoch": 3.6305726270607748,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037687829698810526,
      "loss": 5.7864,
      "mean_token_accuracy": 0.17230567336082458,
      "num_tokens": 62169307.0,
      "step": 18610
    },
    {
      "entropy": 6.290570640563965,
      "epoch": 3.6315481416447177,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003768155809487071,
      "loss": 5.8767,
      "mean_token_accuracy": 0.16765091866254805,
      "num_tokens": 62185665.0,
      "step": 18615
    },
    {
      "entropy": 6.141557121276856,
      "epoch": 3.6325236562286607,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037675285496025315,
      "loss": 5.7822,
      "mean_token_accuracy": 0.17339062243700026,
      "num_tokens": 62199901.0,
      "step": 18620
    },
    {
      "entropy": 6.262485599517822,
      "epoch": 3.6334991708126037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003766901190288729,
      "loss": 5.8539,
      "mean_token_accuracy": 0.16629748493433,
      "num_tokens": 62216813.0,
      "step": 18625
    },
    {
      "entropy": 6.0980305671691895,
      "epoch": 3.6344746853965466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003766273731606966,
      "loss": 5.65,
      "mean_token_accuracy": 0.18137580007314683,
      "num_tokens": 62232578.0,
      "step": 18630
    },
    {
      "entropy": 6.178108358383179,
      "epoch": 3.6354501999804896,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003765646173618557,
      "loss": 5.7822,
      "mean_token_accuracy": 0.1704832136631012,
      "num_tokens": 62248652.0,
      "step": 18635
    },
    {
      "entropy": 6.14487943649292,
      "epoch": 3.6364257145644325,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003765018516384823,
      "loss": 5.6676,
      "mean_token_accuracy": 0.1798950746655464,
      "num_tokens": 62265170.0,
      "step": 18640
    },
    {
      "entropy": 6.061632966995239,
      "epoch": 3.6374012291483755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037643907599670974,
      "loss": 5.5835,
      "mean_token_accuracy": 0.19196642488241195,
      "num_tokens": 62282809.0,
      "step": 18645
    },
    {
      "entropy": 6.2306427478790285,
      "epoch": 3.638376743732319,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003763762904426722,
      "loss": 5.8505,
      "mean_token_accuracy": 0.1641574963927269,
      "num_tokens": 62301036.0,
      "step": 18650
    },
    {
      "entropy": 6.167166423797608,
      "epoch": 3.639352258316262,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003763134949825048,
      "loss": 5.704,
      "mean_token_accuracy": 0.17127352952957153,
      "num_tokens": 62317434.0,
      "step": 18655
    },
    {
      "entropy": 6.172324085235596,
      "epoch": 3.640327772900205,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003762506896223439,
      "loss": 5.7369,
      "mean_token_accuracy": 0.1715108722448349,
      "num_tokens": 62333158.0,
      "step": 18660
    },
    {
      "entropy": 6.195373773574829,
      "epoch": 3.6413032874841478,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003761878743683264,
      "loss": 5.7958,
      "mean_token_accuracy": 0.1724009245634079,
      "num_tokens": 62349075.0,
      "step": 18665
    },
    {
      "entropy": 6.143964004516602,
      "epoch": 3.642278802068091,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003761250492265904,
      "loss": 5.6333,
      "mean_token_accuracy": 0.18354146033525467,
      "num_tokens": 62366322.0,
      "step": 18670
    },
    {
      "entropy": 6.226080465316772,
      "epoch": 3.643254316652034,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037606221420327505,
      "loss": 5.7931,
      "mean_token_accuracy": 0.17247677594423294,
      "num_tokens": 62382140.0,
      "step": 18675
    },
    {
      "entropy": 6.133849954605102,
      "epoch": 3.644229831235977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003759993693045204,
      "loss": 5.7624,
      "mean_token_accuracy": 0.17422409802675248,
      "num_tokens": 62399255.0,
      "step": 18680
    },
    {
      "entropy": 6.092226791381836,
      "epoch": 3.64520534581992,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003759365145364673,
      "loss": 5.6718,
      "mean_token_accuracy": 0.18439730852842331,
      "num_tokens": 62415451.0,
      "step": 18685
    },
    {
      "entropy": 6.113996124267578,
      "epoch": 3.646180860403863,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003758736499052577,
      "loss": 5.661,
      "mean_token_accuracy": 0.17987101078033446,
      "num_tokens": 62432486.0,
      "step": 18690
    },
    {
      "entropy": 6.156524038314819,
      "epoch": 3.647156374987806,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003758107754170346,
      "loss": 5.7429,
      "mean_token_accuracy": 0.17779299318790437,
      "num_tokens": 62450117.0,
      "step": 18695
    },
    {
      "entropy": 6.137366199493409,
      "epoch": 3.648131889571749,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037574789107794175,
      "loss": 5.7181,
      "mean_token_accuracy": 0.1713075190782547,
      "num_tokens": 62465628.0,
      "step": 18700
    },
    {
      "entropy": 6.1253612518310545,
      "epoch": 3.649107404155692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003756849968941241,
      "loss": 5.7072,
      "mean_token_accuracy": 0.18642752468585969,
      "num_tokens": 62483386.0,
      "step": 18705
    },
    {
      "entropy": 6.089823389053345,
      "epoch": 3.6500829187396353,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003756220928717274,
      "loss": 5.6496,
      "mean_token_accuracy": 0.18182309567928315,
      "num_tokens": 62499618.0,
      "step": 18710
    },
    {
      "entropy": 6.215846538543701,
      "epoch": 3.6510584333235783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003755591790168984,
      "loss": 5.7542,
      "mean_token_accuracy": 0.16877429485321044,
      "num_tokens": 62516226.0,
      "step": 18715
    },
    {
      "entropy": 6.201036024093628,
      "epoch": 3.6520339479075212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003754962553357849,
      "loss": 5.7687,
      "mean_token_accuracy": 0.17431914210319518,
      "num_tokens": 62533360.0,
      "step": 18720
    },
    {
      "entropy": 6.17165732383728,
      "epoch": 3.653009462491464,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003754333218345354,
      "loss": 5.7641,
      "mean_token_accuracy": 0.17827552407979966,
      "num_tokens": 62549175.0,
      "step": 18725
    },
    {
      "entropy": 6.114136457443237,
      "epoch": 3.653984977075407,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037537037851929965,
      "loss": 5.7251,
      "mean_token_accuracy": 0.18020714819431305,
      "num_tokens": 62566750.0,
      "step": 18730
    },
    {
      "entropy": 6.178707122802734,
      "epoch": 3.6549604916593506,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037530742539622835,
      "loss": 5.7221,
      "mean_token_accuracy": 0.1756117269396782,
      "num_tokens": 62582692.0,
      "step": 18735
    },
    {
      "entropy": 6.119781875610352,
      "epoch": 3.6559360062432935,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00037524446247147285,
      "loss": 5.7158,
      "mean_token_accuracy": 0.18280628472566604,
      "num_tokens": 62599500.0,
      "step": 18740
    },
    {
      "entropy": 6.183905935287475,
      "epoch": 3.6569115208272365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037518148975118577,
      "loss": 5.7934,
      "mean_token_accuracy": 0.17286243736743928,
      "num_tokens": 62617528.0,
      "step": 18745
    },
    {
      "entropy": 6.1565070152282715,
      "epoch": 3.6578870354111794,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003751185072415206,
      "loss": 5.6675,
      "mean_token_accuracy": 0.18233101665973664,
      "num_tokens": 62634062.0,
      "step": 18750
    },
    {
      "entropy": 6.111159133911133,
      "epoch": 3.6588625499951224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037505551494863165,
      "loss": 5.7184,
      "mean_token_accuracy": 0.1743771180510521,
      "num_tokens": 62650684.0,
      "step": 18755
    },
    {
      "entropy": 6.175211811065674,
      "epoch": 3.6598380645790654,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003749925128786744,
      "loss": 5.7567,
      "mean_token_accuracy": 0.16374513804912566,
      "num_tokens": 62666967.0,
      "step": 18760
    },
    {
      "entropy": 6.20865535736084,
      "epoch": 3.6608135791630083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037492950103780524,
      "loss": 5.7966,
      "mean_token_accuracy": 0.17446136772632598,
      "num_tokens": 62684128.0,
      "step": 18765
    },
    {
      "entropy": 6.139480113983154,
      "epoch": 3.6617890937469513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003748664794321813,
      "loss": 5.7538,
      "mean_token_accuracy": 0.17499858140945435,
      "num_tokens": 62700218.0,
      "step": 18770
    },
    {
      "entropy": 6.170604848861695,
      "epoch": 3.6627646083308947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037480344806796104,
      "loss": 5.7114,
      "mean_token_accuracy": 0.17027602940797806,
      "num_tokens": 62716326.0,
      "step": 18775
    },
    {
      "entropy": 6.242917776107788,
      "epoch": 3.6637401229148376,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037474040695130353,
      "loss": 5.8073,
      "mean_token_accuracy": 0.17059004306793213,
      "num_tokens": 62732534.0,
      "step": 18780
    },
    {
      "entropy": 6.186750602722168,
      "epoch": 3.6647156374987806,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003746773560883688,
      "loss": 5.7785,
      "mean_token_accuracy": 0.17760286778211593,
      "num_tokens": 62749018.0,
      "step": 18785
    },
    {
      "entropy": 6.159440755844116,
      "epoch": 3.6656911520827236,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003746142954853182,
      "loss": 5.7271,
      "mean_token_accuracy": 0.17448296546936035,
      "num_tokens": 62765122.0,
      "step": 18790
    },
    {
      "entropy": 6.168547773361206,
      "epoch": 3.6666666666666665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037455122514831366,
      "loss": 5.6808,
      "mean_token_accuracy": 0.17959486395120622,
      "num_tokens": 62781598.0,
      "step": 18795
    },
    {
      "entropy": 6.071695327758789,
      "epoch": 3.66764218125061,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037448814508351814,
      "loss": 5.6543,
      "mean_token_accuracy": 0.17931485772132874,
      "num_tokens": 62799060.0,
      "step": 18800
    },
    {
      "entropy": 6.198102903366089,
      "epoch": 3.668617695834553,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037442505529709574,
      "loss": 5.731,
      "mean_token_accuracy": 0.1801601842045784,
      "num_tokens": 62815265.0,
      "step": 18805
    },
    {
      "entropy": 6.209441518783569,
      "epoch": 3.669593210418496,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037436195579521124,
      "loss": 5.8603,
      "mean_token_accuracy": 0.16625114977359773,
      "num_tokens": 62832644.0,
      "step": 18810
    },
    {
      "entropy": 6.085378408432007,
      "epoch": 3.670568725002439,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037429884658403065,
      "loss": 5.6654,
      "mean_token_accuracy": 0.18191154599189757,
      "num_tokens": 62850522.0,
      "step": 18815
    },
    {
      "entropy": 5.912379932403565,
      "epoch": 3.6715442395863818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037423572766972054,
      "loss": 5.4518,
      "mean_token_accuracy": 0.21446070671081544,
      "num_tokens": 62868627.0,
      "step": 18820
    },
    {
      "entropy": 6.242873096466065,
      "epoch": 3.6725197541703247,
      "grad_norm": 1.0,
      "learning_rate": 0.0003741725990584489,
      "loss": 5.8141,
      "mean_token_accuracy": 0.17030806988477706,
      "num_tokens": 62884635.0,
      "step": 18825
    },
    {
      "entropy": 6.2476660251617435,
      "epoch": 3.6734952687542677,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003741094607563843,
      "loss": 5.9007,
      "mean_token_accuracy": 0.15804190039634705,
      "num_tokens": 62901345.0,
      "step": 18830
    },
    {
      "entropy": 6.155507278442383,
      "epoch": 3.6744707833382106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037404631276969643,
      "loss": 5.7473,
      "mean_token_accuracy": 0.18135112375020981,
      "num_tokens": 62917594.0,
      "step": 18835
    },
    {
      "entropy": 6.129670095443726,
      "epoch": 3.675446297922154,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037398315510455587,
      "loss": 5.7066,
      "mean_token_accuracy": 0.1758928820490837,
      "num_tokens": 62934705.0,
      "step": 18840
    },
    {
      "entropy": 6.152581024169922,
      "epoch": 3.676421812506097,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037391998776713426,
      "loss": 5.7168,
      "mean_token_accuracy": 0.17955984324216842,
      "num_tokens": 62950720.0,
      "step": 18845
    },
    {
      "entropy": 6.202474546432495,
      "epoch": 3.67739732709004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003738568107636039,
      "loss": 5.8076,
      "mean_token_accuracy": 0.1738844707608223,
      "num_tokens": 62965808.0,
      "step": 18850
    },
    {
      "entropy": 6.170988512039185,
      "epoch": 3.678372841673983,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003737936241001384,
      "loss": 5.7196,
      "mean_token_accuracy": 0.1794581264257431,
      "num_tokens": 62982506.0,
      "step": 18855
    },
    {
      "entropy": 6.214779043197632,
      "epoch": 3.6793483562579263,
      "grad_norm": 1.0625,
      "learning_rate": 0.000373730427782912,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1641818955540657,
      "num_tokens": 62997975.0,
      "step": 18860
    },
    {
      "entropy": 6.194931554794311,
      "epoch": 3.6803238708418693,
      "grad_norm": 1.0,
      "learning_rate": 0.00037366722181810013,
      "loss": 5.707,
      "mean_token_accuracy": 0.16786992251873017,
      "num_tokens": 63014838.0,
      "step": 18865
    },
    {
      "entropy": 6.229346752166748,
      "epoch": 3.6812993854258123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000373604006211879,
      "loss": 5.8083,
      "mean_token_accuracy": 0.16232922971248626,
      "num_tokens": 63031545.0,
      "step": 18870
    },
    {
      "entropy": 6.026651334762573,
      "epoch": 3.682274900009755,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037354078097042576,
      "loss": 5.4919,
      "mean_token_accuracy": 0.19304384291172028,
      "num_tokens": 63048834.0,
      "step": 18875
    },
    {
      "entropy": 6.2151988506317135,
      "epoch": 3.683250414593698,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037347754609991864,
      "loss": 5.7927,
      "mean_token_accuracy": 0.16554353386163712,
      "num_tokens": 63064043.0,
      "step": 18880
    },
    {
      "entropy": 6.139590883255005,
      "epoch": 3.684225929177641,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037341430160653665,
      "loss": 5.6994,
      "mean_token_accuracy": 0.18152327686548234,
      "num_tokens": 63079478.0,
      "step": 18885
    },
    {
      "entropy": 6.204279041290283,
      "epoch": 3.685201443761584,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037335104749645995,
      "loss": 5.8298,
      "mean_token_accuracy": 0.1696286156773567,
      "num_tokens": 63097301.0,
      "step": 18890
    },
    {
      "entropy": 6.1207164287567135,
      "epoch": 3.686176958345527,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003732877837758694,
      "loss": 5.686,
      "mean_token_accuracy": 0.17730615437030792,
      "num_tokens": 63114243.0,
      "step": 18895
    },
    {
      "entropy": 6.166020154953003,
      "epoch": 3.6871524729294705,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003732245104509468,
      "loss": 5.7304,
      "mean_token_accuracy": 0.17168393582105637,
      "num_tokens": 63130810.0,
      "step": 18900
    },
    {
      "entropy": 6.21659369468689,
      "epoch": 3.6881279875134134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037316122752787523,
      "loss": 5.762,
      "mean_token_accuracy": 0.17125632911920546,
      "num_tokens": 63147189.0,
      "step": 18905
    },
    {
      "entropy": 6.224981069564819,
      "epoch": 3.6891035020973564,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003730979350128384,
      "loss": 5.7993,
      "mean_token_accuracy": 0.17505227774381638,
      "num_tokens": 63164027.0,
      "step": 18910
    },
    {
      "entropy": 6.179618310928345,
      "epoch": 3.6900790166812993,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003730346329120208,
      "loss": 5.7876,
      "mean_token_accuracy": 0.17705890387296677,
      "num_tokens": 63182244.0,
      "step": 18915
    },
    {
      "entropy": 6.181617069244385,
      "epoch": 3.6910545312652423,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037297132123160847,
      "loss": 5.783,
      "mean_token_accuracy": 0.17318840324878693,
      "num_tokens": 63198764.0,
      "step": 18920
    },
    {
      "entropy": 6.2584569454193115,
      "epoch": 3.6920300458491857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037290799997778776,
      "loss": 5.7422,
      "mean_token_accuracy": 0.17250111550092698,
      "num_tokens": 63215316.0,
      "step": 18925
    },
    {
      "entropy": 6.170658445358276,
      "epoch": 3.6930055604331287,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037284466915674623,
      "loss": 5.7265,
      "mean_token_accuracy": 0.18013714849948884,
      "num_tokens": 63232241.0,
      "step": 18930
    },
    {
      "entropy": 6.0950898170471195,
      "epoch": 3.6939810750170716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003727813287746723,
      "loss": 5.6485,
      "mean_token_accuracy": 0.17832340002059938,
      "num_tokens": 63249077.0,
      "step": 18935
    },
    {
      "entropy": 6.061636209487915,
      "epoch": 3.6949565896010146,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003727179788377555,
      "loss": 5.648,
      "mean_token_accuracy": 0.18502646684646606,
      "num_tokens": 63266421.0,
      "step": 18940
    },
    {
      "entropy": 6.163741683959961,
      "epoch": 3.6959321041849575,
      "grad_norm": 0.984375,
      "learning_rate": 0.000372654619352186,
      "loss": 5.7352,
      "mean_token_accuracy": 0.17530270516872407,
      "num_tokens": 63281776.0,
      "step": 18945
    },
    {
      "entropy": 6.112560129165649,
      "epoch": 3.6969076187689005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003725912503241552,
      "loss": 5.6923,
      "mean_token_accuracy": 0.17925775349140166,
      "num_tokens": 63298200.0,
      "step": 18950
    },
    {
      "entropy": 6.039791774749756,
      "epoch": 3.6978831333528435,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037252787175985527,
      "loss": 5.5486,
      "mean_token_accuracy": 0.19176419377326964,
      "num_tokens": 63314905.0,
      "step": 18955
    },
    {
      "entropy": 6.112934970855713,
      "epoch": 3.6988586479367864,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003724644836654792,
      "loss": 5.754,
      "mean_token_accuracy": 0.17745828479528428,
      "num_tokens": 63331523.0,
      "step": 18960
    },
    {
      "entropy": 6.242333793640137,
      "epoch": 3.69983416252073,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003724010860472213,
      "loss": 5.7605,
      "mean_token_accuracy": 0.1702016994357109,
      "num_tokens": 63348850.0,
      "step": 18965
    },
    {
      "entropy": 6.14903998374939,
      "epoch": 3.700809677104673,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003723376789112763,
      "loss": 5.6634,
      "mean_token_accuracy": 0.18166787475347518,
      "num_tokens": 63366918.0,
      "step": 18970
    },
    {
      "entropy": 6.194267892837525,
      "epoch": 3.7017851916886158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003722742622638403,
      "loss": 5.7773,
      "mean_token_accuracy": 0.17285212725400925,
      "num_tokens": 63385398.0,
      "step": 18975
    },
    {
      "entropy": 6.200123929977417,
      "epoch": 3.7027607062725587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037221083611111,
      "loss": 5.7904,
      "mean_token_accuracy": 0.17091584652662278,
      "num_tokens": 63402075.0,
      "step": 18980
    },
    {
      "entropy": 6.143798303604126,
      "epoch": 3.7037362208565017,
      "grad_norm": 1.0,
      "learning_rate": 0.00037214740045928333,
      "loss": 5.6372,
      "mean_token_accuracy": 0.1840943694114685,
      "num_tokens": 63418833.0,
      "step": 18985
    },
    {
      "entropy": 6.2149872303009035,
      "epoch": 3.704711735440445,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003720839553145589,
      "loss": 5.7985,
      "mean_token_accuracy": 0.17243146002292634,
      "num_tokens": 63434112.0,
      "step": 18990
    },
    {
      "entropy": 6.078496217727661,
      "epoch": 3.705687250024388,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003720205006831364,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1917930379509926,
      "num_tokens": 63450737.0,
      "step": 18995
    },
    {
      "entropy": 6.191920423507691,
      "epoch": 3.706662764608331,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003719570365712164,
      "loss": 5.7708,
      "mean_token_accuracy": 0.17267039120197297,
      "num_tokens": 63466553.0,
      "step": 19000
    },
    {
      "entropy": 6.092858982086182,
      "epoch": 3.707638279192274,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003718935629850003,
      "loss": 5.6553,
      "mean_token_accuracy": 0.1898433357477188,
      "num_tokens": 63482179.0,
      "step": 19005
    },
    {
      "entropy": 6.117489767074585,
      "epoch": 3.708613793776217,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003718300799306905,
      "loss": 5.6557,
      "mean_token_accuracy": 0.17981858998537065,
      "num_tokens": 63499435.0,
      "step": 19010
    },
    {
      "entropy": 6.086596584320068,
      "epoch": 3.70958930836016,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037176658741449036,
      "loss": 5.7039,
      "mean_token_accuracy": 0.1804579883813858,
      "num_tokens": 63518351.0,
      "step": 19015
    },
    {
      "entropy": 6.115773153305054,
      "epoch": 3.710564822944103,
      "grad_norm": 0.84375,
      "learning_rate": 0.00037170308544260436,
      "loss": 5.5996,
      "mean_token_accuracy": 0.1873631864786148,
      "num_tokens": 63534133.0,
      "step": 19020
    },
    {
      "entropy": 6.2057641506195065,
      "epoch": 3.711540337528046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003716395740212374,
      "loss": 5.7726,
      "mean_token_accuracy": 0.17013194561004638,
      "num_tokens": 63551076.0,
      "step": 19025
    },
    {
      "entropy": 6.244019794464111,
      "epoch": 3.712515852111989,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037157605315659564,
      "loss": 5.8399,
      "mean_token_accuracy": 0.1587700664997101,
      "num_tokens": 63567380.0,
      "step": 19030
    },
    {
      "entropy": 6.234502840042114,
      "epoch": 3.713491366695932,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003715125228548863,
      "loss": 5.8178,
      "mean_token_accuracy": 0.16893676817417144,
      "num_tokens": 63583318.0,
      "step": 19035
    },
    {
      "entropy": 6.2207941055297855,
      "epoch": 3.714466881279875,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037144898312231714,
      "loss": 5.7919,
      "mean_token_accuracy": 0.16880303323268891,
      "num_tokens": 63601667.0,
      "step": 19040
    },
    {
      "entropy": 6.249489831924438,
      "epoch": 3.715442395863818,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003713854339650971,
      "loss": 5.8732,
      "mean_token_accuracy": 0.16644677221775056,
      "num_tokens": 63620335.0,
      "step": 19045
    },
    {
      "entropy": 6.230344581604004,
      "epoch": 3.716417910447761,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037132187538943594,
      "loss": 5.7889,
      "mean_token_accuracy": 0.1803138494491577,
      "num_tokens": 63637337.0,
      "step": 19050
    },
    {
      "entropy": 6.0144425392150875,
      "epoch": 3.7173934250317044,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003712583074015445,
      "loss": 5.5746,
      "mean_token_accuracy": 0.19269518554210663,
      "num_tokens": 63654358.0,
      "step": 19055
    },
    {
      "entropy": 6.24419150352478,
      "epoch": 3.7183689396156474,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037119473000763424,
      "loss": 5.8708,
      "mean_token_accuracy": 0.17024473696947098,
      "num_tokens": 63670448.0,
      "step": 19060
    },
    {
      "entropy": 6.126083326339722,
      "epoch": 3.7193444541995904,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003711311432139179,
      "loss": 5.6306,
      "mean_token_accuracy": 0.18666301518678666,
      "num_tokens": 63687912.0,
      "step": 19065
    },
    {
      "entropy": 6.195589017868042,
      "epoch": 3.7203199687835333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037106754702660884,
      "loss": 5.8143,
      "mean_token_accuracy": 0.1679670825600624,
      "num_tokens": 63703949.0,
      "step": 19070
    },
    {
      "entropy": 6.127406120300293,
      "epoch": 3.7212954833674763,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003710039414519215,
      "loss": 5.7004,
      "mean_token_accuracy": 0.1759282261133194,
      "num_tokens": 63722182.0,
      "step": 19075
    },
    {
      "entropy": 6.161605405807495,
      "epoch": 3.7222709979514192,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000370940326496071,
      "loss": 5.6986,
      "mean_token_accuracy": 0.1788578897714615,
      "num_tokens": 63738055.0,
      "step": 19080
    },
    {
      "entropy": 6.21934142112732,
      "epoch": 3.723246512535362,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037087670216527386,
      "loss": 5.8266,
      "mean_token_accuracy": 0.1655665934085846,
      "num_tokens": 63755570.0,
      "step": 19085
    },
    {
      "entropy": 6.192415237426758,
      "epoch": 3.724222027119305,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003708130684657471,
      "loss": 5.7597,
      "mean_token_accuracy": 0.17815887928009033,
      "num_tokens": 63771032.0,
      "step": 19090
    },
    {
      "entropy": 6.1888463497161865,
      "epoch": 3.7251975417032486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037074942540370876,
      "loss": 5.721,
      "mean_token_accuracy": 0.17397109568119049,
      "num_tokens": 63788402.0,
      "step": 19095
    },
    {
      "entropy": 6.141082286834717,
      "epoch": 3.7261730562871915,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003706857729853778,
      "loss": 5.7069,
      "mean_token_accuracy": 0.17947519570589066,
      "num_tokens": 63805609.0,
      "step": 19100
    },
    {
      "entropy": 6.171285820007324,
      "epoch": 3.7271485708711345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003706221112169742,
      "loss": 5.7569,
      "mean_token_accuracy": 0.1772899955511093,
      "num_tokens": 63821146.0,
      "step": 19105
    },
    {
      "entropy": 6.098960447311401,
      "epoch": 3.7281240854550775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003705584401047186,
      "loss": 5.6696,
      "mean_token_accuracy": 0.17950680255889892,
      "num_tokens": 63838948.0,
      "step": 19110
    },
    {
      "entropy": 6.152759838104248,
      "epoch": 3.729099600039021,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003704947596548329,
      "loss": 5.7331,
      "mean_token_accuracy": 0.17483630627393723,
      "num_tokens": 63854887.0,
      "step": 19115
    },
    {
      "entropy": 6.133314943313598,
      "epoch": 3.730075114622964,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003704310698735395,
      "loss": 5.6825,
      "mean_token_accuracy": 0.18291305154561996,
      "num_tokens": 63873903.0,
      "step": 19120
    },
    {
      "entropy": 6.165752315521241,
      "epoch": 3.7310506292069068,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003703673707670622,
      "loss": 5.7432,
      "mean_token_accuracy": 0.172866028547287,
      "num_tokens": 63890540.0,
      "step": 19125
    },
    {
      "entropy": 6.128530263900757,
      "epoch": 3.7320261437908497,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037030366234162533,
      "loss": 5.6776,
      "mean_token_accuracy": 0.17975136637687683,
      "num_tokens": 63905842.0,
      "step": 19130
    },
    {
      "entropy": 6.15168981552124,
      "epoch": 3.7330016583747927,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003702399446034542,
      "loss": 5.635,
      "mean_token_accuracy": 0.18445043563842772,
      "num_tokens": 63921505.0,
      "step": 19135
    },
    {
      "entropy": 6.122580480575562,
      "epoch": 3.7339771729587357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003701762175587752,
      "loss": 5.7099,
      "mean_token_accuracy": 0.17608538568019866,
      "num_tokens": 63936375.0,
      "step": 19140
    },
    {
      "entropy": 6.1911139488220215,
      "epoch": 3.7349526875426786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003701124812138153,
      "loss": 5.8422,
      "mean_token_accuracy": 0.16344065368175506,
      "num_tokens": 63953129.0,
      "step": 19145
    },
    {
      "entropy": 6.331884908676147,
      "epoch": 3.7359282021266216,
      "grad_norm": 0.875,
      "learning_rate": 0.0003700487355748028,
      "loss": 5.9707,
      "mean_token_accuracy": 0.15672512203454972,
      "num_tokens": 63969806.0,
      "step": 19150
    },
    {
      "entropy": 6.252671146392823,
      "epoch": 3.736903716710565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036998498064796665,
      "loss": 5.8237,
      "mean_token_accuracy": 0.16960125118494035,
      "num_tokens": 63985536.0,
      "step": 19155
    },
    {
      "entropy": 6.218660593032837,
      "epoch": 3.737879231294508,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036992121643953665,
      "loss": 5.8098,
      "mean_token_accuracy": 0.16931730955839158,
      "num_tokens": 64002346.0,
      "step": 19160
    },
    {
      "entropy": 6.256877088546753,
      "epoch": 3.738854745878451,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003698574429557437,
      "loss": 5.8191,
      "mean_token_accuracy": 0.16829742044210433,
      "num_tokens": 64019430.0,
      "step": 19165
    },
    {
      "entropy": 6.018768167495727,
      "epoch": 3.739830260462394,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036979366020281954,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1956884428858757,
      "num_tokens": 64036802.0,
      "step": 19170
    },
    {
      "entropy": 6.176215744018554,
      "epoch": 3.740805775046337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003697298681869967,
      "loss": 5.7798,
      "mean_token_accuracy": 0.171764799952507,
      "num_tokens": 64053818.0,
      "step": 19175
    },
    {
      "entropy": 6.12287220954895,
      "epoch": 3.7417812896302802,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003696660669145088,
      "loss": 5.6837,
      "mean_token_accuracy": 0.17977697402238846,
      "num_tokens": 64070424.0,
      "step": 19180
    },
    {
      "entropy": 6.200584650039673,
      "epoch": 3.742756804214223,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003696022563915903,
      "loss": 5.7865,
      "mean_token_accuracy": 0.16942883133888245,
      "num_tokens": 64089103.0,
      "step": 19185
    },
    {
      "entropy": 6.1454843044281,
      "epoch": 3.743732318798166,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036953843662447636,
      "loss": 5.6335,
      "mean_token_accuracy": 0.18604010641574859,
      "num_tokens": 64105743.0,
      "step": 19190
    },
    {
      "entropy": 6.104838562011719,
      "epoch": 3.744707833382109,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003694746076194034,
      "loss": 5.7345,
      "mean_token_accuracy": 0.17882996946573257,
      "num_tokens": 64124535.0,
      "step": 19195
    },
    {
      "entropy": 6.269287776947022,
      "epoch": 3.745683347966052,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036941076938260854,
      "loss": 5.8059,
      "mean_token_accuracy": 0.17711586952209474,
      "num_tokens": 64142430.0,
      "step": 19200
    },
    {
      "entropy": 6.126835823059082,
      "epoch": 3.746658862549995,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003693469219203297,
      "loss": 5.6989,
      "mean_token_accuracy": 0.17547338902950288,
      "num_tokens": 64159592.0,
      "step": 19205
    },
    {
      "entropy": 6.09461088180542,
      "epoch": 3.747634377133938,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036928306523880593,
      "loss": 5.6331,
      "mean_token_accuracy": 0.17592001259326934,
      "num_tokens": 64176023.0,
      "step": 19210
    },
    {
      "entropy": 6.079236173629761,
      "epoch": 3.748609891717881,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036921919934427704,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1921295404434204,
      "num_tokens": 64192372.0,
      "step": 19215
    },
    {
      "entropy": 6.212511730194092,
      "epoch": 3.7495854063018244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036915532424298383,
      "loss": 5.761,
      "mean_token_accuracy": 0.17496318370103836,
      "num_tokens": 64210420.0,
      "step": 19220
    },
    {
      "entropy": 6.177423334121704,
      "epoch": 3.7505609208857673,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003690914399411678,
      "loss": 5.801,
      "mean_token_accuracy": 0.16946877986192704,
      "num_tokens": 64228281.0,
      "step": 19225
    },
    {
      "entropy": 6.094173002243042,
      "epoch": 3.7515364354697103,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036902754644507177,
      "loss": 5.618,
      "mean_token_accuracy": 0.18642444908618927,
      "num_tokens": 64244002.0,
      "step": 19230
    },
    {
      "entropy": 6.182182550430298,
      "epoch": 3.7525119500536532,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036896364376093896,
      "loss": 5.7431,
      "mean_token_accuracy": 0.17683208733797073,
      "num_tokens": 64259849.0,
      "step": 19235
    },
    {
      "entropy": 6.159491062164307,
      "epoch": 3.753487464637596,
      "grad_norm": 0.859375,
      "learning_rate": 0.00036889973189501373,
      "loss": 5.6823,
      "mean_token_accuracy": 0.18366126269102095,
      "num_tokens": 64278760.0,
      "step": 19240
    },
    {
      "entropy": 6.08899154663086,
      "epoch": 3.7544629792215396,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003688358108535414,
      "loss": 5.6385,
      "mean_token_accuracy": 0.1830077812075615,
      "num_tokens": 64296633.0,
      "step": 19245
    },
    {
      "entropy": 6.201806592941284,
      "epoch": 3.7554384938054826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036877188064276807,
      "loss": 5.7241,
      "mean_token_accuracy": 0.17893355041742326,
      "num_tokens": 64313278.0,
      "step": 19250
    },
    {
      "entropy": 6.055192613601685,
      "epoch": 3.7564140083894255,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003687079412689408,
      "loss": 5.6191,
      "mean_token_accuracy": 0.1883893594145775,
      "num_tokens": 64331521.0,
      "step": 19255
    },
    {
      "entropy": 6.059469413757324,
      "epoch": 3.7573895229733685,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036864399273830745,
      "loss": 5.5484,
      "mean_token_accuracy": 0.19158386141061784,
      "num_tokens": 64347633.0,
      "step": 19260
    },
    {
      "entropy": 6.215648078918457,
      "epoch": 3.7583650375573114,
      "grad_norm": 1.0,
      "learning_rate": 0.00036858003505711685,
      "loss": 5.8518,
      "mean_token_accuracy": 0.16251820474863052,
      "num_tokens": 64363955.0,
      "step": 19265
    },
    {
      "entropy": 6.112931108474731,
      "epoch": 3.7593405521412544,
      "grad_norm": 0.875,
      "learning_rate": 0.0003685160682316188,
      "loss": 5.6553,
      "mean_token_accuracy": 0.18214449137449265,
      "num_tokens": 64382109.0,
      "step": 19270
    },
    {
      "entropy": 6.249288272857666,
      "epoch": 3.7603160667251974,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036845209226806386,
      "loss": 5.8081,
      "mean_token_accuracy": 0.16752441823482514,
      "num_tokens": 64398912.0,
      "step": 19275
    },
    {
      "entropy": 6.177203035354614,
      "epoch": 3.7612915813091403,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003683881071727035,
      "loss": 5.686,
      "mean_token_accuracy": 0.17930564880371094,
      "num_tokens": 64415809.0,
      "step": 19280
    },
    {
      "entropy": 6.015184116363526,
      "epoch": 3.7622670958930837,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003683241129517902,
      "loss": 5.4995,
      "mean_token_accuracy": 0.193967205286026,
      "num_tokens": 64431900.0,
      "step": 19285
    },
    {
      "entropy": 6.186290454864502,
      "epoch": 3.7632426104770267,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036826010961157717,
      "loss": 5.74,
      "mean_token_accuracy": 0.17544918954372407,
      "num_tokens": 64447491.0,
      "step": 19290
    },
    {
      "entropy": 6.157180118560791,
      "epoch": 3.7642181250609696,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036819609715831866,
      "loss": 5.7813,
      "mean_token_accuracy": 0.17075841426849364,
      "num_tokens": 64465121.0,
      "step": 19295
    },
    {
      "entropy": 6.1223071098327635,
      "epoch": 3.7651936396449126,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036813207559826963,
      "loss": 5.6714,
      "mean_token_accuracy": 0.1772956520318985,
      "num_tokens": 64481044.0,
      "step": 19300
    },
    {
      "entropy": 6.024498987197876,
      "epoch": 3.7661691542288556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003680680449376861,
      "loss": 5.5901,
      "mean_token_accuracy": 0.18293003290891646,
      "num_tokens": 64498228.0,
      "step": 19305
    },
    {
      "entropy": 6.120113039016724,
      "epoch": 3.767144668812799,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036800400518282494,
      "loss": 5.7092,
      "mean_token_accuracy": 0.1789330691099167,
      "num_tokens": 64514815.0,
      "step": 19310
    },
    {
      "entropy": 6.165699052810669,
      "epoch": 3.768120183396742,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036793995633994383,
      "loss": 5.7612,
      "mean_token_accuracy": 0.18361404687166213,
      "num_tokens": 64533453.0,
      "step": 19315
    },
    {
      "entropy": 6.220706558227539,
      "epoch": 3.769095697980685,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003678758984153015,
      "loss": 5.8328,
      "mean_token_accuracy": 0.16914756447076798,
      "num_tokens": 64551118.0,
      "step": 19320
    },
    {
      "entropy": 6.1441600799560545,
      "epoch": 3.770071212564628,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003678118314151573,
      "loss": 5.7634,
      "mean_token_accuracy": 0.18139224946498872,
      "num_tokens": 64567472.0,
      "step": 19325
    },
    {
      "entropy": 6.1594444751739506,
      "epoch": 3.771046727148571,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036774775534577174,
      "loss": 5.7872,
      "mean_token_accuracy": 0.17729971557855606,
      "num_tokens": 64583526.0,
      "step": 19330
    },
    {
      "entropy": 6.229633045196533,
      "epoch": 3.7720222417325138,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036768367021340615,
      "loss": 5.7828,
      "mean_token_accuracy": 0.1729291632771492,
      "num_tokens": 64600647.0,
      "step": 19335
    },
    {
      "entropy": 6.072004508972168,
      "epoch": 3.7729977563164567,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003676195760243226,
      "loss": 5.5665,
      "mean_token_accuracy": 0.18876960575580598,
      "num_tokens": 64616963.0,
      "step": 19340
    },
    {
      "entropy": 6.117981481552124,
      "epoch": 3.7739732709003997,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003675554727847842,
      "loss": 5.6846,
      "mean_token_accuracy": 0.18367644846439363,
      "num_tokens": 64634078.0,
      "step": 19345
    },
    {
      "entropy": 6.207468032836914,
      "epoch": 3.774948785484343,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036749136050105485,
      "loss": 5.7667,
      "mean_token_accuracy": 0.1712219938635826,
      "num_tokens": 64650056.0,
      "step": 19350
    },
    {
      "entropy": 6.205478715896606,
      "epoch": 3.775924300068286,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003674272391793994,
      "loss": 5.8077,
      "mean_token_accuracy": 0.1721440151333809,
      "num_tokens": 64667667.0,
      "step": 19355
    },
    {
      "entropy": 5.962297868728638,
      "epoch": 3.776899814652229,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0003673631088260836,
      "loss": 5.5376,
      "mean_token_accuracy": 0.2037075161933899,
      "num_tokens": 64685618.0,
      "step": 19360
    },
    {
      "entropy": 6.201483488082886,
      "epoch": 3.777875329236172,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036729896944737395,
      "loss": 5.762,
      "mean_token_accuracy": 0.1729600742459297,
      "num_tokens": 64702551.0,
      "step": 19365
    },
    {
      "entropy": 6.245392084121704,
      "epoch": 3.7788508438201154,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003672348210495379,
      "loss": 5.8662,
      "mean_token_accuracy": 0.16608483642339705,
      "num_tokens": 64719648.0,
      "step": 19370
    },
    {
      "entropy": 6.186542129516601,
      "epoch": 3.7798263584040583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000367170663638844,
      "loss": 5.8086,
      "mean_token_accuracy": 0.17134677469730378,
      "num_tokens": 64736047.0,
      "step": 19375
    },
    {
      "entropy": 6.091601419448852,
      "epoch": 3.7808018729880013,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003671064972215613,
      "loss": 5.7126,
      "mean_token_accuracy": 0.17681532949209214,
      "num_tokens": 64751832.0,
      "step": 19380
    },
    {
      "entropy": 6.185369443893433,
      "epoch": 3.7817773875719443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036704232180396004,
      "loss": 5.7802,
      "mean_token_accuracy": 0.16665244549512864,
      "num_tokens": 64768744.0,
      "step": 19385
    },
    {
      "entropy": 6.096948146820068,
      "epoch": 3.782752902155887,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000366978137392311,
      "loss": 5.6522,
      "mean_token_accuracy": 0.17886640280485153,
      "num_tokens": 64785133.0,
      "step": 19390
    },
    {
      "entropy": 6.103321218490601,
      "epoch": 3.78372841673983,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036691394399288626,
      "loss": 5.6952,
      "mean_token_accuracy": 0.18349580317735673,
      "num_tokens": 64801965.0,
      "step": 19395
    },
    {
      "entropy": 6.238744831085205,
      "epoch": 3.784703931323773,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036684974161195846,
      "loss": 5.8257,
      "mean_token_accuracy": 0.16856731474399567,
      "num_tokens": 64819821.0,
      "step": 19400
    },
    {
      "entropy": 6.1713825225830075,
      "epoch": 3.785679445907716,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036678553025580124,
      "loss": 5.6874,
      "mean_token_accuracy": 0.17205069214105606,
      "num_tokens": 64836115.0,
      "step": 19405
    },
    {
      "entropy": 6.234222650527954,
      "epoch": 3.7866549604916595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036672130993068927,
      "loss": 5.9097,
      "mean_token_accuracy": 0.15887837409973143,
      "num_tokens": 64853225.0,
      "step": 19410
    },
    {
      "entropy": 6.24246883392334,
      "epoch": 3.7876304750756025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036665708064289774,
      "loss": 5.801,
      "mean_token_accuracy": 0.17483863830566407,
      "num_tokens": 64870400.0,
      "step": 19415
    },
    {
      "entropy": 6.0683502674102785,
      "epoch": 3.7886059896595454,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003665928423987029,
      "loss": 5.5632,
      "mean_token_accuracy": 0.19215108305215836,
      "num_tokens": 64887567.0,
      "step": 19420
    },
    {
      "entropy": 5.966378831863404,
      "epoch": 3.7895815042434884,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0003665285952043819,
      "loss": 5.5475,
      "mean_token_accuracy": 0.18845386505126954,
      "num_tokens": 64903700.0,
      "step": 19425
    },
    {
      "entropy": 6.177962350845337,
      "epoch": 3.7905570188274313,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003664643390662129,
      "loss": 5.7662,
      "mean_token_accuracy": 0.17413062155246734,
      "num_tokens": 64921386.0,
      "step": 19430
    },
    {
      "entropy": 6.142558526992798,
      "epoch": 3.7915325334113748,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036640007399047457,
      "loss": 5.6905,
      "mean_token_accuracy": 0.18111087679862975,
      "num_tokens": 64938180.0,
      "step": 19435
    },
    {
      "entropy": 6.226398706436157,
      "epoch": 3.7925080479953177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036633579998344686,
      "loss": 5.8225,
      "mean_token_accuracy": 0.16604001075029373,
      "num_tokens": 64954489.0,
      "step": 19440
    },
    {
      "entropy": 6.125317239761353,
      "epoch": 3.7934835625792607,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003662715170514103,
      "loss": 5.6608,
      "mean_token_accuracy": 0.18179340213537215,
      "num_tokens": 64971492.0,
      "step": 19445
    },
    {
      "entropy": 6.2093006610870365,
      "epoch": 3.7944590771632036,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003662072252006464,
      "loss": 5.7816,
      "mean_token_accuracy": 0.16787705719470977,
      "num_tokens": 64988322.0,
      "step": 19450
    },
    {
      "entropy": 6.21467604637146,
      "epoch": 3.7954345917471466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003661429244374374,
      "loss": 5.8827,
      "mean_token_accuracy": 0.16956004053354262,
      "num_tokens": 65004973.0,
      "step": 19455
    },
    {
      "entropy": 6.196844434738159,
      "epoch": 3.7964101063310896,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003660786147680668,
      "loss": 5.7807,
      "mean_token_accuracy": 0.1742778405547142,
      "num_tokens": 65021532.0,
      "step": 19460
    },
    {
      "entropy": 6.136039686203003,
      "epoch": 3.7973856209150325,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036601429619881847,
      "loss": 5.6387,
      "mean_token_accuracy": 0.1810624673962593,
      "num_tokens": 65037896.0,
      "step": 19465
    },
    {
      "entropy": 6.215223693847657,
      "epoch": 3.7983611354989755,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003659499687359776,
      "loss": 5.7528,
      "mean_token_accuracy": 0.17306102216243743,
      "num_tokens": 65054000.0,
      "step": 19470
    },
    {
      "entropy": 6.189092063903809,
      "epoch": 3.799336650082919,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003658856323858299,
      "loss": 5.8174,
      "mean_token_accuracy": 0.16936113089323043,
      "num_tokens": 65070818.0,
      "step": 19475
    },
    {
      "entropy": 6.150882434844971,
      "epoch": 3.800312164666862,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036582128715466214,
      "loss": 5.7007,
      "mean_token_accuracy": 0.18467211425304414,
      "num_tokens": 65087129.0,
      "step": 19480
    },
    {
      "entropy": 6.225251245498657,
      "epoch": 3.801287679250805,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003657569330487618,
      "loss": 5.885,
      "mean_token_accuracy": 0.16786122769117356,
      "num_tokens": 65103541.0,
      "step": 19485
    },
    {
      "entropy": 6.197882747650146,
      "epoch": 3.8022631938347478,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036569257007441745,
      "loss": 5.741,
      "mean_token_accuracy": 0.17174642384052277,
      "num_tokens": 65118924.0,
      "step": 19490
    },
    {
      "entropy": 6.1359827518463135,
      "epoch": 3.8032387084186907,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003656281982379185,
      "loss": 5.6917,
      "mean_token_accuracy": 0.18275800794363023,
      "num_tokens": 65134690.0,
      "step": 19495
    },
    {
      "entropy": 6.143512487411499,
      "epoch": 3.804214223002634,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036556381754555497,
      "loss": 5.6948,
      "mean_token_accuracy": 0.17247324883937837,
      "num_tokens": 65150495.0,
      "step": 19500
    },
    {
      "entropy": 6.174287986755371,
      "epoch": 3.805189737586577,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036549942800361793,
      "loss": 5.7473,
      "mean_token_accuracy": 0.17312154173851013,
      "num_tokens": 65166148.0,
      "step": 19505
    },
    {
      "entropy": 6.219706583023071,
      "epoch": 3.80616525217052,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036543502961839937,
      "loss": 5.7686,
      "mean_token_accuracy": 0.1667141556739807,
      "num_tokens": 65182136.0,
      "step": 19510
    },
    {
      "entropy": 6.2411871433258055,
      "epoch": 3.807140766754463,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036537062239619194,
      "loss": 5.7782,
      "mean_token_accuracy": 0.17310404330492019,
      "num_tokens": 65198884.0,
      "step": 19515
    },
    {
      "entropy": 6.181934499740601,
      "epoch": 3.808116281338406,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036530620634328944,
      "loss": 5.7543,
      "mean_token_accuracy": 0.17281676977872848,
      "num_tokens": 65217274.0,
      "step": 19520
    },
    {
      "entropy": 6.138265705108642,
      "epoch": 3.809091795922349,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003652417814659864,
      "loss": 5.7416,
      "mean_token_accuracy": 0.17566783130168914,
      "num_tokens": 65234052.0,
      "step": 19525
    },
    {
      "entropy": 6.236972856521606,
      "epoch": 3.810067310506292,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036517734777057796,
      "loss": 5.7355,
      "mean_token_accuracy": 0.17263084501028061,
      "num_tokens": 65251103.0,
      "step": 19530
    },
    {
      "entropy": 6.175067949295044,
      "epoch": 3.811042825090235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003651129052633605,
      "loss": 5.7361,
      "mean_token_accuracy": 0.17661517411470412,
      "num_tokens": 65266735.0,
      "step": 19535
    },
    {
      "entropy": 6.132193899154663,
      "epoch": 3.8120183396741782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003650484539506312,
      "loss": 5.6658,
      "mean_token_accuracy": 0.18050827234983444,
      "num_tokens": 65283990.0,
      "step": 19540
    },
    {
      "entropy": 6.0823088645935055,
      "epoch": 3.812993854258121,
      "grad_norm": 1.0,
      "learning_rate": 0.00036498399383868787,
      "loss": 5.6984,
      "mean_token_accuracy": 0.18643890470266342,
      "num_tokens": 65301394.0,
      "step": 19545
    },
    {
      "entropy": 6.179421329498291,
      "epoch": 3.813969368842064,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003649195249338294,
      "loss": 5.7206,
      "mean_token_accuracy": 0.17520619183778763,
      "num_tokens": 65317809.0,
      "step": 19550
    },
    {
      "entropy": 6.089091062545776,
      "epoch": 3.814944883426007,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036485504724235535,
      "loss": 5.6425,
      "mean_token_accuracy": 0.18107644468545914,
      "num_tokens": 65335129.0,
      "step": 19555
    },
    {
      "entropy": 6.06548171043396,
      "epoch": 3.81592039800995,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003647905607705664,
      "loss": 5.5769,
      "mean_token_accuracy": 0.1939732626080513,
      "num_tokens": 65352057.0,
      "step": 19560
    },
    {
      "entropy": 6.212250518798828,
      "epoch": 3.8168959125938935,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036472606552476384,
      "loss": 5.7811,
      "mean_token_accuracy": 0.17729623317718507,
      "num_tokens": 65369386.0,
      "step": 19565
    },
    {
      "entropy": 6.135092401504517,
      "epoch": 3.8178714271778365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036466156151125,
      "loss": 5.6861,
      "mean_token_accuracy": 0.1763506516814232,
      "num_tokens": 65385775.0,
      "step": 19570
    },
    {
      "entropy": 6.190204763412476,
      "epoch": 3.8188469417617794,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036459704873632794,
      "loss": 5.807,
      "mean_token_accuracy": 0.16665092557668687,
      "num_tokens": 65401795.0,
      "step": 19575
    },
    {
      "entropy": 6.23859052658081,
      "epoch": 3.8198224563457224,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003645325272063016,
      "loss": 5.8161,
      "mean_token_accuracy": 0.1700226068496704,
      "num_tokens": 65416772.0,
      "step": 19580
    },
    {
      "entropy": 5.932537078857422,
      "epoch": 3.8207979709296653,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036446799692747573,
      "loss": 5.4796,
      "mean_token_accuracy": 0.19493983089923858,
      "num_tokens": 65432485.0,
      "step": 19585
    },
    {
      "entropy": 6.156635093688965,
      "epoch": 3.8217734855136083,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003644034579061561,
      "loss": 5.6983,
      "mean_token_accuracy": 0.18389738500118255,
      "num_tokens": 65449141.0,
      "step": 19590
    },
    {
      "entropy": 6.0514445304870605,
      "epoch": 3.8227490000975513,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003643389101486492,
      "loss": 5.5326,
      "mean_token_accuracy": 0.1980648785829544,
      "num_tokens": 65467753.0,
      "step": 19595
    },
    {
      "entropy": 6.146245431900025,
      "epoch": 3.823724514681494,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003642743536612624,
      "loss": 5.72,
      "mean_token_accuracy": 0.17534969449043275,
      "num_tokens": 65485577.0,
      "step": 19600
    },
    {
      "entropy": 6.170744180679321,
      "epoch": 3.8247000292654376,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000364209788450304,
      "loss": 5.7376,
      "mean_token_accuracy": 0.17869292199611664,
      "num_tokens": 65501824.0,
      "step": 19605
    },
    {
      "entropy": 6.088320207595825,
      "epoch": 3.8256755438493806,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000364145214522083,
      "loss": 5.6555,
      "mean_token_accuracy": 0.18185194730758666,
      "num_tokens": 65518710.0,
      "step": 19610
    },
    {
      "entropy": 6.0475856304168705,
      "epoch": 3.8266510584333235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036408063188290925,
      "loss": 5.5834,
      "mean_token_accuracy": 0.19100072830915452,
      "num_tokens": 65534914.0,
      "step": 19615
    },
    {
      "entropy": 5.970926523208618,
      "epoch": 3.8276265730172665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003640160405390938,
      "loss": 5.5919,
      "mean_token_accuracy": 0.19211532026529313,
      "num_tokens": 65550990.0,
      "step": 19620
    },
    {
      "entropy": 6.170543622970581,
      "epoch": 3.82860208760121,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000363951440496948,
      "loss": 5.7569,
      "mean_token_accuracy": 0.17378720343112947,
      "num_tokens": 65569255.0,
      "step": 19625
    },
    {
      "entropy": 6.205952215194702,
      "epoch": 3.829577602185153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003638868317627845,
      "loss": 5.7477,
      "mean_token_accuracy": 0.1773354545235634,
      "num_tokens": 65585926.0,
      "step": 19630
    },
    {
      "entropy": 6.106066465377808,
      "epoch": 3.830553116769096,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036382221434291665,
      "loss": 5.6757,
      "mean_token_accuracy": 0.18311050832271575,
      "num_tokens": 65602530.0,
      "step": 19635
    },
    {
      "entropy": 6.218972873687744,
      "epoch": 3.831528631353039,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003637575882436585,
      "loss": 5.8224,
      "mean_token_accuracy": 0.16131606101989746,
      "num_tokens": 65619566.0,
      "step": 19640
    },
    {
      "entropy": 6.1237386703491214,
      "epoch": 3.8325041459369817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036369295347132526,
      "loss": 5.6549,
      "mean_token_accuracy": 0.1910523533821106,
      "num_tokens": 65636631.0,
      "step": 19645
    },
    {
      "entropy": 6.145188236236573,
      "epoch": 3.8334796605209247,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036362831003223264,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1660337671637535,
      "num_tokens": 65652536.0,
      "step": 19650
    },
    {
      "entropy": 6.153304100036621,
      "epoch": 3.8344551751048677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003635636579326975,
      "loss": 5.6825,
      "mean_token_accuracy": 0.18126087486743928,
      "num_tokens": 65667934.0,
      "step": 19655
    },
    {
      "entropy": 6.01853723526001,
      "epoch": 3.8354306896888106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036349899717903734,
      "loss": 5.5255,
      "mean_token_accuracy": 0.19336415976285934,
      "num_tokens": 65683508.0,
      "step": 19660
    },
    {
      "entropy": 6.164502763748169,
      "epoch": 3.836406204272754,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036343432777757054,
      "loss": 5.6898,
      "mean_token_accuracy": 0.17403995543718337,
      "num_tokens": 65698198.0,
      "step": 19665
    },
    {
      "entropy": 6.0656617164611815,
      "epoch": 3.837381718856697,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003633696497346164,
      "loss": 5.6864,
      "mean_token_accuracy": 0.18436174243688583,
      "num_tokens": 65715867.0,
      "step": 19670
    },
    {
      "entropy": 6.236215448379516,
      "epoch": 3.83835723344064,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003633049630564951,
      "loss": 5.7585,
      "mean_token_accuracy": 0.16947590708732604,
      "num_tokens": 65732466.0,
      "step": 19675
    },
    {
      "entropy": 6.1683831214904785,
      "epoch": 3.839332748024583,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036324026774952754,
      "loss": 5.7158,
      "mean_token_accuracy": 0.18344138711690902,
      "num_tokens": 65750454.0,
      "step": 19680
    },
    {
      "entropy": 6.212217903137207,
      "epoch": 3.840308262608526,
      "grad_norm": 0.859375,
      "learning_rate": 0.00036317556382003545,
      "loss": 5.85,
      "mean_token_accuracy": 0.17024861872196198,
      "num_tokens": 65767221.0,
      "step": 19685
    },
    {
      "entropy": 5.9470299243927,
      "epoch": 3.8412837771924693,
      "grad_norm": 1.0,
      "learning_rate": 0.0003631108512743417,
      "loss": 5.4676,
      "mean_token_accuracy": 0.20019479542970658,
      "num_tokens": 65783894.0,
      "step": 19690
    },
    {
      "entropy": 6.1057435989379885,
      "epoch": 3.8422592917764122,
      "grad_norm": 0.875,
      "learning_rate": 0.00036304613011876934,
      "loss": 5.6854,
      "mean_token_accuracy": 0.1818847671151161,
      "num_tokens": 65799823.0,
      "step": 19695
    },
    {
      "entropy": 6.064178133010865,
      "epoch": 3.843234806360355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036298140035964313,
      "loss": 5.6298,
      "mean_token_accuracy": 0.1859619602560997,
      "num_tokens": 65816285.0,
      "step": 19700
    },
    {
      "entropy": 6.163671350479126,
      "epoch": 3.844210320944298,
      "grad_norm": 0.9375,
      "learning_rate": 0.000362916662003288,
      "loss": 5.754,
      "mean_token_accuracy": 0.17818135619163514,
      "num_tokens": 65832092.0,
      "step": 19705
    },
    {
      "entropy": 6.150667905807495,
      "epoch": 3.845185835528241,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00036285191505602994,
      "loss": 5.7639,
      "mean_token_accuracy": 0.17163589000701904,
      "num_tokens": 65849232.0,
      "step": 19710
    },
    {
      "entropy": 6.144140768051147,
      "epoch": 3.846161350112184,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036278715952419593,
      "loss": 5.6679,
      "mean_token_accuracy": 0.1872508317232132,
      "num_tokens": 65866723.0,
      "step": 19715
    },
    {
      "entropy": 6.2148045063018795,
      "epoch": 3.847136864696127,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003627223954141136,
      "loss": 5.736,
      "mean_token_accuracy": 0.17718977928161622,
      "num_tokens": 65883311.0,
      "step": 19720
    },
    {
      "entropy": 6.210794258117676,
      "epoch": 3.84811237928007,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003626576227321114,
      "loss": 5.8111,
      "mean_token_accuracy": 0.1719992622733116,
      "num_tokens": 65901087.0,
      "step": 19725
    },
    {
      "entropy": 6.166437721252441,
      "epoch": 3.8490878938640134,
      "grad_norm": 0.875,
      "learning_rate": 0.00036259284148451864,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1814645603299141,
      "num_tokens": 65918404.0,
      "step": 19730
    },
    {
      "entropy": 6.139141035079956,
      "epoch": 3.8500634084479564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036252805167766566,
      "loss": 5.7935,
      "mean_token_accuracy": 0.1785052701830864,
      "num_tokens": 65934512.0,
      "step": 19735
    },
    {
      "entropy": 6.095706272125244,
      "epoch": 3.8510389230318993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003624632533178835,
      "loss": 5.6506,
      "mean_token_accuracy": 0.18396941870450972,
      "num_tokens": 65951553.0,
      "step": 19740
    },
    {
      "entropy": 6.123476457595825,
      "epoch": 3.8520144376158423,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003623984464115038,
      "loss": 5.645,
      "mean_token_accuracy": 0.18098861575126649,
      "num_tokens": 65968573.0,
      "step": 19745
    },
    {
      "entropy": 6.036556768417358,
      "epoch": 3.8529899521997852,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036233363096485946,
      "loss": 5.5456,
      "mean_token_accuracy": 0.19417281150817872,
      "num_tokens": 65985270.0,
      "step": 19750
    },
    {
      "entropy": 6.153480100631714,
      "epoch": 3.8539654667837286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000362268806984284,
      "loss": 5.7844,
      "mean_token_accuracy": 0.17083868831396104,
      "num_tokens": 66002484.0,
      "step": 19755
    },
    {
      "entropy": 6.178623676300049,
      "epoch": 3.8549409813676716,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00036220397447611174,
      "loss": 5.6675,
      "mean_token_accuracy": 0.18482448607683183,
      "num_tokens": 66018724.0,
      "step": 19760
    },
    {
      "entropy": 6.174118995666504,
      "epoch": 3.8559164959516146,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036213913344667795,
      "loss": 5.7115,
      "mean_token_accuracy": 0.17616079300642012,
      "num_tokens": 66035225.0,
      "step": 19765
    },
    {
      "entropy": 6.187944173812866,
      "epoch": 3.8568920105355575,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003620742839023186,
      "loss": 5.7853,
      "mean_token_accuracy": 0.17288946807384492,
      "num_tokens": 66051766.0,
      "step": 19770
    },
    {
      "entropy": 6.257662296295166,
      "epoch": 3.8578675251195005,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036200942584937047,
      "loss": 5.8089,
      "mean_token_accuracy": 0.16777267605066298,
      "num_tokens": 66068360.0,
      "step": 19775
    },
    {
      "entropy": 6.228441333770752,
      "epoch": 3.8588430397034434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036194455929417136,
      "loss": 5.7656,
      "mean_token_accuracy": 0.17529408931732177,
      "num_tokens": 66083224.0,
      "step": 19780
    },
    {
      "entropy": 6.229917669296265,
      "epoch": 3.8598185542873864,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003618796842430599,
      "loss": 5.7532,
      "mean_token_accuracy": 0.18163001984357835,
      "num_tokens": 66098752.0,
      "step": 19785
    },
    {
      "entropy": 6.175240993499756,
      "epoch": 3.8607940688713294,
      "grad_norm": 0.875,
      "learning_rate": 0.00036181480070237524,
      "loss": 5.807,
      "mean_token_accuracy": 0.17053462713956832,
      "num_tokens": 66115556.0,
      "step": 19790
    },
    {
      "entropy": 6.186419343948364,
      "epoch": 3.8617695834552728,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003617499086784577,
      "loss": 5.6953,
      "mean_token_accuracy": 0.178782656788826,
      "num_tokens": 66133195.0,
      "step": 19795
    },
    {
      "entropy": 6.212477731704712,
      "epoch": 3.8627450980392157,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036168500817764833,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1688558578491211,
      "num_tokens": 66150331.0,
      "step": 19800
    },
    {
      "entropy": 6.171803951263428,
      "epoch": 3.8637206126231587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003616200992062888,
      "loss": 5.743,
      "mean_token_accuracy": 0.17714389711618422,
      "num_tokens": 66168948.0,
      "step": 19805
    },
    {
      "entropy": 6.074357509613037,
      "epoch": 3.8646961272071017,
      "grad_norm": 0.875,
      "learning_rate": 0.000361555181770722,
      "loss": 5.6583,
      "mean_token_accuracy": 0.18221284002065657,
      "num_tokens": 66185837.0,
      "step": 19810
    },
    {
      "entropy": 5.990679311752319,
      "epoch": 3.8656716417910446,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003614902558772912,
      "loss": 5.5556,
      "mean_token_accuracy": 0.19660159349441528,
      "num_tokens": 66203872.0,
      "step": 19815
    },
    {
      "entropy": 6.190040206909179,
      "epoch": 3.866647156374988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036142532153234097,
      "loss": 5.7353,
      "mean_token_accuracy": 0.17085922807455062,
      "num_tokens": 66220017.0,
      "step": 19820
    },
    {
      "entropy": 6.094085741043091,
      "epoch": 3.867622670958931,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003613603787422163,
      "loss": 5.6786,
      "mean_token_accuracy": 0.18169263899326324,
      "num_tokens": 66237277.0,
      "step": 19825
    },
    {
      "entropy": 6.168985080718994,
      "epoch": 3.868598185542874,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003612954275132632,
      "loss": 5.6877,
      "mean_token_accuracy": 0.18271084129810333,
      "num_tokens": 66254856.0,
      "step": 19830
    },
    {
      "entropy": 6.284402227401733,
      "epoch": 3.869573700126817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036123046785182855,
      "loss": 5.8123,
      "mean_token_accuracy": 0.16384741812944412,
      "num_tokens": 66271094.0,
      "step": 19835
    },
    {
      "entropy": 6.242619132995605,
      "epoch": 3.87054921471076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003611654997642598,
      "loss": 5.8109,
      "mean_token_accuracy": 0.17231811285018922,
      "num_tokens": 66288600.0,
      "step": 19840
    },
    {
      "entropy": 6.157088851928711,
      "epoch": 3.871524729294703,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003611005232569056,
      "loss": 5.6498,
      "mean_token_accuracy": 0.18206892609596254,
      "num_tokens": 66304395.0,
      "step": 19845
    },
    {
      "entropy": 6.1390540599823,
      "epoch": 3.8725002438786458,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036103553833611505,
      "loss": 5.7351,
      "mean_token_accuracy": 0.17488739639520645,
      "num_tokens": 66320129.0,
      "step": 19850
    },
    {
      "entropy": 6.258793449401855,
      "epoch": 3.873475758462589,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036097054500823847,
      "loss": 5.7949,
      "mean_token_accuracy": 0.16798954010009765,
      "num_tokens": 66337229.0,
      "step": 19855
    },
    {
      "entropy": 6.104214096069336,
      "epoch": 3.874451273046532,
      "grad_norm": 0.875,
      "learning_rate": 0.00036090554327962654,
      "loss": 5.6125,
      "mean_token_accuracy": 0.19088928699493407,
      "num_tokens": 66352842.0,
      "step": 19860
    },
    {
      "entropy": 6.095447587966919,
      "epoch": 3.875426787630475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003608405331566311,
      "loss": 5.705,
      "mean_token_accuracy": 0.17612768560647965,
      "num_tokens": 66369289.0,
      "step": 19865
    },
    {
      "entropy": 6.202578115463257,
      "epoch": 3.876402302214418,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003607755146456047,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1802994892001152,
      "num_tokens": 66385584.0,
      "step": 19870
    },
    {
      "entropy": 6.135389614105224,
      "epoch": 3.877377816798361,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003607104877529007,
      "loss": 5.7056,
      "mean_token_accuracy": 0.18508135676383972,
      "num_tokens": 66404163.0,
      "step": 19875
    },
    {
      "entropy": 6.1013267040252686,
      "epoch": 3.8783533313823044,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003606454524848734,
      "loss": 5.6168,
      "mean_token_accuracy": 0.19095283448696138,
      "num_tokens": 66420111.0,
      "step": 19880
    },
    {
      "entropy": 6.048725938796997,
      "epoch": 3.8793288459662474,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003605804088478777,
      "loss": 5.6168,
      "mean_token_accuracy": 0.18434448838233947,
      "num_tokens": 66437040.0,
      "step": 19885
    },
    {
      "entropy": 6.105166816711426,
      "epoch": 3.8803043605501903,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003605153568482694,
      "loss": 5.6255,
      "mean_token_accuracy": 0.19410260021686554,
      "num_tokens": 66453985.0,
      "step": 19890
    },
    {
      "entropy": 6.157813882827758,
      "epoch": 3.8812798751341333,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003604502964924053,
      "loss": 5.6952,
      "mean_token_accuracy": 0.18352320343255996,
      "num_tokens": 66471703.0,
      "step": 19895
    },
    {
      "entropy": 6.180203819274903,
      "epoch": 3.8822553897180763,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003603852277866427,
      "loss": 5.7955,
      "mean_token_accuracy": 0.1739058807492256,
      "num_tokens": 66488747.0,
      "step": 19900
    },
    {
      "entropy": 6.15335373878479,
      "epoch": 3.8832309043020192,
      "grad_norm": 0.859375,
      "learning_rate": 0.00036032015073734,
      "loss": 5.7489,
      "mean_token_accuracy": 0.181061053276062,
      "num_tokens": 66505858.0,
      "step": 19905
    },
    {
      "entropy": 6.18132004737854,
      "epoch": 3.884206418885962,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003602550653508562,
      "loss": 5.7409,
      "mean_token_accuracy": 0.17194315195083618,
      "num_tokens": 66522157.0,
      "step": 19910
    },
    {
      "entropy": 6.16091022491455,
      "epoch": 3.885181933469905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036018997163355136,
      "loss": 5.6255,
      "mean_token_accuracy": 0.18792558461427689,
      "num_tokens": 66537837.0,
      "step": 19915
    },
    {
      "entropy": 6.157520008087158,
      "epoch": 3.8861574480538486,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000360124869591786,
      "loss": 5.7741,
      "mean_token_accuracy": 0.18002325892448426,
      "num_tokens": 66554236.0,
      "step": 19920
    },
    {
      "entropy": 6.252087068557739,
      "epoch": 3.8871329626377915,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036005975923192183,
      "loss": 5.8193,
      "mean_token_accuracy": 0.16701972782611846,
      "num_tokens": 66569628.0,
      "step": 19925
    },
    {
      "entropy": 6.232841873168946,
      "epoch": 3.8881084772217345,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035999464056032124,
      "loss": 5.7609,
      "mean_token_accuracy": 0.18133064955472947,
      "num_tokens": 66586291.0,
      "step": 19930
    },
    {
      "entropy": 6.120456504821777,
      "epoch": 3.8890839918056774,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035992951358334717,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17783986479043962,
      "num_tokens": 66602350.0,
      "step": 19935
    },
    {
      "entropy": 6.125402450561523,
      "epoch": 3.8900595063896204,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035986437830736373,
      "loss": 5.6694,
      "mean_token_accuracy": 0.18066315352916718,
      "num_tokens": 66619611.0,
      "step": 19940
    },
    {
      "entropy": 6.065525484085083,
      "epoch": 3.891035020973564,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003597992347387358,
      "loss": 5.6223,
      "mean_token_accuracy": 0.1870863437652588,
      "num_tokens": 66637133.0,
      "step": 19945
    },
    {
      "entropy": 6.033091926574707,
      "epoch": 3.8920105355575068,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035973408288382877,
      "loss": 5.5578,
      "mean_token_accuracy": 0.19061252921819688,
      "num_tokens": 66655470.0,
      "step": 19950
    },
    {
      "entropy": 6.1065552711486815,
      "epoch": 3.8929860501414497,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035966892274900925,
      "loss": 5.6944,
      "mean_token_accuracy": 0.17997401505708693,
      "num_tokens": 66672510.0,
      "step": 19955
    },
    {
      "entropy": 6.163587951660157,
      "epoch": 3.8939615647253927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035960375434064426,
      "loss": 5.7367,
      "mean_token_accuracy": 0.1756249859929085,
      "num_tokens": 66687829.0,
      "step": 19960
    },
    {
      "entropy": 6.160969829559326,
      "epoch": 3.8949370793093356,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000359538577665102,
      "loss": 5.6274,
      "mean_token_accuracy": 0.1870764821767807,
      "num_tokens": 66703287.0,
      "step": 19965
    },
    {
      "entropy": 6.183853626251221,
      "epoch": 3.8959125938932786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003594733927287513,
      "loss": 5.7678,
      "mean_token_accuracy": 0.1758470505475998,
      "num_tokens": 66719573.0,
      "step": 19970
    },
    {
      "entropy": 6.215335273742676,
      "epoch": 3.8968881084772216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035940819953796154,
      "loss": 5.8052,
      "mean_token_accuracy": 0.16419077068567275,
      "num_tokens": 66735459.0,
      "step": 19975
    },
    {
      "entropy": 6.096459579467774,
      "epoch": 3.8978636230611645,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035934299809910355,
      "loss": 5.6744,
      "mean_token_accuracy": 0.18353027552366258,
      "num_tokens": 66752522.0,
      "step": 19980
    },
    {
      "entropy": 6.206348180770874,
      "epoch": 3.898839137645108,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003592777884185483,
      "loss": 5.7834,
      "mean_token_accuracy": 0.17712098658084868,
      "num_tokens": 66769380.0,
      "step": 19985
    },
    {
      "entropy": 6.041511726379395,
      "epoch": 3.899814652229051,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003592125705026679,
      "loss": 5.577,
      "mean_token_accuracy": 0.1918059140443802,
      "num_tokens": 66786974.0,
      "step": 19990
    },
    {
      "entropy": 6.156368398666382,
      "epoch": 3.900790166812994,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035914734435783536,
      "loss": 5.6729,
      "mean_token_accuracy": 0.18468571901321412,
      "num_tokens": 66804637.0,
      "step": 19995
    },
    {
      "entropy": 6.203847169876099,
      "epoch": 3.901765681396937,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003590821099904242,
      "loss": 5.7604,
      "mean_token_accuracy": 0.1752881482243538,
      "num_tokens": 66820473.0,
      "step": 20000
    },
    {
      "entropy": 6.110397624969482,
      "epoch": 3.9027411959808798,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003590168674068089,
      "loss": 5.6969,
      "mean_token_accuracy": 0.181009903550148,
      "num_tokens": 66836629.0,
      "step": 20005
    },
    {
      "entropy": 6.194917774200439,
      "epoch": 3.903716710564823,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035895161661336474,
      "loss": 5.758,
      "mean_token_accuracy": 0.17274577915668488,
      "num_tokens": 66853315.0,
      "step": 20010
    },
    {
      "entropy": 6.1691478252410885,
      "epoch": 3.904692225148766,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00035888635761646784,
      "loss": 5.7474,
      "mean_token_accuracy": 0.17266517579555513,
      "num_tokens": 66869862.0,
      "step": 20015
    },
    {
      "entropy": 6.062312507629395,
      "epoch": 3.905667739732709,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035882109042249507,
      "loss": 5.6673,
      "mean_token_accuracy": 0.18536820262670517,
      "num_tokens": 66888081.0,
      "step": 20020
    },
    {
      "entropy": 6.155589962005616,
      "epoch": 3.906643254316652,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035875581503782406,
      "loss": 5.6058,
      "mean_token_accuracy": 0.17840344905853273,
      "num_tokens": 66903312.0,
      "step": 20025
    },
    {
      "entropy": 6.109990167617798,
      "epoch": 3.907618768900595,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003586905314688333,
      "loss": 5.7058,
      "mean_token_accuracy": 0.1777715265750885,
      "num_tokens": 66920078.0,
      "step": 20030
    },
    {
      "entropy": 6.127558994293213,
      "epoch": 3.908594283484538,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00035862523972190216,
      "loss": 5.7151,
      "mean_token_accuracy": 0.17890024781227112,
      "num_tokens": 66938333.0,
      "step": 20035
    },
    {
      "entropy": 6.131604194641113,
      "epoch": 3.909569798068481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035855993980341054,
      "loss": 5.6375,
      "mean_token_accuracy": 0.18728084862232208,
      "num_tokens": 66954857.0,
      "step": 20040
    },
    {
      "entropy": 6.163492107391358,
      "epoch": 3.910545312652424,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035849463171973945,
      "loss": 5.8063,
      "mean_token_accuracy": 0.17550422698259355,
      "num_tokens": 66971268.0,
      "step": 20045
    },
    {
      "entropy": 6.120080375671387,
      "epoch": 3.9115208272363673,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003584293154772705,
      "loss": 5.6542,
      "mean_token_accuracy": 0.1817626476287842,
      "num_tokens": 66987666.0,
      "step": 20050
    },
    {
      "entropy": 6.2336742877960205,
      "epoch": 3.9124963418203103,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003583639910823862,
      "loss": 5.7643,
      "mean_token_accuracy": 0.18082748651504515,
      "num_tokens": 67003250.0,
      "step": 20055
    },
    {
      "entropy": 6.126537466049195,
      "epoch": 3.913471856404253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003582986585414699,
      "loss": 5.7732,
      "mean_token_accuracy": 0.18206633925437926,
      "num_tokens": 67020247.0,
      "step": 20060
    },
    {
      "entropy": 6.130652856826782,
      "epoch": 3.914447370988196,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00035823331786090547,
      "loss": 5.6341,
      "mean_token_accuracy": 0.1852791726589203,
      "num_tokens": 67038443.0,
      "step": 20065
    },
    {
      "entropy": 6.009601736068726,
      "epoch": 3.9154228855721396,
      "grad_norm": 0.875,
      "learning_rate": 0.0003581679690470779,
      "loss": 5.5472,
      "mean_token_accuracy": 0.19643179029226304,
      "num_tokens": 67056207.0,
      "step": 20070
    },
    {
      "entropy": 6.032974624633789,
      "epoch": 3.9163984001560825,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003581026121063728,
      "loss": 5.6308,
      "mean_token_accuracy": 0.1962244600057602,
      "num_tokens": 67073250.0,
      "step": 20075
    },
    {
      "entropy": 6.048186731338501,
      "epoch": 3.9173739147400255,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035803724704517665,
      "loss": 5.6305,
      "mean_token_accuracy": 0.188331039249897,
      "num_tokens": 67091138.0,
      "step": 20080
    },
    {
      "entropy": 6.235982084274292,
      "epoch": 3.9183494293239685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035797187386987664,
      "loss": 5.7913,
      "mean_token_accuracy": 0.1680992692708969,
      "num_tokens": 67108637.0,
      "step": 20085
    },
    {
      "entropy": 6.078702831268311,
      "epoch": 3.9193249439079114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003579064925868608,
      "loss": 5.6067,
      "mean_token_accuracy": 0.19347943365573883,
      "num_tokens": 67126984.0,
      "step": 20090
    },
    {
      "entropy": 6.242213106155395,
      "epoch": 3.9203004584918544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035784110320251806,
      "loss": 5.8438,
      "mean_token_accuracy": 0.16897415965795518,
      "num_tokens": 67144165.0,
      "step": 20095
    },
    {
      "entropy": 6.119971179962159,
      "epoch": 3.9212759730757973,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035777570572323794,
      "loss": 5.6259,
      "mean_token_accuracy": 0.18556030541658403,
      "num_tokens": 67160798.0,
      "step": 20100
    },
    {
      "entropy": 6.137350988388062,
      "epoch": 3.9222514876597403,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035771030015541085,
      "loss": 5.6429,
      "mean_token_accuracy": 0.18499915599822997,
      "num_tokens": 67177104.0,
      "step": 20105
    },
    {
      "entropy": 6.220207405090332,
      "epoch": 3.9232270022436837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003576448865054281,
      "loss": 5.7889,
      "mean_token_accuracy": 0.17780835926532745,
      "num_tokens": 67192746.0,
      "step": 20110
    },
    {
      "entropy": 6.158006238937378,
      "epoch": 3.9242025168276267,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003575794647796815,
      "loss": 5.7271,
      "mean_token_accuracy": 0.1734464779496193,
      "num_tokens": 67208934.0,
      "step": 20115
    },
    {
      "entropy": 6.183045434951782,
      "epoch": 3.9251780314115696,
      "grad_norm": 0.984375,
      "learning_rate": 0.000357514034984564,
      "loss": 5.7333,
      "mean_token_accuracy": 0.1849026769399643,
      "num_tokens": 67224810.0,
      "step": 20120
    },
    {
      "entropy": 6.218660068511963,
      "epoch": 3.9261535459955126,
      "grad_norm": 1.015625,
      "learning_rate": 0.000357448597126469,
      "loss": 5.7765,
      "mean_token_accuracy": 0.16677216291427613,
      "num_tokens": 67240205.0,
      "step": 20125
    },
    {
      "entropy": 6.244700288772583,
      "epoch": 3.9271290605794555,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000357383151211791,
      "loss": 5.7919,
      "mean_token_accuracy": 0.17519795149564743,
      "num_tokens": 67256952.0,
      "step": 20130
    },
    {
      "entropy": 6.216187143325806,
      "epoch": 3.928104575163399,
      "grad_norm": 1.0,
      "learning_rate": 0.000357317697246925,
      "loss": 5.7858,
      "mean_token_accuracy": 0.17409178912639617,
      "num_tokens": 67272953.0,
      "step": 20135
    },
    {
      "entropy": 6.1196118831634525,
      "epoch": 3.929080089747342,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003572522352382671,
      "loss": 5.6599,
      "mean_token_accuracy": 0.18543903827667235,
      "num_tokens": 67289024.0,
      "step": 20140
    },
    {
      "entropy": 6.1691607475280765,
      "epoch": 3.930055604331285,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035718676519221396,
      "loss": 5.6612,
      "mean_token_accuracy": 0.18147649466991425,
      "num_tokens": 67306375.0,
      "step": 20145
    },
    {
      "entropy": 6.128247499465942,
      "epoch": 3.931031118915228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003571212871151629,
      "loss": 5.6767,
      "mean_token_accuracy": 0.17956705689430236,
      "num_tokens": 67322565.0,
      "step": 20150
    },
    {
      "entropy": 6.080391883850098,
      "epoch": 3.932006633499171,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035705580101351247,
      "loss": 5.6293,
      "mean_token_accuracy": 0.18711237609386444,
      "num_tokens": 67338901.0,
      "step": 20155
    },
    {
      "entropy": 6.214396953582764,
      "epoch": 3.9329821480831137,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003569903068936616,
      "loss": 5.8384,
      "mean_token_accuracy": 0.17090129107236862,
      "num_tokens": 67354555.0,
      "step": 20160
    },
    {
      "entropy": 6.122689914703369,
      "epoch": 3.9339576626670567,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035692480476201007,
      "loss": 5.6696,
      "mean_token_accuracy": 0.1815987303853035,
      "num_tokens": 67371620.0,
      "step": 20165
    },
    {
      "entropy": 6.139774513244629,
      "epoch": 3.9349331772509997,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035685929462495866,
      "loss": 5.701,
      "mean_token_accuracy": 0.18016307353973388,
      "num_tokens": 67387336.0,
      "step": 20170
    },
    {
      "entropy": 6.210867643356323,
      "epoch": 3.935908691834943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035679377648890876,
      "loss": 5.7785,
      "mean_token_accuracy": 0.17638000696897507,
      "num_tokens": 67403541.0,
      "step": 20175
    },
    {
      "entropy": 6.249204874038696,
      "epoch": 3.936884206418886,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035672825036026253,
      "loss": 5.8041,
      "mean_token_accuracy": 0.17583739310503005,
      "num_tokens": 67418777.0,
      "step": 20180
    },
    {
      "entropy": 6.2031327247619625,
      "epoch": 3.937859721002829,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0003566627162454229,
      "loss": 5.7534,
      "mean_token_accuracy": 0.17457689940929413,
      "num_tokens": 67435423.0,
      "step": 20185
    },
    {
      "entropy": 6.201024961471558,
      "epoch": 3.938835235586772,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003565971741507938,
      "loss": 5.7033,
      "mean_token_accuracy": 0.1723371312022209,
      "num_tokens": 67452189.0,
      "step": 20190
    },
    {
      "entropy": 6.165621662139893,
      "epoch": 3.939810750170715,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035653162408277953,
      "loss": 5.7169,
      "mean_token_accuracy": 0.18141033500432968,
      "num_tokens": 67470161.0,
      "step": 20195
    },
    {
      "entropy": 6.1662781715393065,
      "epoch": 3.9407862647546583,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003564660660477856,
      "loss": 5.7093,
      "mean_token_accuracy": 0.18418246507644653,
      "num_tokens": 67487392.0,
      "step": 20200
    },
    {
      "entropy": 6.167042303085327,
      "epoch": 3.9417617793386013,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003564005000522181,
      "loss": 5.7349,
      "mean_token_accuracy": 0.1747995674610138,
      "num_tokens": 67505920.0,
      "step": 20205
    },
    {
      "entropy": 6.082456111907959,
      "epoch": 3.9427372939225442,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003563349261024838,
      "loss": 5.6362,
      "mean_token_accuracy": 0.18406520187854766,
      "num_tokens": 67523040.0,
      "step": 20210
    },
    {
      "entropy": 6.158234643936157,
      "epoch": 3.943712808506487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003562693442049904,
      "loss": 5.6994,
      "mean_token_accuracy": 0.17320408821105956,
      "num_tokens": 67538349.0,
      "step": 20215
    },
    {
      "entropy": 6.224424505233765,
      "epoch": 3.94468832309043,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003562037543661464,
      "loss": 5.7962,
      "mean_token_accuracy": 0.1712354004383087,
      "num_tokens": 67554033.0,
      "step": 20220
    },
    {
      "entropy": 6.169460153579712,
      "epoch": 3.945663837674373,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035613815659236086,
      "loss": 5.7362,
      "mean_token_accuracy": 0.17684557139873505,
      "num_tokens": 67571427.0,
      "step": 20225
    },
    {
      "entropy": 6.159450101852417,
      "epoch": 3.946639352258316,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003560725508900438,
      "loss": 5.6857,
      "mean_token_accuracy": 0.1784885570406914,
      "num_tokens": 67588667.0,
      "step": 20230
    },
    {
      "entropy": 6.21240348815918,
      "epoch": 3.947614866842259,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003560069372656061,
      "loss": 5.7287,
      "mean_token_accuracy": 0.17771599143743516,
      "num_tokens": 67604465.0,
      "step": 20235
    },
    {
      "entropy": 6.101442098617554,
      "epoch": 3.9485903814262024,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003559413157254592,
      "loss": 5.6482,
      "mean_token_accuracy": 0.18725232928991317,
      "num_tokens": 67620469.0,
      "step": 20240
    },
    {
      "entropy": 6.1287459373474125,
      "epoch": 3.9495658960101454,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003558756862760153,
      "loss": 5.6927,
      "mean_token_accuracy": 0.18375937789678573,
      "num_tokens": 67636880.0,
      "step": 20245
    },
    {
      "entropy": 6.140260982513428,
      "epoch": 3.9505414105940884,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003558100489236877,
      "loss": 5.6953,
      "mean_token_accuracy": 0.18139811605215073,
      "num_tokens": 67653979.0,
      "step": 20250
    },
    {
      "entropy": 6.096132755279541,
      "epoch": 3.9515169251780313,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035574440367489013,
      "loss": 5.6372,
      "mean_token_accuracy": 0.19405950158834456,
      "num_tokens": 67672751.0,
      "step": 20255
    },
    {
      "entropy": 6.127733898162842,
      "epoch": 3.9524924397619743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003556787505360372,
      "loss": 5.634,
      "mean_token_accuracy": 0.1867621958255768,
      "num_tokens": 67690291.0,
      "step": 20260
    },
    {
      "entropy": 6.189920806884766,
      "epoch": 3.9534679543459177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003556130895135442,
      "loss": 5.7541,
      "mean_token_accuracy": 0.17803787291049958,
      "num_tokens": 67708362.0,
      "step": 20265
    },
    {
      "entropy": 6.254735231399536,
      "epoch": 3.9544434689298607,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003555474206138275,
      "loss": 5.7938,
      "mean_token_accuracy": 0.1745724692940712,
      "num_tokens": 67725624.0,
      "step": 20270
    },
    {
      "entropy": 6.123845148086548,
      "epoch": 3.9554189835138036,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035548174384330395,
      "loss": 5.6118,
      "mean_token_accuracy": 0.19020718485116958,
      "num_tokens": 67742712.0,
      "step": 20275
    },
    {
      "entropy": 6.0196247577667235,
      "epoch": 3.9563944980977466,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035541605920839123,
      "loss": 5.5493,
      "mean_token_accuracy": 0.1894513636827469,
      "num_tokens": 67759162.0,
      "step": 20280
    },
    {
      "entropy": 6.133413505554199,
      "epoch": 3.9573700126816895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003553503667155078,
      "loss": 5.6422,
      "mean_token_accuracy": 0.18239294439554216,
      "num_tokens": 67774206.0,
      "step": 20285
    },
    {
      "entropy": 6.199239826202392,
      "epoch": 3.9583455272656325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035528466637107294,
      "loss": 5.8078,
      "mean_token_accuracy": 0.16554944068193436,
      "num_tokens": 67790793.0,
      "step": 20290
    },
    {
      "entropy": 6.077454853057861,
      "epoch": 3.9593210418495755,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003552189581815065,
      "loss": 5.6124,
      "mean_token_accuracy": 0.18105543106794358,
      "num_tokens": 67809224.0,
      "step": 20295
    },
    {
      "entropy": 6.173920440673828,
      "epoch": 3.9602965564335184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035515324215322947,
      "loss": 5.6936,
      "mean_token_accuracy": 0.17234315127134323,
      "num_tokens": 67825976.0,
      "step": 20300
    },
    {
      "entropy": 6.153704690933227,
      "epoch": 3.961272071017462,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003550875182926633,
      "loss": 5.6367,
      "mean_token_accuracy": 0.17831545621156691,
      "num_tokens": 67843215.0,
      "step": 20305
    },
    {
      "entropy": 6.160950756072998,
      "epoch": 3.9622475856014048,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003550217866062303,
      "loss": 5.7368,
      "mean_token_accuracy": 0.17736884206533432,
      "num_tokens": 67861242.0,
      "step": 20310
    },
    {
      "entropy": 6.0615339279174805,
      "epoch": 3.9632231001853477,
      "grad_norm": 0.828125,
      "learning_rate": 0.0003549560471003535,
      "loss": 5.663,
      "mean_token_accuracy": 0.19000186175107955,
      "num_tokens": 67877141.0,
      "step": 20315
    },
    {
      "entropy": 6.191175699234009,
      "epoch": 3.9641986147692907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035489029978145667,
      "loss": 5.7737,
      "mean_token_accuracy": 0.17252303659915924,
      "num_tokens": 67893322.0,
      "step": 20320
    },
    {
      "entropy": 6.1170081615448,
      "epoch": 3.965174129353234,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003548245446559645,
      "loss": 5.5876,
      "mean_token_accuracy": 0.1869685471057892,
      "num_tokens": 67909498.0,
      "step": 20325
    },
    {
      "entropy": 6.164943361282349,
      "epoch": 3.966149643937177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035475878173030234,
      "loss": 5.6968,
      "mean_token_accuracy": 0.1756229057908058,
      "num_tokens": 67926067.0,
      "step": 20330
    },
    {
      "entropy": 6.203097486495972,
      "epoch": 3.96712515852112,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003546930110108963,
      "loss": 5.7822,
      "mean_token_accuracy": 0.16951508522033693,
      "num_tokens": 67941500.0,
      "step": 20335
    },
    {
      "entropy": 6.199829053878784,
      "epoch": 3.968100673105063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003546272325041733,
      "loss": 5.729,
      "mean_token_accuracy": 0.17436232417821884,
      "num_tokens": 67958351.0,
      "step": 20340
    },
    {
      "entropy": 6.180882930755615,
      "epoch": 3.969076187689006,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003545614462165608,
      "loss": 5.7266,
      "mean_token_accuracy": 0.181679804623127,
      "num_tokens": 67974664.0,
      "step": 20345
    },
    {
      "entropy": 5.970226097106933,
      "epoch": 3.970051702272949,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035449565215448756,
      "loss": 5.5123,
      "mean_token_accuracy": 0.19720131754875184,
      "num_tokens": 67992055.0,
      "step": 20350
    },
    {
      "entropy": 6.219338369369507,
      "epoch": 3.971027216856892,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035442985032438243,
      "loss": 5.7271,
      "mean_token_accuracy": 0.1751032903790474,
      "num_tokens": 68006973.0,
      "step": 20355
    },
    {
      "entropy": 6.099816989898682,
      "epoch": 3.972002731440835,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003543640407326753,
      "loss": 5.6254,
      "mean_token_accuracy": 0.18342766016721726,
      "num_tokens": 68022513.0,
      "step": 20360
    },
    {
      "entropy": 6.188318300247192,
      "epoch": 3.9729782460247782,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035429822338579696,
      "loss": 5.6933,
      "mean_token_accuracy": 0.18188766688108443,
      "num_tokens": 68039090.0,
      "step": 20365
    },
    {
      "entropy": 6.15565299987793,
      "epoch": 3.973953760608721,
      "grad_norm": 1.0,
      "learning_rate": 0.000354232398290179,
      "loss": 5.7803,
      "mean_token_accuracy": 0.17400500327348709,
      "num_tokens": 68055173.0,
      "step": 20370
    },
    {
      "entropy": 6.20079779624939,
      "epoch": 3.974929275192664,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003541665654522534,
      "loss": 5.7527,
      "mean_token_accuracy": 0.17474775463342668,
      "num_tokens": 68071900.0,
      "step": 20375
    },
    {
      "entropy": 6.195204877853394,
      "epoch": 3.975904789776607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003541007248784532,
      "loss": 5.7601,
      "mean_token_accuracy": 0.17435644567012787,
      "num_tokens": 68089206.0,
      "step": 20380
    },
    {
      "entropy": 6.08670015335083,
      "epoch": 3.97688030436055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003540348765752121,
      "loss": 5.6096,
      "mean_token_accuracy": 0.18521000444889069,
      "num_tokens": 68107057.0,
      "step": 20385
    },
    {
      "entropy": 6.084794616699218,
      "epoch": 3.9778558189444935,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035396902054896453,
      "loss": 5.6283,
      "mean_token_accuracy": 0.1897817000746727,
      "num_tokens": 68123499.0,
      "step": 20390
    },
    {
      "entropy": 6.217419385910034,
      "epoch": 3.9788313335284364,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035390315680614573,
      "loss": 5.7699,
      "mean_token_accuracy": 0.1709307923913002,
      "num_tokens": 68140830.0,
      "step": 20395
    },
    {
      "entropy": 6.162506771087647,
      "epoch": 3.9798068481123794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003538372853531916,
      "loss": 5.7444,
      "mean_token_accuracy": 0.17927921265363694,
      "num_tokens": 68156649.0,
      "step": 20400
    },
    {
      "entropy": 6.141341829299927,
      "epoch": 3.9807823626963224,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035377140619653904,
      "loss": 5.7575,
      "mean_token_accuracy": 0.18100358545780182,
      "num_tokens": 68175174.0,
      "step": 20405
    },
    {
      "entropy": 6.137634706497193,
      "epoch": 3.9817578772802653,
      "grad_norm": 1.0,
      "learning_rate": 0.0003537055193426253,
      "loss": 5.5945,
      "mean_token_accuracy": 0.18291160315275193,
      "num_tokens": 68191434.0,
      "step": 20410
    },
    {
      "entropy": 6.229223442077637,
      "epoch": 3.9827333918642083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035363962479788877,
      "loss": 5.7877,
      "mean_token_accuracy": 0.17488571703433992,
      "num_tokens": 68208026.0,
      "step": 20415
    },
    {
      "entropy": 6.138442850112915,
      "epoch": 3.9837089064481512,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003535737225687684,
      "loss": 5.6882,
      "mean_token_accuracy": 0.1828136056661606,
      "num_tokens": 68224515.0,
      "step": 20420
    },
    {
      "entropy": 6.163385343551636,
      "epoch": 3.984684421032094,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035350781266170386,
      "loss": 5.697,
      "mean_token_accuracy": 0.17731484025716782,
      "num_tokens": 68240162.0,
      "step": 20425
    },
    {
      "entropy": 6.224207973480224,
      "epoch": 3.9856599356160376,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003534418950831357,
      "loss": 5.7343,
      "mean_token_accuracy": 0.17710299491882325,
      "num_tokens": 68255884.0,
      "step": 20430
    },
    {
      "entropy": 6.177800559997559,
      "epoch": 3.9866354501999806,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003533759698395051,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1807949051260948,
      "num_tokens": 68272109.0,
      "step": 20435
    },
    {
      "entropy": 6.043172550201416,
      "epoch": 3.9876109647839235,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003533100369372541,
      "loss": 5.5847,
      "mean_token_accuracy": 0.18837495148181915,
      "num_tokens": 68289198.0,
      "step": 20440
    },
    {
      "entropy": 6.078656578063965,
      "epoch": 3.9885864793678665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003532440963828254,
      "loss": 5.6091,
      "mean_token_accuracy": 0.18345313221216203,
      "num_tokens": 68306438.0,
      "step": 20445
    },
    {
      "entropy": 6.145001697540283,
      "epoch": 3.9895619939518094,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035317814818266243,
      "loss": 5.7336,
      "mean_token_accuracy": 0.1775531992316246,
      "num_tokens": 68323380.0,
      "step": 20450
    },
    {
      "entropy": 6.121481466293335,
      "epoch": 3.990537508535753,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003531121923432095,
      "loss": 5.6329,
      "mean_token_accuracy": 0.1862169995903969,
      "num_tokens": 68339568.0,
      "step": 20455
    },
    {
      "entropy": 6.1303479194641115,
      "epoch": 3.991513023119696,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003530462288709114,
      "loss": 5.6832,
      "mean_token_accuracy": 0.18686725050210953,
      "num_tokens": 68357375.0,
      "step": 20460
    },
    {
      "entropy": 6.011633682250976,
      "epoch": 3.9924885377036388,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00035298025777221414,
      "loss": 5.4964,
      "mean_token_accuracy": 0.19338970631361008,
      "num_tokens": 68376231.0,
      "step": 20465
    },
    {
      "entropy": 6.005515670776367,
      "epoch": 3.9934640522875817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035291427905356386,
      "loss": 5.5249,
      "mean_token_accuracy": 0.19066527336835862,
      "num_tokens": 68391877.0,
      "step": 20470
    },
    {
      "entropy": 5.978615570068359,
      "epoch": 3.9944395668715247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000352848292721408,
      "loss": 5.5051,
      "mean_token_accuracy": 0.20084608048200608,
      "num_tokens": 68409793.0,
      "step": 20475
    },
    {
      "entropy": 6.077800846099853,
      "epoch": 3.9954150814554676,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003527822987821945,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1784123122692108,
      "num_tokens": 68427420.0,
      "step": 20480
    },
    {
      "entropy": 6.099447965621948,
      "epoch": 3.9963905960394106,
      "grad_norm": 0.9375,
      "learning_rate": 0.000352716297242372,
      "loss": 5.5629,
      "mean_token_accuracy": 0.1956041172146797,
      "num_tokens": 68445907.0,
      "step": 20485
    },
    {
      "entropy": 6.226896810531616,
      "epoch": 3.9973661106233536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035265028810838977,
      "loss": 5.8192,
      "mean_token_accuracy": 0.1686551094055176,
      "num_tokens": 68462305.0,
      "step": 20490
    },
    {
      "entropy": 6.148355340957641,
      "epoch": 3.998341625207297,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035258427138669824,
      "loss": 5.7163,
      "mean_token_accuracy": 0.17946967482566833,
      "num_tokens": 68478173.0,
      "step": 20495
    },
    {
      "entropy": 6.0275475025177006,
      "epoch": 3.99931713979124,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035251824708374817,
      "loss": 5.5904,
      "mean_token_accuracy": 0.18666137903928756,
      "num_tokens": 68494480.0,
      "step": 20500
    },
    {
      "entropy": 6.066880544026692,
      "epoch": 4.000195102916789,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003524522152059914,
      "loss": 5.5919,
      "mean_token_accuracy": 0.18675533102618325,
      "num_tokens": 68507877.0,
      "step": 20505
    },
    {
      "entropy": 6.190948390960694,
      "epoch": 4.0011706175007316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035238617575988,
      "loss": 5.6604,
      "mean_token_accuracy": 0.17787821292877198,
      "num_tokens": 68523954.0,
      "step": 20510
    },
    {
      "entropy": 6.0406805038452145,
      "epoch": 4.0021461320846745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003523201287518675,
      "loss": 5.5001,
      "mean_token_accuracy": 0.19256115406751634,
      "num_tokens": 68542109.0,
      "step": 20515
    },
    {
      "entropy": 6.183068752288818,
      "epoch": 4.0031216466686175,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035225407418840757,
      "loss": 5.6526,
      "mean_token_accuracy": 0.17801463752985,
      "num_tokens": 68558888.0,
      "step": 20520
    },
    {
      "entropy": 6.108200931549073,
      "epoch": 4.00409716125256,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035218801207595473,
      "loss": 5.6281,
      "mean_token_accuracy": 0.1877203956246376,
      "num_tokens": 68575727.0,
      "step": 20525
    },
    {
      "entropy": 6.159071683883667,
      "epoch": 4.005072675836503,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003521219424209645,
      "loss": 5.582,
      "mean_token_accuracy": 0.18536070585250855,
      "num_tokens": 68593545.0,
      "step": 20530
    },
    {
      "entropy": 6.020578241348266,
      "epoch": 4.006048190420446,
      "grad_norm": 1.0625,
      "learning_rate": 0.000352055865229893,
      "loss": 5.5268,
      "mean_token_accuracy": 0.19720435738563538,
      "num_tokens": 68611564.0,
      "step": 20535
    },
    {
      "entropy": 6.140906476974488,
      "epoch": 4.00702370500439,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035198978050919686,
      "loss": 5.6347,
      "mean_token_accuracy": 0.17583656162023545,
      "num_tokens": 68627079.0,
      "step": 20540
    },
    {
      "entropy": 6.168415164947509,
      "epoch": 4.007999219588333,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035192368826533386,
      "loss": 5.6573,
      "mean_token_accuracy": 0.18345570117235183,
      "num_tokens": 68643633.0,
      "step": 20545
    },
    {
      "entropy": 6.121819257736206,
      "epoch": 4.008974734172276,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035185758850476214,
      "loss": 5.5324,
      "mean_token_accuracy": 0.19330666363239288,
      "num_tokens": 68659570.0,
      "step": 20550
    },
    {
      "entropy": 6.1650965213775635,
      "epoch": 4.009950248756219,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035179148123394083,
      "loss": 5.6587,
      "mean_token_accuracy": 0.1876807764172554,
      "num_tokens": 68674724.0,
      "step": 20555
    },
    {
      "entropy": 6.230097007751465,
      "epoch": 4.010925763340162,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003517253664593297,
      "loss": 5.7123,
      "mean_token_accuracy": 0.17536677569150924,
      "num_tokens": 68689610.0,
      "step": 20560
    },
    {
      "entropy": 6.162482166290284,
      "epoch": 4.011901277924105,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035165924418738916,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1823689579963684,
      "num_tokens": 68706647.0,
      "step": 20565
    },
    {
      "entropy": 6.0833228588104244,
      "epoch": 4.012876792508048,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035159311442458047,
      "loss": 5.5608,
      "mean_token_accuracy": 0.19335795789957047,
      "num_tokens": 68722653.0,
      "step": 20570
    },
    {
      "entropy": 6.213377285003662,
      "epoch": 4.013852307091991,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035152697717736567,
      "loss": 5.6731,
      "mean_token_accuracy": 0.17538556605577468,
      "num_tokens": 68739036.0,
      "step": 20575
    },
    {
      "entropy": 6.0070658206939695,
      "epoch": 4.014827821675934,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035146083245220737,
      "loss": 5.4462,
      "mean_token_accuracy": 0.20016077160835266,
      "num_tokens": 68754607.0,
      "step": 20580
    },
    {
      "entropy": 6.1456409931182865,
      "epoch": 4.015803336259877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003513946802555691,
      "loss": 5.6597,
      "mean_token_accuracy": 0.18546710759401322,
      "num_tokens": 68771880.0,
      "step": 20585
    },
    {
      "entropy": 6.197180891036988,
      "epoch": 4.01677885084382,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035132852059391485,
      "loss": 5.6243,
      "mean_token_accuracy": 0.18197176456451417,
      "num_tokens": 68789321.0,
      "step": 20590
    },
    {
      "entropy": 6.100979518890381,
      "epoch": 4.017754365427763,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003512623534737097,
      "loss": 5.574,
      "mean_token_accuracy": 0.18915597945451737,
      "num_tokens": 68805875.0,
      "step": 20595
    },
    {
      "entropy": 6.181490182876587,
      "epoch": 4.018729880011707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035119617890141914,
      "loss": 5.6932,
      "mean_token_accuracy": 0.17968935519456863,
      "num_tokens": 68821358.0,
      "step": 20600
    },
    {
      "entropy": 6.212922477722168,
      "epoch": 4.01970539459565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003511299968835096,
      "loss": 5.635,
      "mean_token_accuracy": 0.18804911226034166,
      "num_tokens": 68837441.0,
      "step": 20605
    },
    {
      "entropy": 6.128578186035156,
      "epoch": 4.0206809091795925,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000351063807426448,
      "loss": 5.6421,
      "mean_token_accuracy": 0.1836719736456871,
      "num_tokens": 68855122.0,
      "step": 20610
    },
    {
      "entropy": 6.114084672927857,
      "epoch": 4.0216564237635355,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035099761053670233,
      "loss": 5.6319,
      "mean_token_accuracy": 0.18453153669834138,
      "num_tokens": 68873005.0,
      "step": 20615
    },
    {
      "entropy": 6.190891265869141,
      "epoch": 4.0226319383474785,
      "grad_norm": 1.0,
      "learning_rate": 0.00035093140622074093,
      "loss": 5.611,
      "mean_token_accuracy": 0.1817262127995491,
      "num_tokens": 68887996.0,
      "step": 20620
    },
    {
      "entropy": 6.10423469543457,
      "epoch": 4.023607452931421,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003508651944850332,
      "loss": 5.6159,
      "mean_token_accuracy": 0.18680131137371064,
      "num_tokens": 68903696.0,
      "step": 20625
    },
    {
      "entropy": 6.1226503372192385,
      "epoch": 4.024582967515364,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035079897533604905,
      "loss": 5.6421,
      "mean_token_accuracy": 0.1852834165096283,
      "num_tokens": 68920507.0,
      "step": 20630
    },
    {
      "entropy": 6.14814248085022,
      "epoch": 4.025558482099307,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003507327487802591,
      "loss": 5.5722,
      "mean_token_accuracy": 0.1885313555598259,
      "num_tokens": 68937363.0,
      "step": 20635
    },
    {
      "entropy": 6.104361057281494,
      "epoch": 4.02653399668325,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035066651482413495,
      "loss": 5.5866,
      "mean_token_accuracy": 0.18285648822784423,
      "num_tokens": 68954544.0,
      "step": 20640
    },
    {
      "entropy": 6.0779565334320065,
      "epoch": 4.027509511267193,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003506002734741486,
      "loss": 5.5194,
      "mean_token_accuracy": 0.19793642163276673,
      "num_tokens": 68970991.0,
      "step": 20645
    },
    {
      "entropy": 6.059228706359863,
      "epoch": 4.028485025851136,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00035053402473677307,
      "loss": 5.5292,
      "mean_token_accuracy": 0.19917955100536347,
      "num_tokens": 68988438.0,
      "step": 20650
    },
    {
      "entropy": 6.084825658798218,
      "epoch": 4.029460540435079,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003504677686184817,
      "loss": 5.5728,
      "mean_token_accuracy": 0.1862540513277054,
      "num_tokens": 69004367.0,
      "step": 20655
    },
    {
      "entropy": 6.110918045043945,
      "epoch": 4.030436055019022,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0003504015051257491,
      "loss": 5.5721,
      "mean_token_accuracy": 0.19005829840898514,
      "num_tokens": 69021714.0,
      "step": 20660
    },
    {
      "entropy": 6.180917358398437,
      "epoch": 4.031411569602966,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0003503352342650501,
      "loss": 5.643,
      "mean_token_accuracy": 0.18118324130773544,
      "num_tokens": 69037501.0,
      "step": 20665
    },
    {
      "entropy": 6.191648721694946,
      "epoch": 4.032387084186909,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003502689560428604,
      "loss": 5.6591,
      "mean_token_accuracy": 0.17825592756271363,
      "num_tokens": 69054521.0,
      "step": 20670
    },
    {
      "entropy": 6.055092525482178,
      "epoch": 4.033362598770852,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035020267046565666,
      "loss": 5.5025,
      "mean_token_accuracy": 0.19795837253332138,
      "num_tokens": 69073028.0,
      "step": 20675
    },
    {
      "entropy": 6.116227102279663,
      "epoch": 4.034338113354795,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000350136377539916,
      "loss": 5.526,
      "mean_token_accuracy": 0.1869712218642235,
      "num_tokens": 69090670.0,
      "step": 20680
    },
    {
      "entropy": 6.153650093078613,
      "epoch": 4.035313627938738,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035007007727211635,
      "loss": 5.6336,
      "mean_token_accuracy": 0.18426549732685088,
      "num_tokens": 69106388.0,
      "step": 20685
    },
    {
      "entropy": 6.182142305374145,
      "epoch": 4.036289142522681,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035000376966873623,
      "loss": 5.6356,
      "mean_token_accuracy": 0.18109919428825377,
      "num_tokens": 69121681.0,
      "step": 20690
    },
    {
      "entropy": 6.145299339294434,
      "epoch": 4.037264657106624,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003499374547362551,
      "loss": 5.6595,
      "mean_token_accuracy": 0.18228716254234315,
      "num_tokens": 69138183.0,
      "step": 20695
    },
    {
      "entropy": 6.097837448120117,
      "epoch": 4.038240171690567,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034987113248115297,
      "loss": 5.5806,
      "mean_token_accuracy": 0.1840936839580536,
      "num_tokens": 69156739.0,
      "step": 20700
    },
    {
      "entropy": 6.212952613830566,
      "epoch": 4.03921568627451,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034980480290991065,
      "loss": 5.6459,
      "mean_token_accuracy": 0.17989733070135117,
      "num_tokens": 69173208.0,
      "step": 20705
    },
    {
      "entropy": 6.214990663528442,
      "epoch": 4.040191200858453,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003497384660290095,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1836091548204422,
      "num_tokens": 69189597.0,
      "step": 20710
    },
    {
      "entropy": 5.953159046173096,
      "epoch": 4.041166715442396,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0003496721218449319,
      "loss": 5.4405,
      "mean_token_accuracy": 0.20098259299993515,
      "num_tokens": 69207238.0,
      "step": 20715
    },
    {
      "entropy": 6.055395078659058,
      "epoch": 4.0421422300263385,
      "grad_norm": 0.859375,
      "learning_rate": 0.00034960577036416066,
      "loss": 5.5393,
      "mean_token_accuracy": 0.1970365896821022,
      "num_tokens": 69223760.0,
      "step": 20720
    },
    {
      "entropy": 6.1625330448150635,
      "epoch": 4.0431177446102815,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00034953941159317946,
      "loss": 5.6403,
      "mean_token_accuracy": 0.17684833854436874,
      "num_tokens": 69242244.0,
      "step": 20725
    },
    {
      "entropy": 6.065440988540649,
      "epoch": 4.044093259194225,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003494730455384727,
      "loss": 5.5032,
      "mean_token_accuracy": 0.19977523386478424,
      "num_tokens": 69259637.0,
      "step": 20730
    },
    {
      "entropy": 6.073290634155273,
      "epoch": 4.045068773778168,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034940667220652517,
      "loss": 5.5232,
      "mean_token_accuracy": 0.1921641632914543,
      "num_tokens": 69275721.0,
      "step": 20735
    },
    {
      "entropy": 6.1370100498199465,
      "epoch": 4.046044288362111,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000349340291603823,
      "loss": 5.6071,
      "mean_token_accuracy": 0.18548605740070342,
      "num_tokens": 69292638.0,
      "step": 20740
    },
    {
      "entropy": 6.066116571426392,
      "epoch": 4.047019802946054,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034927390373685243,
      "loss": 5.496,
      "mean_token_accuracy": 0.19953154474496843,
      "num_tokens": 69309070.0,
      "step": 20745
    },
    {
      "entropy": 6.069885349273681,
      "epoch": 4.047995317529997,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003492075086121006,
      "loss": 5.5347,
      "mean_token_accuracy": 0.18767193406820298,
      "num_tokens": 69326225.0,
      "step": 20750
    },
    {
      "entropy": 6.117251348495484,
      "epoch": 4.04897083211394,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003491411062360557,
      "loss": 5.5608,
      "mean_token_accuracy": 0.19111579954624175,
      "num_tokens": 69343360.0,
      "step": 20755
    },
    {
      "entropy": 6.159803199768066,
      "epoch": 4.049946346697883,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034907469661520595,
      "loss": 5.6898,
      "mean_token_accuracy": 0.17910227477550505,
      "num_tokens": 69361143.0,
      "step": 20760
    },
    {
      "entropy": 6.108432722091675,
      "epoch": 4.050921861281826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034900827975604097,
      "loss": 5.6296,
      "mean_token_accuracy": 0.186154405772686,
      "num_tokens": 69378039.0,
      "step": 20765
    },
    {
      "entropy": 6.086580038070679,
      "epoch": 4.051897375865769,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034894185566505066,
      "loss": 5.5651,
      "mean_token_accuracy": 0.1941751703619957,
      "num_tokens": 69395383.0,
      "step": 20770
    },
    {
      "entropy": 6.1585094928741455,
      "epoch": 4.052872890449712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003488754243487257,
      "loss": 5.6058,
      "mean_token_accuracy": 0.18754274696111678,
      "num_tokens": 69412870.0,
      "step": 20775
    },
    {
      "entropy": 6.183681011199951,
      "epoch": 4.053848405033655,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0003488089858135576,
      "loss": 5.6519,
      "mean_token_accuracy": 0.18200846910476684,
      "num_tokens": 69430510.0,
      "step": 20780
    },
    {
      "entropy": 6.114499092102051,
      "epoch": 4.054823919617598,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00034874254006603847,
      "loss": 5.5999,
      "mean_token_accuracy": 0.18740893304347991,
      "num_tokens": 69447381.0,
      "step": 20785
    },
    {
      "entropy": 6.218741369247437,
      "epoch": 4.055799434201541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034867608711266107,
      "loss": 5.7217,
      "mean_token_accuracy": 0.17837779968976974,
      "num_tokens": 69464505.0,
      "step": 20790
    },
    {
      "entropy": 6.123658037185669,
      "epoch": 4.056774948785485,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034860962695991914,
      "loss": 5.5162,
      "mean_token_accuracy": 0.190896575152874,
      "num_tokens": 69479696.0,
      "step": 20795
    },
    {
      "entropy": 6.011943006515503,
      "epoch": 4.057750463369428,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003485431596143068,
      "loss": 5.547,
      "mean_token_accuracy": 0.1854204460978508,
      "num_tokens": 69495522.0,
      "step": 20800
    },
    {
      "entropy": 6.183997631072998,
      "epoch": 4.058725977953371,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00034847668508231896,
      "loss": 5.6921,
      "mean_token_accuracy": 0.17923343777656556,
      "num_tokens": 69511123.0,
      "step": 20805
    },
    {
      "entropy": 6.197345113754272,
      "epoch": 4.059701492537314,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003484102033704513,
      "loss": 5.6055,
      "mean_token_accuracy": 0.18395028859376908,
      "num_tokens": 69527207.0,
      "step": 20810
    },
    {
      "entropy": 6.20032057762146,
      "epoch": 4.060677007121257,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003483437144852003,
      "loss": 5.7299,
      "mean_token_accuracy": 0.1762184903025627,
      "num_tokens": 69544927.0,
      "step": 20815
    },
    {
      "entropy": 5.96566743850708,
      "epoch": 4.0616525217051995,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0003482772184330629,
      "loss": 5.404,
      "mean_token_accuracy": 0.21025486141443253,
      "num_tokens": 69561712.0,
      "step": 20820
    },
    {
      "entropy": 6.168039989471436,
      "epoch": 4.0626280362891425,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003482107152205368,
      "loss": 5.5979,
      "mean_token_accuracy": 0.18337129950523376,
      "num_tokens": 69578564.0,
      "step": 20825
    },
    {
      "entropy": 6.191647720336914,
      "epoch": 4.0636035508730854,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003481442048541206,
      "loss": 5.6494,
      "mean_token_accuracy": 0.18667061775922775,
      "num_tokens": 69594454.0,
      "step": 20830
    },
    {
      "entropy": 6.0811464309692385,
      "epoch": 4.064579065457028,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034807768734031343,
      "loss": 5.5303,
      "mean_token_accuracy": 0.1898088350892067,
      "num_tokens": 69612256.0,
      "step": 20835
    },
    {
      "entropy": 5.957011556625366,
      "epoch": 4.065554580040971,
      "grad_norm": 1.046875,
      "learning_rate": 0.000348011162685615,
      "loss": 5.4488,
      "mean_token_accuracy": 0.20397056341171266,
      "num_tokens": 69630135.0,
      "step": 20840
    },
    {
      "entropy": 6.116321420669555,
      "epoch": 4.066530094624914,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034794463089652604,
      "loss": 5.6115,
      "mean_token_accuracy": 0.1880885362625122,
      "num_tokens": 69647613.0,
      "step": 20845
    },
    {
      "entropy": 6.039923191070557,
      "epoch": 4.067505609208857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003478780919795477,
      "loss": 5.5455,
      "mean_token_accuracy": 0.20030542612075805,
      "num_tokens": 69665884.0,
      "step": 20850
    },
    {
      "entropy": 6.0164697647094725,
      "epoch": 4.068481123792801,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00034781154594118197,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1945302039384842,
      "num_tokens": 69682688.0,
      "step": 20855
    },
    {
      "entropy": 6.181608057022094,
      "epoch": 4.069456638376744,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00034774499278793155,
      "loss": 5.6956,
      "mean_token_accuracy": 0.17893177568912505,
      "num_tokens": 69699990.0,
      "step": 20860
    },
    {
      "entropy": 6.071349859237671,
      "epoch": 4.070432152960687,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003476784325262996,
      "loss": 5.5109,
      "mean_token_accuracy": 0.19042532294988632,
      "num_tokens": 69718250.0,
      "step": 20865
    },
    {
      "entropy": 6.108017873764038,
      "epoch": 4.07140766754463,
      "grad_norm": 1.0,
      "learning_rate": 0.0003476118651627903,
      "loss": 5.6206,
      "mean_token_accuracy": 0.18539695888757707,
      "num_tokens": 69734371.0,
      "step": 20870
    },
    {
      "entropy": 6.219491624832154,
      "epoch": 4.072383182128573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034754529070390826,
      "loss": 5.6884,
      "mean_token_accuracy": 0.180024915933609,
      "num_tokens": 69750299.0,
      "step": 20875
    },
    {
      "entropy": 6.149170970916748,
      "epoch": 4.073358696712516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000347478709156159,
      "loss": 5.6931,
      "mean_token_accuracy": 0.17836014777421952,
      "num_tokens": 69765956.0,
      "step": 20880
    },
    {
      "entropy": 6.033831310272217,
      "epoch": 4.074334211296459,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003474121205260486,
      "loss": 5.4811,
      "mean_token_accuracy": 0.18801862448453904,
      "num_tokens": 69782859.0,
      "step": 20885
    },
    {
      "entropy": 6.2068071365356445,
      "epoch": 4.075309725880402,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003473455248200839,
      "loss": 5.6575,
      "mean_token_accuracy": 0.17730453759431838,
      "num_tokens": 69799378.0,
      "step": 20890
    },
    {
      "entropy": 6.062890529632568,
      "epoch": 4.076285240464345,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034727892204477237,
      "loss": 5.5491,
      "mean_token_accuracy": 0.19725158363580703,
      "num_tokens": 69816917.0,
      "step": 20895
    },
    {
      "entropy": 6.1121100902557375,
      "epoch": 4.077260755048288,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00034721231220662216,
      "loss": 5.5796,
      "mean_token_accuracy": 0.19516492038965225,
      "num_tokens": 69834095.0,
      "step": 20900
    },
    {
      "entropy": 6.1904154300689695,
      "epoch": 4.078236269632231,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003471456953121421,
      "loss": 5.6409,
      "mean_token_accuracy": 0.18729536086320878,
      "num_tokens": 69850951.0,
      "step": 20905
    },
    {
      "entropy": 6.130520963668824,
      "epoch": 4.079211784216174,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034707907136784195,
      "loss": 5.5976,
      "mean_token_accuracy": 0.19040365219116212,
      "num_tokens": 69867318.0,
      "step": 20910
    },
    {
      "entropy": 5.9196878433227536,
      "epoch": 4.080187298800117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003470124403802318,
      "loss": 5.3766,
      "mean_token_accuracy": 0.2154011011123657,
      "num_tokens": 69883415.0,
      "step": 20915
    },
    {
      "entropy": 6.183736896514892,
      "epoch": 4.0811628133840605,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034694580235582255,
      "loss": 5.6469,
      "mean_token_accuracy": 0.18028580397367477,
      "num_tokens": 69898731.0,
      "step": 20920
    },
    {
      "entropy": 6.175652313232422,
      "epoch": 4.0821383279680035,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003468791573011261,
      "loss": 5.6202,
      "mean_token_accuracy": 0.1915121018886566,
      "num_tokens": 69916255.0,
      "step": 20925
    },
    {
      "entropy": 6.182034301757812,
      "epoch": 4.083113842551946,
      "grad_norm": 1.0,
      "learning_rate": 0.00034681250522265435,
      "loss": 5.7029,
      "mean_token_accuracy": 0.1859193801879883,
      "num_tokens": 69933773.0,
      "step": 20930
    },
    {
      "entropy": 6.120853614807129,
      "epoch": 4.084089357135889,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034674584612692067,
      "loss": 5.6069,
      "mean_token_accuracy": 0.18527360409498214,
      "num_tokens": 69949529.0,
      "step": 20935
    },
    {
      "entropy": 6.18389401435852,
      "epoch": 4.085064871719832,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003466791800204386,
      "loss": 5.6726,
      "mean_token_accuracy": 0.18746664971113206,
      "num_tokens": 69965262.0,
      "step": 20940
    },
    {
      "entropy": 6.216898584365845,
      "epoch": 4.086040386303775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003466125069097225,
      "loss": 5.6936,
      "mean_token_accuracy": 0.17249835878610612,
      "num_tokens": 69981185.0,
      "step": 20945
    },
    {
      "entropy": 6.059428405761719,
      "epoch": 4.087015900887718,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003465458268012875,
      "loss": 5.5926,
      "mean_token_accuracy": 0.18933395445346832,
      "num_tokens": 69997876.0,
      "step": 20950
    },
    {
      "entropy": 6.1891217708587645,
      "epoch": 4.087991415471661,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034647913970164924,
      "loss": 5.6319,
      "mean_token_accuracy": 0.17918548882007598,
      "num_tokens": 70013375.0,
      "step": 20955
    },
    {
      "entropy": 6.189695596694946,
      "epoch": 4.088966930055604,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003464124456173243,
      "loss": 5.6693,
      "mean_token_accuracy": 0.18615681380033494,
      "num_tokens": 70029231.0,
      "step": 20960
    },
    {
      "entropy": 6.1277268409729,
      "epoch": 4.089942444639547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003463457445548296,
      "loss": 5.6069,
      "mean_token_accuracy": 0.18197508305311202,
      "num_tokens": 70045313.0,
      "step": 20965
    },
    {
      "entropy": 6.145070362091064,
      "epoch": 4.09091795922349,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000346279036520683,
      "loss": 5.5878,
      "mean_token_accuracy": 0.18685027360916137,
      "num_tokens": 70061599.0,
      "step": 20970
    },
    {
      "entropy": 6.110782718658447,
      "epoch": 4.091893473807433,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000346212321521403,
      "loss": 5.604,
      "mean_token_accuracy": 0.19175069481134416,
      "num_tokens": 70078573.0,
      "step": 20975
    },
    {
      "entropy": 5.977773237228393,
      "epoch": 4.092868988391376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003461455995635088,
      "loss": 5.4661,
      "mean_token_accuracy": 0.20286388397216798,
      "num_tokens": 70094845.0,
      "step": 20980
    },
    {
      "entropy": 6.225396156311035,
      "epoch": 4.09384450297532,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00034607887065352024,
      "loss": 5.7044,
      "mean_token_accuracy": 0.17619217187166214,
      "num_tokens": 70112269.0,
      "step": 20985
    },
    {
      "entropy": 6.061975765228271,
      "epoch": 4.094820017559263,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034601213479795766,
      "loss": 5.4944,
      "mean_token_accuracy": 0.19834715873003006,
      "num_tokens": 70128827.0,
      "step": 20990
    },
    {
      "entropy": 6.102264404296875,
      "epoch": 4.095795532143206,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003459453920033424,
      "loss": 5.5868,
      "mean_token_accuracy": 0.1841529577970505,
      "num_tokens": 70145203.0,
      "step": 20995
    },
    {
      "entropy": 6.014283037185669,
      "epoch": 4.096771046727149,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003458786422761963,
      "loss": 5.4293,
      "mean_token_accuracy": 0.20151807367801666,
      "num_tokens": 70164015.0,
      "step": 21000
    },
    {
      "epoch": 4.096771046727149,
      "eval_entropy": 5.986044418392007,
      "eval_loss": 5.814789295196533,
      "eval_mean_token_accuracy": 0.18220698355269846,
      "eval_num_tokens": 70164015.0,
      "eval_runtime": 15.4016,
      "eval_samples_per_second": 1076.318,
      "eval_steps_per_second": 134.597,
      "step": 21000
    }
  ],
  "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": 9.48552847488e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}